RM0402

Reference manual
STM32F412 advanced Arm®-based 32-bit MCUs

Introduction
This reference manual targets application developers. It provides complete information on
how to use the memory and the peripherals of the STM32F412 microcontrollers.
The STM32F412 is a line of microcontrollers with different memory sizes, packages and
peripherals.
For ordering information, mechanical and electrical device characteristics refer to the
datasheet.
For information on the Arm Cortex®-M4 with FPU core, refer to the Cortex®-M4 with FPU
Technical Reference Manual.

Related documents
Available from STMicroelectronics web site www.st.com:
• STM32F412xE/xG datasheet
• PM0214 “STM32F3 and STM32F4 Series Cortex®-M4 with FPU-M4 programming
manual” for information on the ArmCortex®-M4 with FPU.

October 2020 RM0402 Rev 6 1/1163

www.st.com 1
Contents RM0402

Contents

1 Documentation conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.1 General information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.2 List of abbreviations for registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1.3 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.4 Availability of peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

2 System and memory overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

2.1 System architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.1.1 I-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.2 D-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.3 S-bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.4 DMA memory bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.5 DMA peripheral bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.6 BusMatrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.1.7 AHB/APB bridges (APB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2 Memory organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.2.2 Memory map and register boundary addresses . . . . . . . . . . . . . . . . . . 50
2.3 Embedded SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.4 Flash memory overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.5 Bit banding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.6 Boot configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3 Embedded Flash memory interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2 Main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.3 Embedded Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.4 Read interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.4.1 Relation between CPU clock frequency and Flash memory read time . 60
3.4.2 Adaptive real-time memory accelerator (ART Accelerator™) . . . . . . . . 61
3.5 Erase and program operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.5.1 Unlocking the Flash control register . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.5.2 Program/erase parallelism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

2/1163 RM0402 Rev 6

RM0402 Contents

3.5.3 Erase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.5.4 Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.5.5 Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.6 Option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.6.1 Description of user option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.6.2 Programming user option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.6.3 Read protection (RDP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3.6.4 Write protections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.6.5 Proprietary code readout protection (PCROP) . . . . . . . . . . . . . . . . . . . 71
3.7 One-time programmable bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.8 Flash interface registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.8.1 Flash access control register (FLASH_ACR) . . . . . . . . . . . . . . . . . . . . 74
3.8.2 Flash key register (FLASH_KEYR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.8.3 Flash option key register (FLASH_OPTKEYR) . . . . . . . . . . . . . . . . . . . 75
3.8.4 Flash status register (FLASH_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.8.5 Flash control register (FLASH_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.8.6 Flash option control register (FLASH_OPTCR) . . . . . . . . . . . . . . . . . . . 78
3.8.7 Flash interface register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

4 CRC calculation unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

4.1 CRC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.2 CRC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3 CRC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.4 CRC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.4.1 Data register (CRC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.4.2 Independent data register (CRC_IDR) . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.4.3 Control register (CRC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.4.4 CRC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

5 Power controller (PWR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

5.1 Power supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.1.1 Independent A/D converter supply and reference voltage . . . . . . . . . . . 87
5.1.2 Battery backup domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.1.3 Voltage regulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2 Power supply supervisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.2.1 Power-on reset (POR)/power-down reset (PDR) . . . . . . . . . . . . . . . . . . 90

RM0402 Rev 6 3/1163

31
Contents RM0402

5.2.2 Brownout reset (BOR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

5.2.3 Programmable voltage detector (PVD) . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3 Low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.3.1 Slowing down system clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.3.2 Peripheral clock gating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.3.3 Sleep mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.3.4 Batch acquisition mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.3.5 Stop mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.3.6 Standby mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.3.7 Programming the RTC alternate functions to wake up the device from
the Stop and Standby modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.4 Power control registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.4.1 PWR power control register (PWR_CR) . . . . . . . . . . . . . . . . . . . . . . . 104
5.4.2 PWR power control/status register (PWR_CSR) . . . . . . . . . . . . . . . . . 106
5.5 PWR register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

6 Reset and clock control (RCC) for STM32F412xx . . . . . . . . . . . . . . . 109

6.1 Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.1.1 System reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.1.2 Power reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.1.3 Backup domain reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.2 Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.2.1 HSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.2.2 HSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.2.3 PLL configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.2.4 LSE clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.2.5 LSI clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.2.6 System clock (SYSCLK) selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.2.7 Clock security system (CSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
6.2.8 RTC/AWU clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.2.9 Watchdog clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.2.10 Clock-out capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.2.11 Internal/external clock measurement using TIM5/TIM11 . . . . . . . . . . . 118
6.3 RCC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.3.1 RCC clock control register (RCC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.3.2 RCC PLL configuration register (RCC_PLLCFGR) . . . . . . . . . . . . . . . 123
6.3.3 RCC clock configuration register (RCC_CFGR) . . . . . . . . . . . . . . . . . 125

4/1163 RM0402 Rev 6

RM0402 Contents

6.3.4 RCC clock interrupt register (RCC_CIR) . . . . . . . . . . . . . . . . . . . . . . . 128

6.3.5 RCC AHB1 peripheral reset register (RCC_AHB1RSTR) . . . . . . . . . . 130
6.3.6 RCC AHB2 peripheral reset register (RCC_AHB2RSTR) . . . . . . . . . 132
6.3.7 RCC AHB3 peripheral reset register (RCC_AHB3RSTR) . . . . . . . . . . 133
6.3.8 RCC APB1 peripheral reset register for (RCC_APB1RSTR) . . . . . . . . 133
6.3.9 RCC APB2 peripheral reset register (RCC_APB2RSTR) . . . . . . . . . . 136
6.3.10 RCC AHB1 peripheral clock enable register (RCC_AHB1ENR) . . . . . 138
6.3.11 RCC AHB2 peripheral clock enable register (RCC_AHB2ENR) . . . . . 140
6.3.12 RCC AHB3 peripheral clock enable register (RCC_AHB3ENR) . . . . . 141
6.3.13 RCC APB1 peripheral clock enable register (RCC_APB1ENR) . . . . . 141
6.3.14 RCC APB2 peripheral clock enable register
(RCC_APB2ENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6.3.15 RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
6.3.16 RCC AHB2 peripheral clock enable in low power mode register
(RCC_AHB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.3.17 RCC AHB3 peripheral clock enable in low power mode register
(RCC_AHB3LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.3.18 RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.3.19 RCC APB2 peripheral clock enabled in low power mode register
(RCC_APB2LPENR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.3.20 RCC Backup domain control register (RCC_BDCR) . . . . . . . . . . . . . . 155
6.3.21 RCC clock control & status register (RCC_CSR) . . . . . . . . . . . . . . . . 156
6.3.22 RCC spread spectrum clock generation register (RCC_SSCGR) . . . . 158
6.3.23 RCC PLLI2S configuration register (RCC_PLLI2SCFGR) . . . . . . . . . 159
6.3.24 RCC Dedicated Clocks Configuration Register (RCC_DCKCFGR) . . 161
6.3.25 RCC clocks gated enable register (CKGATENR) . . . . . . . . . . . . . . . . 162
6.3.26 RCC Dedicated Clocks Configuration Register (RCC_DCKCFGR2) . 163
6.3.27 RCC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

7 General-purpose I/Os (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

7.1 GPIO introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.2 GPIO main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.3 GPIO functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.3.1 General-purpose I/O (GPIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.3.2 I/O pin multiplexer and mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
7.3.3 I/O port control registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
7.3.4 I/O port data registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

RM0402 Rev 6 5/1163

31
Contents RM0402

7.3.5 I/O data bitwise handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

7.3.6 GPIO locking mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
7.3.7 I/O alternate function input/output . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
7.3.8 External interrupt/wakeup lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
7.3.9 Input configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
7.3.10 Output configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
7.3.11 Alternate function configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
7.3.12 Analog configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7.3.13 Using the OSC32_IN/OSC32_OUT pins as GPIO PC14/PC15
port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7.3.14 Using the OSC_IN/OSC_OUT pins as GPIO PH0/PH1 port pins . . . . 177
7.3.15 Selection of RTC additional functions . . . . . . . . . . . . . . . . . . . . . . . . . 178
7.4 GPIO registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
7.4.1 GPIO port mode register (GPIOx_MODER) (x = A...H) . . . . . . . . . . . . 179
7.4.2 GPIO port output type register (GPIOx_OTYPER)
(x = A...H) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
7.4.3 GPIO port output speed register (GPIOx_OSPEEDR)
(x = A...H) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
7.4.4 GPIO port pull-up/pull-down register (GPIOx_PUPDR)
(x = A...H) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
7.4.5 GPIO port input data register (GPIOx_IDR) (x = A...H) . . . . . . . . . . . . 181
7.4.6 GPIO port output data register (GPIOx_ODR) (x = A...H) . . . . . . . . . . 181
7.4.7 GPIO port bit set/reset register (GPIOx_BSRR) (x = A...H) . . . . . . . . . 181
7.4.8 GPIO port configuration lock register (GPIOx_LCKR)
(x = A...H) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
7.4.9 GPIO alternate function low register (GPIOx_AFRL) (x = A...H) . . . . . 183
7.4.10 GPIO alternate function high register (GPIOx_AFRH)
(x = A...H) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
7.4.11 GPIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184

8 System configuration controller (SYSCFG) . . . . . . . . . . . . . . . . . . . . 187

8.1 I/O compensation cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
8.2 SYSCFG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
8.2.1 SYSCFG memory remap register (SYSCFG_MEMRMP) . . . . . . . . . . 187
8.2.2 SYSCFG peripheral mode configuration register (SYSCFG_PMC) . . 188
8.2.3 SYSCFG external interrupt configuration register 1
(SYSCFG_EXTICR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.2.4 SYSCFG external interrupt configuration register 2
(SYSCFG_EXTICR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

6/1163 RM0402 Rev 6

RM0402 Contents

8.2.5 SYSCFG external interrupt configuration register 3

(SYSCFG_EXTICR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
8.2.6 SYSCFG external interrupt configuration register 4
(SYSCFG_EXTICR4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
8.2.7 SYSCFG configuration register 2 (SYSCFG_CFGR2) . . . . . . . . . . . . 191
8.2.8 Compensation cell control register (SYSCFG_CMPCR) . . . . . . . . . . . 192
8.2.9 SYSCFG configuration register (SYSCFG_CFGR) . . . . . . . . . . . . . . . 193
8.2.10 SYSCFG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

9 Direct memory access controller (DMA) . . . . . . . . . . . . . . . . . . . . . . . 195

9.1 DMA introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
9.2 DMA main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
9.3 DMA functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
9.3.1 DMA block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
9.3.2 DMA overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
9.3.3 DMA transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
9.3.4 Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
9.3.5 Arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
9.3.6 DMA streams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
9.3.7 Source, destination and transfer modes . . . . . . . . . . . . . . . . . . . . . . . 200
9.3.8 Pointer incrementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
9.3.9 Circular mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
9.3.10 Double-buffer mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
9.3.11 Programmable data width, packing/unpacking, endianness . . . . . . . . 206
9.3.12 Single and burst transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
9.3.13 FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
9.3.14 DMA transfer completion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
9.3.15 DMA transfer suspension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
9.3.16 Flow controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
9.3.17 Summary of the possible DMA configurations . . . . . . . . . . . . . . . . . . . 214
9.3.18 Stream configuration procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
9.3.19 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.4 DMA interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
9.5 DMA registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
9.5.1 DMA low interrupt status register (DMA_LISR) . . . . . . . . . . . . . . . . . . 217
9.5.2 DMA high interrupt status register (DMA_HISR) . . . . . . . . . . . . . . . . . 218
9.5.3 DMA low interrupt flag clear register (DMA_LIFCR) . . . . . . . . . . . . . . 219

RM0402 Rev 6 7/1163

31
Contents RM0402

9.5.4 DMA high interrupt flag clear register (DMA_HIFCR) . . . . . . . . . . . . . 219

9.5.5 DMA stream x configuration register (DMA_SxCR) . . . . . . . . . . . . . . . 220
9.5.6 DMA stream x number of data register (DMA_SxNDTR) . . . . . . . . . . 223
9.5.7 DMA stream x peripheral address register (DMA_SxPAR) . . . . . . . . . 224
9.5.8 DMA stream x memory 0 address register
(DMA_SxM0AR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
9.5.9 DMA stream x memory 1 address register
(DMA_SxM1AR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
9.5.10 DMA stream x FIFO control register (DMA_SxFCR) . . . . . . . . . . . . . . 225
9.5.11 DMA register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

10 Interrupts and events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

10.1 Nested vectored interrupt controller (NVIC) . . . . . . . . . . . . . . . . . . . . . . 231
10.1.1 NVIC features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
10.1.2 SysTick calibration value register . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
10.1.3 Interrupt and exception vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
10.2 External interrupt/event controller (EXTI) . . . . . . . . . . . . . . . . . . . . . . . . 231
10.2.1 EXTI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
10.2.2 EXTI block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
10.2.3 Wakeup event management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
10.2.4 Functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
10.2.5 External interrupt/event line mapping . . . . . . . . . . . . . . . . . . . . . . . . . 238
10.3 EXTI registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
10.3.1 Interrupt mask register (EXTI_IMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
10.3.2 Event mask register (EXTI_EMR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
10.3.3 Rising trigger selection register (EXTI_RTSR) . . . . . . . . . . . . . . . . . . 241
10.3.4 Falling trigger selection register (EXTI_FTSR) . . . . . . . . . . . . . . . . . . 242
10.3.5 Software interrupt event register (EXTI_SWIER) . . . . . . . . . . . . . . . . 243
10.3.6 Pending register (EXTI_PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
10.3.7 EXTI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245

11 Flexible static memory controller (FSMC) . . . . . . . . . . . . . . . . . . . . . 246

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
11.2 FSMC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
11.3 FMC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
11.4 AHB interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
11.4.1 Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . . 248

8/1163 RM0402 Rev 6

RM0402 Contents

11.5 External device address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249

11.5.1 NOR/PSRAM address mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
11.6 NOR Flash/PSRAM controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
11.6.1 External memory interface signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.6.2 Supported memories and transactions . . . . . . . . . . . . . . . . . . . . . . . . 253
11.6.3 General timing rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
11.6.4 NOR Flash/PSRAM controller asynchronous transactions . . . . . . . . . 255
11.6.5 Synchronous transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
11.6.6 NOR/PSRAM controller registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
11.6.7 FSMC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286

12 Quad-SPI interface (QUADSPI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288

12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
12.2 QUADSPI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
12.3 QUADSPI functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
12.3.1 QUADSPI block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
12.3.2 QUADSPI pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
12.3.3 QUADSPI command sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
12.3.4 QUADSPI signal interface protocol modes . . . . . . . . . . . . . . . . . . . . . 292
12.3.5 QUADSPI indirect mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
12.3.6 QUADSPI status flag polling mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
12.3.7 QUADSPI memory-mapped mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
12.3.8 QUADSPI Flash memory configuration . . . . . . . . . . . . . . . . . . . . . . . . 297
12.3.9 QUADSPI delayed data sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
12.3.10 QUADSPI configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
12.3.11 QUADSPI usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
12.3.12 Sending the instruction only once . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
12.3.13 QUADSPI error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
12.3.14 QUADSPI busy bit and abort functionality . . . . . . . . . . . . . . . . . . . . . . 300
12.3.15 nCS behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
12.4 QUADSPI interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
12.5 QUADSPI registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
12.5.1 QUADSPI control register (QUADSPI_CR) . . . . . . . . . . . . . . . . . . . . . 304
12.5.2 QUADSPI device configuration register (QUADSPI_DCR) . . . . . . . . . 307
12.5.3 QUADSPI status register (QUADSPI_SR) . . . . . . . . . . . . . . . . . . . . . 308
12.5.4 QUADSPI flag clear register (QUADSPI_FCR) . . . . . . . . . . . . . . . . . . 309

RM0402 Rev 6 9/1163

31
Contents RM0402

12.5.5 QUADSPI data length register (QUADSPI_DLR) . . . . . . . . . . . . . . . . 309

12.5.6 QUADSPI communication configuration register (QUADSPI_CCR) . . 310
12.5.7 QUADSPI address register (QUADSPI_AR) . . . . . . . . . . . . . . . . . . . . 312
12.5.8 QUADSPI alternate bytes registers (QUADSPI_ABR) . . . . . . . . . . . . 313
12.5.9 QUADSPI data register (QUADSPI_DR) . . . . . . . . . . . . . . . . . . . . . . . 313
12.5.10 QUADSPI polling status mask register (QUADSPI_PSMKR) . . . . . . . 314
12.5.11 QUADSPI polling status match register (QUADSPI_PSMAR) . . . . . . 314
12.5.12 QUADSPI polling interval register (QUADSPI_PIR) . . . . . . . . . . . . . . 315
12.5.13 QUADSPI low-power timeout register (QUADSPI_LPTR) . . . . . . . . . . 315
12.5.14 QUADSPI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316

13 Analog-to-digital converter (ADC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

13.1 ADC introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
13.2 ADC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
13.3 ADC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
13.3.1 ADC on-off control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
13.3.2 ADC clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
13.3.3 Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
13.3.4 Single conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
13.3.5 Continuous conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
13.3.6 Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
13.3.7 Analog watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
13.3.8 Scan mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
13.3.9 Injected channel management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
13.3.10 Discontinuous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
13.4 Data alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
13.5 Channel-wise programmable sampling time . . . . . . . . . . . . . . . . . . . . . 326
13.6 Conversion on external trigger and trigger polarity . . . . . . . . . . . . . . . . 326
13.7 Fast conversion mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
13.8 Data management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
13.8.1 Using the DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
13.8.2 Managing a sequence of conversions without using the DMA . . . . . . 329
13.8.3 Conversions without DMA and without overrun detection . . . . . . . . . . 330
13.9 Temperature sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
13.10 Battery charge monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
13.11 ADC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332

10/1163 RM0402 Rev 6

RM0402 Contents

13.12 ADC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333

13.12.1 ADC status register (ADC_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
13.12.2 ADC control register 1 (ADC_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
13.12.3 ADC control register 2 (ADC_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
13.12.4 ADC sample time register 1 (ADC_SMPR1) . . . . . . . . . . . . . . . . . . . . 338
13.12.5 ADC sample time register 2 (ADC_SMPR2) . . . . . . . . . . . . . . . . . . . . 339
13.12.6 ADC injected channel data offset register x (ADC_JOFRx) (x=1..4) . . 339
13.12.7 ADC watchdog higher threshold register (ADC_HTR) . . . . . . . . . . . . . 339
13.12.8 ADC watchdog lower threshold register (ADC_LTR) . . . . . . . . . . . . . . 340
13.12.9 ADC regular sequence register 1 (ADC_SQR1) . . . . . . . . . . . . . . . . . 340
13.12.10 ADC regular sequence register 2 (ADC_SQR2) . . . . . . . . . . . . . . . . . 341
13.12.11 ADC regular sequence register 3 (ADC_SQR3) . . . . . . . . . . . . . . . . . 342
13.12.12 ADC injected sequence register (ADC_JSQR) . . . . . . . . . . . . . . . . . . 343
13.12.13 ADC injected data register x (ADC_JDRx) (x= 1..4) . . . . . . . . . . . . . . 343
13.12.14 ADC regular data register (ADC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . 344
13.12.15 ADC Common status register (ADC_CSR) . . . . . . . . . . . . . . . . . . . . . 344
13.12.16 ADC common control register (ADC_CCR) . . . . . . . . . . . . . . . . . . . . . 345
13.13 ADC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346

14 Digital filter for sigma delta modulators (DFSDM) . . . . . . . . . . . . . . . 348

14.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
14.2 DFSDM main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
14.3 DFSDM implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
14.4 DFSDM functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
14.4.1 DFSDM block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
14.4.2 DFSDM pins and internal signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
14.4.3 DFSDM reset and clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
14.4.4 Serial channel transceivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
14.4.5 Configuring the input serial interface . . . . . . . . . . . . . . . . . . . . . . . . . . 363
14.4.6 Parallel data inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
14.4.7 Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
14.4.8 Digital filter configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
14.4.9 Integrator unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
14.4.10 Analog watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
14.4.11 Short-circuit detector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
14.4.12 Extreme detector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370

RM0402 Rev 6 11/1163

31
Contents RM0402

14.4.13 Data unit block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

14.4.14 Signed data format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
14.4.15 Launching conversions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
14.4.16 Continuous and fast continuous modes . . . . . . . . . . . . . . . . . . . . . . . . 373
14.4.17 Request precedence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
14.4.18 Power optimization in run mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
14.5 DFSDM interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
14.6 DFSDM DMA transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
14.7 DFSDM channel y registers (y=0..3) . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
14.7.1 DFSDM channel y configuration register (DFSDM_CHyCFGR1) . . . . 376
14.7.2 DFSDM channel y configuration register (DFSDM_CHyCFGR2) . . . . 379
14.7.3 DFSDM channel y analog watchdog and short-circuit detector register
(DFSDM_CHyAWSCDR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
14.7.4 DFSDM channel y watchdog filter data register
(DFSDM_CHyWDATR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
14.7.5 DFSDM channel y data input register (DFSDM_CHyDATINR) . . . . . . 381
14.8 DFSDM filter x module registers (x=0..1) . . . . . . . . . . . . . . . . . . . . . . . . 382
14.8.1 DFSDM filter x control register 1 (DFSDM_FLTxCR1) . . . . . . . . . . . . 382
14.8.2 DFSDM filter x control register 2 (DFSDM_FLTxCR2) . . . . . . . . . . . . 384
14.8.3 DFSDM filter x interrupt and status register (DFSDM_FLTxISR) . . . . . 386
14.8.4 DFSDM filter x interrupt flag clear register (DFSDM_FLTxICR) . . . . . 387
14.8.5 DFSDM filter x injected channel group selection register
(DFSDM_FLTxJCHGR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
14.8.6 DFSDM filter x control register (DFSDM_FLTxFCR) . . . . . . . . . . . . . . 389
14.8.7 DFSDM filter x data register for injected group
(DFSDM_FLTxJDATAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390
14.8.8 DFSDM filter x data register for the regular channel
(DFSDM_FLTxRDATAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
14.8.9 DFSDM filter x analog watchdog high threshold register
(DFSDM_FLTxAWHTR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
14.8.10 DFSDM filter x analog watchdog low threshold register
(DFSDM_FLTxAWLTR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
14.8.11 DFSDM filter x analog watchdog status register
(DFSDM_FLTxAWSR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
14.8.12 DFSDM filter x analog watchdog clear flag register
(DFSDM_FLTxAWCFR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
14.8.13 DFSDM filter x extremes detector maximum register
(DFSDM_FLTxEXMAX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
14.8.14 DFSDM filter x extremes detector minimum register
(DFSDM_FLTxEXMIN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394

12/1163 RM0402 Rev 6

RM0402 Contents

14.8.15 DFSDM filter x conversion timer register (DFSDM_FLTxCNVTIMR) . . 395

14.8.16 DFSDM register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396

15 True random number generator (RNG) . . . . . . . . . . . . . . . . . . . . . . . . 401

15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
15.2 RNG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
15.3 RNG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
15.3.1 RNG block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
15.3.2 RNG internal signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
15.3.3 Random number generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
15.3.4 RNG initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
15.3.5 RNG operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
15.3.6 RNG clocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
15.3.7 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
15.3.8 RNG low-power usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
15.4 RNG interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
15.5 RNG processing time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
15.6 RNG entropy source validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
15.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
15.6.2 Validation conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
15.6.3 Data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
15.7 RNG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
15.7.1 RNG control register (RNG_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
15.7.2 RNG status register (RNG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
15.7.3 RNG data register (RNG_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
15.7.4 RNG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412

16 Advanced-control timers (TIM1&TIM8) . . . . . . . . . . . . . . . . . . . . . . . . 413

16.1 TIM1&TIM8 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
16.2 TIM1&TIM8 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
16.3 TIM1&TIM8 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
16.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
16.3.2 Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
16.3.3 Repetition counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
16.3.4 Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
16.3.5 Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

RM0402 Rev 6 13/1163

31
Contents RM0402

16.3.6 Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434

16.3.7 PWM input mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
16.3.8 Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
16.3.9 Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
16.3.10 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
16.3.11 Complementary outputs and dead-time insertion . . . . . . . . . . . . . . . . 440
16.3.12 Using the break function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
16.3.13 Clearing the OCxREF signal on an external event . . . . . . . . . . . . . . . 445
16.3.14 6-step PWM generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
16.3.15 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
16.3.16 Encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
16.3.17 Timer input XOR function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
16.3.18 Interfacing with Hall sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
16.3.19 TIMx and external trigger synchronization . . . . . . . . . . . . . . . . . . . . . . 453
16.3.20 Timer synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
16.3.21 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
16.4 TIM1&TIM8 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
16.4.1 TIM1&TIM8 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . 457
16.4.2 TIM1&TIM8 control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . . . . 458
16.4.3 TIM1&TIM8 slave mode control register (TIMx_SMCR) . . . . . . . . . . . 460
16.4.4 TIM1&TIM8 DMA/interrupt enable register (TIMx_DIER) . . . . . . . . . . 462
16.4.5 TIM1&TIM8 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . 464
16.4.6 TIM1&TIM8 event generation register (TIMx_EGR) . . . . . . . . . . . . . . 465
16.4.7 TIM1&TIM8 capture/compare mode register 1 (TIMx_CCMR1) . . . . . 467
16.4.8 TIM1&TIM8 capture/compare mode register 2 (TIMx_CCMR2) . . . . . 470
16.4.9 TIM1&TIM8 capture/compare enable register (TIMx_CCER) . . . . . . . 471
16.4.10 TIM1&TIM8 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
16.4.11 TIM1&TIM8 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
16.4.12 TIM1&TIM8 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . 475
16.4.13 TIM1&TIM8 repetition counter register (TIMx_RCR) . . . . . . . . . . . . . . 476
16.4.14 TIM1&TIM8 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . 476
16.4.15 TIM1&TIM8 capture/compare register 2 (TIMx_CCR2) . . . . . . . . . . . . 477
16.4.16 TIM1&TIM8 capture/compare register 3 (TIMx_CCR3) . . . . . . . . . . . . 477
16.4.17 TIM1&TIM8 capture/compare register 4 (TIMx_CCR4) . . . . . . . . . . . . 478
16.4.18 TIM1&TIM8 break and dead-time register (TIMx_BDTR) . . . . . . . . . . 478
16.4.19 TIM1&TIM8 DMA control register (TIMx_DCR) . . . . . . . . . . . . . . . . . . 480
16.4.20 TIM1&TIM8 DMA address for full transfer (TIMx_DMAR) . . . . . . . . . . 481

14/1163 RM0402 Rev 6

RM0402 Contents

16.4.21 TIM1&TIM8 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482

17 General-purpose timers (TIM2 to TIM5) . . . . . . . . . . . . . . . . . . . . . . . . 484

17.1 TIM2 to TIM5 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
17.2 TIM2 to TIM5 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
17.3 TIM2 to TIM5 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
17.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
17.3.2 Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
17.3.3 Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
17.3.4 Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
17.3.5 Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
17.3.6 PWM input mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
17.3.7 Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
17.3.8 Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
17.3.9 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
17.3.10 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
17.3.11 Clearing the OCxREF signal on an external event . . . . . . . . . . . . . . . 509
17.3.12 Encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
17.3.13 Timer input XOR function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
17.3.14 Timers and external trigger synchronization . . . . . . . . . . . . . . . . . . . . 513
17.3.15 Timer synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
17.3.16 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
17.4 TIM2 to TIM5 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
17.4.1 TIMx control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . 522
17.4.2 TIMx control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . 524
17.4.3 TIMx slave mode control register (TIMx_SMCR) . . . . . . . . . . . . . . . . . 525
17.4.4 TIMx DMA/Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . . . 527
17.4.5 TIMx status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
17.4.6 TIMx event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . . . . 530
17.4.7 TIMx capture/compare mode register 1 (TIMx_CCMR1) . . . . . . . . . . . 531
17.4.8 TIMx capture/compare mode register 2 (TIMx_CCMR2) . . . . . . . . . . . 534
17.4.9 TIMx capture/compare enable register (TIMx_CCER) . . . . . . . . . . . . 535
17.4.10 TIMx counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
17.4.11 TIMx prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
17.4.12 TIMx auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . . . . 537
17.4.13 TIMx capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . . . . . . 538
17.4.14 TIMx capture/compare register 2 (TIMx_CCR2) . . . . . . . . . . . . . . . . . 538

RM0402 Rev 6 15/1163

31
Contents RM0402

17.4.15 TIMx capture/compare register 3 (TIMx_CCR3) . . . . . . . . . . . . . . . . . 539

17.4.16 TIMx capture/compare register 4 (TIMx_CCR4) . . . . . . . . . . . . . . . . . 539
17.4.17 TIMx DMA control register (TIMx_DCR) . . . . . . . . . . . . . . . . . . . . . . . 540
17.4.18 TIMx DMA address for full transfer (TIMx_DMAR) . . . . . . . . . . . . . . . 540
17.4.19 TIM2 option register (TIM2_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
17.4.20 TIM5 option register (TIM5_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
17.4.21 TIMx register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543

18 General-purpose timers (TIM9 to TIM14) . . . . . . . . . . . . . . . . . . . . . . . 545

18.1 TIM9 to TIM14 introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
18.2 TIM9 to TIM14 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
18.2.1 TIM9/TIM12 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
18.2.2 TIM10/TIM11 and TIM13/TIM14 main features . . . . . . . . . . . . . . . . . . 546
18.3 TIM9 to TIM14 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
18.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
18.3.2 Counter modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
18.3.3 Clock selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553
18.3.4 Capture/compare channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
18.3.5 Input capture mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
18.3.6 PWM input mode (only for TIM9/12) . . . . . . . . . . . . . . . . . . . . . . . . . . 557
18.3.7 Forced output mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
18.3.8 Output compare mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
18.3.9 PWM mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
18.3.10 One-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
18.3.11 TIM9/12 external trigger synchronization . . . . . . . . . . . . . . . . . . . . . . . 563
18.3.12 Timer synchronization (TIM9/12) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
18.3.13 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
18.4 TIM9 and TIM12 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
18.4.1 TIM9/12 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . 566
18.4.2 TIM9/12 slave mode control register (TIMx_SMCR) . . . . . . . . . . . . . . 568
18.4.3 TIM9/12 Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . . . . 569
18.4.4 TIM9/12 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
18.4.5 TIM9/12 event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . 572
18.4.6 TIM9/12 capture/compare mode register 1 (TIMx_CCMR1) . . . . . . . . 572
18.4.7 TIM9/12 capture/compare enable register (TIMx_CCER) . . . . . . . . . . 576
18.4.8 TIM9/12 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
18.4.9 TIM9/12 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577

16/1163 RM0402 Rev 6

RM0402 Contents

18.4.10 TIM9/12 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . 577

18.4.11 TIM9/12 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . . . . . . 578
18.4.12 TIM9/12 capture/compare register 2 (TIMx_CCR2) . . . . . . . . . . . . . . 578
18.4.13 TIM9/12 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
18.5 TIM10/11/13/14 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 581
18.5.1 TIM10/11/13/14 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . 581
18.5.2 TIM10/11/13/14 Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . 582
18.5.3 TIM10/11/13/14 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . 582
18.5.4 TIM10/11/13/14 event generation register (TIMx_EGR) . . . . . . . . . . . 583
18.5.5 TIM10/11/13/14 capture/compare mode register 1
(TIMx_CCMR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
18.5.6 TIM10/11/13/14 capture/compare enable register
(TIMx_CCER) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
18.5.7 TIM10/11/13/14 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . 588
18.5.8 TIM10/11/13/14 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . 588
18.5.9 TIM10/11/13/14 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . 588
18.5.10 TIM10/11/13/14 capture/compare register 1 (TIMx_CCR1) . . . . . . . . . 589
18.5.11 TIM11 option register 1 (TIM11_OR) . . . . . . . . . . . . . . . . . . . . . . . . . . 589
18.5.12 TIM10/11/13/14 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590

19 Basic timers (TIM6/7) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592

19.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
19.2 TIM6/7 main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
19.3 TIM6/7 functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
19.3.1 Time-base unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
19.3.2 Counting mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
19.3.3 Clock source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
19.3.4 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
19.4 TIM6/7 registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
19.4.1 TIM6/7 control register 1 (TIMx_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . 600
19.4.2 TIM6/7 control register 2 (TIMx_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . 601
19.4.3 TIM6/7 DMA/Interrupt enable register (TIMx_DIER) . . . . . . . . . . . . . . 601
19.4.4 TIM6/7 status register (TIMx_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
19.4.5 TIM6/7 event generation register (TIMx_EGR) . . . . . . . . . . . . . . . . . . 602
19.4.6 TIM6/7 counter (TIMx_CNT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
19.4.7 TIM6/7 prescaler (TIMx_PSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
19.4.8 TIM6/7 auto-reload register (TIMx_ARR) . . . . . . . . . . . . . . . . . . . . . . 603

RM0402 Rev 6 17/1163

31
Contents RM0402

19.4.9 TIM6/7 register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604

20 Independent watchdog (IWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605

20.1 IWDG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
20.2 IWDG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
20.3 IWDG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
20.3.1 Hardware watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
20.3.2 Register access protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
20.3.3 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
20.4 IWDG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
20.4.1 Key register (IWDG_KR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
20.4.2 Prescaler register (IWDG_PR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
20.4.3 Reload register (IWDG_RLR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
20.4.4 Status register (IWDG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
20.4.5 IWDG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610

21 Window watchdog (WWDG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611

21.1 WWDG introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .611
21.2 WWDG main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .611
21.3 WWDG functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .611
21.4 How to program the watchdog timeout . . . . . . . . . . . . . . . . . . . . . . . . . . 613
21.5 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
21.6 WWDG registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
21.6.1 Control register (WWDG_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
21.6.2 Configuration register (WWDG_CFR) . . . . . . . . . . . . . . . . . . . . . . . . . 616
21.6.3 Status register (WWDG_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616
21.6.4 WWDG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617

22 Real-time clock (RTC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618

22.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
22.2 RTC main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618
22.3 RTC functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
22.3.1 Clock and prescalers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
22.3.2 Real-time clock and calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
22.3.3 Programmable alarms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
22.3.4 Periodic auto-wakeup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621

18/1163 RM0402 Rev 6

RM0402 Contents

22.3.5 RTC initialization and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . 622

22.3.6 Reading the calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
22.3.7 Resetting the RTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
22.3.8 RTC synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
22.3.9 RTC reference clock detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
22.3.10 RTC coarse digital calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626
22.3.11 RTC smooth digital calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627
22.3.12 Timestamp function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 629
22.3.13 Tamper detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
22.3.14 Calibration clock output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
22.3.15 Alarm output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
22.4 RTC and low power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
22.5 RTC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
22.6 RTC registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
22.6.1 RTC time register (RTC_TR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
22.6.2 RTC date register (RTC_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635
22.6.3 RTC control register (RTC_CR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636
22.6.4 RTC initialization and status register (RTC_ISR) . . . . . . . . . . . . . . . . . 638
22.6.5 RTC prescaler register (RTC_PRER) . . . . . . . . . . . . . . . . . . . . . . . . . 640
22.6.6 RTC wakeup timer register (RTC_WUTR) . . . . . . . . . . . . . . . . . . . . . . 641
22.6.7 RTC calibration register (RTC_CALIBR) . . . . . . . . . . . . . . . . . . . . . . . 641
22.6.8 RTC alarm A register (RTC_ALRMAR) . . . . . . . . . . . . . . . . . . . . . . . . 643
22.6.9 RTC alarm B register (RTC_ALRMBR) . . . . . . . . . . . . . . . . . . . . . . . . 644
22.6.10 RTC write protection register (RTC_WPR) . . . . . . . . . . . . . . . . . . . . . 645
22.6.11 RTC sub second register (RTC_SSR) . . . . . . . . . . . . . . . . . . . . . . . . . 645
22.6.12 RTC shift control register (RTC_SHIFTR) . . . . . . . . . . . . . . . . . . . . . . 646
22.6.13 RTC time stamp time register (RTC_TSTR) . . . . . . . . . . . . . . . . . . . . 647
22.6.14 RTC time stamp date register (RTC_TSDR) . . . . . . . . . . . . . . . . . . . . 647
22.6.15 RTC timestamp sub second register (RTC_TSSSR) . . . . . . . . . . . . . . 648
22.6.16 RTC calibration register (RTC_CALR) . . . . . . . . . . . . . . . . . . . . . . . . . 648
22.6.17 RTC tamper and alternate function configuration register
(RTC_TAFCR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
22.6.18 RTC alarm A sub second register (RTC_ALRMASSR) . . . . . . . . . . . . 651
22.6.19 RTC alarm B sub second register (RTC_ALRMBSSR) . . . . . . . . . . . . 652
22.6.20 RTC backup registers (RTC_BKPxR) . . . . . . . . . . . . . . . . . . . . . . . . . 653
22.6.21 RTC register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653

RM0402 Rev 6 19/1163

31
Contents RM0402

23 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface . . . . . . . 656

23.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
23.2 FMPI2C main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
23.3 FMPI2C implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
23.4 FMPI2C functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
23.4.1 FMPI2C block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
23.4.2 FMPI2C pins and internal signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
23.4.3 FMPI2C clock requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
23.4.4 Mode selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
23.4.5 FMPI2C initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
23.4.6 Software reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
23.4.7 Data transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
23.4.8 FMPI2C slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
23.4.9 FMPI2C master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
23.4.10 FMPI2C_TIMINGR register configuration examples . . . . . . . . . . . . . . 689
23.4.11 SMBus specific features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690
23.4.12 SMBus initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693
23.4.13 SMBus: FMPI2C_TIMEOUTR register configuration examples . . . . . 695
23.4.14 SMBus slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
23.4.15 Error conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
23.4.16 DMA requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
23.4.17 Debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
23.5 FMPI2C low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
23.6 FMPI2C interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
23.7 FMPI2C registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707
23.7.1 FMPI2C control register 1 (FMPI2C_CR1) . . . . . . . . . . . . . . . . . . . . . 707
23.7.2 FMPI2C control register 2 (FMPI2C_CR2) . . . . . . . . . . . . . . . . . . . . . 709
23.7.3 FMPI2C own address 1 register (FMPI2C_OAR1) . . . . . . . . . . . . . . . 712
23.7.4 FMPI2C own address 2 register (FMPI2C_OAR2) . . . . . . . . . . . . . . . 713
23.7.5 FMPI2C timing register (FMPI2C_TIMINGR) . . . . . . . . . . . . . . . . . . . 714
23.7.6 FMPI2C timeout register (FMPI2C_TIMEOUTR) . . . . . . . . . . . . . . . . 715
23.7.7 FMPI2C interrupt and status register (FMPI2C_ISR) . . . . . . . . . . . . . 716
23.7.8 FMPI2C interrupt clear register (FMPI2C_ICR) . . . . . . . . . . . . . . . . . . 718
23.7.9 FMPI2C PEC register (FMPI2C_PECR) . . . . . . . . . . . . . . . . . . . . . . . 719
23.7.10 FMPI2C receive data register (FMPI2C_RXDR) . . . . . . . . . . . . . . . . . 720
23.7.11 FMPI2C transmit data register (FMPI2C_TXDR) . . . . . . . . . . . . . . . . 720

20/1163 RM0402 Rev 6

RM0402 Contents

23.7.12 FMPI2C register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721

24 Inter-integrated circuit (I2C) interface . . . . . . . . . . . . . . . . . . . . . . . . . 723

24.1 I2C introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723
24.2 I2C main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
24.3 I2C functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
24.3.1 Mode selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
24.3.2 I2C slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726
24.3.3 I2C master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
24.3.4 Error conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735
24.3.5 Programmable noise filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736
24.3.6 SDA/SCL line control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
24.3.7 SMBus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
24.3.8 DMA requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
24.3.9 Packet error checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741
24.4 I2C interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
24.5 I2C debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
24.6 I2C registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
24.6.1 I2C control register 1 (I2C_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
24.6.2 I2C control register 2 (I2C_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
24.6.3 I2C own address register 1 (I2C_OAR1) . . . . . . . . . . . . . . . . . . . . . . . 748
24.6.4 I2C own address register 2 (I2C_OAR2) . . . . . . . . . . . . . . . . . . . . . . . 748
24.6.5 I2C data register (I2C_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
24.6.6 I2C status register 1 (I2C_SR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
24.6.7 I2C status register 2 (I2C_SR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 753
24.6.8 I2C clock control register (I2C_CCR) . . . . . . . . . . . . . . . . . . . . . . . . . . 754
24.6.9 I2C TRISE register (I2C_TRISE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
24.6.10 I2C FLTR register (I2C_FLTR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
24.6.11 I2C register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757

25 Universal synchronous receiver transmitter (USART)

/universal asynchronous receiver transmitter (UART) . . . . . . . . . . . 758
25.1 USART introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
25.2 USART main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759
25.3 USART implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
25.4 USART functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760

RM0402 Rev 6 21/1163

31
Contents RM0402

25.4.1 USART character description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763

25.4.2 Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 764
25.4.3 Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
25.4.4 Fractional baud rate generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
25.4.5 USART receiver tolerance to clock deviation . . . . . . . . . . . . . . . . . . . . 782
25.4.6 Multiprocessor communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
25.4.7 Parity control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
25.4.8 LIN (local interconnection network) mode . . . . . . . . . . . . . . . . . . . . . . 786
25.4.9 USART synchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788
25.4.10 Single-wire half-duplex communication . . . . . . . . . . . . . . . . . . . . . . . . 790
25.4.11 Smartcard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
25.4.12 IrDA SIR ENDEC block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793
25.4.13 Continuous communication using DMA . . . . . . . . . . . . . . . . . . . . . . . . 795
25.4.14 Hardware flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
25.5 USART interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799
25.6 USART registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
25.6.1 Status register (USART_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
25.6.2 Data register (USART_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803
25.6.3 Baud rate register (USART_BRR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 803
25.6.4 Control register 1 (USART_CR1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
25.6.5 Control register 2 (USART_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806
25.6.6 Control register 3 (USART_CR3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807
25.6.7 Guard time and prescaler register (USART_GTPR) . . . . . . . . . . . . . . 809
25.6.8 USART register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810

26 Serial peripheral interface/ inter-IC sound (SPI/I2S) . . . . . . . . . . . . . 811

26.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .811
26.1.1 SPI main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812
26.1.2 SPI extended features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
26.1.3 I2S features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
26.2 SPI/I2S implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
26.3 SPI functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
26.3.1 General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
26.3.2 Communications between one master and one slave . . . . . . . . . . . . . 815
26.3.3 Standard multi-slave communication . . . . . . . . . . . . . . . . . . . . . . . . . . 818
26.3.4 Multi-master communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 819
26.3.5 Slave select (NSS) pin management . . . . . . . . . . . . . . . . . . . . . . . . . . 819

22/1163 RM0402 Rev 6

RM0402 Contents

26.3.6 Communication formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821

26.3.7 SPI configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823
26.3.8 Procedure for enabling SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823
26.3.9 Data transmission and reception procedures . . . . . . . . . . . . . . . . . . . 824
26.3.10 Procedure for disabling the SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826
26.3.11 Communication using DMA (direct memory addressing) . . . . . . . . . . . 827
26.3.12 SPI status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829
26.3.13 SPI error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830
26.4 SPI special features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
26.4.1 TI mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 831
26.4.2 CRC calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
26.5 SPI interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
26.6 I2S functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 835
26.6.1 I2S general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 835
26.6.2 I2S full-duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
26.6.3 Supported audio protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837
26.6.4 Clock generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
26.6.5 I2S master mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846
26.6.6 I2S slave mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
26.6.7 I2S status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 849
26.6.8 I2S error flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850
26.6.9 I2S interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
26.6.10 DMA features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
26.7 SPI and I2S registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 852
26.7.1 SPI control register 1 (SPI_CR1) (not used in I2S mode) . . . . . . . . . . 852
26.7.2 SPI control register 2 (SPI_CR2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854
26.7.3 SPI status register (SPI_SR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855
26.7.4 SPI data register (SPI_DR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857
26.7.5 SPI CRC polynomial register (SPI_CRCPR) (not used in I2S
mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857
26.7.6 SPI RX CRC register (SPI_RXCRCR) (not used in I2S mode) . . . . . . 858
26.7.7 SPI TX CRC register (SPI_TXCRCR) (not used in I2S mode) . . . . . . . 858
26.7.8 SPI_I2S configuration register (SPI_I2SCFGR) . . . . . . . . . . . . . . . . . . 859
26.7.9 SPI_I2S prescaler register (SPI_I2SPR) . . . . . . . . . . . . . . . . . . . . . . . 860
26.7.10 SPI register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862

27 Secure digital input/output interface (SDIO) . . . . . . . . . . . . . . . . . . . . 863

RM0402 Rev 6 23/1163

31
Contents RM0402

27.1 SDIO main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863

27.2 SDIO bus topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863
27.3 SDIO functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
27.3.1 SDIO adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
27.3.2 SDIO APB2 interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 878
27.4 Card functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
27.4.1 Card identification mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879
27.4.2 Card reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
27.4.3 Operating voltage range validation . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
27.4.4 Card identification process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880
27.4.5 Block write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
27.4.6 Block read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882
27.4.7 Stream access, stream write and stream read
(MultiMediaCard only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882
27.4.8 Erase: group erase and sector erase . . . . . . . . . . . . . . . . . . . . . . . . . 884
27.4.9 Wide bus selection or deselection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884
27.4.10 Protection management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884
27.4.11 Card status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
27.4.12 SD status register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
27.4.13 SD I/O mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895
27.4.14 Commands and responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896
27.5 Response formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
27.5.1 R1 (normal response command) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
27.5.2 R1b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
27.5.3 R2 (CID, CSD register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
27.5.4 R3 (OCR register) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
27.5.5 R4 (Fast I/O) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
27.5.6 R4b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
27.5.7 R5 (interrupt request) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
27.5.8 R6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
27.6 SDIO I/O card-specific operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903
27.6.1 SDIO I/O read wait operation by SDIO_D2 signalling . . . . . . . . . . . . . 903
27.6.2 SDIO read wait operation by stopping SDIO_CK . . . . . . . . . . . . . . . . 904
27.6.3 SDIO suspend/resume operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
27.6.4 SDIO interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904
27.7 HW flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904

24/1163 RM0402 Rev 6

RM0402 Contents

27.8 SDIO registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905

27.8.1 SDIO power control register (SDIO_POWER) . . . . . . . . . . . . . . . . . . . 905
27.8.2 SDIO clock control register (SDIO_CLKCR) . . . . . . . . . . . . . . . . . . . . 905
27.8.3 SDIO argument register (SDIO_ARG) . . . . . . . . . . . . . . . . . . . . . . . . . 907
27.8.4 SDIO command register (SDIO_CMD) . . . . . . . . . . . . . . . . . . . . . . . . 907
27.8.5 SDIO command response register (SDIO_RESPCMD) . . . . . . . . . . . 908
27.8.6 SDIO response 1..4 register (SDIO_RESPx) . . . . . . . . . . . . . . . . . . . 908
27.8.7 SDIO data timer register (SDIO_DTIMER) . . . . . . . . . . . . . . . . . . . . . 909
27.8.8 SDIO data length register (SDIO_DLEN) . . . . . . . . . . . . . . . . . . . . . . 910
27.8.9 SDIO data control register (SDIO_DCTRL) . . . . . . . . . . . . . . . . . . . . . 910
27.8.10 SDIO data counter register (SDIO_DCOUNT) . . . . . . . . . . . . . . . . . . 913
27.8.11 SDIO status register (SDIO_STA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913
27.8.12 SDIO interrupt clear register (SDIO_ICR) . . . . . . . . . . . . . . . . . . . . . . 914
27.8.13 SDIO mask register (SDIO_MASK) . . . . . . . . . . . . . . . . . . . . . . . . . . . 916
27.8.14 SDIO FIFO counter register (SDIO_FIFOCNT) . . . . . . . . . . . . . . . . . . 918
27.8.15 SDIO data FIFO register (SDIO_FIFO) . . . . . . . . . . . . . . . . . . . . . . . . 919
27.8.16 SDIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920

28 Controller area network (bxCAN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922

28.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922
28.2 bxCAN main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922
28.3 bxCAN general description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923
28.3.1 CAN 2.0B active core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923
28.3.2 Control, status and configuration registers . . . . . . . . . . . . . . . . . . . . . 923
28.3.3 Tx mailboxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923
28.3.4 Acceptance filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 924
28.4 bxCAN operating modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925
28.4.1 Initialization mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925
28.4.2 Normal mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 925
28.4.3 Sleep mode (low-power) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 926
28.5 Test mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927
28.5.1 Silent mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927
28.5.2 Loop back mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927
28.5.3 Loop back combined with silent mode . . . . . . . . . . . . . . . . . . . . . . . . . 928
28.6 Behavior in debug mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928
28.7 bxCAN functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928

RM0402 Rev 6 25/1163

31
Contents RM0402

28.7.1 Transmission handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928

28.7.2 Time triggered communication mode . . . . . . . . . . . . . . . . . . . . . . . . . . 930
28.7.3 Reception handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
28.7.4 Identifier filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 932
28.7.5 Message storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
28.7.6 Error management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938
28.7.7 Bit timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938
28.8 bxCAN interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941
28.9 CAN registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 942
28.9.1 Register access protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 942
28.9.2 CAN control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 942
28.9.3 CAN mailbox registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 952
28.9.4 CAN filter registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 959
28.9.5 bxCAN register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963

29 USB on-the-go full-speed (OTG_FS) . . . . . . . . . . . . . . . . . . . . . . . . . . 967

29.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 967
29.2 OTG_FS main features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968
29.2.1 General features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968
29.2.2 Host-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969
29.2.3 Peripheral-mode features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969
29.2.4 Split rail for USB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969
29.3 OTG_FS implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 970
29.4 OTG_FS functional description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
29.4.1 OTG_FS block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
29.4.2 OTG_FS pin and internal signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
29.4.3 OTG_FS core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 972
29.4.4 Embedded full-speed OTG PHY connected to OTG_FS . . . . . . . . . . . 972
29.4.5 OTG detections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
29.5 OTG_FS dual role device (DRD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
29.5.1 ID line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
29.5.2 HNP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974
29.5.3 SRP dual role device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974
29.6 OTG_FS as a USB peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974
29.6.1 SRP-capable peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975
29.6.2 Peripheral states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975

26/1163 RM0402 Rev 6

RM0402 Contents

29.6.3 Peripheral endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 976

29.7 OTG_FS as a USB host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 978
29.7.1 SRP-capable host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 979
29.7.2 USB host states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 979
29.7.3 Host channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 981
29.7.4 Host scheduler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 982
29.8 OTG_FS SOF trigger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
29.8.1 Host SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
29.8.2 Peripheral SOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 983
29.9 OTG_FS low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984
29.10 OTG_FS Dynamic update of the OTG_HFIR register . . . . . . . . . . . . . . 985
29.11 OTG_FS data FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985
29.11.1 Peripheral FIFO architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 986
29.11.2 Host FIFO architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 987
29.11.3 FIFO RAM allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 988
29.12 OTG_FS system performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
29.13 OTG_FS interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 990
29.14 OTG_FS control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 992
29.14.1 CSR memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 992
29.15 OTG_FS registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
29.15.1 OTG control and status register (OTG_GOTGCTL) . . . . . . . . . . . . . . 997
29.15.2 OTG interrupt register (OTG_GOTGINT) . . . . . . . . . . . . . . . . . . . . . 1000
29.15.3 OTG AHB configuration register (OTG_GAHBCFG) . . . . . . . . . . . . . 1001
29.15.4 OTG USB configuration register (OTG_GUSBCFG) . . . . . . . . . . . . . 1002
29.15.5 OTG reset register (OTG_GRSTCTL) . . . . . . . . . . . . . . . . . . . . . . . . 1004
29.15.6 OTG core interrupt register (OTG_GINTSTS) . . . . . . . . . . . . . . . . . . 1006
29.15.7 OTG interrupt mask register (OTG_GINTMSK) . . . . . . . . . . . . . . . . . 1010
29.15.8 OTG receive status debug read register (OTG_GRXSTSR) . . . . . . . 1013
29.15.9 OTG receive status debug read [alternate] (OTG_GRXSTSR) . . . . . 1014
29.15.10 OTG status read and pop registers (OTG_GRXSTSP) . . . . . . . . . . . 1015
29.15.11 OTG status read and pop registers [alternate] (OTG_GRXSTSP) . . 1016
29.15.12 OTG receive FIFO size register (OTG_GRXFSIZ) . . . . . . . . . . . . . . 1017
29.15.13 OTG host non-periodic transmit FIFO size register
(OTG_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size
(OTG_DIEPTXF0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1017
29.15.14 OTG non-periodic transmit FIFO/queue status register
(OTG_HNPTXSTS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018

RM0402 Rev 6 27/1163

31
Contents RM0402

29.15.15 OTG general core configuration register (OTG_GCCFG) . . . . . . . . . 1019

29.15.16 OTG core ID register (OTG_CID) . . . . . . . . . . . . . . . . . . . . . . . . . . . 1021
29.15.17 OTG core LPM configuration register (OTG_GLPMCFG) . . . . . . . . . 1021
29.15.18 OTG host periodic transmit FIFO size register
(OTG_HPTXFSIZ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025
29.15.19 OTG device IN endpoint transmit FIFO x size register
(OTG_DIEPTXFx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025
29.15.20 Host-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1026
29.15.21 OTG host configuration register (OTG_HCFG) . . . . . . . . . . . . . . . . . 1026
29.15.22 OTG host frame interval register (OTG_HFIR) . . . . . . . . . . . . . . . . . 1027
29.15.23 OTG host frame number/frame time remaining register
(OTG_HFNUM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
29.15.24 OTG_Host periodic transmit FIFO/queue status register
(OTG_HPTXSTS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1028
29.15.25 OTG host all channels interrupt register (OTG_HAINT) . . . . . . . . . . 1029
29.15.26 OTG host all channels interrupt mask register
(OTG_HAINTMSK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1030
29.15.27 OTG host port control and status register (OTG_HPRT) . . . . . . . . . . 1031
29.15.28 OTG host channel x characteristics register (OTG_HCCHARx) . . . . 1033
29.15.29 OTG host channel x interrupt register (OTG_HCINTx) . . . . . . . . . . . 1034
29.15.30 OTG host channel x interrupt mask register (OTG_HCINTMSKx) . . 1035
29.15.31 OTG host channel x transfer size register (OTG_HCTSIZx) . . . . . . . 1036
29.15.32 Device-mode registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1037
29.15.33 OTG device configuration register (OTG_DCFG) . . . . . . . . . . . . . . . 1037
29.15.34 OTG device control register (OTG_DCTL) . . . . . . . . . . . . . . . . . . . . 1038
29.15.35 OTG device status register (OTG_DSTS) . . . . . . . . . . . . . . . . . . . . . 1041
29.15.36 OTG device IN endpoint common interrupt mask register
(OTG_DIEPMSK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1042
29.15.37 OTG device OUT endpoint common interrupt mask register
(OTG_DOEPMSK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1043
29.15.38 OTG device all endpoints interrupt register (OTG_DAINT) . . . . . . . . 1044
29.15.39 OTG all endpoints interrupt mask register
(OTG_DAINTMSK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045
29.15.40 OTG device VBUS discharge time register
(OTG_DVBUSDIS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045
29.15.41 OTG device VBUS pulsing time register
(OTG_DVBUSPULSE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1046
29.15.42 OTG device IN endpoint FIFO empty interrupt mask register
(OTG_DIEPEMPMSK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1046
29.15.43 OTG device control IN endpoint 0 control register
(OTG_DIEPCTL0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1047

28/1163 RM0402 Rev 6

RM0402 Contents

29.15.44 OTG device IN endpoint x control register (OTG_DIEPCTLx) . . . . . 1048

29.15.45 OTG device IN endpoint x interrupt register (OTG_DIEPINTx) . . . . . 1051
29.15.46 OTG device IN endpoint 0 transfer size register
(OTG_DIEPTSIZ0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1052
29.15.47 OTG device IN endpoint transmit FIFO status register
(OTG_DTXFSTSx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053
29.15.48 OTG device IN endpoint x transfer size register (OTG_DIEPTSIZx) . 1054
29.15.49 OTG device control OUT endpoint 0 control register
(OTG_DOEPCTL0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1055
29.15.50 OTG device OUT endpoint x interrupt register (OTG_DOEPINTx) . . 1056
29.15.51 OTG device OUT endpoint 0 transfer size register
(OTG_DOEPTSIZ0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058
29.15.52 OTG device OUT endpoint x control register
(OTG_DOEPCTLx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1059
29.15.53 OTG device OUT endpoint x transfer size register
(OTG_DOEPTSIZx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1061
29.15.54 OTG power and clock gating control register (OTG_PCGCCTL) . . . 1062
29.15.55 OTG_FS register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1063
29.16 OTG_FS programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1071
29.16.1 Core initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1071
29.16.2 Host initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1072
29.16.3 Device initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1072
29.16.4 Host programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1073
29.16.5 Device programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1094
29.16.6 Worst case response time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1115
29.16.7 OTG programming model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1117

30 Debug support (DBG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123

30.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1123
30.2 Reference Arm® documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1124
30.3 SWJ debug port (serial wire and JTAG) . . . . . . . . . . . . . . . . . . . . . . . . .1124
30.3.1 Mechanism to select the JTAG-DP or the SW-DP . . . . . . . . . . . . . . . 1125
30.4 Pinout and debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1125
30.4.1 SWJ debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126
30.4.2 Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126
30.4.3 Internal pull-up and pull-down on JTAG pins . . . . . . . . . . . . . . . . . . . 1127
30.4.4 Using serial wire and releasing the unused debug pins as GPIOs . . 1128
30.5 JTAG TAP connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1128

RM0402 Rev 6 29/1163

31
Contents RM0402

30.6 ID codes and locking mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1130

30.6.1 MCU device ID code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
30.6.2 Boundary scan TAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
30.6.3 Cortex®-M4 with FPU TAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1130
30.6.4 Cortex®-M4 with FPU JEDEC-106 ID code . . . . . . . . . . . . . . . . . . . . 1131
30.7 JTAG debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1131
30.8 SW debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1133
30.8.1 SW protocol introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133
30.8.2 SW protocol sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133
30.8.3 SW-DP state machine (reset, idle states, ID code) . . . . . . . . . . . . . . 1134
30.8.4 DP and AP read/write accesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1134
30.8.5 SW-DP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135
30.8.6 SW-AP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136
30.9 AHB-AP (AHB access port) - valid for both JTAG-DP
and SW-DP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1136
30.10 Core debug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1137
30.11 Capability of the debugger host to connect under system reset . . . . . .1138
30.12 FPB (Flash patch breakpoint) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1138
30.13 DWT (data watchpoint trigger) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1139
30.14 ITM (instrumentation trace macrocell) . . . . . . . . . . . . . . . . . . . . . . . . . .1139
30.14.1 General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139
30.14.2 Time stamp packets, synchronization and overflow packets . . . . . . . 1139
30.15 ETM (Embedded trace macrocell) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1141
30.15.1 General description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1141
30.15.2 Signal protocol, packet types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1141
30.15.3 Main ETM registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1141
30.15.4 Configuration example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1142
30.16 MCU debug component (DBGMCU) . . . . . . . . . . . . . . . . . . . . . . . . . . .1142
30.16.1 Debug support for low-power modes . . . . . . . . . . . . . . . . . . . . . . . . . 1142
30.16.2 Debug support for timers, watchdog, bxCAN and I2C . . . . . . . . . . . . 1143
30.16.3 Debug MCU configuration register . . . . . . . . . . . . . . . . . . . . . . . . . . 1143
30.16.4 Debug MCU APB1 freeze register (DBGMCU_APB1_FZ) . . . . . . . . 1144
30.16.5 Debug MCU APB2 Freeze register (DBGMCU_APB2_FZ) . . . . . . . . 1146
30.17 TPIU (trace port interface unit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1146
30.17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1146
30.17.2 TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1148

30/1163 RM0402 Rev 6

RM0402 Contents

30.17.3 TPUI formatter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149

30.17.4 TPUI frame synchronization packets . . . . . . . . . . . . . . . . . . . . . . . . . 1150
30.17.5 Transmission of the synchronization frame packet . . . . . . . . . . . . . . 1150
30.17.6 Synchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1150
30.17.7 Asynchronous mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151
30.17.8 TRACECLKIN connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151
30.17.9 TPIU registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151
30.17.10 Example of configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1152
30.18 DBG register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1153

31 Device electronic signature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154

31.1 Unique device ID register (96 bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1154
31.2 Flash size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1155
31.3 Package data register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1155

32 Revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1157

RM0402 Rev 6 31/1163

31
List of tables RM0402

List of tables

Table 1. Register boundary addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Table 2. Boot modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Table 3. Embedded bootloader interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Table 4. Memory mapping vs. Boot mode/physical remap in STM32F412xx. . . . . . . . . . . . . . . . . . 56
Table 5. Flash module organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Table 6. Number of wait states according to CPU clock (HCLK) frequency . . . . . . . . . . . . . . . . . . . 60
Table 7. Program/erase parallelism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Table 8. Flash interrupt request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Table 9. Option byte organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Table 10. Description of the option bytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Table 11. Access versus read protection level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Table 12. OTP area organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Table 13. Flash register map and reset values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Table 14. CRC calculation unit register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Table 15. Low-power mode summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Table 16. Sleep-now entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Table 17. Sleep-on-exit entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Table 18. BAM-now entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Table 19. BAM-on-exit entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Table 20. Stop operating modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Table 21. Stop mode entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Table 22. Standby mode entry and exit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Table 23. PWR - register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Table 24. RCC register map and reset values for STM32F412xx . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Table 25. Port bit configuration table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Table 26. Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Table 27. RTC additional functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Table 28. GPIO register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Table 29. SYSCFG register map and reset values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Table 30. DMA1 request mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Table 31. DMA2 request mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Table 32. Source and destination address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Table 33. Source and destination address registers in double-buffer mode (DBM = 1) . . . . . . . . . . 206
Table 34. Packing/unpacking and endian behavior (bit PINC = MINC = 1) . . . . . . . . . . . . . . . . . . . 207
Table 35. Restriction on NDT versus PSIZE and MSIZE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Table 36. FIFO threshold configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Table 37. Possible DMA configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Table 38. DMA interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Table 39. DMA register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Table 40. Vector table for STM32F412xx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
Table 41. External interrupt/event controller register map and reset values. . . . . . . . . . . . . . . . . . . 245
Table 42. NOR/PSRAM bank selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Table 43. NOR/PSRAM External memory address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Table 44. Programmable NOR/PSRAM access parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Table 45. Non-multiplexed I/O NOR Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Table 46. 16-bit multiplexed I/O NOR Flash memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Table 47. Non-multiplexed I/Os PSRAM/SRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Table 48. 16-Bit multiplexed I/O PSRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

32/1163 RM0402 Rev 6

RM0402 List of tables

Table 49. NOR Flash/PSRAM: example of supported memories

and transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Table 50. FSMC_BCRx bitfields (mode 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Table 51. FSMC_BTRx bitfields (mode 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Table 52. FSMC_BCRx bitfields (mode A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Table 53. FSMC_BTRx bitfields (mode A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Table 54. FSMC_BWTRx bitfields (mode A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Table 55. FSMC_BCRx bitfields (mode 2/B) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Table 56. FSMC_BTRx bitfields (mode 2/B). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Table 57. FSMC_BWTRx bitfields (mode 2/B) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Table 58. FSMC_BCRx bitfields (mode C) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Table 59. FSMC_BTRx bitfields (mode C) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Table 60. FSMC_BWTRx bitfields (mode C) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Table 61. FSMC_BCRx bitfields (mode D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Table 62. FSMC_BTRx bitfields (mode D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Table 63. FSMC_BWTRx bitfields (mode D) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Table 64. FSMC_BCRx bitfields (Muxed mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Table 65. FSMC_BTRx bitfields (Muxed mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
Table 66. FSMC_BCRx bitfields (Synchronous multiplexed read mode) . . . . . . . . . . . . . . . . . . . . . 275
Table 67. FSMC_BTRx bitfields (Synchronous multiplexed read mode) . . . . . . . . . . . . . . . . . . . . . 276
Table 68. FSMC_BCRx bitfields (Synchronous multiplexed write mode) . . . . . . . . . . . . . . . . . . . . . 277
Table 69. FSMC_BTRx bitfields (Synchronous multiplexed write mode) . . . . . . . . . . . . . . . . . . . . . 278
Table 70. FSMC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Table 71. QUADSPI pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Table 72. QUADSPI interrupt requests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Table 73. QUADSPI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
Table 74. ADC pins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
Table 75. Analog watchdog channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
Table 76. Configuring the trigger polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
Table 77. External trigger for regular channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Table 78. External trigger for injected channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
Table 79. ADC interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
Table 80. ADC global register map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Table 81. ADC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Table 82. ADC register map and reset values (common ADC registers) . . . . . . . . . . . . . . . . . . . . . 347
Table 83. DFSDM1 implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Table 84. DFSDM external pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Table 85. DFSDM internal signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Table 86. DFSDM triggers connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
Table 87. DFSDM break connection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Table 88. Filter maximum output resolution (peak data values from filter output)
for some FOSR values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Table 89. Integrator maximum output resolution (peak data values from integrator
output) for some IOSR values and FOSR = 256 and Sinc3 filter type (largest data) . . . . 367
Table 90. DFSDM interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Table 91. DFSDM register map and reset values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
Table 92. RNG internal input/output signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Table 93. RNG interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
Table 94. RNG register map and reset map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
Table 95. Counting direction versus encoder signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
Table 96. TIMx Internal trigger connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Table 97. Output control bits for complementary OCx and OCxN channels

RM0402 Rev 6 33/1163

36
List of tables RM0402

with break feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474

Table 98. TIM1&TIM8 register map and reset values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
Table 99. Counting direction versus encoder signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
Table 100. TIMx internal trigger connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
Table 101. Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
Table 102. TIM2 to TIM5 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Table 103. TIMx internal trigger connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
Table 104. Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
Table 105. TIM9/12 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Table 106. Output control bit for standard OCx channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
Table 107. TIM10/11/13/14 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
Table 108. TIM6/7 register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
Table 109. Min/max IWDG timeout period at 32 kHz (LSI). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Table 110. IWDG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
Table 111. WWDG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
Table 112. Effect of low power modes on RTC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
Table 113. Interrupt control bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
Table 114. RTC register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
Table 115. STM32F412xx FMPI2C implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
Table 116. FMPI2C input/output pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Table 117. FMPI2C internal input/output signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Table 118. Comparison of analog vs. digital filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Table 119. I2C-SMBus specification data setup and hold times . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Table 120. FMPI2C configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
Table 121. I2C-SMBus specification clock timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679
Table 122. Examples of timing settings for fI2CCLK = 8 MHz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
Table 123. Examples of timings settings for fI2CCLK = 16 MHz . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
Table 124. SMBus timeout specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
Table 125. SMBus with PEC configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694
Table 126. Examples of TIMEOUTA settings for various FMPI2CCLK frequencies
(max tTIMEOUT = 25 ms) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
Table 127. Examples of TIMEOUTB settings for various FMPI2CCLK frequencies. . . . . . . . . . . . . . 695
Table 128. Examples of TIMEOUTA settings for various FMPI2CCLK frequencies
(max tIDLE = 50 µs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695
Table 129. Effect of low-power modes on the FMPI2C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705
Table 130. FMPI2C Interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
Table 131. FMPI2C register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
Table 132. Maximum DNF[3:0] value to be compliant with Thd:dat(max) . . . . . . . . . . . . . . . . . . . . . 736
Table 133. SMBus vs. I2C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738
Table 134. I2C Interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Table 135. I2C register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757
Table 136. USART features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
Table 137. Noise detection from sampled data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
Table 138. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Table 139. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 774
Table 140. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
Table 141. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776
Table 142. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,

34/1163 RM0402 Rev 6

RM0402 List of tables

oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776

Table 143. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 777
Table 144. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778
Table 145. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 778
Table 146. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779
Table 147. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780
Table 148. Error calculation for programmed baud rates at fPCLK = 100 MHz or fPCLK = 50 MHz,
oversampling by 16. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
Table 149. Error calculation for programmed baud rates at fPCLK = 100 MHz or fPCLK = 50 MHz,
oversampling by 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 782
Table 150. USART receiver tolerance when DIV fraction is 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
Table 151. USART receiver tolerance when DIV_Fraction is different from 0 . . . . . . . . . . . . . . . . . . 783
Table 152. Frame formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
Table 153. USART interrupt requests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799
Table 154. USART register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810
Table 155. STM32F412xx SPI implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 813
Table 156. SPI interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834
Table 157. Audio-frequency precision using standard 8 MHz HSE . . . . . . . . . . . . . . . . . . . . . . . . . . 845
Table 158. I2S interrupt requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
Table 159. SPI register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862
Table 160. SDIO I/O definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866
Table 161. Command format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
Table 162. Short response format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872
Table 163. Long response format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872
Table 164. Command path status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872
Table 165. Data token format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875
Table 166. DPSM flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876
Table 167. Transmit FIFO status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877
Table 168. Receive FIFO status flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877
Table 169. Card status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 888
Table 170. SD status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 891
Table 171. Speed class code field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892
Table 172. Performance move field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
Table 173. AU_SIZE field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
Table 174. Maximum AU size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893
Table 175. Erase size field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894
Table 176. Erase timeout field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894
Table 177. Erase offset field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894
Table 178. Block-oriented write commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897
Table 179. Block-oriented write protection commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Table 180. Erase commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Table 181. I/O mode commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Table 182. Lock card . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
Table 183. Application-specific commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899
Table 184. R1 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
Table 185. R2 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900
Table 186. R3 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901

RM0402 Rev 6 35/1163

36
List of tables RM0402

Table 187. R4 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901

Table 188. R4b response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
Table 189. R5 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902
Table 190. R6 response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 903
Table 191. Response type and SDIO_RESPx registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909
Table 192. SDIO register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920
Table 193. Transmit mailbox mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 937
Table 194. Receive mailbox mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 937
Table 195. bxCAN register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963
Table 196. OTG_FS speeds supported . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 967
Table 197. OTG_FS implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 970
Table 198. OTG_FS input/output pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
Table 199. OTG_FS input/output signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 972
Table 200. Compatibility of STM32 low power modes with the OTG . . . . . . . . . . . . . . . . . . . . . . . . . 984
Table 201. Core global control and status registers (CSRs). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 992
Table 202. Host-mode control and status registers (CSRs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 993
Table 203. Device-mode control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 994
Table 204. Data FIFO (DFIFO) access register map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
Table 205. Power and clock gating control and status registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 996
Table 206. TRDT values (FS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003
Table 207. Minimum duration for soft disconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1040
Table 208. OTG_FS register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1063
Table 209. SWJ debug port pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126
Table 210. Flexible SWJ-DP pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1126
Table 211. JTAG debug port data registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1131
Table 212. 32-bit debug port registers addressed through the shifted value A[3:2] . . . . . . . . . . . . . 1132
Table 213. Packet request (8-bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1133
Table 214. ACK response (3 bits). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1134
Table 215. DATA transfer (33 bits) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1134
Table 216. SW-DP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135
Table 217. Cortex®-M4 with FPU AHB-AP registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136
Table 218. Core debug registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137
Table 219. Main ITM registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1140
Table 220. Main ETM registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1142
Table 221. Asynchronous TRACE pin assignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1148
Table 222. Synchronous TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1148
Table 223. Flexible TRACE pin assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1149
Table 224. Important TPIU registers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1151
Table 225. DBG register map and reset values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1153
Table 226. Document revision history . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1157

36/1163 RM0402 Rev 6

RM0402 List of figures

List of figures

Figure 1. System architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Figure 2. Memory map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Figure 3. Flash memory interface connection inside system architecture . . . . . . . . . . . . . . . . . . . . . 58
Figure 4. Sequential 32-bit instruction execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Figure 5. RDP levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Figure 6. PCROP levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Figure 7. CRC calculation unit block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Figure 8. Power supply overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Figure 9. Power-on reset/power-down reset waveform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Figure 10. BOR thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Figure 11. PVD thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Figure 12. Simplified diagram of the reset circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Figure 13. Clock tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Figure 14. HSE/ LSE clock sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Figure 15. Frequency measurement with TIM5 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . . 119
Figure 16. Frequency measurement with TIM11 in Input capture mode . . . . . . . . . . . . . . . . . . . . . . 120
Figure 17. Basic structure of a five-volt tolerant I/O port bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Figure 18. Selecting an alternate function on STM32F412xx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
Figure 19. Input floating/pull up/pull down configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Figure 20. Output configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Figure 21. Alternate function configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Figure 22. High impedance-analog configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Figure 23. DMA block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Figure 24. Channel selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Figure 25. Peripheral-to-memory mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Figure 26. Memory-to-peripheral mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Figure 27. Memory-to-memory mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
Figure 28. FIFO structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Figure 29. External interrupt/event controller block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Figure 30. External interrupt/event GPIO mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Figure 31. FSMC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Figure 32. FSMC memory banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Figure 33. Mode 1 read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Figure 34. Mode 1 write access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Figure 35. Mode A read access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Figure 36. Mode A write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Figure 37. Mode 2 and mode B read access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Figure 38. Mode 2 write access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Figure 39. Mode B write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Figure 40. Mode C read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
Figure 41. Mode C write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
Figure 42. Mode D read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Figure 43. Mode D write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Figure 44. Muxed read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Figure 45. Muxed write access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Figure 46. Asynchronous wait during a read access waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Figure 47. Asynchronous wait during a write access waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
Figure 48. Wait configuration waveforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

RM0402 Rev 6 37/1163

44
List of figures RM0402

Figure 49. Synchronous multiplexed read mode waveforms - NOR, PSRAM (CRAM) . . . . . . . . . . . 275
Figure 50. Synchronous multiplexed write mode waveforms - PSRAM (CRAM). . . . . . . . . . . . . . . . 277
Figure 51. QUADSPI block diagram when dual-flash mode is disabled . . . . . . . . . . . . . . . . . . . . . . 288
Figure 52. QUADSPI block diagram when dual-flash mode is enabled . . . . . . . . . . . . . . . . . . . . . . 289
Figure 53. An example of a read command in quad mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
Figure 54. An example of a DDR command in quad mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Figure 55. nCS when CKMODE = 0 (T = CLK period). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Figure 56. nCS when CKMODE = 1 in SDR mode (T = CLK period) . . . . . . . . . . . . . . . . . . . . . . . . 301
Figure 57. nCS when CKMODE = 1 in DDR mode (T = CLK period) . . . . . . . . . . . . . . . . . . . . . . . . 302
Figure 58. nCS when CKMODE = 1 with an abort (T = CLK period) . . . . . . . . . . . . . . . . . . . . . . . . . 302
Figure 59. Single ADC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
Figure 60. Timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Figure 61. Analog watchdog’s guarded area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
Figure 62. Injected conversion latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Figure 63. Right alignment of 12-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Figure 64. Left alignment of 12-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Figure 65. Left alignment of 6-bit data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Figure 66. Temperature sensor and VREFINT channel block diagram . . . . . . . . . . . . . . . . . . . . . . 330
Figure 67. Single DFSDM block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Figure 68. Input channel pins redirection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
Figure 69. Channel transceiver timing diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Figure 70. Clock absence timing diagram for SPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
Figure 71. Clock absence timing diagram for Manchester coding . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
Figure 72. First conversion for Manchester coding (Manchester synchronization) . . . . . . . . . . . . . . 361
Figure 73. DFSDM_CHyDATINR registers operation modes and assignment . . . . . . . . . . . . . . . . . 364
Figure 74. Example: Sinc3 filter response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
Figure 75. RNG block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Figure 76. Entropy source model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
Figure 77. Advanced-control timer block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
Figure 78. Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 416
Figure 79. Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 416
Figure 80. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Figure 81. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
Figure 82. Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
Figure 83. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
Figure 84. Counter timing diagram, update event when ARPE=0
(TIMx_ARR not preloaded) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Figure 85. Counter timing diagram, update event when ARPE=1
(TIMx_ARR preloaded) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Figure 86. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Figure 87. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Figure 88. Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
Figure 89. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
Figure 90. Counter timing diagram, update event when repetition counter is not used . . . . . . . . . . . 423
Figure 91. Counter timing diagram, internal clock divided by 1, TIMx_ARR = 0x6 . . . . . . . . . . . . . . 424
Figure 92. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
Figure 93. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36 . . . . . . . . . . . . . . 425
Figure 94. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
Figure 95. Counter timing diagram, update event with ARPE=1 (counter underflow) . . . . . . . . . . . . 426
Figure 96. Counter timing diagram, update event with ARPE=1 (counter overflow) . . . . . . . . . . . . . 426
Figure 97. Update rate examples depending on mode and TIMx_RCR register settings . . . . . . . . . 427
Figure 98. Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 428

38/1163 RM0402 Rev 6

RM0402 List of figures

Figure 99. TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

Figure 100. Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Figure 101. External trigger input block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
Figure 102. Control circuit in external clock mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Figure 103. Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 432
Figure 104. Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Figure 105. Output stage of capture/compare channel (channels 1 to 3) . . . . . . . . . . . . . . . . . . . . . . 433
Figure 106. Output stage of capture/compare channel (channel 4). . . . . . . . . . . . . . . . . . . . . . . . . . . 433
Figure 107. PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Figure 108. Output compare mode, toggle on OC1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
Figure 109. Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Figure 110. Center-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
Figure 111. Complementary output with dead-time insertion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
Figure 112. Dead-time waveforms with delay greater than the negative pulse. . . . . . . . . . . . . . . . . . 441
Figure 113. Dead-time waveforms with delay greater than the positive pulse. . . . . . . . . . . . . . . . . . . 441
Figure 114. Output behavior in response to a break.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
Figure 115. Clearing TIMx OCxREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
Figure 116. 6-step generation, COM example (OSSR=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
Figure 117. Example of one pulse mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
Figure 118. Example of counter operation in encoder interface mode. . . . . . . . . . . . . . . . . . . . . . . . . 450
Figure 119. Example of encoder interface mode with TI1FP1 polarity inverted. . . . . . . . . . . . . . . . . . 450
Figure 120. Example of Hall sensor interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
Figure 121. Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Figure 122. Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
Figure 123. Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Figure 124. Control circuit in external clock mode 2 + trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . 456
Figure 125. General-purpose timer block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 485
Figure 126. Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 486
Figure 127. Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 487
Figure 128. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
Figure 129. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
Figure 130. Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
Figure 131. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Figure 132. Counter timing diagram, Update event when ARPE=0 (TIMx_ARR not preloaded). . . . . 489
Figure 133. Counter timing diagram, Update event when ARPE=1 (TIMx_ARR preloaded). . . . . . . . 490
Figure 134. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Figure 135. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Figure 136. Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Figure 137. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
Figure 138. Counter timing diagram, Update event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
Figure 139. Counter timing diagram, internal clock divided by 1, TIMx_ARR=0x6 . . . . . . . . . . . . . . . 493
Figure 140. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
Figure 141. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36 . . . . . . . . . . . . . . 494
Figure 142. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
Figure 143. Counter timing diagram, Update event with ARPE=1 (counter underflow). . . . . . . . . . . . 495
Figure 144. Counter timing diagram, Update event with ARPE=1 (counter overflow) . . . . . . . . . . . . . 495
Figure 145. Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 496
Figure 146. TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
Figure 147. Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
Figure 148. External trigger input block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
Figure 149. Control circuit in external clock mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
Figure 150. Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 500

RM0402 Rev 6 39/1163

44
List of figures RM0402

Figure 151. Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500

Figure 152. Output stage of capture/compare channel (channel 1). . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Figure 153. PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
Figure 154. Output compare mode, toggle on OC1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
Figure 155. Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
Figure 156. Center-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
Figure 157. Example of one-pulse mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
Figure 158. Clearing TIMx OCxREF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
Figure 159. Example of counter operation in encoder interface mode . . . . . . . . . . . . . . . . . . . . . . . . 512
Figure 160. Example of encoder interface mode with TI1FP1 polarity inverted . . . . . . . . . . . . . . . . . 512
Figure 161. Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
Figure 162. Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
Figure 163. Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Figure 164. Control circuit in external clock mode 2 + trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . 516
Figure 165. Master/Slave timer example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
Figure 166. Gating timer 2 with OC1REF of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Figure 167. Gating timer 2 with Enable of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
Figure 168. Triggering timer 2 with update of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
Figure 169. Triggering timer 2 with Enable of timer 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
Figure 170. Triggering timer 1 and 2 with timer 1 TI1 input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
Figure 171. General-purpose timer block diagram (TIM9 and TIM12) . . . . . . . . . . . . . . . . . . . . . . . . 546
Figure 172. General-purpose timer block diagram (TIM10/11/13/14) . . . . . . . . . . . . . . . . . . . . . . . . . 547
Figure 173. Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 549
Figure 174. Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 549
Figure 175. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
Figure 176. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Figure 177. Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Figure 178. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Figure 179. Counter timing diagram, update event when ARPE=0
(TIMx_ARR not preloaded) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
Figure 180. Counter timing diagram, update event when ARPE=1
(TIMx_ARR preloaded) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
Figure 181. Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 553
Figure 182. TI2 external clock connection example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
Figure 183. Control circuit in external clock mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
Figure 184. Capture/compare channel (example: channel 1 input stage) . . . . . . . . . . . . . . . . . . . . . . 555
Figure 185. Capture/compare channel 1 main circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Figure 186. Output stage of capture/compare channel (channel 1). . . . . . . . . . . . . . . . . . . . . . . . . . . 556
Figure 187. PWM input mode timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
Figure 188. Output compare mode, toggle on OC1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
Figure 189. Edge-aligned PWM waveforms (ARR=8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
Figure 190. Example of one pulse mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
Figure 191. Control circuit in reset mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564
Figure 192. Control circuit in gated mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Figure 193. Control circuit in trigger mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Figure 194. Basic timer block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
Figure 195. Counter timing diagram with prescaler division change from 1 to 2 . . . . . . . . . . . . . . . . . 594
Figure 196. Counter timing diagram with prescaler division change from 1 to 4 . . . . . . . . . . . . . . . . . 594
Figure 197. Counter timing diagram, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Figure 198. Counter timing diagram, internal clock divided by 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
Figure 199. Counter timing diagram, internal clock divided by 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
Figure 200. Counter timing diagram, internal clock divided by N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597

40/1163 RM0402 Rev 6

RM0402 List of figures

Figure 201. Counter timing diagram, update event when ARPE = 0 (TIMx_ARR not
preloaded). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Figure 202. Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
Figure 203. Control circuit in normal mode, internal clock divided by 1 . . . . . . . . . . . . . . . . . . . . . . . . 599
Figure 204. Independent watchdog block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Figure 205. Watchdog block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
Figure 206. Window watchdog timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
Figure 207. RTC block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
Figure 208. FMPI2C block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
Figure 209. I2C bus protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
Figure 210. Setup and hold timings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
Figure 211. FMPI2C initialization flowchart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
Figure 212. Data reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
Figure 213. Data transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
Figure 214. Slave initialization flowchart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
Figure 215. Transfer sequence flowchart for FMPI2C slave transmitter,
NOSTRETCH= 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
Figure 216. Transfer sequence flowchart for FMPI2C slave transmitter,
NOSTRETCH= 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673
Figure 217. Transfer bus diagrams for FMPI2C slave transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
Figure 218. Transfer sequence flowchart for slave receiver with NOSTRETCH=0 . . . . . . . . . . . . . . 675
Figure 219. Transfer sequence flowchart for slave receiver with NOSTRETCH=1 . . . . . . . . . . . . . . 676
Figure 220. Transfer bus diagrams for FMPI2C slave receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676
Figure 221. Master clock generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678
Figure 222. Master initialization flowchart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
Figure 223. 10-bit address read access with HEAD10R=0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
Figure 224. 10-bit address read access with HEAD10R=1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
Figure 225. Transfer sequence flowchart for FMPI2C master transmitter for N≤255 bytes. . . . . . . . . 682
Figure 226. Transfer sequence flowchart for FMPI2C master transmitter for N>255 bytes. . . . . . . . . 683
Figure 227. Transfer bus diagrams for FMPI2C master transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . 684
Figure 228. Transfer sequence flowchart for FMPI2C master receiver for N≤255 bytes . . . . . . . . . . 686
Figure 229. Transfer sequence flowchart for FMPI2C master receiver for N >255 bytes . . . . . . . . . . 687
Figure 230. Transfer bus diagrams for FMPI2C master receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688
Figure 231. Timeout intervals for tLOW:SEXT, tLOW:MEXT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
Figure 232. Transfer sequence flowchart for SMBus slave transmitter N bytes + PEC. . . . . . . . . . . . 696
Figure 233. Transfer bus diagrams for SMBus slave transmitter (SBC=1) . . . . . . . . . . . . . . . . . . . . . 697
Figure 234. Transfer sequence flowchart for SMBus slave receiver N Bytes + PEC . . . . . . . . . . . . . 698
Figure 235. Bus transfer diagrams for SMBus slave receiver (SBC=1). . . . . . . . . . . . . . . . . . . . . . . . 699
Figure 236. Bus transfer diagrams for SMBus master transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700
Figure 237. Bus transfer diagrams for SMBus master receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
Figure 238. I2C bus protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Figure 239. I2C block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726
Figure 240. Transfer sequence diagram for slave transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
Figure 241. Transfer sequence diagram for slave receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729
Figure 242. Transfer sequence diagram for master transmitter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732
Figure 243. Transfer sequence diagram for master receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
Figure 244. I2C interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 743
Figure 245. USART block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
Figure 246. Word length programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763
Figure 247. Configurable stop bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
Figure 248. TC/TXE behavior when transmitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766

RM0402 Rev 6 41/1163

44
List of figures RM0402

Figure 249. Start bit detection when oversampling by 16 or 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767

Figure 250. Data sampling when oversampling by 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770
Figure 251. Data sampling when oversampling by 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
Figure 252. Mute mode using Idle line detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
Figure 253. Mute mode using address mark detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785
Figure 254. Break detection in LIN mode (11-bit break length - LBDL bit is set) . . . . . . . . . . . . . . . . . 787
Figure 255. Break detection in LIN mode vs. Framing error detection. . . . . . . . . . . . . . . . . . . . . . . . . 788
Figure 256. USART example of synchronous transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
Figure 257. USART data clock timing diagram (M=0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789
Figure 258. USART data clock timing diagram (M=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790
Figure 259. RX data setup/hold time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790
Figure 260. ISO 7816-3 asynchronous protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
Figure 261. Parity error detection using the 1.5 stop bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 792
Figure 262. IrDA SIR ENDEC- block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
Figure 263. IrDA data modulation (3/16) -Normal mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794
Figure 264. Transmission using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 796
Figure 265. Reception using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
Figure 266. Hardware flow control between 2 USARTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797
Figure 267. RTS flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
Figure 268. CTS flow control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798
Figure 269. USART interrupt mapping diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800
Figure 270. SPI block diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 814
Figure 271. Full-duplex single master/ single slave application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815
Figure 272. Half-duplex single master/ single slave application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
Figure 273. Simplex single master/single slave application (master in transmit-only/
slave in receive-only mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
Figure 274. Master and three independent slaves. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818
Figure 275. Multi-master application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 819
Figure 276. Hardware/software slave select management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820
Figure 277. Data clock timing diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 822
Figure 278. TXE/RXNE/BSY behavior in master / full-duplex mode (BIDIMODE=0,
RXONLY=0) in the case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825
Figure 279. TXE/RXNE/BSY behavior in slave / full-duplex mode (BIDIMODE=0,
RXONLY=0) in the case of continuous transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826
Figure 280. Transmission using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 828
Figure 281. Reception using DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829
Figure 282. TI mode transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
Figure 283. I2S block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 835
Figure 284. I2S full-duplex block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Figure 285. I2S Philips protocol waveforms (16/32-bit full accuracy, CPOL = 0). . . . . . . . . . . . . . . . . 838
Figure 286. I2S Philips standard waveforms (24-bit frame with CPOL = 0) . . . . . . . . . . . . . . . . . . . . . 838
Figure 287. Transmitting 0x8EAA33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838
Figure 288. Receiving 0x8EAA33 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839
Figure 289. I2S Philips standard (16-bit extended to 32-bit packet frame with CPOL = 0) . . . . . . . . . 839
Figure 290. Example of 16-bit data frame extended to 32-bit channel frame . . . . . . . . . . . . . . . . . . . 839
Figure 291. MSB Justified 16-bit or 32-bit full-accuracy length with CPOL = 0 . . . . . . . . . . . . . . . . . . 840
Figure 292. MSB justified 24-bit frame length with CPOL = 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
Figure 293. MSB justified 16-bit extended to 32-bit packet frame with CPOL = 0 . . . . . . . . . . . . . . . . 840
Figure 294. LSB justified 16-bit or 32-bit full-accuracy with CPOL = 0 . . . . . . . . . . . . . . . . . . . . . . . . 841
Figure 295. LSB justified 24-bit frame length with CPOL = 0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
Figure 296. Operations required to transmit 0x3478AE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
Figure 297. Operations required to receive 0x3478AE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842

42/1163 RM0402 Rev 6

RM0402 List of figures

Figure 298. LSB justified 16-bit extended to 32-bit packet frame with CPOL = 0 . . . . . . . . . . . . . . . . 842
Figure 299. Example of 16-bit data frame extended to 32-bit channel frame . . . . . . . . . . . . . . . . . . . 842
Figure 300. PCM standard waveforms (16-bit) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843
Figure 301. PCM standard waveforms (16-bit extended to 32-bit packet frame). . . . . . . . . . . . . . . . . 843
Figure 302. Audio sampling frequency definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844
Figure 303. I2S clock generator architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844
Figure 304. “No response” and “no data” operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864
Figure 305. (Multiple) block read operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864
Figure 306. (Multiple) block write operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864
Figure 307. Sequential read operation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
Figure 308. Sequential write operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
Figure 309. SDIO block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
Figure 310. SDIO adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867
Figure 311. Control unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868
Figure 312. SDIO_CK clock dephasing (BYPASS = 0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868
Figure 313. SDIO adapter command path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869
Figure 314. Command path state machine (SDIO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
Figure 315. SDIO command transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 871
Figure 316. Data path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873
Figure 317. Data path state machine (DPSM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874
Figure 318. CAN network topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923
Figure 319. Dual-CAN block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 924
Figure 320. bxCAN operating modes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 926
Figure 321. bxCAN in silent mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927
Figure 322. bxCAN in loop back mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927
Figure 323. bxCAN in combined mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928
Figure 324. Transmit mailbox states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 930
Figure 325. Receive FIFO states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 931
Figure 326. Filter bank scale configuration - register organization . . . . . . . . . . . . . . . . . . . . . . . . . . . 934
Figure 327. Example of filter numbering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
Figure 328. Filtering mechanism - example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936
Figure 329. CAN error state diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 937
Figure 330. Bit timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939
Figure 331. CAN frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940
Figure 332. Event flags and interrupt generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941
Figure 333. CAN mailbox registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 952
Figure 334. OTG_FS full-speed block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 971
Figure 335. OTG_FS A-B device connection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 973
Figure 336. OTG_FS peripheral-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975
Figure 337. OTG_FS host-only connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 979
Figure 338. SOF connectivity (SOF trigger output to TIM and ITR1 connection) . . . . . . . . . . . . . . . . 983
Figure 339. Updating OTG_HFIR dynamically (RLDCTRL = 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985
Figure 340. Device-mode FIFO address mapping and AHB FIFO access mapping . . . . . . . . . . . . . . 986
Figure 341. Host-mode FIFO address mapping and AHB FIFO access mapping . . . . . . . . . . . . . . . . 987
Figure 342. Interrupt hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 991
Figure 343. Transmit FIFO write task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1074
Figure 344. Receive FIFO read task . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1075
Figure 345. Normal bulk/control OUT/SETUP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1076
Figure 346. Bulk/control IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1080
Figure 347. Normal interrupt OUT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1083
Figure 348. Normal interrupt IN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1088
Figure 349. Isochronous OUT transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1090

RM0402 Rev 6 43/1163

44
List of figures RM0402

Figure 350. Isochronous IN transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1093

Figure 351. Receive FIFO packet read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1097
Figure 352. Processing a SETUP packet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1099
Figure 353. Bulk OUT transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1106
Figure 354. TRDT max timing case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1116
Figure 355. A-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1117
Figure 356. B-device SRP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1118
Figure 357. A-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1119
Figure 358. B-device HNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1121
Figure 359. Block diagram of STM32 MCU and Cortex®-M4 with FPU-level
debug support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1123
Figure 360. SWJ debug port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125
Figure 361. JTAG TAP connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1129
Figure 362. TPIU block diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1147

44/1163 RM0402 Rev 6

RM0402 Documentation conventions

1 Documentation conventions

1.1 General information

The STM32F412xx devices have an Arm®(a) Cortex®-M4 with FPU core.

1.2 List of abbreviations for registers

The following abbreviations(b) are used in register descriptions:

read/write (rw) Software can read and write to this bit.

read-only (r) Software can only read this bit.
write-only (w) Software can only write to this bit. Reading this bit returns the reset value.
read/clear write0 (rc_w0) Software can read as well as clear this bit by writing 0. Writing 1 has no
effect on the bit value.
read/clear write1 (rc_w1) Software can read as well as clear this bit by writing 1. Writing 0 has no
effect on the bit value.
read/clear write (rc_w) Software can read as well as clear this bit by writing to the register. The
value written to this bit is not important.
read/clear by read (rc_r) Software can read this bit. Reading this bit automatically clears it to 0.
Writing this bit has no effect on the bit value.
read/set by read (rs_r) Software can read this bit. Reading this bit automatically sets it to 1.
Writing this bit has no effect on the bit value.
read/set (rs) Software can read as well as set this bit. Writing 0 has no effect on the bit
value.
read/write once (rwo) Software can only write once to this bit and can also read it at any time.
Only a reset can return the bit to its reset value.
toggle (t) The software can toggle this bit by writing 1. Writing 0 has no effect.
read-only write trigger (rt_w1) Software can read this bit. Writing 1 triggers an event but has no effect on
the bit value.
Reserved (Res.) Reserved bit, must be kept at reset value.

a. Arm is a registered trademark of Arm Limited (or its subsidiaries) in the US and/or elsewhere.
b. This is an exhaustive list of all abbreviations applicable to STMicroelectronics microcontrollers, some of
them may not be used in the current document.

RM0402 Rev 6 45/1163

46
Documentation conventions RM0402

1.3 Glossary
This section gives a brief definition of acronyms and abbreviations used in this document:
• Word: data of 32-bit length.
• Half-word: data of 16-bit length.
• Byte: data of 8-bit length.
• IAP (in-application programming): IAP is the ability to re-program the Flash memory
of a microcontroller while the user program is running.
• ICP (in-circuit programming): ICP is the ability to program the Flash memory of a
microcontroller using the JTAG protocol, the SWD protocol or the bootloader while the
device is mounted on the user application board.
• Option bytes: product configuration bits stored in the Flash memory.
• AHB: advanced high-performance bus.

1.4 Availability of peripherals

For availability of peripherals and their number across all sales types, refer to the particular
device datasheet.

46/1163 RM0402 Rev 6

RM0402 System and memory overview

2 System and memory overview

2.1 System architecture

In STM32F412xx, the main system consists of 32-bit multilayer AHB bus matrix that
interconnects:
• Six masters:
– Cortex®-M4 with FPU core I-bus, D-bus and S-bus
– DMA1 memory bus
– DMA2 memory bus
– DMA2 peripheral bus
• Six slaves:
– Internal Flash memory ICode bus
– Internal Flash memory DCode bus
– Main internal SRAM
– AHB1 peripherals including AHB to APB bridges and APB peripherals
– AHB2 peripherals
– FSMC / QuadSPI
The bus matrix provides access from a master to a slave, enabling concurrent access and
efficient operation even when several high-speed peripherals work simultaneously. This
architecture is shown in Figure 1.

Figure 1. System architecture

ARM GP GP
Cortex-M4 DMA1 DMA2
DMA_MEM1

DMA_MEM2

DMA_P2
DMA_PI
D-bus

S-bus
I-bus

S0 S1 S2 S3 S4 S5
M0 ICODE
ACCEL

Flash
Up to 1MB
M1 DCODE
SRAM1
M2
256 KB
M3 AHB APB1
periph. 1

M4 AHB APB2
periph. 2
FSMC external
M5 MemCtrl/
Bus matrix-S QuadSPI
MSv37276V1

RM0402 Rev 6 47/1163

48
System and memory overview RM0402

2.1.1 I-bus
This bus connects the Instruction bus of the Cortex®-M4 with FPU core to the BusMatrix.
This bus is used by the core to fetch instructions. The target of this bus is a memory
containing code (internal Flash memory/SRAM1).

2.1.2 D-bus
This bus connects the databus of the Cortex®-M4 with FPU to the BusMatrix. This bus is
used by the core for literal load and debug access. The target of this bus is a memory
containing code or data (internal Flash memory/SRAM1).

2.1.3 S-bus
This bus connects the system bus of the Cortex®-M4 with FPU core to a BusMatrix. This
bus is used to access data located in a peripheral or in SRAM1. Instructions may also be
fetch on this bus (less efficient than ICode). The targets of this bus are the internal SRAM1,
the AHB1 peripherals including the APB peripherals, the AHB2 peripherals and the external
memories through the FSMC and the QUADSPI.

2.1.4 DMA memory bus

This bus connects the DMA memory bus master interface to the BusMatrix. It is used by the
DMA to perform transfer to/from memories. The targets of this bus are data memories:
internal Flash memory, internal SRAM1 and additionally for S4 the AHB1/AHB2 peripherals
including the APB peripherals.

2.1.5 DMA peripheral bus

This bus connects the DMA peripheral master bus interface to the BusMatrix. This bus is
used by the DMA to access AHB peripherals or to perform memory-to-memory transfers.
The targets of this bus are the AHB and APB peripherals plus data memories: Flash
memory and internal SRAM1.

2.1.6 BusMatrix
The BusMatrix manages the access arbitration between masters. The arbitration uses a
round-robin algorithm.

2.1.7 AHB/APB bridges (APB)

The two AHB/APB bridges, APB1 and APB2, provide full synchronous connections between
the AHB and the two APB buses, allowing flexible selection of the peripheral frequency.
Refer to the device datasheets for more details on APB1 and APB2 maximum frequencies,
and to Table 1 for the address mapping of AHB and APB peripherals.
After each device reset, all peripheral clocks are disabled (except for the SRAM and Flash
memory interface). Before using a peripheral you have to enable its clock in the
RCC_AHBxENR or RCC_APBxENR register.
Note: When a 16- or an 8-bit access is performed on an APB register, the access is transformed
into a 32-bit access: the bridge duplicates the 16- or 8-bit data to feed the 32-bit vector.

48/1163 RM0402 Rev 6

RM0402

2.2 Memory organization

2.2.1 Introduction
Program memory, data memory, registers and I/O ports are organized within the same linear
4-Gbyte address space.
The bytes are coded in memory in Little Endian format. The lowest numbered byte in a word
is considered the word’s least significant byte and the highest numbered byte the most
significant.
The addressable memory space is divided into eight main blocks, of 512 Mbytes each.

RM0402 Rev 6 49/1163

49
 RM0402

2.2.2 Memory map and register boundary addresses

Figure 2. Memory map

Reserved 0xE010 0000 - 0xFFFF FFFF

Cortex¾-M4 internal 0xE000 0000 - 0xE00F FFFF
peripherals
0xA000 2000 – 0DFFF FFF
Reserved
0xA000 1FFF
AHB3
0x6000 0000
0x5FFF FFFF
AHB2
0x5000 0000
Reserved 0x4002 6800 - 0x4FFF FFFF
0x4002 67FF

0xFFFF FFFF 512-Mbyte

block 7
internal
Peripherals AHB1
0xE000 0000
0xDFFF FFFF
512-Mbyte
block 6
Not used
0xC000 0000 0x4002 0000
0xBFFF FFFF
Reserved 0x4001 7400 - 0x4001 FFFF
Reserved
0x4001 63FF
0xA000 2000
0xA000 1FFF
FSMC
and
QuadSPI
0x6000 0000
0x5FFF FFFF APB2
512-Mbyte
block 2
Peripherals
0x4000 0000
0x3FFF FFFF
512-Mbyte
block 1 Reserved 0x2004 0000 - 0x3FFF FFFF
0x4001 0000
Peripherals
SRAM (256 KB aliased 0x2000 0000 - 0x2003 FFFF Reserved 0x4000 7400 - 0x4000 FFFF
0x2000 0000 by bit-banding) 0x4000 73FF
0x1FFF FFFF 0x1FFF C008 - 0x1FFF FFFF
Reserved
512-Mbyte 0x1FFF C000 - 0x1FFF C007
block 0 Option bytes
0x1FFF 7A10 - 0x1FFF BFFF
Peripherals Reserved 0x1FFF 7800 - 0x1FFF 7A0F
0x0000 0000 OTP area + lock 0x1FFF 0000 - 0x1FFF 77FF
System memory 0x0808 0000 - 0x1FFE FFFF
Reserved
0x0800 0000 - 0x080F FFFF APB1
Flash memory
0x0010 0000 - 0x07FF FFFF
Reserved

Aliased to Flash, system,

memory or SRAM depending, 0x0000 0000 - 0x000F FFFF
on the BOOT pins

0x4000 0000
MSv37284V4

All the memory map areas that are not allocated to on-chip memories and peripherals are
considered “Reserved”. For the detailed mapping of available memory and register areas,
refer to the following table.

50/1163 RM0402 Rev 6

RM0402

The following table gives the boundary addresses of the peripherals available in the
devices.

Table 1. Register boundary addresses

Bus Boundary address Peripheral

- 0xE010 0000 - 0xFFFF FFFF Reserved

Cortex®-M4 0xE000 0000 - 0xE00F FFFF Cortex-M4 internal peripherals
0xA000 2000 - 0xDFFF FFFF Reserved
0xA000 1000 - 0xA000 1FFF QuadSPI control register
0xA000 0000 - 0xA000 0FFF FSMC control register
AHB3
0x9000 0000 -0x9FFF FFFF QUADSPI
0x7000 0000 - 0x8FFF FFFF Reserved
0x6000 0000 - 0x6FFF FFFF FSMC
0x5006 0C00 - 0x5FFF FFFF Reserved
0x5006 0800 - 0x5006 0BFF RNG
AHB2
0x5004 0000 - 0x5006 07FF Reserved
0x5000 0000 - 0x5003 FFFF USB OTG FS
0x4002 6800 - 0x4FFF FFFF Reserved
0x4002 6400 - 0x4002 67FF DMA2
0x4002 6000 - 0x4002 63FF DMA1
0x4002 5000 - 0x4002 4FFFF Reserved
0x4002 3C00 - 0x4002 3FFF Flash interface register
0x4002 3800 - 0x4002 3BFF RCC
0x4002 3400 - 0x4002 37FF Reserved
0x4002 3000 - 0x4002 33FF CRC
AHB1 0x4002 2000 - 0x4002 2FFF Reserved
0x4002 1C00 - 0x4002 1FFF GPIOH
0x4002 1800 - 0x4002 1BFF GPIOG
0x4002 1400 - 0x4002 17FF GPIOF
0x4002 1000 - 0x4002 13FF GPIOE
0x4002 0C00 - 0x4002 0FFF GPIOD
0x4002 0800 - 0x4002 0BFF GPIOC
0x4002 0400 - 0x4002 07FF GPIOB
0x4002 0000 - 0x4002 03FF GPIOA

RM0402 Rev 6 51/1163

57
 RM0402

Table 1. Register boundary addresses (continued)

Bus Boundary address Peripheral

0x4001 6400- 0x4001 FFFF Reserved

0x4001 6000 - 0x4001 63FF DFSDM1
0x4001 5400 - 0x4001 5FFF Reserved
0x4001 5000 - 0x4001 53FF SPI5/I2S5
0x4001 4800 - 0x4001 4BFF TIM11
0x4001 4400 - 0x4001 47FF TIM10
0x4001 4000 - 0x4001 43FF TIM9
0x4001 3C00 - 0x4001 3FFF EXTI
0x4001 3800 - 0x4001 3BFF SYSCFG
0x4001 3400 - 0x4001 37FF SPI4/I2S4
APB2 0x4001 3000 - 0x4001 33FF SPI1/I2S1
0x4001 2C00 - 0x4001 2FFF SDIO
0x4001 2400 - 0x4001 2BFF Reserved
0x4001 2000 - 0x4001 23FF ADC1
0x4001 1800 - 0x4001 1FFF Reserved
0x4001 1400 - 0x4001 17FF USART6
0x4001 1000 - 0x4001 13FF USART1
0x4001 0800 - 0x4001 0FFF Reserved
0x4001 0400 - 0x4001 07FF TIM8
0x4001 0000 - 0x4001 03FF TIM1
0x4000 7400 - 0x4000 FFFF Reserved

52/1163 RM0402 Rev 6

RM0402

Table 1. Register boundary addresses (continued)

Bus Boundary address Peripheral

0x4000 7000 - 0x4000 73FF PWR

0x4000 6C00 - 0x4000 6FFF Reserved
0x4000 6800 - 0x4000 6BFF CAN2
0x4000 6400 - 0x4000 67FF CAN1
0x4000 6000 - 0x4000 63FF I2CFMP1
0x4000 5C00 - 0x4000 5FFF I2C3
0x4000 5800 - 0x4000 5BFF I2C2
0x4000 5400 - 0x4000 57FF I2C1
0x4000 4C00 - 0x4000 53FF Reserved
0x4000 4800 - 0x4000 4BFF USART3
0x4000 4400 - 0x4000 47FF USART2
0x4000 4000 - 0x4000 3FFF I2S3ext
0x4000 3C00 - 0x4000 3FFF SPI3 / I2S3
0x4000 3800 - 0x4000 3BFF SPI2 / I2S2
APB1
0x4000 3400 - 0x4000 37FF I2S2ext
0x4000 3000 - 0x4000 33FF IWDG
0x4000 2C00 - 0x4000 2FFF WWDG
0x4000 2800 - 0x4000 2BFF RTC & BKP Registers
0x4000 2400 - 0x4000 27FF Reserved
0x4000 2000 - 0x4000 23FF TIM14
0x4000 1C00 - 0x4000 1FFF TIM13
0x4000 1800 - 0x4000 1BFF TIM12
0x4000 1400 - 0x4000 17FF TIM7
0x4000 1000 - 0x4000 13FF TIM6
0x4000 0C00 - 0x4000 0FFF TIM5
0x4000 0800 - 0x4000 0BFF TIM4
0x4000 0400 - 0x4000 07FF TIM3
0x4000 0000 - 0x4000 03FF TIM2

2.3 Embedded SRAM

STM32F412xx devices feature 256 Kbytes of system SRAM.
The embedded SRAM can be accessed as bytes, half-words (16 bits) or full words (32 bits).
Read and write operations are performed at CPU speed with 0 wait state.
The CPU can access the embedded SRAM1, through the System Bus or through the I-
Code/D-Code buses when boot from SRAM is selected or when physical remap is selected

RM0402 Rev 6 53/1163

57
 RM0402

(Section 7.2.1: SYSCFG memory remap register (SYSCFG_MEMRMP) in the SYSCFG

controller). To get the max performance on SRAM execution, physical remap should be
selected (boot or software selection).

2.4 Flash memory overview

The Flash memory interface manages CPU AHB I-Code and D-Code accesses to the Flash
memory. It implements the erase and program Flash memory operations and the read and
write protection mechanisms. It accelerates code execution with a system of instruction
prefetch and cache lines.
The Flash memory is organized as follows:
• A main memory block divided into sectors.
• System memory from which the device boots in System memory boot mode
• 512 OTP (one-time programmable) bytes for user data.
• Option bytes to configure read and write protection, BOR level, watchdog
software/hardware and reset when the device is in Standby or Stop mode.
Refer to Section 3: Embedded Flash memory interface for more details.

2.5 Bit banding

The Cortex®-M4 with FPU memory map includes two bit-band regions. These regions map
each word in an alias region of memory to a bit in a bit-band region of memory. Writing to a
word in the alias region has the same effect as a read-modify-write operation on the
targeted bit in the bit-band region.
In the STM32F412xx devices both the peripheral registers and the SRAM1 are mapped to a
bit-band region, so that single bit-band write and read operations are allowed. The
operations are only available for Cortex®-M4 with FPU accesses, and not from other bus
masters (e.g. DMA).
A mapping formula shows how to reference each word in the alias region to a corresponding
bit in the bit-band region. The mapping formula is:
bit_word_addr = bit_band_base + (byte_offset x 32) + (bit_number × 4)
where:
– bit_word_addr is the address of the word in the alias memory region that maps to
the targeted bit
– bit_band_base is the starting address of the alias region
– byte_offset is the number of the byte in the bit-band region that contains the
targeted bit
– bit_number is the bit position (0-7) of the targeted bit

Example
The following example shows how to map bit 2 of the byte located at SRAM1 address
0x20000300 to the alias region:
0x22006008 = 0x22000000 + (0x300*32) + (2*4)
Writing to address 0x22006008 has the same effect as a read-modify-write operation on bit
2 of the byte at SRAM1 address 0x20000300.

54/1163 RM0402 Rev 6

RM0402

Reading address 0x22006008 returns the value (0x01 or 0x00) of bit 2 of the byte at SRAM1
address 0x20000300 (0x01: bit set; 0x00: bit reset).
For more information on bit-banding, refer to the Cortex®-M4 with FPU programming
manual (see Related documents on page 1).

2.6 Boot configuration

Due to its fixed memory map, the code area starts from address 0x0000 0000 (accessed
through the ICode/DCode buses) while the data area (SRAM) starts from address
0x2000 0000 (accessed through the system bus). The Cortex®-M4 with FPU CPU always
fetches the reset vector on the ICode bus, which implies to have the boot space available
only in the code area (typically, Flash memory). STM32F4xx microcontrollers implement a
special mechanism to be able to boot from other memories (like the internal SRAM).
In the STM32F412xx, three different boot modes can be selected through the BOOT[1:0]
pins as shown in Table 2.

Table 2. Boot modes

Boot mode selection pins
Boot mode Aliasing
BOOT1 BOOT0

x 0 Main Flash memory Main Flash memory is selected as the boot space
0 1 System memory System memory is selected as the boot space
1 1 Embedded SRAM Embedded SRAM is selected as the boot space

The values on the BOOT pins are latched on the 4th rising edge of SYSCLK after a reset. It
is up to the user to set the BOOT1 and BOOT0 pins after reset to select the required boot
mode.
BOOT0 is a dedicated pin while BOOT1 is shared with a GPIO pin. Once BOOT1 has been
sampled, the corresponding GPIO pin is free and can be used for other purposes.
The BOOT pins are also resampled when the device exits the Standby mode. Consequently,
they must be kept in the required Boot mode configuration when the device is in the Standby
mode. After this startup delay is over, the CPU fetches the top-of-stack value from address
0x0000 0000, then starts code execution from the boot memory starting from 0x0000 0004.
Note: When the device boots from SRAM, in the application initialization code, you have to
relocate the vector table in SRAM using the NVIC exception table and the offset register.

RM0402 Rev 6 55/1163

57
 RM0402

Embedded bootloader
The embedded bootloader mode is used to reprogram the Flash memory using one of the
interface described in the table below. The availability of the interface is package
dependent.

Table 3. Embedded bootloader interfaces

SPI1 SPI3 SPI4
I2C
USART1 USART2 USART3 I2C1 I2C2 I2C3 PA4/ PA15/ PE11/ CAN2 USB
FMP1
Package PA9/ PD6/ PB11/ PB6/ PF0/ PA8/ PA5/ PC10/ PE12/ PB5/ PA11
PB14/
PA10 PD5 PB10 PB7 PF1 PB4 PA6/ PC11/ PE13/ PB13 /P12
PB15
PA7 PC12 PE14

UFQFPN48 Y - - Y - Y Y Y - - Y Y
WLCSP64 Y - - Y - Y Y Y Y - Y Y
LQFP64 Y - - Y - Y Y Y Y - Y Y
LQFP100 Y Y - Y - Y Y Y Y Y Y Y
LQFP144 Y Y Y Y Y Y Y Y Y Y Y Y
UFBGA100 Y Y Y Y - Y Y Y Y Y Y Y
UFBGA144 Y Y Y Y Y Y Y Y Y Y Y Y

The USART peripherals operate at the internal 16 MHz oscillator (HSI) frequency, while the
CAN and USB OTG FS require an external clock (HSE) multiple of 1 MHz (ranging from 4 to
26 MHz).
The embedded bootloader code is located in system memory. It is programmed by ST
during production. For additional information, refer to application note AN2606.

Physical remap in STM32F412xx

Once the boot pins are selected, the application software can modify the memory
accessible in the code area (in this way the code can be executed through the ICode bus in
place of the System bus). This modification is performed by programming the Section 7.2.1:
SYSCFG memory remap register (SYSCFG_MEMRMP) in the SYSCFG controller.
The following memories can thus be remapped:
• Main Flash memory
• System memory
• Embedded SRAM

Table 4. Memory mapping vs. Boot mode/physical remap in STM32F412xx

Boot/Remap in main Boot/Remap in Boot/Remap in
Addresses
Flash memory embedded SRAM System memory

0x2000 0000 - 0x2003 FFFF SRAM (256 KB) SRAM (256KB) SRAM (256KB)
0x1FFF 0000 - 0x1FFF 77FF System memory System memory System memory
0x0802 0000 - 0x1FFE FFFF Reserved Reserved Reserved
0x0800 0000 - 0x080F FFFF Flash memory Flash memory Flash memory

56/1163 RM0402 Rev 6

RM0402

Table 4. Memory mapping vs. Boot mode/physical remap in STM32F412xx

Boot/Remap in main Boot/Remap in Boot/Remap in
Addresses
Flash memory embedded SRAM System memory

0x0400 000 - 0x07FF FFFF Reserved Reserved Reserved

SRAM1 (256 KB) System memory
0x0000 0000 - 0x0003 FFFF(1) Flash (1M) Aliased
Aliased (30 KB) Aliased
1. Even when aliased in the boot memory space, the related memory is still accessible at its original memory
space.

RM0402 Rev 6 57/1163

57
Embedded Flash memory interface RM0402

3 Embedded Flash memory interface

3.1 Introduction
The Flash memory interface manages CPU AHB I-Code and D-Code accesses to the Flash
memory. It implements the erase and program Flash memory operations and the read and
write protection mechanisms.
The Flash memory interface accelerates code execution with a system of instruction
prefetch and cache lines.

3.2 Main features

• Flash memory read operations
• Flash memory program/erase operations
• Read / write protections
• Prefetch on I-Code
• 64 cache lines of 128 bits on I-Code
• 8 cache lines of 128 bits on D-Code
Figure 3 shows the Flash memory interface connection inside the system architecture.

Figure 3. Flash memory interface connection inside system architecture

AHB Flash
Cortex-M4 with FPU 32-bit memory
instruction Flash interface bus
I-Code bus I-Code bus 128 bits
Flash
Cortex D-Code
memory
core
S bus AHB
D-code bus 32-bit FLITF registers
data bus

AHB
AHB periph1
32-bit
DMA1 system bus SRAM and
External
DMA2 memories

AHB
periph2

Access to instruction in Flash memory

Access to data and literal pool in Flash memory
FLITF register access

MS31423V1

58/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

3.3 Embedded Flash memory

The Flash memory has the following main features:
• Capacity up to 1 Mbyte
• 128 bits wide data read
• Byte, half-word, word and double word write
• Sector and mass erase
• Memory organization
The Flash memory is organized as follows:
– A main memory block divided into 4 sectors of 16 Kbyte, plus 1 sector of 64 Kbyte
and plus 7 sector of 128 Kbyte.
– System memory from which the device boots in System memory boot mode
– 512 OTP (one-time programmable) bytes for user data
The OTP area contains 32 additional bits used to lock the corresponding OTP
data block.
– Option bytes to configure read and write protection, BOR level, watchdog
software/hardware and reset when the device is in Standby or Stop mode.
• Low-power modes (for details refer to the Power control (PWR) section of the reference
manual)

Table 5. Flash module organization

Block Name Block base addresses Size

Sector 0 0x0800 0000 - 0x0800 3FFF 16 Kbyte

Sector 1 0x0800 4000 - 0x0800 7FFF 16 Kbyte
Sector 2 0x0800 8000 - 0x0800 BFFF 16 Kbyte
Sector 3 0x0800 C000 - 0x0800 FFFF 16 Kbyte
Sector 4 0x0801 0000 - 0x0801 FFFF 64 Kbyte

Sector 5 0x0802 0000 - 0x0803 FFFF 128 Kbyte

Main memory Sector 6 0x0804 0000 - 0x0805 FFFF 128 Kbyte

Sector 7 0x0806 0000 - 0x0807 FFFF 128 Kbyte

Sector 8(1) 0x0808 0000 - 0x0809 FFFF 128 Kbyte

Sector 9(1) 0x080A 0000 - 0x080B FFFF 128 Kbyte

Sector 10(1) 0x080C 0000 - 0x080D FFFF 128 Kbyte

Sector 11(1) 0x080E 0000 - 0x080F FFFF 128 Kbyte

System memory 0x1FFF 0000 - 0x1FFF 77FF 30 Kbyte

OTP area 0x1FFF 7800 - 0x1FFF 7A0F 528 byte

Option bytes 0x1FFF C000 - 0x1FFF C00F 16 byte
1. Not available for STM32F412xE products

RM0402 Rev 6 59/1163

81
Embedded Flash memory interface RM0402

3.4 Read interface

3.4.1 Relation between CPU clock frequency and Flash memory read time
To correctly read data from Flash memory, the number of wait states (LATENCY) must be
correctly programmed in the Flash access control register (FLASH_ACR) according to the
frequency of the CPU clock (HCLK) and the supply voltage of the device.
The prefetch buffer must be disabled when the supply voltage is below 2.1 V. The
correspondence between wait states and CPU clock frequency is given inTable 8:
• When VOS[1:0] = 0x01, the maximum value of fHCLK = 64 MHz.
• When VOS[1:0] = 0x10, the maximum value of fHCLK = 84 MHz.
• When VOS[1:0] = 0x11, the maximum value of fHCLK = 100 MHz.

Table 6. Number of wait states according to CPU clock (HCLK) frequency

HCLK (MHz)
Wait states (WS)
(LATENCY) Voltage range Voltage range Voltage range Voltage range
2.7 V - 3.6 V 2.4 V - 2.7 V 2.1 V - 2.4 V 1.7 V - 2.1 V

0 WS (1 CPU cycle) 0 < HCLK ≤ 30 0 < HCLK ≤ 24 0 < HCLK ≤ 18 0 < HCLK ≤ 16
1 WS (2 CPU cycles) 30 < HCLK ≤ 64 24 < HCLK ≤ 48 18 < HCLK ≤ 36 16 <HCLK ≤ 32
2 WS (3 CPU cycles) 64 < HCLK ≤ 90 48 < HCLK ≤ 72 36 < HCLK ≤ 54 32 < HCLK ≤ 48
3 WS (4 CPU cycles) 90 < HCLK ≤ 100 72 < HCLK ≤ 96 54 < HCLK ≤ 72 48 < HCLK ≤ 64
4 WS (5 CPU cycles) - 96 < HCLK ≤ 100 72 < HCLK ≤ 90 64 < HCLK ≤ 80
5 WS (6 CPU cycles) - - 90 < HCLK ≤ 100 80 < HCLK ≤ 96
6 WS (7 CPU cycles) - - - 96 < HCLK ≤ 100

After reset, the CPU clock frequency is 16 MHz and 0 wait state (WS) is configured in the
FLASH_ACR register.
It is highly recommended to use the following software sequences to tune the number of
wait states needed to access the Flash memory with the CPU frequency.

Increasing the CPU frequency

1. Program the new number of wait states to the LATENCY bits in the FLASH_ACR
register
2. Check that the new number of wait states is taken into account to access the Flash
memory by reading the FLASH_ACR register
3. Modify the CPU clock source by writing the SW bits in the RCC_CFGR register
4. If needed, modify the CPU clock prescaler by writing the HPRE bits in RCC_CFGR
5. Check that the new CPU clock source or/and the new CPU clock prescaler value is/are
taken into account by reading the clock source status (SWS bits) or/and the AHB
prescaler value (HPRE bits), respectively, in the RCC_CFGR register.

60/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Decreasing the CPU frequency

1. Modify the CPU clock source by writing the SW bits in the RCC_CFGR register
2. If needed, modify the CPU clock prescaler by writing the HPRE bits in RCC_CFGR
3. Check that the new CPU clock source or/and the new CPU clock prescaler value is/are
taken into account by reading the clock source status (SWS bits) or/and the AHB
prescaler value (HPRE bits), respectively, in the RCC_CFGR register
4. Program the new number of wait states to the LATENCY bits in FLASH_ACR
5. Check that the new number of wait states is used to access the Flash memory by
reading the FLASH_ACR register
Note: A change in CPU clock configuration or wait state (WS) configuration may not be effective
straight away. To make sure that the current CPU clock frequency is the one you have
configured, you can check the AHB prescaler factor and clock source status values. To
make sure that the number of WS you have programmed is effective, you can read the
FLASH_ACR register.

3.4.2 Adaptive real-time memory accelerator (ART Accelerator™)

The proprietary Adaptive real-time (ART) memory accelerator is optimized for STM32
industry-standard Arm® Cortex®-M4 with FPU processors. It balances the inherent
performance advantage of the Arm® Cortex®-M4 with FPU over Flash memory
technologies, which normally requires the processor to wait for the Flash memory at higher
operating frequencies.
To release the processor full performance, the accelerator implements an instruction
prefetch queue and branch cache which increases program execution speed from the 128-
bit Flash memory. Based on CoreMark benchmark, the performance achieved thanks to the
ART accelerator is equivalent to 0 wait state program execution from Flash memory at a
CPU frequency up to 100 MHz.

Instruction prefetch
Each Flash memory read operation provides 128 bits from either four instructions of 32 bits
or 8 instructions of 16 bits according to the program launched. So, in case of sequential
code, at least four CPU cycles are needed to execute the previous read instruction line.
Prefetch on the I-Code bus can be used to read the next sequential instruction line from the
Flash memory while the current instruction line is being requested by the CPU. Prefetch is
enabled by setting the PRFTEN bit in the FLASH_ACR register. This feature is useful if at
least one wait state is needed to access the Flash memory.
Figure 4 shows the execution of sequential 32-bit instructions with and without prefetch
when 3 WSs are needed to access the Flash memory.

RM0402 Rev 6 61/1163

81
Embedded Flash memory interface RM0402

Figure 4. Sequential 32-bit instruction execution

@ F D E
WAIT
1 1 1 1
Without prefetch
@ F D E
2 2 2 2
@ F D E
3 3 3 3
@ F D E
4 4 4 4
@ F D E
WAIT
5 5 5 5
@ F D E
6 6 6 6
@ F D
7 7 7
@ F
8 8

ins 1 ins 2 ins 3 ins 4 ins 5 ins 6 ins 7 ins 8

fetch fetch fetch fetch fetch fetch fetch fetch

Read ins 1, 2, 3, 4 Gives ins 1, 2, 3, 4 Read ins 5, 6, 7, 8 Gives ins 5, 6, 7, 8

@ Wait data F D E
1 1 1 1 With prefetch
@ F D E
2 2 2 2
@ F D E
3 3 3 3
@ F D E
4 4 4 4
@ F D E
5 5 5 5
@ F D E
6 6 6
@ F D Cortex-M4 pipeline
7 7 7
@ F
@ F D E
8 8

AHB protocol
ins 1 ins 2 ins 3 ins 4 ins 5 ins 6 ins 7 ins 8
fetch fetch fetch fetch fetch fetch fetch fetch @ : address requested
F: Fetch stage
D: Decode stage
Read ins 1, 2, 3, 4 Gives ins 1, 2, 3, 4 Gives ins 5, 6, 7, 8 E: Execute stage

Read ins 5, 6, 7, 8 Read ins 9, 10, ...

MS31831V1

When the code is not sequential (branch), the instruction may not be present in the currently
used instruction line or in the prefetched instruction line. In this case (miss), the penalty in
terms of number of cycles is at least equal to the number of wait states.

62/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Instruction cache memory

To limit the time lost due to jumps, it is possible to retain 64 lines of 128 bits in an instruction
cache memory. This feature can be enabled by setting the instruction cache enable (ICEN)
bit in the FLASH_ACR register. Each time a miss occurs (requested data not present in the
currently used instruction line, in the prefetched instruction line or in the instruction cache
memory), the line read is copied into the instruction cache memory. If some data contained
in the instruction cache memory are requested by the CPU, they are provided without
inserting any delay. Once all the instruction cache memory lines have been filled, the LRU
(least recently used) policy is used to determine the line to replace in the instruction memory
cache. This feature is particularly useful in case of code containing loops.

Data management
Literal pools are fetched from Flash memory through the D-Code bus during the execution
stage of the CPU pipeline. The CPU pipeline is consequently stalled until the requested
literal pool is provided. To limit the time lost due to literal pools, accesses through the AHB
databus D-Code have priority over accesses through the AHB instruction bus I-Code.
If some literal pools are frequently used, the data cache memory can be enabled by setting
the data cache enable (DCEN) bit in the FLASH_ACR register. This feature works like the
instruction cache memory, but the retained data size is limited to 8 rows of 128 bits.
Note: Data in user configuration sector are not cacheable.

3.5 Erase and program operations

For any Flash memory program operation (erase or program), the CPU clock frequency
(HCLK) must be at least 1 MHz. The contents of the Flash memory are not guaranteed if a
device reset occurs during a Flash memory operation.
Any attempt to read the Flash memory on STM32F4xx while it is being written or erased,
causes the bus to stall. Read operations are processed correctly once the program
operation has completed. This means that code or data fetches cannot be performed while
a write/erase operation is ongoing.

3.5.1 Unlocking the Flash control register

After reset, write is not allowed in the Flash control register (FLASH_CR) to protect the
Flash memory against possible unwanted operations due, for example, to electric
disturbances. The following sequence is used to unlock this register:
1. Write KEY1 = 0x45670123 in the Flash key register (FLASH_KEYR)
2. Write KEY2 = 0xCDEF89AB in the Flash key register (FLASH_KEYR)
Any wrong sequence will return a bus error and lock up the FLASH_CR register until the
next reset.
The FLASH_CR register can be locked again by software by setting the LOCK bit in the
FLASH_CR register.
Note: The FLASH_CR register is not accessible in write mode when the BSY bit in the FLASH_SR
register is set. Any attempt to write to it with the BSY bit set will cause the AHB bus to stall
until the BSY bit is cleared.

RM0402 Rev 6 63/1163

81
Embedded Flash memory interface RM0402

3.5.2 Program/erase parallelism

The Parallelism size is configured through the PSIZE field in the FLASH_CR register. It
represents the number of bytes to be programmed each time a write operation occurs to the
Flash memory. PSIZE is limited by the supply voltage and by whether the external VPP
supply is used or not. It must therefore be correctly configured in the FLASH_CR register
before any programming/erasing operation.
A Flash memory erase operation can only be performed by sector or for the whole Flash
memory (mass erase). The erase time depends on PSIZE programmed value. For more
details on the erase time, refer to the electrical characteristics section of the device
datasheet.
Table 7 provides the correct PSIZE values.

Table 7. Program/erase parallelism

Voltage range 2.7 - 3.6 V Voltage range Voltage range Voltage range Voltage range
with External VPP 2.7 - 3.6 V 2.4 - 2.7 V 2.1 - 2.4 V 1.7 V - 2.1 V

Parallelism size x64 x32 x16 x8

PSIZE(1:0) 11 10 01 00

Note: Any program or erase operation started with inconsistent program parallelism/voltage range
settings may lead to unpredicted results. Even if a subsequent read operation indicates that
the logical value was effectively written to the memory, this value may not be retained.
To use VPP, an external high-voltage supply (between 8 and 9 V) must be applied to the VPP
pad. The external supply must be able to sustain this voltage range even if the DC
consumption exceeds 10 mA. It is advised to limit the use of VPP to initial programming on
the factory line. The VPP supply must not be applied for more than an hour, otherwise the
Flash memory might be damaged.

3.5.3 Erase
The Flash memory erase operation can be performed at sector level or on the whole Flash
memory (Mass Erase). Mass Erase does not affect the OTP sector or the configuration
sector.

Sector Erase
To erase a sector, follow the procedure below:
1. Check that no Flash memory operation is ongoing by checking the BSY bit in the
FLASH_SR register
2. Set the SER bit and select the sector out of the 12 sectors in the main memory block
you wish to erase (SNB) in the FLASH_CR register
3. Set the STRT bit in the FLASH_CR register
4. Wait for the BSY bit to be cleared

64/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Mass Erase
To perform Mass Erase, the following sequence is recommended:
1. Check that no Flash memory operation is ongoing by checking the BSY bit in the
FLASH_SR register
2. Set the MER bit in the FLASH_CR register
3. Set the STRT bit in the FLASH_CR register
4. Wait for the BSY bit to be cleared
Note: If MERx and SER bits are both set in the FLASH_CR register, mass erase is performed.
If both MERx and SER bits are reset and the STRT bit is set, an unpredictable behavior may
occur without generating any error flag. This condition should be forbidden.

3.5.4 Programming
Standard programming
The Flash memory programming sequence is as follows:
1. Check that no main Flash memory operation is ongoing by checking the BSY bit in the
FLASH_SR register.
2. Set the PG bit in the FLASH_CR register
3. Perform the data write operation(s) to the desired memory address (inside main
memory block or OTP area):
– Byte access in case of x8 parallelism
– Half-word access in case of x16 parallelism
– Word access in case of x32 parallelism
– Double word access in case of x64 parallelism
4. Wait for the BSY bit to be cleared.
Note: Successive write operations are possible without the need of an erase operation when
changing bits from ‘1’ to ‘0’. Writing ‘1’ requires a Flash memory erase operation.
If an erase and a program operation are requested simultaneously, the erase operation is
performed first.

Programming errors
It is not allowed to program data to the Flash memory that would cross the 128-bit row
boundary. In such a case, the write operation is not performed and a program alignment
error flag (PGAERR) is set in the FLASH_SR register.
The write access type (byte, half-word, word or double word) must correspond to the type of
parallelism chosen (x8, x16, x32 or x64). If not, the write operation is not performed and a
program parallelism error flag (PGPERR) is set in the FLASH_SR register.
If the standard programming sequence is not respected (for example, if there is an attempt
to write to a Flash memory address when the PG bit is not set), the operation is aborted and
a program sequence error flag (PGSERR) is set in the FLASH_SR register.

RM0402 Rev 6 65/1163

81
Embedded Flash memory interface RM0402

Programming and caches

If a Flash memory write access concerns some data in the data cache, the Flash write
access modifies the data in the Flash memory and the data in the cache.
If an erase operation in Flash memory also concerns data in the data or instruction cache,
you have to make sure that these data are rewritten before they are accessed during code
execution. If this cannot be done safely, it is recommended to flush the caches by setting the
DCRST and ICRST bits in the FLASH_CR register.
Note: The I/D cache should be flushed only when it is disabled (I/DCEN = 0).

3.5.5 Interrupts
Setting the end of operation interrupt enable bit (EOPIE) in the FLASH_CR register enables
interrupt generation when an erase or program operation ends, that is when the busy bit
(BSY) in the FLASH_SR register is cleared (operation completed, correctly or not). In this
case, the end of operation (EOP) bit in the FLASH_SR register is set.
If an error occurs during a program, an erase, or a read operation request, one of the
following error flags is set in the FLASH_SR register:
• PGAERR, PGPERR, PGSERR (Program error flags)
• WRPERR (Protection error flag)
In this case, if the error interrupt enable bit (ERRIE) is set in the FLASH_CR register, an
interrupt is generated and the operation error bit (OPERR) is set in the FLASH_SR register.
Note: If several successive errors are detected (for example, in case of DMA transfer to the Flash
memory), the error flags cannot be cleared until the end of the successive write requests.

Table 8. Flash interrupt request

Interrupt event Event flag Enable control bit

End of operation EOP EOPIE

Write protection error WRPERR ERRIE
Programming error PGAERR, PGPERR, PGSERR ERRIE

3.6 Option bytes

3.6.1 Description of user option bytes

The option bytes are configured by the end user depending on the application requirements.
Table 9 shows the organization of these bytes inside the user configuration sector.

Table 9. Option byte organization

Address [63:16] [15:0]

0x1FFF C0000 Reserved ROP & user option bytes (RDP & USER)
0x1FFF C008 Reserved Write protection nWRP bits for sectors 0 to 11

66/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Table 10. Description of the option bytes

Option bytes (word, address 0x1FFF C000)
RDP: Read protection option byte.
The read protection is used to protect the software code stored in Flash memory.
0xAA: Level 0, no protection
Bits 15:8 0xCC: Level 2, chip protection (debug and boot from RAM features disabled)
Others: Level 1, read protection of memories (debug features limited)
USER: User option byte
This byte is used to configure the following features:
– Select the watchdog event: Hardware or software
– Reset event when entering the Stop mode
– Reset event when entering the Standby mode
nRST_STDBY
Bit 7 0: Reset generated when entering the Standby mode
1: No reset generated
nRST_STOP
Bit 6 0: Reset generated when entering the Stop mode
1: No reset generated
WDG_SW
Bit 5 0: Hardware independent watchdog
1: Software independent watchdog
Bit 4 0x1: Not used
BOR_LEV: BOR reset Level
These bits contain the supply level threshold that activates/releases the reset.
They can be written to program a new BOR level value into Flash memory.
00: BOR Level 3 (VBOR3), brownout threshold level 3
Bits 3:2 01: BOR Level 2 (VBOR2), brownout threshold level 2
10: BOR Level 1 (VBOR1), brownout threshold level 1
11: BOR off, POR/PDR reset threshold level is applied
Note: For full details on BOR characteristics, refer to the “Electrical characteristics”
section of the product datasheet.
Bits 1:0 0x1: Not used
Option bytes (word, address 0x1FFF C008)
SPRMOD: Selection of Protection Mode of nWPRi bits
Bit 15 0: nWPRi bits used for sector i write protection (Default)
1: nWPRi bits used for sector i PCROP protection (Sector)
Bits 14:12 Reserved

RM0402 Rev 6 67/1163

81
Embedded Flash memory interface RM0402

Table 10. Description of the option bytes

nWRP: Flash memory write protection option bytes
sector 0 to 11 can be write protected
nWRPi
If SPRMOD is reset (default value):
0: Write protection active on sector i.
Bits 11:0 1: Write protection not active on sector i.
If SPRMOD is set (active):
0: PCROP protection not active on sector i.
1: PCROP protection active on sector i.

3.6.2 Programming user option bytes

To run any operation on this sector, the option lock bit (OPTLOCK) in the Flash option
control register (FLASH_OPTCR) must be cleared. To be allowed to clear this bit, you have
to perform the following sequence:
1. Write OPTKEY1 = 0x0819 2A3B in the Flash option key register (FLASH_OPTKEYR)
2. Write OPTKEY2 = 0x4C5D 6E7F in the Flash option key register (FLASH_OPTKEYR)
The user option bytes can be protected against unwanted erase/program operations by
setting the OPTLOCK bit by software.

Modifying user option bytes

To modify the user option value, follow the sequence below:
1. Check that no Flash memory operation is ongoing by checking the BSY bit in the
FLASH_SR register
2. Write the desired option value in the FLASH_OPTCR register.
3. Set the option start bit (OPTSTRT) in the FLASH_OPTCR register
4. Wait for the BSY bit to be cleared.
Note: The value of an option is automatically modified by first erasing the user configuration sector
and then programming all the option bytes with the values contained in the FLASH_OPTCR
register.

3.6.3 Read protection (RDP)

The user area in the Flash memory can be protected against read operations by an
entrusted code. Three read protection levels are defined:
• Level 0: no read protection
When the read protection level is set to Level 0 by writing 0xAA into the read protection
option byte (RDP), all read/write operations (if no write protection is set) from/to the

68/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Flash memory are possible in all boot configurations (Flash user boot, debug or boot
from RAM).
• Level 1: read protection enabled
It is the default read protection level after option byte erase. The read protection Level
1 is activated by writing any value (except for 0xAA and 0xCC used to set Level 0 and
Level 2, respectively) into the RDP option byte. When the read protection Level 1 is set:
– No access (read, erase, program) to Flash memory can be performed while the
debug feature is connected or while booting from RAM or system memory
bootloader. A bus error is generated in case of read request.
– When booting from Flash memory, accesses (read, erase, program) to Flash
memory from user code are allowed.
When Level 1 is active, programming the protection option byte (RDP) to Level 0
causes the Flash memory to be mass-erased. As a result the user code area is cleared
before the read protection is removed. The mass erase only erases the user code area.
The other option bytes including write protections remain unchanged from before the
mass-erase operation. The OTP area is not affected by mass erase and remains
unchanged. Mass erase is performed only when Level 1 is active and Level 0
requested. When the protection level is increased (0->1, 1->2, 0->2) there is no mass
erase.
• Level 2: debug/chip read protection disabled
The read protection Level 2 is activated by writing 0xCC to the RDP option byte. When
the read protection Level 2 is set:
– All protections provided by Level 1 are active.
– Booting from RAM or system memory bootloader is no more allowed.
– JTAG, SWV (single-wire viewer), ETM, and boundary scan are disabled.
– User option bytes can no longer be changed.
– When booting from Flash memory, accesses (read, erase and program) to Flash
memory from user code are allowed.
Memory read protection Level 2 is an irreversible operation. When Level 2 is activated,
the level of protection cannot be decreased to Level 0 or Level 1.
Note: The JTAG port is permanently disabled when Level 2 is active (acting as a JTAG fuse). As a
consequence, boundary scan cannot be performed. STMicroelectronics is not able to
perform analysis on defective parts on which the Level 2 protection has been set.

RM0402 Rev 6 69/1163

81
Embedded Flash memory interface RM0402

--

Table 11. Access versus read protection level

Debug features, Boot from RAM or
Booting from Flash memory
Protection from System memory bootloader
Memory area
Level
Read Write Erase Read Write Erase

Level 1 NO NO(1) YES

Main Flash Memory
Level 2 NO YES
Level 1 YES YES
Option Bytes
Level 2 NO NO
Level 1 NO NA YES NA
OTP
Level 2 NO NA YES NA
1. The main Flash memory is only erased when the RDP changes from level 1 to 0. The OTP area remains unchanged.

Figure 5 shows how to go from one RDP level to another.

Figure 5. RDP levels

RDP /= AAh & /= CCh

Others options modified

Level 1
RDP /= AAh
RDP /= CCh
Write options default Write options
including including
RDP = CCh Write optionsincluding RDP = AAh
RDP /= CCh & /= AAh

L ev e l 2 L ev e l 0
RDP = CCh RDP = AA h
Write options
including
RDP = CCh

Options write (RDP level increase) includes

RDP = AAh
- Options erase Others option(s) modified
- New options program

Options write (RDP level decrease) includes Options write (RDP level identical) includes
- Mass erase - Options erase
- Options erase - New options program
- New options program
ai16045

3.6.4 Write protections

Up to 12 user sectors in Flash memory can be protected against unwanted write operations
due to loss of program counter contexts. When the non-write protection nWRPi bit
(0 ≤ i ≤ 11) in the FLASH_OPTCR registers is low, the corresponding sector cannot be
erased or programmed. Consequently, a mass erase cannot be performed if one of the
sectors is write-protected.

70/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

If an erase/program operation to a write-protected part of the Flash memory is attempted

(sector protected by write protection bit, OTP part locked or part of the Flash memory that
can never be written like the ICP), the write protection error flag (WRPERR) is set in the
FLASH_SR register.
Note: When the memory read protection level is selected (RDP level = 1), it is not possible to
program or erase Flash memory sector i if the CPU debug features are connected (JTAG or
single wire) or boot code is being executed from RAM, even if nWRPi = 1.

Write protection error flag

If an erase/program operation to a write protected area of the Flash memory is performed,
the Write Protection Error flag (WRPERR) is set in the FLASH_SR register.
If an erase operation is requested, the WRPERR bit is set when:
• Mass, sector erase are configured (MER or MER/MER1 and SER = 1)
• A sector erase is requested and the Sector Number SNB field is not valid
• A mass erase is requested while at least one of the user sector is write protected by
option bit (MER or MER/MER1 = 1 and nWRPi = 0 with 0 ≤ i ≤ 11 bits in the
FLASH_OPTCRx register
• A sector erase is requested on a write protected sector. (SER = 1, SNB = i and
nWRPi = 0 with 0 ≤ i ≤ 11 bits in the FLASH_OPTCRx register)
• The Flash memory is readout protected and an intrusion is detected.
If a program operation is requested, the WRPERR bit is set when:
• A write operation is performed on system memory or on the reserved part of the user
specific sector.
• A write operation is performed to the user configuration sector
• A write operation is performed on a sector write protected by option bit.
• A write operation is requested on an OTP area which is already locked
• The Flash memory is read protected and an intrusion is detected.

3.6.5 Proprietary code readout protection (PCROP)

Flash memory user sectors (0 to 11) can be protected against D-bus read accesses by
using the proprietary readout protection (PCROP).
The PCROP protection is selected as follows, through the SPRMOD option bit in the
FLASH_CR register:
• SPRMOD = 0: nWRPi control the write protection of respective user sectors
• SPRMOD = 1: nWRPi control the read and write protection (PCROP) of respective
user sectors.
When a sector is readout protected (PCROP mode activated), it can only be accessed for
code fetch through ICODE Bus on Flash interface:
• Any read access performed through the D-bus triggers a RDERR flag error.
• Any program/erase operation on a PCROPed sector triggers a WRPERR flag error.

RM0402 Rev 6 71/1163

81
Embedded Flash memory interface RM0402

Figure 6. PCROP levels

Write options
SPMOD = active
and valid nWRPi*

Level 1
RDP /= 0xAA
RDP /= 0xCC
Write options default Write options No restriction on
SPMOD = active SPMOD = active Write options
and valid nWRPi* and valid nWRPi*

Level 2 Level 0
RDP = 0xCC RDP = 0xAA
Write options
SPMOD = active
Write options
and valid nWRPi*
SPMOD = active
User option sector erase and valid nWRPi*
Program new options

Global mass erase

User option section erase
Program new options
* Valid nWRPi means that none of the nWRP bits set can be reset (transition from 1 to 0)
MS30467V1

The deactivation of the SPRMOD and/or the unprotection of PCROPed user sectors can
only occur when, at the same time, the RDP level changes from 1 to 0. If this condition is not
respected, the user option byte modification is canceled and the write error WRPERR flag is
set. The modification of the users option bytes (BOR_LEV, RST_STDBY, ..) is allowed since
none of the active nWRPi bits is reset and SPRMOD is kept active.
Note: The active value of nWRPi bits is inverted when PCROP mode is active (SPRMOD =1).

72/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

3.7 One-time programmable bytes

Table 12 shows the organization of the one-time programmable (OTP) part of the OTP area.

Table 12. OTP area organization

Block [128:96] [95:64] [63:32] [31:0] Address byte 0

OTP0 OTP0 OTP0 OTP0 0x1FFF 7800

0
OTP0 OTP0 OTP0 OTP0 0x1FFF 7810
OTP1 OTP1 OTP1 OTP1 0x1FFF 7820
1
OTP1 OTP1 OTP1 OTP1 0x1FFF 7830
. . .
. . .
. . .
OTP15 OTP15 OTP15 OTP15 0x1FFF 79E0
15
OTP15 OTP15 OTP15 OTP15 0x1FFF 79F0
LOCKB15 ... LOCKB11 ... LOCKB7 ... LOCKB3 ...
Lock block 0x1FFF 7A00
LOCKB12 LOCKB8 LOCKB4 LOCKB0

The OTP area is divided into 16 OTP data blocks of 32 bytes and one lock OTP block of 16
bytes. The OTP data and lock blocks cannot be erased. The lock block contains 16 bytes
LOCKBi (0 ≤ i ≤ 15) to lock the corresponding OTP data block (blocks 0 to 15). Each OTP
data block can be programmed until the value 0x00 is programmed in the corresponding
OTP lock byte. The lock bytes must only contain 0x00 and 0xFF values, otherwise the OTP
bytes might not be taken into account correctly.

RM0402 Rev 6 73/1163

81
Embedded Flash memory interface RM0402

3.8 Flash interface registers

3.8.1 Flash access control register (FLASH_ACR)

The Flash access control register is used to enable/disable the acceleration features and
control the Flash memory access time according to CPU frequency.
Address offset: 0x00
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. DCRST ICRST DCEN ICEN PRFTEN Res. Res. Res. Res. LATENCY

rw w rw rw rw rw rw rw rw

Bits 31:13 Reserved, must be kept cleared.

Bit 12 DCRST: Data cache reset
0: Data cache is not reset
1: Data cache is reset
This bit can be written only when the D cache is disabled.
Bit 11 ICRST: Instruction cache reset
0: Instruction cache is not reset
1: Instruction cache is reset
This bit can be written only when the I cache is disabled.
Bit 10 DCEN: Data cache enable
0: Data cache is disabled
1: Data cache is enabled
Bit 9 ICEN: Instruction cache enable
0: Instruction cache is disabled
1: Instruction cache is enabled
Bit 8 PRFTEN: Prefetch enable
0: Prefetch is disabled
1: Prefetch is enabled
Bits 7:4 Reserved, must be kept cleared.
Bits 3:0 LATENCY: Latency
These bits represent the ratio of the CPU clock period to the Flash memory access time.
0000: Zero wait state
0001: One wait state
0010: Two wait states
-
-
-
1110: Fourteen wait states
1111: Fifteen wait states

74/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

3.8.2 Flash key register (FLASH_KEYR)

The Flash key register is used to allow access to the Flash control register and so, to allow
program and erase operations.
Address offset: 0x04
Reset value: 0x0000 0000
Access: no wait state, word access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
KEY[31:16]
w w w w w w w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
KEY[15:0]
w w w w w w w w w w w w w w w w

Bits 31:0 FKEYR: FPEC key

The following values must be programmed consecutively to unlock the FLASH_CR register
and allow programming/erasing it:
a) KEY1 = 0x45670123
b) KEY2 = 0xCDEF89AB

3.8.3 Flash option key register (FLASH_OPTKEYR)

The Flash option key register is used to allow program and erase operations in the user
configuration sector.
Address offset: 0x08
Reset value: 0x0000 0000
Access: no wait state, word access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OPTKEYR[31:16
w w w w w w w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OPTKEYR[15:0]
w w w w w w w w w w w w w w w w

Bits 31:0 OPTKEYR: Option byte key

The following values must be programmed consecutively to unlock the FLASH_OPTCR
register and allow programming it:
a) OPTKEY1 = 0x08192A3B
b) OPTKEY2 = 0x4C5D6E7F

RM0402 Rev 6 75/1163

81
Embedded Flash memory interface RM0402

3.8.4 Flash status register (FLASH_SR)

The Flash status register gives information on ongoing program and erase operations.
Address offset: 0x0C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. BSY

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. RDERR PGSERR PGPERR PGAERR WRPERR Res. Res. OPERR EOP

rw rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bits 31:17 Reserved, must be kept cleared.

Bit 16 BSY: Busy
This bit indicates that a Flash memory operation is in progress. It is set at the beginning of a
Flash memory operation and cleared when the operation finishes or an error occurs.
0: no Flash memory operation ongoing
1: Flash memory operation ongoing
Bits 15:9 Reserved, must be kept cleared.
Bit 8 RDERR: Read Protection Error (PCROP)
Set by hardware when an address to be read through the Dbus belongs to a read protected
part of the flash.
Reset by writing 1.
Bit 7 PGSERR: Programming sequence error
Set by hardware when a write access to the Flash memory is performed by the code while
the control register has not been correctly configured.
Cleared by writing 1.
Bit 6 PGPERR: Programming parallelism error
Set by hardware when the size of the access (byte, half-word, word, double word) during the
program sequence does not correspond to the parallelism configuration PSIZE (x8, x16,
x32, x64).
Cleared by writing 1.
Bit 5 PGAERR: Programming alignment error
Set by hardware when the data to program cannot be contained in the same 128-bit Flash
memory row.
Cleared by writing 1.
Bit 4 WRPERR: Write protection error
Set by hardware when an address to be erased/programmed belongs to a write-protected
part of the Flash memory.
Cleared by writing 1.

76/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Bits 3:2 Reserved, must be kept cleared.

Bit 1 OPERR: Operation error
Set by hardware when a flash operation (programming / erase /read) request is detected and
can not be run because of parallelism, alignment, or write protection error. This bit is set only
if error interrupts are enabled (ERRIE = 1).
Bit 0 EOP: End of operation
Set by hardware when one or more Flash memory operations (program/erase) has/have
completed successfully. It is set only if the end of operation interrupts are enabled (EOPIE =
1).
Cleared by writing a 1.

3.8.5 Flash control register (FLASH_CR)

The Flash control register is used to configure and start Flash memory operations.
Address offset: 0x10
Reset value: 0x8000 0000
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LOCK Res. Res. Res. Res. Res. ERRIE EOPIE Res. Res. Res. Res. Res. Res. Res. STRT

rs rw rw rs

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. PSIZE[1:0] Res. SNB[3:0] MER SER PG

rw rw rw rw rw rw rw rw rw

Bit 31 LOCK: Lock

Write to 1 only. When it is set, this bit indicates that the FLASH_CR register is locked. It is
cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.
Bits 30:26 Reserved, must be kept cleared.
Bit 25 ERRIE: Error interrupt enable
This bit enables the interrupt generation when the OPERR bit in the FLASH_SR register is
set to 1.
0: Error interrupt generation disabled
1: Error interrupt generation enabled
Bit 24 EOPIE: End of operation interrupt enable
This bit enables the interrupt generation when the EOP bit in the FLASH_SR register goes
to 1.
0: Interrupt generation disabled
1: Interrupt generation enabled
Bits 23:17 Reserved, must be kept cleared.
Bit 16 STRT: Start
This bit triggers an erase operation when set. It is set only by software and cleared when the
BSY bit is cleared.
Bits 15:10 Reserved, must be kept cleared.

RM0402 Rev 6 77/1163

81
Embedded Flash memory interface RM0402

Bits 9:8 PSIZE: Program size

These bits select the program parallelism.
00 program x8
01 program x16
10 program x32
11 program x64
Bit 7 Reserved, must be kept cleared.
Bits 6:3 SNB: Sector number
These bits select the sector to erase.
0000 sector 0
0001 sector 1
...
1010 sector 10
1011 sector 11
1100 user specific sector
1101 user configuration sector
1110 not allowed
1111 not allowed
Bit 2 MER: Mass Erase
Erase activated for all user sectors.
Bit 1 SER: Sector Erase
Sector Erase activated.
Bit 0 PG: Programming
Flash programming activated.

3.8.6 Flash option control register (FLASH_OPTCR)

The FLASH_OPTCR register is used to modify the user option bytes.
Address offset: 0x14
Reset value: 0x0FFF FFED. The option bits are loaded with values from Flash memory at
reset release.
Access: no wait state when no Flash memory operation is ongoing, word, half-word and
byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

SPR
Res. Res. Res. nWRP[11:0]
MOD

rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
nRST_ nRST_ WDG_ OPT OPT
RDP[7:0] Res. BOR_LEV
STDBY STOP SW STRT LOCK
rw rw rw rw rw rw rw rw rw rw rw rw rw rs rs

78/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

Bit 31 SPRMOD: Selection of Protection Mode of nWPRi bits

0: PCROP disabled, nWPRi bits used for Write Protection on sector i
1: PCROP enabled, nWPRi bits used for PCROP Protection on sector i
Bits 30:28 Reserved, must be kept cleared.
Bits 27:16 nWRP[11:0]: Not write protect
These bits contain the value of the write-protection option bytes of sectors after reset. They
can be written to program a new write protect value into Flash memory.
0: Write protection active on selected sector
1: Write protection not active on selected sector
These bits contain the value of the write-protection and read-protection (PCROP) option
bytes for sectors 0 to 11 after reset. They can be written to program a new write-protect or
PCROP value into Flash memory.
If SPRMOD is reset:
0: Write protection active on sector i
1: Write protection not active on sector i
If SPRMOD is set:
0: PCROP protection not active on sector i
1: PCROP protection active on sector i
Bits 15:8 RDP: Read protect
These bits contain the value of the read-protection option level after reset. They can be
written to program a new read protection value into Flash memory.
0xAA: Level 0, read protection not active
0xCC: Level 2, chip read protection active
Others: Level 1, read protection of memories active
Bits 7:5 USER: User option bytes
These bits contain the value of the user option byte after reset. They can be written to
program a new user option byte value into Flash memory.
Bit 7: nRST_STDBY
Bit 6: nRST_STOP
Bit 5: WDG_SW
Note: When changing the WDG mode from hardware to software or from software to
hardware, a system reset is required to make the change effective.
Bit 4 Reserved, must be kept cleared. Always read as “0”.

RM0402 Rev 6 79/1163

81
Embedded Flash memory interface RM0402

Bits 3:2 BOR_LEV: BOR reset Level

These bits contain the supply level threshold that activates/releases the reset. They can be
written to program a new BOR level. By default, BOR is off. When the supply voltage (VDD)
drops below the selected BOR level, a device reset is generated.
00: BOR Level 3 (VBOR3), brownout threshold level 3
01: BOR Level 2 (VBOR2), brownout threshold level 2
10: BOR Level 1 (VBOR1), brownout threshold level 1
11: BOR off, POR/PDR reset threshold level is applied
Note: For full details about BOR characteristics, refer to the “Electrical characteristics” section
in the device datasheet.
Bit 1 OPTSTRT: Option start
This bit triggers a user option operation when set. It is set only by software and cleared when
the BSY bit is cleared.
Bit 0 OPTLOCK: Option lock
Write to 1 only. When this bit is set, it indicates that the FLASH_OPTCR register is locked.
This bit is cleared by hardware after detecting the unlock sequence.
In the event of an unsuccessful unlock operation, this bit remains set until the next reset.

80/1163 RM0402 Rev 6

RM0402 Embedded Flash memory interface

3.8.7 Flash interface register map

Table 13. Flash register map and reset values

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
PRFTEN
DCRST
ICRST
DCEN
ICEN
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
FLASH_ACR LATENCY
0x00

Reset value 0 0 0 0 0 0 0 0 0

FLASH_KEYR KEY[31:16] KEY[15:0]

0x04

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

FLASH_
OPTKEYR[31:16] OPTKEYR[15:0]
0x08 OPTKEYR

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

WRPERR
PGSERR
PGPERR
PGAERR
RDERR

OPERR
EOP
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
BSY

FLASH_SR
0x0C

Reset value 0 0 0 0 0 0 0 0

PSIZE[1:0]
ERRIE
EOPIE
LOCK

STRT

MER
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.

Res.

SER
PG
FLASH_CR SNB[3:0]
0x10

Reset value 1 0 0 0 0 0 0 0 0 0 0 0 0

nRST_STOP
nRST_STDB

OPTLOCK
OPTSTRT
WDG_SW

BOR_LEV
SPRMOD
Res.
Res.
Res.

Res.
FLASH_OPTCR nWRP[11:0] RDP[7:0]
0x14

Reset value 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 0 1 0 1 1 1 1 1 0 1

Refer to Section 2.2 on page 49 for the register boundary addresses.

RM0402 Rev 6 81/1163

81
CRC calculation unit RM0402

4 CRC calculation unit

4.1 CRC introduction

The CRC (cyclic redundancy check) calculation unit is used to get a CRC code from a 32-bit
data word and a fixed generator polynomial.
Among other applications, CRC-based techniques are used to verify data transmission or
storage integrity. In the scope of the EN/IEC 60335-1 standard, they offer a way of verifying
the Flash memory integrity. The CRC calculation unit helps compute a signature of the
software during runtime, to be compared with a reference signature generated at link-time
and stored at a given memory location.

4.2 CRC main features

• Uses CRC-32 (Ethernet) polynomial: 0x4C11DB7
– X32 + X26 + X23 + X22 + X16 + X12 + X11 + X10 +X8 + X7 + X5 + X4 + X2+ X +1
• Single input/output 32-bit data register
• CRC computation done in four AHB clock cycles (HCLK)
• General-purpose 8-bit register (can be used for temporary storage)
The block diagram is shown in Figure 7.

Figure 7. CRC calculation unit block diagram

AHB bus

32-bit (read access)

Data register (output)

CRC computation (polynomial: 0x4C11DB7)

32-bit (write access)

Data register (input)

ai14968

4.3 CRC functional description

The CRC calculation unit mainly consists of a single 32-bit data register, which:
• is used as an input register to enter new data in the CRC calculator (when writing into
the register)
• holds the result of the previous CRC calculation (when reading the register)

82/1163 RM0402 Rev 6

RM0402 CRC calculation unit

Each write operation into the data register creates a combination of the previous CRC value
and the new one (CRC computation is done on the whole 32-bit data word, and not byte per
byte).
The write operation is stalled until the end of the CRC computation, thus allowing back-to-
back write accesses or consecutive write and read accesses.
The CRC calculator can be reset to 0xFFFF FFFF with the RESET control bit in the
CRC_CR register. This operation does not affect the contents of the CRC_IDR register.

4.4 CRC registers

The CRC calculation unit contains two data registers and a control register.The peripheral
The CRC registers have to be accessed by words (32 bits).

4.4.1 Data register (CRC_DR)

Address offset: 0x00
Reset value: 0xFFFF FFFF

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DR [31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DR [15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 Data register bits

Used as an input register when writing new data into the CRC calculator.
Holds the previous CRC calculation result when it is read.

RM0402 Rev 6 83/1163

85
CRC calculation unit RM0402

4.4.2 Independent data register (CRC_IDR)

Address offset: 0x04
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. IDR[7:0]
rw rw rw rw rw rw rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bits 7:0 General-purpose 8-bit data register bits
Can be used as a temporary storage location for one byte.
This register is not affected by CRC resets generated by the RESET bit in the CRC_CR
register.

4.4.3 Control register (CRC_CR)

Address offset: 0x08
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. RESET

Bits 31:1 Reserved, must be kept at reset value.

Bit 0 RESET bit
Resets the CRC calculation unit and sets the data register to 0xFFFF FFFF.
This bit can only be set, it is automatically cleared by hardware.

84/1163 RM0402 Rev 6

RM0402 CRC calculation unit

4.4.4 CRC register map

Table 14. CRC calculation unit register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
CRC_DR Data register
0x00

Reset value 0xFFFF FFFF

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
CRC_IDR Independent data register
0x04

Reset value 0x0000

RESET
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
CRC_CR
0x08

Reset value 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

RM0402 Rev 6 85/1163

85
Power controller (PWR) RM0402

5 Power controller (PWR)

5.1 Power supplies

There are two main power supply schemes:
• VDD = 1.7 to 3.6 V: external power supply for I/Os with the internal regulator disabled,
provided externally through VDD pins. Requires the use of an external power supply
supervisor connected to the VDD and PDR_ON pins.
• VDD = 1.8 to 3.6 V: external power supply for I/Os and the internal regulator (when
enabled), provided externally through VDD pins.
• VDD_USB = 3.0 to 3.6 V
VDD_USB is a dedicated independent USB power supply for full speed transceivers.
Note: VDD_USB value does not dependent on VDD and VDDA. However, VDD_USB must be the last
supply to be delivered to the device and the first to be switched off. When the three power
supplies are shut down, if VDD_USB remains active for a short period of time and VDDA/VDDIO
fall below the functional range, the device is not be damaged.
The device is still functional when VDD_USB is switched off.
The real-time clock (RTC), and the RTC backup registers can be powered from the VBAT
voltage when the main VDD supply is powered off.
Note: Depending on the operating power supply range, some peripheral may be used with limited
functionality and performance. For more details refer to section "General operating
conditions" in the datasheet.

86/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Figure 8. Power supply overview

VBAT

Backup circuitry
VBAT = Power (OSC32K,RTC,
1.65 to 3.6V switch Wakeup logic
Backup registers)

OUT

Level shifter
IO
GPIOs
IN Logic

VCAP_1 Kernel logic

2 × 2.2 μF VCAP_2 (CPU, digital
& RAM)
VDD VDD
1/2/...11/12 Voltage
11 × 100 nF VSS regulator
+ 1 × 4.7 μF 1/2/...11/12

VDD_USB BYPASS_REG Flash memory

OTG
VDD_USB
FS
100 nF PHY
+ 1 μF Reset
PDR_ON controller
VDD
VDDA
VREF
VREF+
Analog:
100 nF 100 nF VREF- ADC RCs,
+ 1 μF + 1 μF PLL,..
VSSA

MSv39022V1

1. VDDA and VSSA must be connected to VDD and VSS, respectively.

5.1.1 Independent A/D converter supply and reference voltage

To improve conversion accuracy, the ADC has an independent power supply which can be
separately filtered and shielded from noise on the PCB.
• The ADC voltage supply input is available on a separate VDDA pin.
• An isolated supply ground connection is provided on pin VSSA.
To ensure a better accuracy of low voltage inputs, the user can connect a separate external
reference voltage ADC input on VREF. The voltage on VREF ranges from 1.7 V to VDDA.

5.1.2 Battery backup domain

Backup domain description
To retain the content of the RTC backup registers and supply the RTC when VDD is turned
off, VBAT pin can be connected to an optional standby voltage supplied by a battery or by
another source.

RM0402 Rev 6 87/1163

108
Power controller (PWR) RM0402

To allow the RTC to operate even when the main digital supply (VDD) is turned off, the VBAT
pin powers the following blocks:
• The RTC
• The LSE oscillator
• PC13 to PC15 I/Os
The switch to the VBAT supply is controlled by the power-down reset embedded in the Reset
block.

Warning: During tRSTTEMPO (temporization at VDD startup) or after a PDR

is detected, the power switch between VBAT and VDD remains
connected to VBAT.
During the startup phase, if VDD is established in less than
tRSTTEMPO (Refer to the datasheet for the value of tRSTTEMPO)
and VDD > VBAT + 0.6 V, a current may be injected into VBAT
through an internal diode connected between VDD and the
power switch (VBAT).
If the power supply/battery connected to the VBAT pin cannot
support this current injection, it is strongly recommended to
connect an external low-drop diode between this power
supply and the VBAT pin.

If no external battery is used in the application, it is recommended to connect the VBAT pin to
VDD with a 100 nF external decoupling ceramic capacitor in parallel.
When the backup domain is supplied by VDD (analog switch connected to VDD), the
following functions are available:
• PC14 and PC15 can be used as either GPIO or LSE pins
• PC13 can be used as a GPIO or additional functions can be configured (refer to
Table 27: RTC additional functions for more details about this pin configuration)
Note: Due to the fact that the switch only sinks a limited amount of current (3 mA), the use of
PC13 to PC15 GPIOs in output mode is restricted: the speed has to be limited to 2 MHz with
a maximum load of 30 pF and these I/Os must not be used as a current source (e.g. to drive
an LED).
When the backup domain is supplied by VBAT (analog switch connected to VBAT because
VDD is not present), the following functions are available:
• PC14 and PC15 can be used as LSE pins only
• PC13 can be used as the RTC additional function pin (refer to Table 27: RTC additional
functions for more details about this pin configuration)

88/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Backup domain access

After reset, the backup domain (RTC registers, and RTC backup register) is protected
against possible unwanted write accesses. To enable access to the backup domain,
proceed as follows:
• Access to the RTC and RTC backup registers
1. Enable the power interface clock by setting the PWREN bits in the RCC_APB1ENR
register (see Section 6.3.12: RCC AHB3 peripheral clock enable register
(RCC_AHB3ENR))
2. Set the DBP bit in the Section 5.4.1 to enable access to the backup domain
3. Select the RTC clock source: see Section 6.2.8: RTC/AWU clock
4. Enable the RTC clock by programming the RTCEN [15] bit in the Section 6.3.20: RCC
Backup domain control register (RCC_BDCR)

RTC and RTC backup registers

The real-time clock (RTC) is an independent BCD timer/counter. The RTC provides a time-
of-day clock/calendar, two programmable alarm interrupts, and a periodic programmable
wakeup flag with interrupt capability. The RTC contains 20 backup data registers (80 bytes)
which are reset when a tamper detection event occurs. For more details refer to Section 22:
Real-time clock (RTC).

5.1.3 Voltage regulator

An embedded linear voltage regulator supplies all the digital circuitries except for the backup
domain and the Standby circuitry. The regulator output voltage is around 1.2 V.
This voltage regulator requires one or two external capacitors to be connected to one or two
dedicated pins, VCAP_1 and for some packages VCAP_2. Specific pins must be connected
either to VSS or VDD to activate or deactivate the voltage regulator. These pins depend on
the package.
When activated by software, the voltage regulator is always enabled after Reset. It works in
three different modes depending on the application modes.
• In Run mode, the regulator supplies full power to the 1.2 V domain (core, memories
and digital peripherals). In this mode, the regulator output voltage (around 1.2 V) can
be scaled by software to different voltage values, Scale 1, scale 2, or scale 3, that can
be configured through the VOS[1:0] bits of the PWR_CR register. After reset the VOS
register is set to scale 2. When the PLL is OFF, the voltage regulator is set to scale 3
independently of the VOS register content. The VOS register content is only taken into
account once the PLL is activated and the HSI or HSE is selected as clock source.
The voltage scaling allows optimizing the power consumption when the device is
clocked below the maximum system frequency.
• In Stop mode, the main regulator or the low-power regulator supplies low power to the
1.2 V domain, thus preserving the content of registers and internal SRAM. The voltage
regulator can be put either in main regulator mode (MR) or in low-power mode (LPR).
The programmed voltage scale remains the same during Stop mode:
Voltage scale 3 is automatically selected when the microcontroller enters Stop mode
(see Section 5.4.1: PWR power control register (PWR_CR)).
• In Standby mode, the regulator is powered down. The content of the registers and
SRAM are lost except for the Standby circuitry and the backup domain.

RM0402 Rev 6 89/1163

108
Power controller (PWR) RM0402

Note: For more details, refer to the voltage regulator section in the STM32F412xx datasheet.

5.2 Power supply supervisor

5.2.1 Power-on reset (POR)/power-down reset (PDR)

The device has an integrated POR/PDR circuitry that allows proper operation starting
from 1.8 V.
To use the device below 1.8 V, the internal power supervisor must be switched off using the
PDR_ON pin (please refer to section Power supply supervisor of the STM32F412xx
datasheet). The device remains in Reset mode when VDD/VDDA is below a specified
threshold, VPOR/PDR, without the need for an external reset circuit. For more details
concerning the power on/power-down reset threshold, refer to the electrical characteristics
of the datasheet.

Figure 9. Power-on reset/power-down reset waveform

VDD/VDDA

PDR

40 mV
hysteresis PDR

Temporization
tRSTTEMPO

Reset
MS30431V1

5.2.2 Brownout reset (BOR)

During power on, the Brownout reset (BOR) keeps the device under reset until the supply
voltage reaches the specified VBOR threshold.
VBOR is configured through device option bytes. By default, BOR is off. 3 programmable
VBOR threshold levels can be selected:
• BOR Level 3 (VBOR3). Brownout threshold level 3.
• BOR Level 2 (VBOR2). Brownout threshold level 2.
• BOR Level 1 (VBOR1). Brownout threshold level 1.
Note: For full details about BOR characteristics, refer to the "Electrical characteristics" section in
the device datasheet.

90/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

When the supply voltage (VDD) drops below the selected VBOR threshold, a device reset is
generated.
The BOR can be disabled by programming the device option bytes. In this case, the
power-on and power-down is then monitored by the POR/ PDR or by an external power
supervisor if the PDR is switched off through the PDR_ON pin (see Section 5.2.1: Power-on
reset (POR)/power-down reset (PDR)).
The BOR threshold hysteresis is ~100 mV (between the rising and the falling edge of the
supply voltage).

Figure 10. BOR thresholds

VDD/VDDA

BOR threshold 100 mV

hysteresis

Reset

MS30433V1

5.2.3 Programmable voltage detector (PVD)

You can use the PVD to monitor the VDD power supply by comparing it to a threshold
selected by the PLS[2:0] bits in the PWR power control register (PWR_CR).
The PVD is enabled by setting the PVDE bit.
A PVDO flag is available, in the PWR power control/status register (PWR_CSR), to indicate
if VDD is higher or lower than the PVD threshold. This event is internally connected to the
EXTI line16 and can generate an interrupt if enabled through the EXTI registers. The PVD
output interrupt can be generated when VDD drops below the PVD threshold and/or when
VDD rises above the PVD threshold depending on EXTI line16 rising/falling edge
configuration. As an example the service routine could perform emergency shutdown tasks.

RM0402 Rev 6 91/1163

108
Power controller (PWR) RM0402

Figure 11. PVD thresholds

VDD

PVD threshold 100 mV

hysteresis

PVD output

MS30432V2

5.3 Low-power modes

By default, the microcontroller is in Run mode after a system or a power-on reset. In Run
mode the CPU is clocked by HCLK and the program code is executed. Several low-power
modes are available to save power when the CPU does not need to be kept running, for
example when waiting for an external event. It is up to the user to select the mode that gives
the best compromise between low-power consumption, short startup time and available
wakeup sources.
The devices feature four low-power modes:
• Sleep mode (Cortex®-M4 with FPU core stopped, peripherals kept running)
• Stop mode (all clocks are stopped)
• Standby mode (1.2 V domain powered off)
• Batch acquisition mode (BAM): the devices are in Sleep mode, the Flash memory is off,
needed peripheral are kept running, data transfer are still possible through DMA.
In addition, the power consumption in Run mode can be reduce by one of the following
means:
• Slowing down the system clocks
• Gating the clocks to the APBx and AHBx peripherals when they are unused.

Entering low-power mode

Low-power modes are entered by the MCU by executing the WFI (Wait For Interrupt), or
WFE (Wait for Event) instructions, or when the SLEEPONEXIT bit in the Cortex®-M4 with
FPU System Control register is set on Return from ISR.
Entering Low-power mode through WFI or WFE will be executed only if no interrupt is
pending or no event is pending.

92/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Exiting low-power mode

The MCU exits from Sleep and Stop modes low-power mode depending on the way the low-
power mode was entered:
• If the WFI instruction or Return from ISR was used to enter the low-power mode, any
peripheral interrupt acknowledged by the NVIC can wake up the device.
• If the WFE instruction is used to enter the low-power mode, the MCU exits the low-
power mode as soon as an event occurs. The wakeup event can be generated either
by:
– NVIC IRQ interrupt:
When SEVONPEND = 0 in the Cortex®-M4 with FPU System Control register: by
enabling an interrupt in the peripheral control register and in the NVIC. When the
MCU resumes from WFE, the peripheral interrupt pending bit and the NVIC
peripheral IRQ channel pending bit (in the NVIC interrupt clear pending register)
have to be cleared. Only NVIC interrupts with sufficient priority will wakeup and
interrupt the MCU.
When SEVONPEND = 1 in the Cortex®-M4 with FPU System Control register: by
enabling an interrupt in the peripheral control register and optionally in the NVIC.
When the MCU resumes from WFE, the peripheral interrupt pending bit and when
enabled the NVIC peripheral IRQ channel pending bit (in the NVIC interrupt clear
pending register) have to be cleared. All NVIC interrupts will wakeup the MCU,
even the disabled ones.Only enabled NVIC interrupts with sufficient priority will
wakeup and interrupt the MCU.
– Event
This is done by configuring a EXTI line in event mode. When the CPU resumes
from WFE, it is not necessary to clear the EXTI peripheral interrupt pending bit or
the NVIC IRQ channel pending bit as the pending bits corresponding to the event
line is not set. It may be necessary to clear the interrupt flag in the peripheral.
The MCU exits from Standby low-power mode through an external reset (NRST pin), an
IWDG reset, a rising edge on one of the enabled WKUPx pins or a RTC event occurs (see
Figure 207: RTC block diagram).
After waking up from Standby mode, program execution restarts in the same way as after a
Reset (boot pin sampling, option bytes loading, reset vector is fetched, etc.).

RM0402 Rev 6 93/1163

108
Power controller (PWR) RM0402

Only enabled NVIC interrupts with sufficient priority will wakeup and interrupt the MCU.

Table 15. Low-power mode summary

Effect on
Effect on 1.2 V VDD
Mode name Entry Wakeup Voltage regulator
domain clocks domain
clocks

Sleep and WFI or Return

Any interrupt CPU CLK OFF
BAM(1) from ISR
no effect on other
(Sleep now or None ON
clocks or analog
Sleep-on- WFE Wakeup event clock sources
exit)

Main regulator or
Low-Power
SLEEPDEEP bit Any EXTI line (configured
regulator (depends
Stop + WFI, Return in the EXTI registers,
on PWR power
from ISR or WFE internal and external lines)
control register
HSI and (PWR_CR)
All 1.2 V domain HSE
WKUP pin rising edge, clocks OFF oscillator
RTC alarm (Alarm A or s OFF
PDDS bit +
Alarm B), RTC Wakeup
SLEEPDEEP bit
Standby event, RTC tamper OFF
+ WFI, Return
events, RTC time stamp
from ISR or WFE
event, external reset in
NRST pin, IWDG reset
1. Refer to Section 5.3.4: Batch acquisition mode for specific BAM entry and exit requirements.

5.3.1 Slowing down system clocks

In Run mode the speed of the system clocks (SYSCLK, HCLK, PCLK1, PCLK2) can be
reduced by programming the prescaler registers. These prescalers can also be used to slow
down peripherals before entering Sleep mode.
For more details refer to Section 6.3.3: RCC clock configuration register (RCC_CFGR).

5.3.2 Peripheral clock gating

In Run mode, the HCLKx and PCLKx for individual peripherals and memories can be
stopped at any time to reduce power consumption.
To further reduce power consumption in Sleep mode the peripheral clocks can be disabled
prior to executing the WFI or WFE instructions.
Peripheral clock gating is controlled by the AHB1 peripheral clock enable register
(RCC_AHB1ENR), AHB2 peripheral clock enable register (RCC_AHB2ENR) (see
Section 6.3.10: RCC AHB1 peripheral clock enable register (RCC_AHB1ENR),
Section 6.3.11: RCC AHB2 peripheral clock enable register (RCC_AHB2ENR) and
Section 6.3.12: RCC AHB3 peripheral clock enable register (RCC_AHB3ENR)).
Disabling the peripherals clocks in Sleep mode can be performed automatically by resetting
the corresponding bit in RCC_AHBxLPENR and RCC_APBxLPENR registers.

94/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

5.3.3 Sleep mode

Entering Sleep mode
The Sleep mode is entered according to Entering low-power mode, when the SLEEPDEEP
bit in the Cortex®-M4 with FPU System Control register is cleared.
Refer to Table 16 and Table 17 for details on how to enter Sleep mode.
Note: All interrupt pending bits must be cleared before the sleep mode entry.

Exiting Sleep mode

The Sleep mode is exited according to Exiting low-power mode.
Refer to Table 16 and Table 17 for more details on how to exit Sleep mode.

Table 16. Sleep-now entry and exit

Sleep-now mode Description

WFI (Wait for Interrupt) or WFE (Wait for Event) while:

– SLEEPDEEP = 0, and
– No interrupt (for WFI) or event (for WFE) is pending.
Refer to the Cortex®-M4 with FPU System Control register.
Mode entry On Return from ISR while:
– SLEEPDEEP = 0 and
– SLEEPONEXIT = 1,
– No interrupt is pending.
Refer to the Cortex®-M4 with FPU System Control register.
If WFI or Return from ISR was used for entry:
Interrupt: Refer to Table 40: Vector table for STM32F412xx
If WFE was used for entry and SEVONPEND = 0
Mode exit Wakeup event: Refer to Section 10.2.3: Wakeup event management
f WFE was used for entry and SEVONPEND = 1
Interrupt even when disabled in NVIC: refer to Table 40: Vector table for
STM32F412xx or Wakeup event (see Section 10.2.3: Wakeup event
management).
Wakeup latency None

Table 17. Sleep-on-exit entry and exit

Sleep-on-exit Description

WFI (Wait for Interrupt) or WFE (Wait for Event) while:

– SLEEPDEEP = 0, and
– No interrupt (for WFI) or event (for WFE) is pending.
Refer to the Cortex®-M4 with FPU System Control register.
Mode entry On Return from ISR while:
– SLEEPDEEP = 0, and
– SLEEPONEXIT = 1, and
– No interrupt is pending.
Refer to the Cortex®-M4 with FPU System Control register.

RM0402 Rev 6 95/1163

108
Power controller (PWR) RM0402

Table 17. Sleep-on-exit entry and exit (continued)

Sleep-on-exit Description

Mode exit Interrupt: refer to Table 40: Vector table for STM32F412xx
Wakeup latency None

5.3.4 Batch acquisition mode

Entering BAM
The BAM is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is cleared.
Refer to Table 18 and Table 19 for details on how to enter Sleep mode.
Before entering Sleep mode, the Flash memory must be configured by software to operate
in the required low- power mode. If data need to be transferred from peripheral to RAM
during BAM, the DMA must be enabled before entering Sleep mode.

Exiting BAM
The BAM is exited according to Section : Exiting low-power mode.
Refer to Table 18 and Table 19 for more details on how to exit Sleep mode.
After waking up from BAM, the Flash memory must first to be waked up if code execution
restarts from Flash memory.
This wakeup time must be managed by software running from the internal SRAM.

Table 18. BAM-now entry and exit

Sleep-now mode Description

Set the Flash memory in low-power mode:

– FISSR/FMSSR and FPDS bits of the PWR_CR register
WFI (Wait for Interrupt) or WFE (Wait for Event) while:
Mode entry
– SLEEPDEEP = 0 and
– SLEEPONEXIT = 0
Refer to the Cortex®-M4 with FPU System Control register.
If WFI was used for entry:
Interrupt: Refer to Table 40: Vector table for STM32F412xx
If WFE was used for entry
Mode exit
Wakeup event: Refer to Section 10.2.3: Wakeup event management
If Flash memory wakeup time is needed, FISSR/FMSSR bits of PWR_CR
register must be set
None if code executed from RAM
Low-power mode Flash memory wakeup time, before restarting code
Wakeup latency
execution from Flash memory (refer to the Flash memory wakeup time in
the Electrical characteristics section of the datasheet).

96/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Table 19. BAM-on-exit entry and exit

Sleep-on-exit Description

Set the Flash memory in low-power mode:

– FISSR/FMSSR and FPDS bits of the PWR_CR register
WFI (wait for interrupt) while:
Mode entry
– SLEEPDEEP = 0 and
– SLEEPONEXIT = 1
Refer to the Cortex®-M4 with FPU System Control register.
Interrupt: refer to Table 40: Vector table for STM32F412xx
Mode exit If Flash memory wakeup time is needed, FISSR/FMSSR bits of PWR_CR
register must be set
None when code executed from internal SRAM
Low-power mode Flash memory wakeup time, before restarting code
Wakeup latency
execution from Flash memory (refer to the Flash memory wakeup time in
the Electrical characteristics section of the datasheet).

5.3.5 Stop mode

The Stop mode is based on the Cortex®-M4 with FPU deepsleep mode combined with
peripheral clock gating. The voltage regulator can be configured either in normal or low-
power mode. In Stop mode, all clocks in the 1.2 V domain are stopped, the PLLs, the HSI
and the HSE RC oscillators are disabled. Internal SRAM and register contents are
preserved.
Some settings in the PWR_CR register allow to further reduce the power consumption.
When the Flash memory is in power-down mode, an additional startup delay is incurred
when waking up from Stop mode (see Table 20: Stop operating modes and Section 5.4.1:
PWR power control register (PWR_CR)).

RM0402 Rev 6 97/1163

108
Power controller (PWR) RM0402

Table 20. Stop operating modes

Stop mode MRLV bit LPLV bit FPDS bit LPDS bit Wakeup latency

STOP MR 0 - 0 0 HSI RC startup time

HSI RC startup time +

STOP MRFPD 0 - 1 0 Flash wakeup time from Deep
Power Down mode
HSI RC startup time +
STOP LP 0 0 0 1 regulator wakeup time from LP
mode
HSI RC startup time +
Normal mode

Flash wakeup time from Deep

STOP LPFPD - 0 1 1 Power Down mode +
regulator wakeup time from LP
mode
HSI RC startup time +
Flash wakeup time from Deep
STOP MRLV 1 - - 0 Power Down mode +
Main regulator from low voltage
mode
HSI RC startup time +
Flash wakeup time from Deep
STOP LPLV - 1 - 1 Power Down mode +
regulator wakeup time from Low
Voltage LP mode

Entering Stop mode

The Stop mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is set.
Refer to Table 21 for details on how to enter the Stop mode.
To further reduce power consumption in Stop mode, the internal voltage regulator can be put
in low-power mode. This is configured by the LPDS bit of the PWR power control register
(PWR_CR).
If Flash memory programming is ongoing, the Stop mode entry is delayed until the memory
access is finished.
If an access to the APB domain is ongoing, The Stop mode entry is delayed until the APB
access is finished.

98/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

In Stop mode, the following features can be selected by programming individual control bits:
• Independent watchdog (IWDG): the IWDG is started by writing to its Key register or by
hardware option. Once started it cannot be stopped except by a Reset. See
Section 21.3 in Section 21: Window watchdog (WWDG).
• Real-time clock (RTC): this is configured by the RTCEN bit in the Section 6.3.20: RCC
Backup domain control register (RCC_BDCR)
• Internal RC oscillator (LSI RC): this is configured by the LSION bit in the
Section 6.3.21: RCC clock control & status register (RCC_CSR).
• External 32.768 kHz oscillator (LSE OSC): this is configured by the LSEON bit in the
Section 6.3.20: RCC Backup domain control register (RCC_BDCR).
The ADC can also consume power during the Stop mode, unless it is disabled before
entering it. To disable it, the ADON bit in the ADC_CR2 register must be written to 0.
Note: If the application needs to disable the external clock before entering Stop mode, the HSEON
bit must first be disabled and the system clock switched to HSI.
Otherwise, if the HSEON bit is kept enabled while the external clock (external oscillator) can
be removed before entering stop mode, the clock security system (CSS) feature must be
enabled to detect any external oscillator failure and avoid a malfunction behavior when
entering stop mode.

Exiting Stop mode

The Stop mode is exited according to Section : Exiting low-power mode.
Refer to Table 21 for more details on how to exit Stop mode.
When exiting Stop mode by issuing an interrupt or a wakeup event, the HSI RC oscillator is
selected as system clock.
When the voltage regulator operates in low-power mode, an additional startup delay is
incurred when waking up from Stop mode. By keeping the internal regulator ON during Stop
mode, the consumption is higher although the startup time is reduced.

Table 21. Stop mode entry and exit

Stop mode Description

WFI (Wait for Interrupt) or WFE (Wait for Event) while:

– No interrupt (for WFI) or event (for WFE) is pending,
– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is cleared in Power Control register (PWR_CR),
– Select the voltage regulator mode by configuring LPDS bit in PWR_CR.
On Return from ISR:
– No interrupt is pending,
Mode entry – SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
– SLEEPONEXIT = 1,
– PDDS bit is cleared in Power Control register (PWR_CR).
Note: To enter Stop mode, all EXTI Line pending bits (in Section 10.3.6:
Pending register (EXTI_PR)), all peripheral interrupts pending bits,
the RTC Alarm (Alarm A and Alarm B), RTC wakeup, RTC tamper,
and RTC time stamp flags, must be reset. Otherwise, the Stop
mode entry procedure is ignored and program execution continues.

RM0402 Rev 6 99/1163

108
Power controller (PWR) RM0402

Table 21. Stop mode entry and exit

Stop mode Description

If WFI or Return from ISR was used for entry:

Any EXTI lines configured in Interrupt mode (the corresponding EXTI
Interrupt vector must be enabled in the NVIC). The interrupt source can
be external interrupts or peripherals with wakeup capability. Refer to
Table 40: Vector table for STM32F412xx.
If WFE was used for entry and SEVONPEND = 0
Any EXTI lines configured in event mode. Refer to Section 10.2.3:
Mode exit
Wakeup event management.
If WFE was used for entry and SEVONPEND = 1:
– Any EXTI lines configured in Interrupt mode (even if the corresponding
EXTI Interrupt vector is disabled in the NVIC). The interrupt source can
be an external interrupt or a peripheral with wakeup capability. Refer to
Table 40: Vector table for STM32F412xx.
– Wakeup event: refer to Section 10.2.3: Wakeup event management.
Wakeup latency See Table 20: Stop operating modes

5.3.6 Standby mode

The Standby mode allows to achieve the lowest power consumption. It is based on the
Cortex®-M4 with FPU deepsleep mode, with the voltage regulator disabled. The 1.2 V
domain is consequently powered off. The PLLs, the HSI oscillator and the HSE oscillator are
also switched off. SRAM and register contents are lost except for registers in the backup
domain (RTC registers and RTC backup register), and Standby circuitry (see Figure 8).

Entering Standby mode

The Standby mode is entered according to Section : Entering low-power mode, when the
SLEEPDEEP bit in the Cortex®-M4 with FPU System Control register is set.
Refer to Table 22 for more details on how to enter Standby mode.
In Standby mode, the following features can be selected by programming individual control
bits:
• Independent watchdog (IWDG): the IWDG is started by writing to its Key register or by
hardware option. Once started it cannot be stopped except by a reset. See
Section 20.3 in Section 20: Independent watchdog (IWDG).
• Real-time clock (RTC): this is configured by the RTCEN bit in the backup domain
control register (RCC_BDCR)
• Internal RC oscillator (LSI RC): this is configured by the LSION bit in the Control/status
register (RCC_CSR).
• External 32.768 kHz oscillator (LSE OSC): this is configured by the LSEON bit in the
backup domain control register (RCC_BDCR)

Exiting Standby mode

The Standby mode is exited according to Section : Exiting low-power mode. The SBF status
flag in PWR_CR (see Section 5.4.2: PWR power control/status register (PWR_CSR))
indicates that the MCU was in Standby mode. All registers are reset after wakeup from
Standby except for PWR_CR.

100/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Refer to Table 22 for more details on how to exit Standby mode.

Table 22. Standby mode entry and exit

Standby mode Description

WFI (Wait for Interrupt) or WFE (Wait for Event) while:

– SLEEPDEEP is set in Cortex®-M4 with FPU System Control register,
– PDDS bit is set in Power Control register (PWR_CR),
– No interrupt (for WFI) or event (for WFE) is pending,
– WUF bit is cleared in Power Control register (PWR_CR),
– the RTC flag corresponding to the chosen wakeup source (RTC Alarm A,
RTC Alarm B, RTC wakeup, Tamper or Timestamp flags) is cleared

Mode entry On return from ISR while:

– SLEEPDEEP bit is set in Cortex®-M4 with FPU System Control register,
and
– SLEEPONEXIT = 1, and
– PDDS bit is set in Power Control register (PWR_CR), and
– No interrupt is pending,
– WUF bit is cleared in Power Control/Status register (PWR_SR),
– The RTC flag corresponding to the chosen wakeup source (RTC Alarm
A, RTC Alarm B, RTC wakeup, Tamper or Timestamp flags) is cleared.
WKUP pin rising edge, RTC alarm (Alarm A and Alarm B), RTC wakeup,
Mode exit
tamper event, time stamp event, external reset in NRST pin, IWDG reset.
Wakeup latency Reset phase.

I/O states in Standby mode

In Standby mode, all I/O pins are high impedance except for:
• Reset pad (still available)
• RTC_AF1 pin (PC13) if configured for tamper, time stamp, RTC Alarm out, or RTC
clock calibration out
• WKUP pin (PA0/PC0/PC1), if enabled

Debug mode
By default, the debug connection is lost if the application puts the MCU in Stop or Standby
mode while the debug features are used. This is due to the fact that the Cortex®-M4 with
FPU core is no longer clocked.
However, by setting some configuration bits in the DBGMCU_CR register, the software can
be debugged even when using the low-power modes extensively. For more details, refer to
Section 30.16.1: Debug support for low-power modes.

5.3.7 Programming the RTC alternate functions to wake up the device from
the Stop and Standby modes
The MCU can be woken up from a low-power mode by an RTC alternate function.
The RTC alternate functions are the RTC alarms (Alarm A and Alarm B), RTC wakeup, RTC
tamper event detection and RTC time stamp event detection.

RM0402 Rev 6 101/1163

108
Power controller (PWR) RM0402

These RTC alternate functions can wake up the system from the Stop and Standby low-
power modes.
The system can also wake up from low-power modes without depending on an external
interrupt (Auto-wakeup mode), by using the RTC alarm or the RTC wakeup events.
The RTC provides a programmable time base for waking up from the Stop or Standby mode
at regular intervals.
For this purpose, two of the three alternate RTC clock sources can be selected by
programming the RTCSEL[1:0] bits in the Section 6.3.20: RCC Backup domain control
register (RCC_BDCR):
• Low-power 32.768 kHz external crystal oscillator (LSE OSC)
This clock source provides a precise time base with a very low-power consumption
(additional consumption of less than 1 µA under typical conditions)
• Low-power internal RC oscillator (LSI RC)
This clock source has the advantage of saving the cost of the 32.768 kHz crystal. This
internal RC oscillator is designed to use minimum power.

RTC alternate functions to wake up the device from the Stop mode
• To wake up the device from the Stop mode with an RTC alarm event, it is necessary to:
a) Configure the EXTI Line 17 to be sensitive to rising edges (Interrupt or Event
modes)
b) Enable the RTC Alarm Interrupt in the RTC_CR register
c) Configure the RTC to generate the RTC alarm
• To wake up the device from the Stop mode with an RTC tamper or time stamp event, it
is necessary to:
a) Configure the EXTI Line 21 to be sensitive to rising edges (Interrupt or Event
modes)
b) Enable the RTC time stamp Interrupt in the RTC_CR register or the RTC tamper
interrupt in the RTC_TAFCR register
c) Configure the RTC to detect the tamper or time stamp event
• To wake up the device from the Stop mode with an RTC wakeup event, it is necessary
to:
a) Configure the EXTI Line 22 to be sensitive to rising edges (Interrupt or Event
modes)
b) Enable the RTC wakeup interrupt in the RTC_CR register
c) Configure the RTC to generate the RTC Wakeup event

RTC alternate functions to wake up the device from the Standby mode
• To wake up the device from the Standby mode with an RTC alarm event, it is necessary
to:
a) Enable the RTC alarm interrupt in the RTC_CR register
b) Configure the RTC to generate the RTC alarm
• To wake up the device from the Standby mode with an RTC tamper or time stamp
event, it is necessary to:
a) Enable the RTC time stamp interrupt in the RTC_CR register or the RTC tamper
interrupt in the RTC_TAFCR register

102/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

b) Configure the RTC to detect the tamper or time stamp event

• To wake up the device from the Standby mode with an RTC wakeup event, it is
necessary to:
a) Enable the RTC wakeup interrupt in the RTC_CR register
b) Configure the RTC to generate the RTC wakeup event

Safe RTC alternate function wakeup flag clearing sequence

If the selected RTC alternate function is set before the PWR wakeup flag (WUTF) is cleared,
it will not be detected on the next event as detection is made once on the rising edge.
To avoid bouncing on the pins onto which the RTC alternate functions are mapped, and exit
correctly from the Stop and Standby modes, it is recommended to follow the sequence
below before entering the Standby mode:
• When using RTC alarm to wake up the device from the low-power modes:
a) Disable the RTC alarm interrupt (ALRAIE or ALRBIE bits in the RTC_CR register)
b) Clear the RTC alarm (ALRAF/ALRBF) flag
c) Clear the PWR Wakeup (WUF) flag
d) Enable the RTC alarm interrupt
e) Re-enter the low-power mode
• When using RTC wakeup to wake up the device from the low-power modes:
a) Disable the RTC Wakeup interrupt (WUTIE bit in the RTC_CR register)
b) Clear the RTC Wakeup (WUTF) flag
c) Clear the PWR Wakeup (WUF) flag
d) Enable the RTC Wakeup interrupt
e) Re-enter the low-power mode
• When using RTC tamper to wake up the device from the low-power modes:
a) Disable the RTC tamper interrupt (TAMPIE bit in the RTC_TAFCR register)
b) Clear the Tamper (TAMP1F/TSF) flag
c) Clear the PWR Wakeup (WUF) flag
d) Enable the RTC tamper interrupt
e) Re-enter the low-power mode
• When using RTC time stamp to wake up the device from the low-power modes:
a) Disable the RTC time stamp interrupt (TSIE bit in RTC_CR)
b) Clear the RTC time stamp (TSF) flag
c) Clear the PWR Wakeup (WUF) flag
d) Enable the RTC TimeStamp interrupt
e) Re-enter the low-power mode

RM0402 Rev 6 103/1163

108
Power controller (PWR) RM0402

5.4 Power control registers

5.4.1 PWR power control register (PWR_CR)

Address offset: 0x00
Reset value: 0x0000 8000 (reset by wakeup from Standby mode)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. FISSR FMSSR Res. Res. Res. Res.
rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MRLV LPLV
VOS ADCDC1 Res. FPDS DBP PLS[2:0] PVDE CSBF CWUF PDDS LPDS
DS DS
rw rw rw rw rw rw rw rw rw rw rw w w rw rw

Bits 31:22 Reserved, must be kept at reset value.

Bit 21 FISSR: Flash Interface Stop while System Run
0: Flash Interface clock run (Default value).
1: Flash Interface clock off.
Note: This bit could not be set while executing with the Flash itself. It should be done with
specific routine executed from RAM.
Bit 20 FMSSR: Flash Memory Sleep System Run.
0: Flash standard mode (Default value)
1: Flash forced to be in STOP or DeepPower Down mode (depending of FPDS value bit) by
hardware.
Note: This bit could not be set while executing with the Flash itself. It should be done with
specific routine executed from RAM.
Bits 19:16 Reserved, must be kept at reset value.
Bits 15:14 VOS[1:0]: Regulator voltage scaling output selection
These bits control the main internal voltage regulator output voltage to achieve a trade-off
between performance and power consumption when the device does not operate at the
maximum frequency (refer to the corresponding datasheet for more details).
These bits can be modified only when the PLL is OFF. The new value programmed is active
only when the PLL is ON. When the PLL is OFF, the voltage regulator is set to scale 3
independently of the VOS register content.
00: Reserved (Scale 3 mode selected)
01: Scale 3 mode <= 64 MHz
10: Scale 2 mode (reset value) <= 84 MHz
11: Scale 1 mode <= 100 MHz
Bit 13 ADCDC1:
0: No effect.
1: Refer to AN4073 for details on how to use this bit.
Note: This bit can only be set when operating at supply voltage range 2.7 to 3.6V and when
the Prefetch is OFF.

104/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Bit 12 Reserved, must be kept at reset value.

Bit 11 MRLVDS: Main regulator Low Voltage in Deep Sleep
0: Main regulator in Voltage scale 3 when the device is in Stop mode.
1: Main regulator in Low Voltage and Flash memory in Deep Sleep mode when the device is
in Stop mode.
Bit 10 LPLVDS: Low-power regulator Low Voltage in Deep Sleep
0: Low-power regulator on if LPDS bit is set when the device is in Stop mode.
1: Low-power regulator in Low Voltage and Flash memory in Deep Sleep mode if LPDS bit is
set when device is in Stop mode.
Bit 9 FPDS: Flash power-down in Stop mode
When set, the Flash memory enters power-down mode when the device enters Stop mode.
This allows to achieve a lower consumption in stop mode but a longer restart time.
0: Flash memory not in power-down when the device is in Stop mode
1: Flash memory in power-down when the device is in Stop mode
Bit 8 DBP: Disable backup domain write protection
In reset state, the RCC_BDCR register, the RTC registers (including the backup registers),
and the BRE bit of the PWR_CSR register, are protected against parasitic write access. This
bit must be set to enable write access to these registers.
0: Access to RTC and RTC Backup registers.
1: Access to RTC and RTC Backup registers.
Bits 7:5 PLS[2:0]: PVD level selection
These bits are written by software to select the voltage threshold detected by the Power
Voltage Detector
000: 2.2 V
001: 2.3 V
010: 2.4 V
011: 2.5 V
100: 2.6 V
101: 2.7 V
110: 2.8 V
111: 2.9 V
Note: Refer to the electrical characteristics of the datasheet for more details.
Bit 4 PVDE: Power voltage detector enable
This bit is set and cleared by software.
0: PVD disabled
1: PVD enabled
Bit 3 CSBF: Clear standby flag
This bit is always read as 0.
0: No effect.
1: Clear the SBF Standby Flag (write).
Bit 2 CWUF: Clear wakeup flag
This bit is always read as 0.
0: No effect.
1: Clear the WUF Wakeup Flag after 2 System clock cycles.

RM0402 Rev 6 105/1163

108
Power controller (PWR) RM0402

Bit 1 PDDS: Power-down deepsleep

This bit is set and cleared by software. It works together with the LPDS bit.
0: Enter Stop mode when the CPU enters deepsleep. The regulator status depends on the
LPDS bit.
1: Enter Standby mode when the CPU enters deepsleep.
Bit 0 LPDS: Low-power deepsleep
This bit is set and cleared by software. It works together with the PDDS bit.
0: Voltage regulator on during Stop mode.
1: Low-power Voltage regulator on during Stop mode.

5.4.2 PWR power control/status register (PWR_CSR)

Address offset: 0x04
Reset value: 0x0000 0000 (not reset by wakeup from Standby mode)
Additional APB cycles are needed to read this register versus a standard APB read.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
VOS EWUP EWUP EWUP
Res. Res. Res. Res. Res. BRE Res. Res. BRR PVDO SBF WUF
RDY 1 2 3
r rw rw rw rw r r r r

Bits 31:15 Reserved, must be kept at reset value.

Bit 14 VOSRDY: Regulator voltage scaling output selection ready bit
0: Not ready
1: Ready
Bits 13:10 Reserved, must be kept at reset value.
Bit 9 BRE: Backup regulator enable
When set, the Backup regulator (used to maintain the backup domain content) is enabled. If
BRE is reset, the backup regulator is switched off. Once set, the application must wait that
the Backup Regulator Ready flag (BRR) is set to indicate that the data written into the
backup registers will be maintained in the Standby and VBAT modes.
0: Backup regulator disabled
1: Backup regulator enabled
Note: This bit is not reset when the device wakes up from Standby mode, by a system reset,
or by a power reset.
Bit 8 EWUP1: Enable WKUP1 pin (PA0)
This bit is set and cleared by software.
0: WKUP1 pin is used for general purpose I/O. An event on the WKUP1 pin does not
wakeup the device from Standby mode.
1: WKUP1 pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP1 pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.

106/1163 RM0402 Rev 6

RM0402 Power controller (PWR)

Bit 7 EWUP2: Enable WKUP2 pin (PC0)

This bit is set and cleared by software.
0: WKUP2 pin is used for general purpose I/O. An event on the WKUP2 pin does not
wakeup the device from Standby mode.
1: WKUP2 pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP2 pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.
Bit 6 EWUP3: Enable WKUP3 pin (PC1)
This bit is set and cleared by software.
0: WKUP3 pin is used for general purpose I/O. An event on the WKUP3 pin does not
wakeup the device from Standby mode.
1: WKUP3 pin is used for wakeup from Standby mode and forced in input pull down
configuration (rising edge on WKUP3 pin wakes-up the system from Standby mode).
Note: This bit is reset by a system reset.
Bits 5:4 Reserved, must be kept at reset value.
Bit 3 BRR: Backup regulator ready
Set by hardware to indicate that the Backup Regulator is ready.
0: Backup Regulator not ready
1: Backup Regulator ready
Note: This bit is not reset when the device wakes up from Standby mode or by a system reset
or power reset.
Bit 2 PVDO: PVD output
This bit is set and cleared by hardware. It is valid only if PVD is enabled by the PVDE bit.
0: VDD is higher than the PVD threshold selected with the PLS[2:0] bits.
1: VDD is lower than the PVD threshold selected with the PLS[2:0] bits.
Note: The PVD is stopped by Standby mode. For this reason, this bit is equal to 0 after
Standby or reset until the PVDE bit is set.
Bit 1 SBF: Standby flag
This bit is set by hardware and cleared only by a POR/PDR (power-on reset/power-down
reset) or by setting the CSBF bit in the PWR_CR register.
0: Device has not been in Standby mode
1: Device has been in Standby mode
Bit 0 WUF: Wakeup flag
This bit is set by hardware and cleared either by a system reset or by setting the CWUF bit in
the PWR_CR register.
0: No wakeup event occurred
1: A wakeup event was received from the WKUP pin or from the RTC alarm (Alarm A or
Alarm B), RTC Tamper event, RTC TimeStamp event or RTC Wakeup).
Note: An additional wakeup event is detected if the WKUP pin is enabled (by setting the
EWUP bit) when the WKUP pin level is already high.

RM0402 Rev 6 107/1163

108
Power controller (PWR) RM0402

5.5 PWR register map

The following table summarizes the PWR registers.

Table 23. PWR - register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
ADCDC1

MRLVDS
VOS[1:0]

LPLVDS
FMSSR

CWUF
FISSR

PDDS
PVDE
FPDS

CSBF

LPDS
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

Res.

DBP
PWR_CR PLS[2:0]
0x000

Reset value 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0

VOSRDY

EWUP1
EWUP2
EWUP3

PVDO

WUF
BRR
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

Res.
Res.
BRE

SBF
PWR_CSR
0x004

Reset value 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

108/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6 Reset and clock control (RCC) for STM32F412xx

6.1 Reset
There are three types of reset, defined as system Reset, power Reset and backup domain
Reset.

6.1.1 System reset

A system reset sets all registers to their reset values except the reset flags in the clock
controller CSR register and the registers in the Backup domain.
A system reset is generated when one of the following events occurs:
1. A low level on the NRST pin (external reset)
2. Window watchdog end of count condition (WWDG reset)
3. Independent watchdog end of count condition (IWDG reset)
4. A software reset (SW reset) (see Software reset)
5. Low-power management reset (see Low-power management reset)

Software reset
The reset source can be identified by checking the reset flags in the RCC clock control &
status register (RCC_CSR).
The SYSRESETREQ bit in Cortex®-M4 with FPU Application Interrupt and Reset Control
Register must be set to force a software reset on the device. Refer to the Cortex®-M4 with
FPU technical reference manual for more details.

RM0402 Rev 6 109/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Low-power management reset

There are two ways of generating a low-power management reset:
1. Reset generated when entering the Standby mode:
This type of reset is enabled by resetting the nRST_STDBY bit in the user option bytes.
In this case, whenever a Standby mode entry sequence is successfully executed, the
device is reset instead of entering the Standby mode.
2. Reset when entering the Stop mode:
This type of reset is enabled by resetting the nRST_STOP bit in the user option bytes.
In this case, whenever a Stop mode entry sequence is successfully executed, the
device is reset instead of entering the Stop mode.
For further information on the user option bytes, refer to the STM32F412xx Flash
programming manual available from your ST sales office.

6.1.2 Power reset

A power reset is generated when one of the following events occurs:
1. Power-on/power-down reset (POR/PDR reset) or brownout (BOR) reset
2. When exiting the Standby mode
A power reset sets all registers to their reset values except the Backup domain.
These sources act on the NRST pin and it is always kept low during the delay phase. The
RESET service routine vector is fixed at address 0x0000_0004 in the memory map.
The system reset signal provided to the device is output on the NRST pin. The pulse
generator guarantees a minimum reset pulse duration of 20 µs for each internal reset
source. In case of an external reset, the reset pulse is generated while the NRST pin is
asserted low.

Figure 12. Simplified diagram of the reset circuit

VDD/VDDA

RPU
External System reset
reset Filter
NRST

WWDG reset
Pulse
IWDG reset
generator
Power reset
(min 20 μs)
Software reset
Low-power management reset

ai16095c

110/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.1.3 Backup domain reset

The backup domain reset sets all RTC registers and the RCC_BDCR register to their reset
values.
A backup domain reset is generated when one of the following events occurs:
1. Software reset, triggered by setting the BDRST bit in the RCC Backup domain control
register (RCC_BDCR).
2. VDD or VBAT power on, if both supplies have previously been powered off.

6.2 Clocks
Three different clock sources can be used to drive the system clock (SYSCLK):
• HSI oscillator clock
• HSE oscillator clock
• Main PLL (PLL) clock
The devices have the two following secondary clock sources:
• 32 kHz low-speed internal RC (LSI RC) which drives the independent watchdog and,
optionally, the RTC used for Auto-wakeup from the Stop/Standby mode.
• 32.768 kHz low-speed external crystal (LSE crystal) which optionally drives the RTC
clock (RTCCLK)
Each clock source can be switched on or off independently when it is not used, to optimize
power consumption.

RM0402 Rev 6 111/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Figure 13. Clock tree

Enable IWDGCLK Not (sleep or deep sleep)

LSI RC LSI watchdog CPU clock
32 kHz
RTC / AWU Not deepsleep
FCLK Cortex
enable RTC / AWU
free-running clock
OSC32_IN clock
LSE OSC LSE Peripheral
AHB AHB peripheral
32.768 kHz clock enable
PRESC clocks
OSC32_OUT /1,2,..512
MCO2 Clock enable
SysTick
SYSCLK /8 clock
HSE_RTC Peripheral
MCO2 /1 to 5 PLLI2SCLK APB1 clock enable
HSE AHB1 peripheral
PRESC
PLLCLK /1,2,4,8,16 clocks
/2 to 31 Peripheral
LSE clock enable
If (APB1 presc = 1) x1 APB1 timer
else x2 clocks
MCO1 /1 to 5 HSI Peripheral
clock enable
APB2 APB2 peripheral
PRESC clocks
MCO1 /1,2,4,8,16 Peripheral
clock enable
SW APB2 timer
If (APB2 presc = 1) x1 clocks
HSI RC HSI HSI else x2
16 MHz HSE SYSCLK CKDFSDMSEL
PLLCLK DFSDM_ck_enable
CK_DFSDM
OSC_OUT
HSE
4-26 MHz CKSDIOSEL
HSE OSC SDIO_ck_enable
OSC_IN
SDIO

PLLSRC CK48MSEL
USB_ck_enable
PLLM HSI PLL48CK
HS_ck USB FS
/2 to 63

RNG_ck_enable
PLLCLK RNG
VCO /P I2S1RC
/Q
xN HS_ck
/R R1 I2S clock
PLL For IPs on APB1
R2 I2S_ck_enable1
I2S_CKIN

I2S2RC I2S_ck_enable2
VCO /P
HS_ck I2S clock
/Q R1 For IPs on APB2
xN CKDFSDAMSEL
PLLI2SCLK R2
/R CKDFSDM_ck_enable
PLLI2S I2S_CKIN
PLLI2SSRC CK_DFSDM_Audio
N multiplier range: 50 to 432
PLLI2SM HS_ck Q divider range: 2 to 15 I2C4SEL
/2 to 63 R divider range: 2 to 7 I2C4_ck_enable
P divider options: 2, 4, 6, 8 HSI
SYSCLK I2C4 clock
Clock for I2C FM+
I2S_CKIN ck_APB1

MSv39615V2

1. For full details about the internal and external clock source characteristics, refer to the Electrical characteristics section in
the device datasheet.

112/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

The clock controller provides a high degree of flexibility to the application in the choice of the
external crystal or the oscillator to run the core and peripherals at the highest frequency
and, guarantee the appropriate frequency for peripherals that need a specific clock like USB
OTG FS, I2S and SDIO.
Several prescalers are used to configure the AHB frequency, the high-speed APB (APB2)
and the low-speed APB (APB1) domains. The maximum frequency of the AHB domain is
100 MHz. The maximum allowed frequency of the high-speed APB2 domain is 100 MHz.
The maximum allowed frequency of the low-speed APB1 domain is 50 MHz
All peripheral clocks are derived from the system clock (SYSCLK) except for:
• The USB OTG FS clock (48 MHz) and the SDIO clock (≤ 48 MHz) which are coming
from a specific output of PLL (PLL48CLK)
• The I2S clock
To achieve high-quality audio performance, the I2S clock can be derived either from a
specific PLL (PLLI2S) or from an external clock mapped on the I2S_CKIN pin. For
more information about I2S clock frequency and precision, refer to Section 26.6.4:
Clock generator.
• I2CFMP1 clock which can also be generated from HSI, SYSCLK or APB1 clock.
The RCC feeds the external clock of the Cortex System Timer (SysTick) with the AHB clock
(HCLK) divided by 8. The SysTick can work either with this clock or with the Cortex clock
(HCLK), configurable in the SysTick control and status register.
The timer clock frequencies are automatically set by hardware. There are two cases
depending on the value of TIMPRE bit in RCC_DCKCFGR register:
• If TIMPRE bit is reset:
If the APB prescaler is configured to a division factor of 1, the timer clock frequencies
(TIMxCLK) are set to HCLK. Otherwise, the timer clock frequencies are twice the
frequency of the APB domain to which the timers are connected: TIMxCLK = 2xPCLKx.
• If TIMPRE bit is set:
If the APB prescaler is configured to a division factor of 1 or 2, the timer clock
frequencies (TIMxCLK) are set to HCLK. Otherwise, the timer clock frequencies is four
times the frequency of the APB domain to which the timers are connected: TIMxCLK =
4xPCLKx.
FCLK acts as Cortex®-M4 with FPU free-running clock. For more details, refer to the
Cortex®-M4 with FPU technical reference manual.

6.2.1 HSE clock

The high speed external clock signal (HSE) can be generated from two possible clock
sources:
• HSE external crystal/ceramic resonator
• HSE external user clock
The resonator and the load capacitors have to be placed as close as possible to the
oscillator pins in order to minimize output distortion and startup stabilization time. The
loading capacitance values must be adjusted according to the selected oscillator.

RM0402 Rev 6 113/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Figure 14. HSE/ LSE clock sources

Hardware configuration

OSC_OUT

External clock
(HI-Z)
External
source

OSC_IN OSC_OUT

Crystal/ceramic
resonators

CL1 CL2
Load
capacitors

External source (HSE bypass)

In this mode, an external clock source must be provided. You select this mode by setting the
HSEBYP and HSEON bits in the RCC clock control register (RCC_CR). The external clock
signal (square, sinus or triangle) with ~50% duty cycle has to drive the OSC_IN pin while the
OSC_OUT pin should be left HI-Z. See Figure 14.

External crystal/ceramic resonator (HSE crystal)

The HSE has the advantage of producing a very accurate rate on the main clock.
The associated hardware configuration is shown in Figure 14. Refer to the electrical
characteristics section of the datasheet for more details.
The HSERDY flag in the RCC clock control register (RCC_CR) indicates if the high-speed
external oscillator is stable or not. At startup, the clock is not released until this bit is set by
hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).
The HSE Crystal can be switched on and off using the HSEON bit in the RCC clock control
register (RCC_CR).

6.2.2 HSI clock

The HSI clock signal is generated from an internal 16 MHz RC oscillator and can be used
directly as a system clock, or used as PLL input.
The HSI RC oscillator has the advantage of providing a clock source at low cost (no external
components). It also has a faster startup time than the HSE crystal oscillator however, even
with calibration the frequency is less accurate than an external crystal oscillator or ceramic
resonator.

114/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Calibration
RC oscillator frequencies can vary from one chip to another due to manufacturing process
variations, this is why each device is factory calibrated by ST for 1% accuracy at TA= 25 °C.
After reset, the factory calibration value is loaded in the HSICAL[7:0] bits in the RCC clock
control register (RCC_CR).
If the application is subject to voltage or temperature variations this may affect the RC
oscillator speed. You can trim the HSI frequency in the application using the HSITRIM[4:0]
bits in the RCC clock control register (RCC_CR).
The HSIRDY flag in the RCC clock control register (RCC_CR) indicates if the HSI RC is
stable or not. At startup, the HSI RC output clock is not released until this bit is set by
hardware.
The HSI RC can be switched on and off using the HSION bit in the RCC clock control
register (RCC_CR).
The HSI signal can also be used as a backup source (Auxiliary clock) if the HSE crystal
oscillator fails. Refer to Section 6.2.7: Clock security system (CSS) on page 116.

6.2.3 PLL configuration

The STM32F412xx devices feature two PLLs:
• A main PLL (PLL) clocked by the HSE or HSI oscillator and featuring two different
output clocks:
– The first output is used to generate the high speed system clock (up to 100 MHz)
– The second output is used to generate the clock for the USB OTG FS (48 MHz),
RNG and the SDIO (≤ 50 MHz).
• A dedicated PLL (PLLI2S) used to generate an accurate clock to achieve high-quality
audio performance on the I2S interface.
Since the main-PLL configuration parameters cannot be changed once PLL is enabled, it is
recommended to configure PLL before enabling it (selection of the HSI or HSE oscillator as
PLL clock source, and configuration of division factors M, P, Q and multiplication factor N).
The PLLI2S uses the same input clock as the main PLL (HSI or HSE). However, the PLLI2S
has dedicated enable/disable and division factors configuration bits. Refer to Section 6.3.1:
RCC clock control register (RCC_CR), Section 6.3.2: RCC PLL configuration register
(RCC_PLLCFGR) and Section 6.3.23: RCC PLLI2S configuration register
(RCC_PLLI2SCFGR). Once the PLLI2S is enabled, the configuration parameters cannot be
changed.
The two PLLs are disabled by hardware when entering Stop and Standby modes, or when
an HSE failure occurs when HSE or PLL (clocked by HSE) are used as system clock. RCC
PLL configuration register (RCC_PLLCFGR) and RCC clock configuration register
(RCC_CFGR) can be used to configure PLL and PLLI2S, respectively.

6.2.4 LSE clock

The LSE clock is generated using a 32.768kHz low speed external crystal or ceramic
resonator. It has the advantage providing a low-power but highly accurate clock source to
the real-time clock peripheral (RTC) for clock/calendar or other timing functions.

RM0402 Rev 6 115/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

The LSE oscillator is switched on and off using the LSEON bit in RCC Backup domain
control register (RCC_BDCR).
The LSERDY flag in the RCC Backup domain control register (RCC_BDCR) indicates if the
LSE crystal is stable or not. At startup, the LSE crystal output clock signal is not released
until this bit is set by hardware. An interrupt can be generated if enabled in the RCC clock
interrupt register (RCC_CIR).

External source (LSE bypass)

In this mode, an external clock source must be provided. It must have a frequency up to
1 MHz. You select this mode by setting the LSEBYP and LSEON bits in the RCC Backup
domain control register (RCC_BDCR). The external clock signal (square, sinus or triangle)
with ~50% duty cycle has to drive the OSC32_IN pin while the OSC32_OUT pin should be
left HI-Z. See Figure 14.

6.2.5 LSI clock

The LSI RC acts as an low-power clock source that can be kept running in Stop and
Standby mode for the independent watchdog (IWDG) and Auto-wakeup unit (AWU). The
clock frequency is around 32 kHz. For more details, refer to the electrical characteristics
section of the datasheets.
The LSI RC can be switched on and off using the LSION bit in the RCC clock control &
status register (RCC_CSR).
The LSIRDY flag in the RCC clock control & status register (RCC_CSR) indicates if the low-
speed internal oscillator is stable or not. At startup, the clock is not released until this bit is
set by hardware. An interrupt can be generated if enabled in the RCC clock interrupt register
(RCC_CIR).

6.2.6 System clock (SYSCLK) selection

After a system reset, the HSI oscillator is selected as the system clock. When a clock source
is used directly or through PLL as the system clock, it is not possible to stop it.
A switch from one clock source to another occurs only if the target clock source is ready
(clock stable after startup delay or PLL locked). If a clock source that is not yet ready is
selected, the switch occurs when the clock source is ready. Status bits in the RCC clock
control register (RCC_CR) indicate which clock(s) is (are) ready and which clock is currently
used as the system clock.

6.2.7 Clock security system (CSS)

The clock security system can be activated by software. In this case, the clock detector is
enabled after the HSE oscillator startup delay, and disabled when this oscillator is stopped.
If a failure is detected on the HSE clock, this oscillator is automatically disabled, a clock
failure event is sent to the break inputs of advanced-control timer TIM1, and an interrupt is
generated to inform the software about the failure (clock security system interrupt CSSI),
allowing the MCU to perform rescue operations. The CSSI is linked to the Cortex®-M4 with
FPU NMI (non-maskable interrupt) exception vector.
Note: When the CSS is enabled, if the HSE clock happens to fail, the CSS generates an interrupt,
which causes the automatic generation of an NMI. The NMI is executed indefinitely unless
the CSS interrupt pending bit is cleared. As a consequence, the application has to clear the

116/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

CSS interrupt in the NMI ISR by setting the CSSC bit in the Clock interrupt register
(RCC_CIR).
If the HSE oscillator is used directly or indirectly as the system clock (indirectly meaning that
it is directly used as PLL input clock, and that PLL clock is the system clock) and a failure is
detected, then the system clock switches to the HSI oscillator and the HSE oscillator is
disabled.
If the HSE oscillator clock was the clock source of PLL used as the system clock when the
failure occurred, PLL is also disabled. In this case, if the PLLI2S was enabled, it is also
disabled when the HSE fails.

6.2.8 RTC/AWU clock

Once the RTCCLK clock source has been selected, the only possible way of modifying the
selection is to reset the power domain.
The RTCCLK clock source can be either the HSE 1 MHz (HSE divided by a programmable
prescaler), the LSE or the LSI clock. This is selected by programming the RTCSEL[1:0] bits
in the RCC Backup domain control register (RCC_BDCR) and the RTCPRE[4:0] bits in RCC
clock configuration register (RCC_CFGR). This selection cannot be modified without
resetting the Backup domain.
If the LSE is selected as the RTC clock, the RTC will work normally if the backup or the
system supply disappears. If the LSI is selected as the AWU clock, the AWU state is not
guaranteed if the system supply disappears. If the HSE oscillator divided by a value
between 2 and 31 is used as the RTC clock, the RTC state is not guaranteed if the backup
or the system supply disappears.
The LSE clock is in the Backup domain, whereas the HSE and LSI clocks are not. As a
consequence:
• If LSE is selected as the RTC clock:
– The RTC continues to work even if the VDD supply is switched off, provided the
VBAT supply is maintained.
– The RTC remains clocked and functional under system reset.
• If LSI is selected as the Auto-wakeup unit (AWU) clock:
– The AWU state is not guaranteed if the VDD supply is powered off. Refer to
Section 6.2.5: LSI clock for more details on LSI calibration.
• If the HSE clock is used as the RTC clock:
– The RTC state is not guaranteed if the VDD supply is powered off or if the internal
voltage regulator is powered off (removing power from the 1.2 V domain).
Note: To read the RTC calendar register when the APB1 clock frequency is less than seven times
the RTC clock frequency (fAPB1 < 7xfRTCLCK), the software must read the calendar time and
date registers twice. The data are correct if the second read access to RTC_TR gives the
same result than the first one. Otherwise a third read access must be performed.

6.2.9 Watchdog clock

If the independent watchdog (IWDG) is started by either hardware option or software
access, the LSI oscillator is forced ON and cannot be disabled. After the LSI oscillator
temporization, the clock is provided to the IWDG.

RM0402 Rev 6 117/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.2.10 Clock-out capability

Two microcontroller clock output (MCO) pins are available:
• MCO1
You can output four different clock sources onto the MCO1 pin (PA8) using the
configurable prescaler (from 1 to 5):
– HSI clock
– LSE clock
– HSE clock
– PLL clock
The desired clock source is selected using the MCO1PRE[2:0] and MCO1[1:0] bits in
the RCC clock configuration register (RCC_CFGR).
• MCO2
You can output four different clock sources onto the MCO2 pin (PC9) using the
configurable prescaler (from 1 to 5):
– HSE clock
– PLL clock
– System clock (SYSCLK)
– PLLI2S clock
The desired clock source is selected using the MCO2PRE[2:0] and MCO2 bits in the
RCC clock configuration register (RCC_CFGR).
For the different MCO pins, the corresponding GPIO port has to be programmed in alternate
function mode.
The selected clock to output onto MCO must not exceed 100 MHz (the maximum I/O
speed).

6.2.11 Internal/external clock measurement using TIM5/TIM11

It is possible to indirectly measure the frequencies of all on-board clock source generators
by means of the input capture of TIM5 channel4 and TIM11 channel1 as shown in Figure 15
and Figure 16.

Internal/external clock measurement using TIM5 channel4

TIM5 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through the TI4_RMP [1:0] bits
in the TIM5_OR register.
The primary purpose of having the LSE connected to the channel4 input capture is to be
able to precisely measure the HSI (this requires to have the HSI used as the system clock
source). The number of HSI clock counts between consecutive edges of the LSE signal
provides a measurement of the internal clock period. Taking advantage of the high precision
of LSE crystals (typically a few tens of ppm) we can determine the internal clock frequency
with the same resolution, and trim the source to compensate for manufacturing-process
and/or temperature- and voltage-related frequency deviations.
The HSI oscillator has dedicated, user-accessible calibration bits for this purpose.

118/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

The basic concept consists in providing a relative measurement (e.g. HSI/LSE ratio): the
precision is therefore tightly linked to the ratio between the two clock sources. The greater
the ratio, the better the measurement.
It is also possible to measure the LSI frequency: this is useful for applications that do not
have a crystal. The ultralow-power LSI oscillator has a large manufacturing process
deviation: by measuring it versus the HSI clock source, it is possible to determine its
frequency with the precision of the HSI. The measured value can be used to have more
accurate RTC time base timeouts (when LSI is used as the RTC clock source) and/or an
IWDG timeout with an acceptable accuracy.
Use the following procedure to measure the LSI frequency:
1. Enable the TIM5 timer and configure channel4 in Input capture mode.
2. Set the TI4_RMP bits in the TIM5_OR register to 0x01 to connect the LSI clock
internally to TIM5 channel4 input capture for calibration purposes.
3. Measure the LSI clock frequency using the TIM5 capture/compare 4 event or interrupt.
4. Use the measured LSI frequency to update the prescaler of the RTC depending on the
desired time base and/or to compute the IWDG timeout.

Figure 15. Frequency measurement with TIM5 in Input capture mode

TIM5

TI4_RMP[1:0]
GPIO
RTC_WakeUp_IT TI4
LSE
LSI

ai17741V2

RM0402 Rev 6 119/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Internal/external clock measurement using TIM11 channel1

TIM11 has an input multiplexer which allows choosing whether the input capture is triggered
by the I/O or by an internal clock. This selection is performed through TI1_RMP [1:0] bits in
the TIM11_OR register. The HSE_RTC clock (HSE divided by a programmable prescaler) is
connected to channel 1 input capture to have a rough indication of the external crystal
frequency. This requires that the HSI is the system clock source. This can be useful for
instance to ensure compliance with the IEC 60730/IEC 61335 standards which require to be
able to determine harmonic or subharmonic frequencies (–50/+100% deviations).

Figure 16. Frequency measurement with TIM11 in Input capture mode

TIM11

TI1_RMP[1:0]
GPIO
TI1

HSE_RTC(1 MHz)
ai18433

120/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3 RCC registers

Refer to Section 1.2: List of abbreviations for registers for a list of abbreviations used in
register descriptions.

6.3.1 RCC clock control register (RCC_CR)

Address offset: 0x00
Reset value: 0x0000 XX81 where X is undefined.
Access: no wait state, word, half-word and byte access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PLLI2S PLLI2S CSS HSE HSE
Res. Res. Res. Res. PLLRDY PLLON Res. Res. Res. Res. HSE ON
RDY ON ON BYP RDY
r rw r rw rw rw r rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HSI
HSICAL[7:0] HSITRIM[4:0] Res. HSION
RDY
r r r r r r r r rw rw rw rw rw r rw

Bits 31:28 Reserved, must be kept at reset value.

Bit 27 PLLI2SRDY: PLLI2S clock ready flag
Set by hardware to indicate that the PLLI2S is locked.
0: PLLI2S unlocked
1: PLLI2S locked
Bit 26 PLLI2SON: PLLI2S enable
Set and cleared by software to enable PLLI2S.
Cleared by hardware when entering Stop or Standby mode.
0: PLLI2S OFF
1: PLLI2S ON
Bit 25 PLLRDY: Main PLL (PLL) clock ready flag
Set by hardware to indicate that PLL is locked.
0: PLL unlocked
1: PLL locked
Bit 24 PLLON: Main PLL (PLL) enable
Set and cleared by software to enable PLL.
Cleared by hardware when entering Stop or Standby mode. This bit cannot be reset if PLL
clock is used as the system clock.
0: PLL OFF
1: PLL ON
Bits 23:20 Reserved, must be kept at reset value.
Bit 19 CSSON: Clock security system enable
Set and cleared by software to enable the clock security system. When CSSON is set, the
clock detector is enabled by hardware when the HSE oscillator is ready, and disabled by
hardware if an oscillator failure is detected.
0: Clock security system OFF (Clock detector OFF)
1: Clock security system ON (Clock detector ON if HSE oscillator is stable, OFF if not)

RM0402 Rev 6 121/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 18 HSEBYP: HSE clock bypass

Set and cleared by software to bypass the oscillator with an external clock. The external
clock must be enabled with the HSEON bit, to be used by the device.
The HSEBYP bit can be written only if the HSE oscillator is disabled.
0: HSE oscillator not bypassed
1: HSE oscillator bypassed with an external clock
Bit 17 HSERDY: HSE clock ready flag
Set by hardware to indicate that the HSE oscillator is stable. After the HSEON bit is cleared,
HSERDY goes low after 6 HSE oscillator clock cycles.
0: HSE oscillator not ready
1: HSE oscillator ready
Bit 16 HSEON: HSE clock enable
Set and cleared by software.
Cleared by hardware to stop the HSE oscillator when entering Stop or Standby mode. This
bit cannot be reset if the HSE oscillator is used directly or indirectly as the system clock.
0: HSE oscillator OFF
1: HSE oscillator ON
Bits 15:8 HSICAL[7:0]: Internal high-speed clock calibration
These bits are initialized automatically at startup.
Bits 7:3 HSITRIM[4:0]: Internal high-speed clock trimming
These bits provide an additional user-programmable trimming value that is added to the
HSICAL[7:0] bits. It can be programmed to adjust to variations in voltage and temperature
that influence the frequency of the internal HSI RC.
Bit 2 Reserved, must be kept at reset value.
Bit 1 HSIRDY: Internal high-speed clock ready flag
Set by hardware to indicate that the HSI oscillator is stable. After the HSION bit is cleared,
HSIRDY goes low after 6 HSI clock cycles.
0: HSI oscillator not ready
1: HSI oscillator ready
Bit 0 HSION: Internal high-speed clock enable
Set and cleared by software.
Set by hardware to force the HSI oscillator ON when leaving the Stop or Standby mode or in
case of a failure of the HSE oscillator used directly or indirectly as the system clock. This bit
cannot be cleared if the HSI is used directly or indirectly as the system clock.
0: HSI oscillator OFF
1: HSI oscillator ON

122/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.2 RCC PLL configuration register (RCC_PLLCFGR)

Address offset: 0x04
Reset value: 0x2400 3010
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLL clock outputs according to the formulas:
• f(VCO clock) = f(PLL clock input) × (PLLN / PLLM)
• f(PLL general clock output) = f(VCO clock) / PLLP
• f(USB OTG FS, SDIO, RNG clock output) = f(VCO clock) / PLLQ
• f(I2S, DFSDM clock output) = f(VCO clock)/ PLLR

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. PLLR[2:0] PLLQ[3:0] Res. PLLSRC Res. Res. Res. Res. PLLP[1:0]
rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. PLLN[8:0] PLLM[5:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.

Bits 30:28 PLLR[2:0]: Main PLL (PLL) division factor for I2S, DFSDM clocks
Set and cleared by software to control the frequency of the clock. These bits should be
written only if PLL is disabled.
Clock frequency = VCO frequency / PLLR with 2 ≤ PLLR ≤ 7
000: PLLR = 0, wrong configuration
001: PLLR = 1, wrong configuration
010: PLLR = 2
011: PLLR = 3
...
111: PLLR = 7
Bits 27:24 PLLQ[3:0]: Main PLL (PLL) division factor for USB OTG FS, SDIO and random number
generator clocks.
Set and cleared by software to control the frequency of USB OTG FS clock, the random
number generator clock and the SDIO clock. These bits should be written only if PLL is
disabled.
Caution: The USB OTG FS requires a 48 MHz clock to work correctly. The SDIO and the
random number generator need a frequency lower than or equal to 48 MHz to work
correctly.
USB OTG FS clock frequency = VCO frequency / PLLQ with 2 ≤ PLLQ ≤ 15
0000: PLLQ = 0, wrong configuration
0001: PLLQ = 1, wrong configuration
0010: PLLQ = 2
0011: PLLQ = 3
0100: PLLQ = 4
...
1111: PLLQ = 15
Bit 23 Reserved, must be kept at reset value.

RM0402 Rev 6 123/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 22 PLLSRC: Main PLL(PLL) and audio PLL (PLLI2S) entry clock source
Set and cleared by software to select PLL and PLLI2S clock source. This bit can be written
only when PLL and PLLI2S are disabled.
0: HSI clock selected as PLL and PLLI2S clock entry
1: HSE oscillator clock selected as PLL and PLLI2S clock entry
Bits 21:18 Reserved, must be kept at reset value.
Bits 17:16 PLLP[1:0]: Main PLL (PLL) division factor for main system clock
Set and cleared by software to control the frequency of the general PLL output clock. These
bits can be written only if PLL is disabled.
Caution: The software has to set these bits correctly not to exceed 100 MHz on this domain.
PLL output clock frequency = VCO frequency / PLLP with PLLP = 2, 4, 6, or 8
00: PLLP = 2
01: PLLP = 4
10: PLLP = 6
11: PLLP = 8
Bits 14:6 PLLN[8:0]: Main PLL (PLL) multiplication factor for VCO
Set and cleared by software to control the multiplication factor of the VCO. These bits can
be written only when PLL is disabled. Only half-word and word accesses are allowed to
write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output
frequency is between 100 and 432 MHz. (check also Section 6.3.23: RCC PLLI2S
configuration register (RCC_PLLI2SCFGR))
VCO output frequency = VCO input frequency × PLLN with 50 ≤ PLLN ≤ 432
000000000: PLLN = 0, wrong configuration
000000001: PLLN = 1, wrong configuration
...
000110010: PLLN = 50
...
001100011: PLLN = 99
001100100: PLLN = 100
...
110110000: PLLN = 432
110110001: PLLN = 433, wrong configuration
...
111111111: PLLN = 511, wrong configuration
Note: Multiplication factors possible for VCO input frequency higher than 1 MHz but care
must be taken to fulfill the VCO output frequency range as specified above.

124/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bits 5:0 PLLM[5:0]: Division factor for the main PLL (PLL) input clock
Set and cleared by software to divide the PLL and PLLI2S input clock before the VCO.
These bits can be written only when the PLL and PLLI2S are disabled.
Caution: The software has to set these bits correctly to ensure that the VCO input frequency
ranges from 1 to 2 MHz. It is recommended to select a frequency of 2 MHz to limit
PLL jitter.
VCO input frequency = PLL input clock frequency / PLLM with 2 ≤ PLLM ≤ 63
000000: PLLM = 0, wrong configuration
000001: PLLM = 1, wrong configuration
000010: PLLM = 2
000011: PLLM = 3
000100: PLLM = 4
...
111110: PLLM = 62
111111: PLLM = 63

6.3.3 RCC clock configuration register (RCC_CFGR)

Address offset: 0x08
Reset value: 0x0000 0000
Access: 0 ≤ wait state ≤ 2, word, half-word and byte access
1 or 2 wait states inserted only if the access occurs during a clock source switch.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

MCO2[1:0] MCO2 PRE[2:0] MCO1 PRE[2:0] Res. MCO1[1:0] RTCPRE[4:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

PPRE2[2:0] PPRE1[2:0] Res. Res. HPRE[3:0] SWS[1:0] SW[1:0]

rw rw rw rw rw rw rw rw rw rw r r rw rw

Bits 31:30 MCO2[1:0]: Microcontroller clock output 2

Set and cleared by software. Clock source selection may generate glitches on MCO2. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and the PLLs.
00: System clock (SYSCLK) selected
01: PLLI2S clock selected
10: HSE oscillator clock selected
11: PLL clock selected

RM0402 Rev 6 125/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bits 29:27 MCO2PRE[1:0]: MCO2 prescaler

Set and cleared by software to configure the prescaler of the MCO2. Modification of this
prescaler may generate glitches on MCO2. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLLs.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bits 26:24 MCO1PRE[1:0]: MCO1 prescaler
Set and cleared by software to configure the prescaler of the MCO1. Modification of this
prescaler may generate glitches on MCO1. It is highly recommended to change this
prescaler only after reset before enabling the external oscillators and the PLL.
0xx: no division
100: division by 2
101: division by 3
110: division by 4
111: division by 5
Bit 23 Reserved, always read as 0.
Bits 22:21 MCO1[1:0]: Microcontroller clock output 1
Set and cleared by software. Clock source selection may generate glitches on MCO1. It is
highly recommended to configure these bits only after reset before enabling the external
oscillators and PLL.
00: HSI clock selected
01: LSE oscillator selected
10: HSE oscillator clock selected
11: PLL clock selected
Bits 20:16 RTCPRE[4:0]: HSE division factor for RTC clock
Set and cleared by software to divide the HSE clock input clock to generate a 1 MHz clock
for RTC.
Caution: The software has to set these bits correctly to ensure that the clock supplied to the
RTC is 1 MHz. These bits must be configured if needed before selecting the RTC
clock source.
00000: no clock
00001: no clock
00010: HSE/2
00011: HSE/3
00100: HSE/4
...
11110: HSE/30
11111: HSE/31
Bits 15:13 PPRE2[2:0]: APB high-speed prescaler (APB2)
Set and cleared by software to control APB high-speed clock division factor.
Caution: The software has to set these bits correctly not to exceed 100 MHz on this domain.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE2 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
101: AHB clock divided by 4
110: AHB clock divided by 8
111: AHB clock divided by 16

126/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bits 12:10 PPRE1[2:0]: APB Low speed prescaler (APB1)

Set and cleared by software to control APB low-speed clock division factor.
Caution: The software has to set these bits correctly not to exceed 50 MHz on this domain.
The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
PPRE1 write.
0xx: AHB clock not divided
100: AHB clock divided by 2
101: AHB clock divided by 4
110: AHB clock divided by 8
111: AHB clock divided by 16
Bits 9:8 Reserved, must be kept at reset value.
Bits 7:4 HPRE[3:0]: AHB prescaler
Set and cleared by software to control AHB clock division factor.
Caution: The clocks are divided with the new prescaler factor from 1 to 16 AHB cycles after
HPRE write.
Caution: The AHB clock frequency must be at least 25 MHz when the Ethernet is used.
0xxx: system clock not divided
1000: system clock divided by 2
1001: system clock divided by 4
1010: system clock divided by 8
1011: system clock divided by 16
1100: system clock divided by 64
1101: system clock divided by 128
1110: system clock divided by 256
1111: system clock divided by 512
Bits 3:2 SWS[1:0]: System clock switch status
Set and cleared by hardware to indicate which clock source is used as the system clock.
00: HSI oscillator used as the system clock
01: HSE oscillator used as the system clock
10: PLL used as the system clock
11: not applicable
Bits 1:0 SW[1:0]: System clock switch
Set and cleared by software to select the system clock source.
Set by hardware to force the HSI selection when leaving the Stop or Standby mode or in
case of failure of the HSE oscillator used directly or indirectly as the system clock.
00: HSI oscillator selected as system clock
01: HSE oscillator selected as system clock
10: PLL selected as system clock
11: not allowed

RM0402 Rev 6 127/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.4 RCC clock interrupt register (RCC_CIR)

Address offset: 0x0C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PLLI2S PLL HSE HSI LSE LSI

Res. Res. Res. Res. Res. Res. Res. Res. CSSC Res.
RDYC RDYC RDYC RDYC RDYC RDYC

w w w w w w w

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

PLLI2S PLL HSE HSI LSE LSI PLLI2S PLL HSE HSI LSE LSI
Res. Res. CSSF Res.
RDYIE RDYIE RDYIE RDYIE RDYIE RDYIE RDYF RDYF RDYF RDYF RDYF RDYF

rw rw rw rw rw rw r r r r r r r

Bits 31:24 Reserved, must be kept at reset value.

Bit 23 CSSC: Clock security system interrupt clear
This bit is set by software to clear the CSSF flag.
0: No effect
1: Clear CSSF flag
Bit 22 Reserved, must be kept at reset value.
Bit 21 PLLI2SRDYC: PLLI2S ready interrupt clear
This bit is set by software to clear the PLLI2SRDYF flag.
0: No effect
1: PLLI2SRDYF cleared
Bit 20 PLLRDYC: Main PLL(PLL) ready interrupt clear
This bit is set by software to clear the PLLRDYF flag.
0: No effect
1: PLLRDYF cleared
Bit 19 HSERDYC: HSE ready interrupt clear
This bit is set by software to clear the HSERDYF flag.
0: No effect
1: HSERDYF cleared
Bit 18 HSIRDYC: HSI ready interrupt clear
This bit is set software to clear the HSIRDYF flag.
0: No effect
1: HSIRDYF cleared
Bit 17 LSERDYC: LSE ready interrupt clear
This bit is set by software to clear the LSERDYF flag.
0: No effect
1: LSERDYF cleared
Bit 16 LSIRDYC: LSI ready interrupt clear
This bit is set by software to clear the LSIRDYF flag.
0: No effect
1: LSIRDYF cleared

128/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bits 15:14 Reserved, must be kept at reset value.

Bit 13 PLLI2SRDYIE: PLLI2S ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by PLLI2S lock.
0: PLLI2S lock interrupt disabled
1: PLLI2S lock interrupt enabled
Bit 12 PLLRDYIE: Main PLL (PLL) ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by PLL lock.
0: PLL lock interrupt disabled
1: PLL lock interrupt enabled
Bit 11 HSERDYIE: HSE ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the HSE oscillator
stabilization.
0: HSE ready interrupt disabled
1: HSE ready interrupt enabled
Bit 10 HSIRDYIE: HSI ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the HSI oscillator
stabilization.
0: HSI ready interrupt disabled
1: HSI ready interrupt enabled
Bit 9 LSERDYIE: LSE ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by the LSE oscillator
stabilization.
0: LSE ready interrupt disabled
1: LSE ready interrupt enabled
Bit 8 LSIRDYIE: LSI ready interrupt enable
Set and cleared by software to enable/disable interrupt caused by LSI oscillator
stabilization.
0: LSI ready interrupt disabled
1: LSI ready interrupt enabled
Bit 7 CSSF: Clock security system interrupt flag
Set by hardware when a failure is detected in the HSE oscillator.
Cleared by software setting the CSSC bit.
0: No clock security interrupt caused by HSE clock failure
1: Clock security interrupt caused by HSE clock failure
Bit 6 Reserved, must be kept at reset value.
Bit 5 PLLI2SRDYF: PLLI2S ready interrupt flag
Set by hardware when the PLLI2S locks and PLLI2SRDYDIE is set.
Cleared by software setting the PLLRI2SDYC bit.
0: No clock ready interrupt caused by PLLI2S lock
1: Clock ready interrupt caused by PLLI2S lock
Bit 4 PLLRDYF: Main PLL (PLL) ready interrupt flag
Set by hardware when PLL locks and PLLRDYDIE is set.
Cleared by software setting the PLLRDYC bit.
0: No clock ready interrupt caused by PLL lock
1: Clock ready interrupt caused by PLL lock

RM0402 Rev 6 129/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 3 HSERDYF: HSE ready interrupt flag

Set by hardware when External High Speed clock becomes stable and HSERDYDIE is set.
Cleared by software setting the HSERDYC bit.
0: No clock ready interrupt caused by the HSE oscillator
1: Clock ready interrupt caused by the HSE oscillator
Bit 2 HSIRDYF: HSI ready interrupt flag
Set by hardware when the Internal High Speed clock becomes stable and HSIRDYDIE is
set.
Cleared by software setting the HSIRDYC bit.
0: No clock ready interrupt caused by the HSI oscillator
1: Clock ready interrupt caused by the HSI oscillator
Bit 1 LSERDYF: LSE ready interrupt flag
Set by hardware when the External Low Speed clock becomes stable and LSERDYDIE is
set.
Cleared by software setting the LSERDYC bit.
0: No clock ready interrupt caused by the LSE oscillator
1: Clock ready interrupt caused by the LSE oscillator
Bit 0 LSIRDYF: LSI ready interrupt flag
Set by hardware when the internal low speed clock becomes stable and LSIRDYDIE is set.
Cleared by software setting the LSIRDYC bit.
0: No clock ready interrupt caused by the LSI oscillator
1: Clock ready interrupt caused by the LSI oscillator

6.3.5 RCC AHB1 peripheral reset register (RCC_AHB1RSTR)

Address offset: 0x10
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DMA2 DMA1
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
RST RST
rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GPIOH GPIOG GPIOF GPIOE GPIOD GPIOC GPIOB GPIOA
Res. Res. Res. CRCRST Res. Res. Res. Res.
RST RST RST RST RST RST RST RST
rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 DMA2RST: DMA2 reset
Set and cleared by software.
0: does not reset DMA2
1: resets DMA2
Bit 21 DMA1RST: DMA1 reset
Set and cleared by software.
0: does not reset DMA1
1: resets DMA1
Bits 20:13 Reserved, must be kept at reset value.

130/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 12 CRCRST: CRC reset

Set and cleared by software.
0: does not reset CRC
1: resets CRC
Bits 11:8 Reserved, must be kept at reset value.
Bit 7 GPIOHRST: IO port H reset
Set and cleared by software.
0: does not reset IO port H
1: resets IO port H
Bit 6 GPIOGRST: IO port G reset
Set and cleared by software.
0: does not reset IO port G
1: resets IO port G
Bit 5 GPIOFRST: IO port F reset
Set and cleared by software.
0: does not reset IO port F
1: resets IO port F
Bit 4 GPIOERST: IO port E reset
Set and cleared by software.
0: does not reset IO port E
1: resets IO port E
Bit 3 GPIODRST: IO port D reset
Set and cleared by software.
0: does not reset IO port D
1: resets IO port D
Bit 2 GPIOCRST: IO port C reset
Set and cleared by software.
0: does not reset IO port C
1: resets IO port C
Bit 1 GPIOBRST: IO port B reset
Set and cleared by software.
0: does not reset IO port B
1:resets IO port B
Bit 0 GPIOARST: IO port A reset
Set and cleared by software.
0: does not reset IO port A
1: resets IO port A

RM0402 Rev 6 131/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.6 RCC AHB2 peripheral reset register (RCC_AHB2RSTR)

Address offset: 0x14
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OTGFS RNG
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
RST RST
rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 OTGFSRST: USB OTG FS module reset
Set and cleared by software.
0: does not reset the USB OTG FS module
1: resets the USB OTG FS module
Bit 6 RNGSRST: RNG module reset
Set and cleared by software.
0: does not reset RNG module
1: resets RNG module
Bits 5:0 Reserved, must be kept at reset value.

132/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.7 RCC AHB3 peripheral reset register (RCC_AHB3RSTR)

Address offset: 0x18
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSMC
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. QSPIRST
RST
rw rw

Bits 31:2 Reserved, must be kept at reset value.

Bit 1 QSPIRST: QUADSPI module reset
Set and cleared by software.
0: does not reset QUADSPI module
1: resets QUADSPI module
Bit 0 FSMCRST: Flexible memory controller module reset
Set and cleared by software.
0: does not reset the FSMC module
1: resets the FSMC module

6.3.8 RCC APB1 peripheral reset register for (RCC_APB1RSTR)

Address offset: 0x20
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PWR CAN2 CAN1 I2CFMP1 I2C3 I2C2 I2C1 USART3 USART2

Res. Res. Res. Res. Res. Res. Res.
RST RST RST RST RST RST RST RST RST

rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

SPI3 SPI2 WWDG TIM14 TIM13 TIM12 TIM7 TIM6 TIM5 TIM4 TIM3 TIM2
Res. Res. Res. Res.
RST RST RST RST RST RST RST RST RST RST RST RST

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:29 Reserved, must be kept at reset value.

Bit 28 PWRRST: Power interface reset
Set and reset by software.
0: does not reset the power interface
1: resets the power interface
Bit 27 Reserved, must be kept at reset value.

RM0402 Rev 6 133/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 26 CAN2RST: CAN2 reset

Set and cleared by software.
0: does not reset CAN2
1: resets CAN2
Bit 25 CAN1RST: CAN1 reset
Set and cleared by software.
0: does not reset CAN1
1: resets CAN1
Bit 24 I2CFMP1RST: I2CFMP1 reset
Set and cleared by software.
0: does not reset I2CFMP1
1: resets I2CFMP1
Bit 23 I2C3RST: I2C3 reset
Set and reset by software.
0: does not reset I2C3
1: resets I2C3
Bit 22 I2C2RST: I2C2 reset
Set and cleared by software.
0: does not reset I2C2
1: resets I2C2
Bit 21 I2C1RST: I2C1 reset
Set and reset by software.
0: does not reset I2C1
1: resets I2C1
Bits 20:19 Reserved, must be kept at reset value.
Bit 18 USART3RST: USART3 reset
Set and cleared by software.
0: does not reset USART3
1: resets USART3
Bit 17 USART2RST: USART2 reset
Set and cleared by software.
0: does not reset USART2
1: resets USART2
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3RST: SPI3 reset
Set and cleared by software.
0: does not reset SPI3
1: resets SPI3
Bit 14 SPI2RST: SPI2 reset
Set and cleared by software.
0: does not reset SPI2
1: resets SPI2
Bits 13:12 Reserved, must be kept at reset value.

134/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 11 WWDGRST: Window watchdog reset

Set and cleared by software.
0: does not reset the window watchdog
1: resets the window watchdog
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 TIM14RST: TIM14 reset
Set and cleared by software.
0: does not reset TIM14
1: resets TIM14
Bit 7 TIM13RST: TIM13 reset
Set and cleared by software.
0: does not reset TIM13
1: resets TIM13
Bit 6 TIM12RST: TIM12 reset
Set and cleared by software.
0: does not reset TIM12
1: resets TIM12
Bit 5 TIM7RST: TIM7 reset
Set and cleared by software.
0: does not reset TIM7
1: resets TIM7
Bit 4 TIM6RST: TIM6 reset
Set and cleared by software.
0: does not reset TIM6
1: resets TIM6
Bit 3 TIM5RST: TIM5 reset
Set and cleared by software.
0: does not reset TIM5
1: resets TIM5
Bit 2 TIM4RST: TIM4 reset
Set and cleared by software.
0: does not reset TIM4
1: resets TIM4
Bit 1 TIM3RST: TIM3 reset
Set and cleared by software.
0: does not reset TIM3
1: resets TIM3
Bit 0 TIM2RST: TIM2 reset
Set and cleared by software.
0: does not reset TIM2
1: resets TIM2

RM0402 Rev 6 135/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.9 RCC APB2 peripheral reset register (RCC_APB2RSTR)

Address offset: 0x24
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DFSDM1 SPI5 TIM11 TIM10 TIM9
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
RST RST RST RST RST
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 Res. 2 1 0
SYSCFG SPI4 SPI1 SDIO ADC1 USART6 USART1 TIM8 TIM1
Res. Res. Res. Res. Res. Res. Res.
RST RST RST RST RST RST RST RST RST
rw rw rw rw rw rw rw rw rw

Bits 31:25 Reserved, must be kept at reset value.

Bit 24 DFSDM1RST: DFSDM1 reset
Set and cleared by software.
0: does not reset DFSDM1
1: resets DFSDM1
Bits 23:21 Reserved, always read as 0.
Bit 20 SPI5RST: SPI5RST
This bit is set and cleared by software.
0: does not reset SPI5
1: resets SPI5
Bit 19 Reserved, must be kept at reset value.
Bit 18 TIM11RST: TIM11 reset
Set and cleared by software.
0: does not reset TIM11
1: resets TIM11
Bit 17 TIM10RST: TIM10 reset
Set and cleared by software.
0: does not reset TIM10
1: resets TIM10
Bit 16 TIM9RST: TIM9 reset
Set and cleared by software.
0: does not reset TIM9
1: resets TIM9
Bit 15 Reserved, must be kept at reset value.
Bit 14 SYSCFGRST: System configuration controller reset
Set and cleared by software.
0: does not reset the System configuration controller
1: resets the System configuration controller
Bit 13 SPI4RST: SPI4 reset
Set and reset by software.
0: does not reset SPI4
1: resets SPI4

136/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 12 SPI1RST: SPI1 reset

Set and cleared by software.
0: does not reset SPI1
1: resets SPI1
Bit 11 SDIORST: SDIO reset
Set and cleared by software.
0: does not reset the SDIO module
1: resets the SDIO module
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 ADC1RST: ADC interface reset
Set and cleared by software.
0: does not reset the ADC interface
1: resets the ADC interface
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6RST: USART6 reset
Set and cleared by software.
0: does not reset USART6
1: resets USART6
Bit 4 USART1RST: USART1 reset
Set and cleared by software.
0: does not reset USART1
1: resets USART1
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8RST: TIM8 reset
Set and cleared by software.
0: does not reset TIM8
1: resets TIM8
Bit 0 TIM1RST: TIM1 reset
Set and cleared by software.
0: does not reset TIM1
1: resets TIM1

RM0402 Rev 6 137/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.10 RCC AHB1 peripheral clock enable register (RCC_AHB1ENR)

Address offset: 0x30
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. DMA2EN DMA1EN Res. Res. Res. Res. Res.
rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GPIOH GPIOG GPIOF GPIOE GPIOD GPIOC GPIOB GPIOA
Res. Res. Res. CRCEN Res. Res. Res. Res.
EN EN EN EN EN EN EN EN
rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 DMA2EN: DMA2 clock enable
Set and cleared by software.
0: DMA2 clock disabled
1: DMA2 clock enabled
Bit 21 DMA1EN: DMA1 clock enable
Set and cleared by software.
0: DMA1 clock disabled
1: DMA1 clock enabled
Bits 20:13 Reserved, must be kept at reset value.
Bit 12 CRCEN: CRC clock enable
Set and cleared by software.
0: CRC clock disabled
1: CRC clock enabled
Bits 11:8 Reserved, must be kept at reset value.
Bit 7 GPIOHEN: IO port H clock enable
Set and reset by software.
0: IO port H clock disabled
1: IO port H clock enabled
Bit 6 GPIOGEN: IO port G clock enable
Set and cleared by software.
0: IO port G clock disabled
1: IO port G clock enabled
Bit 5 GPIOFEN: IO port F clock enable
Set and cleared by software.
0: IO port F clock disabled
1: IO port F clock enabled
Bit 4 GPIOEEN: IO port E clock enable
Set and cleared by software.
0: IO port E clock disabled
1: IO port E clock enabled

138/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 3 GPIODEN: IO port D clock enable

Set and cleared by software.
0: IO port D clock disabled
1: IO port D clock enabled
Bit 2 GPIOCEN: IO port C clock enable
Set and cleared by software.
0: IO port C clock disabled
1: IO port C clock enabled
Bit 1 GPIOBEN: IO port B clock enable
Set and cleared by software.
0: IO port B clock disabled
1: IO port B clock enabled
Bit 0 GPIOAEN: IO port A clock enable
Set and cleared by software.
0: IO port A clock disabled
1: IO port A clock enabled

RM0402 Rev 6 139/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.11 RCC AHB2 peripheral clock enable register (RCC_AHB2ENR)

Address offset: 0x34
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OTGFS RNG
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
EN EN
rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 OTGFSEN: USB OTG FS clock enable
Set and cleared by software.
0: USB OTG FS clock disabled
1: USB OTG FS clock enabled
Bit 6 RNGEN: RNG clock enable
Set and cleared by software.
0: RNG clock disabled
1: RNG clock enabled
Bits 5:0 Reserved, always read as 0.

140/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.12 RCC AHB3 peripheral clock enable register (RCC_AHB3ENR)

Address offset: 0x38
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
QSPI FSMC
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
EN EN
rw rw

Bits 31:2 Reserved, must be kept at reset value.

Bit 1 QSPIEN: QUADSPI memory controller module clock enable
Set and cleared by software.
0: QUADSPI clock disabled
1: QUADSPI clock enabled
Bit 0 FSMCEN: Flexible memory controller module clock enable
Set and cleared by software.
0: FSMC module clock disabled
1: FSMC module clock enabled

6.3.13 RCC APB1 peripheral clock enable register (RCC_APB1ENR)

Address offset: 0x40
Reset value: 0x0000 0400
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PWR CAN2 CAN1 I2CFMP1 I2C3 I2C2 I2C1 USART3 USART2
Res. Res. Res. Res. Res. Res. Res.
EN EN EN EN EN EN EN EN EN
rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SPI3 SPI2 WWDG TIM14 TIM13 TIM12 TIM7 TIM6 TIM5 TIM4 TIM3 TIM2
Res. Res. RTCAPB Res.
EN EN EN EN EN EN EN EN EN EN EN EN
rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:29 Reserved, must be kept at reset value.

Bit 28 PWREN: Power interface clock enable
Set and cleared by software.
0: Power interface clock disabled
1: Power interface clock enable
Bit 27 Reserved, must be kept at reset value.

RM0402 Rev 6 141/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 26 CAN2EN: CAN 2 clock enable

This bit is set and cleared by software.
0: CAN 2 clock disabled
1: CAN 2 clock enabled
Bit 25 CAN1EN: CAN 1 clock enable
This bit is set and cleared by software.
0: CAN 1 clock disabled
1: CAN 1 clock enabled
Bit 24 I2CFMP1EN: I2CFMP1 clock enable
This bit is set and cleared by software.
0: I2CFMP1 clock disabled
1: I2CFMP1 clock enabled
Bit 23 I2C3EN: I2C3 clock enable
Set and cleared by software.
0: I2C3 clock disabled
1: I2C3 clock enabled
Bit 22 I2C2EN: I2C2 clock enable
Set and cleared by software.
0: I2C2 clock disabled
1: I2C2 clock enabled
Bit 21 I2C1EN: I2C1 clock enable
Set and cleared by software.
0: I2C1 clock disabled
1: I2C1 clock enabled
Bits 20:17 Reserved, must be kept at reset value.
Bit 18 USART3EN: USART3 clock enable
Set and cleared by software.
0: USART3 clock disabled
1: USART3 clock enabled
Bit 17 USART2EN: USART2 clock enable
Set and cleared by software.
0: USART2 clock disabled
1: USART2 clock enabled
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3EN: SPI3 clock enable
Set and cleared by software.
0: SPI3 clock disabled
1: SPI3 clock enabled
Bit 14 SPI2EN: SPI2 clock enable
Set and cleared by software.
0: SPI2 clock disabled
1: SPI2 clock enabled
Bits 13:12 Reserved, must be kept at reset value.

142/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 11 WWDGEN: Window watchdog clock enable

Set and cleared by software.
0: Window watchdog clock disabled
1: Window watchdog clock enabled
Bit 10 RTC APB: clock enable
Set and cleared by software.
0: RTC APB clock disabled
1: RTC APB clock enabled (default value).
Bit 9 Reserved, must be kept at reset value.
Bit 8 TIM14EN: TIM14 reset
Set and cleared by software.
0: does not reset TIM14
1: resets TIM14
Bit 7 TIM13EN: TIM13 reset
Set and cleared by software.
0: does not reset TIM13
1: resets TIM13
Bit 6 TIM12EN: TIM12 reset
Set and cleared by software.
0: does not reset TIM12
1: resets TIM12
Bit 5 TIM7EN: TIM7 reset
Set and cleared by software.
0: does not reset TIM7
1: resets TIM7
Bit 4 TIM6EN: TIM6 reset
Set and cleared by software.
0: does not reset TIM6
1: resets TIM6
Bit 3 TIM5EN: TIM5 clock enable
Set and cleared by software.
0: TIM5 clock disabled
1: TIM5 clock enabled
Bit 2 TIM4EN: TIM4 clock enable
Set and cleared by software.
0: TIM4 clock disabled
1: TIM4 clock enabled
Bit 1 TIM3EN: TIM3 clock enable
Set and cleared by software.
0: TIM3 clock disabled
1: TIM3 clock enabled
Bit 0 TIM2EN: TIM2 clock enable
Set and cleared by software.
0: TIM2 clock disabled
1: TIM2 clock enabled

RM0402 Rev 6 143/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.14 RCC APB2 peripheral clock enable register

(RCC_APB2ENR)
Address offset: 0x44
Reset value: 0x0000 8000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DFSDM1 TIM11 TIM10 TIM9
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. SPI5EN Res.
EN EN EN EN
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SYSCF SPI1 SDIO ADC1 USART6 USART1 TIM8 TIM1
Res. SPI4EN Res. Res. Res. Res. Res. Res.
G EN EN EN EN EN EN EN EN
rw rw rw rw rw rw rw rw rw

Bits 31:25 Reserved, must be kept at reset value.

Bit 24 DFSDM1EN: DFSDM1 clock enable
Set and cleared by software
0: DFSDM1 clock disabled
1: DFSDM1 clock enabled
Bits 23:21 Reserved, must be kept at reset value.
Bit 20 SPI5EN:SPI5 clock enable
Set and cleared by software
0: SPI5 clock disabled
1: SPI5 clock enabled
Bit 19 Reserved, must be kept at reset value.
Bit 18 TIM11EN: TIM11 clock enable
Set and cleared by software.
0: TIM11 clock disabled
1: TIM11 clock enabled
Bit 17 TIM10EN: TIM10 clock enable
Set and cleared by software.
0: TIM10 clock disabled
1: TIM10 clock enabled
Bit 16 TIM9EN: TIM9 clock enable
Set and cleared by software.
0: TIM9 clock disabled
1: TIM9 clock enabled
Bit 15 Reserved, must be kept at reset value.
Bit 14 SYSCFGEN: System configuration controller clock enable
Set and cleared by software.
0: System configuration controller clock disabled
1: System configuration controller clock enabled

144/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 13 SPI4EN: SPI4 clock enable

Set and reset by software.
0: SPI4 clock disabled
1: SPI4 clock enable
Bit 12 SPI1EN: SPI1 clock enable
Set and cleared by software.
0: SPI1 clock disabled
1: SPI1 clock enabled
Bit 11 SDIOEN: SDIO clock enable
Set and cleared by software.
0: SDIO module clock disabled
1: SDIO module clock enabled
Bit 8 ADC1EN: ADC1 clock enable
Set and cleared by software.
0: ADC1 clock disabled
1: ADC1 clock disabled
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6EN: USART6 clock enable
Set and cleared by software.
0: USART6 clock disabled
1: USART6 clock enabled
Bit 4 USART1EN: USART1 clock enable
Set and cleared by software.
0: USART1 clock disabled
1: USART1 clock enabled
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8EN: TIM8 clock enable
Set and cleared by software.
0: TIM8 clock disabled
1: TIM8 clock enabled
Bit 0 TIM1EN: TIM1 clock enable
Set and cleared by software.
0: TIM1 clock disabled
1: TIM1 clock enabled

RM0402 Rev 6 145/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.15 RCC AHB1 peripheral clock enable in low power mode register
(RCC_AHB1LPENR)
Address offset: 0x50
Reset value: 0x0061 90FF
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DMA2 DMA1 SRAM1
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
LPEN LPEN LPEN
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FLITF CRC GPIOH GPIOG GPIOF GPIOE GPIOD GPIOC GPIOB GPIOA
Res. Res. Res. Res. Res. Res.
LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN
rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 DMA2LPEN: DMA2 clock enable during Sleep mode
Set and cleared by software.
0: DMA2 clock disabled during Sleep mode
1: DMA2 clock enabled during Sleep mode
Bit 21 DMA1LPEN: DMA1 clock enable during Sleep mode
Set and cleared by software.
0: DMA1 clock disabled during Sleep mode
1: DMA1 clock enabled during Sleep mode
Bits 20:17 Reserved, must be kept at reset value.
Bit 16 SRAM1LPEN: SRAM1interface clock enable during Sleep mode
Set and cleared by software.
0: SRAM1 interface clock disabled during Sleep mode
1: SRAM1 interface clock enabled during Sleep mode
Bit 15 FLITFLPEN: Flash interface clock enable during Sleep mode
Set and cleared by software.
0: Flash interface clock disabled during Sleep mode
1: Flash interface clock enabled during Sleep mode
Bits 14:13 Reserved, must be kept at reset value.
Bit 12 CRCLPEN: CRC clock enable during Sleep mode
Set and cleared by software.
0: CRC clock disabled during Sleep mode
1: CRC clock enabled during Sleep mode
Bits 11:8 Reserved, must be kept at reset value.
Bit 7 GPIOHLPEN: IO port H clock enable during sleep mode
Set and reset by software.
0: IO port H clock disabled during sleep mode
1: IO port H clock enabled during sleep mode

146/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 6 GPIOGLPEN: IO port G clock enable during Sleep mode

Set and cleared by software.
0: IO port G clock disabled during Sleep mode
1: IO port G clock enabled during Sleep mode
Bit 5 GPIOFLPEN: IO port F clock enable during Sleep mode
Set and cleared by software.
0: IO port F clock disabled during Sleep mode
1: IO port F clock enabled during Sleep mode
Bit 4 GPIOELPEN: IO port E clock enable during Sleep mode
Set and cleared by software.
0: IO port E clock disabled during Sleep mode
1: IO port E clock enabled during Sleep mode
Bit 3 GPIODLPEN: IO port D clock enable during Sleep mode
Set and cleared by software.
0: IO port D clock disabled during Sleep mode
1: IO port D clock enabled during Sleep mode
Bit 2 GPIOCLPEN: IO port C clock enable during Sleep mode
Set and cleared by software.
0: IO port C clock disabled during Sleep mode
1: IO port C clock enabled during Sleep mode
Bit 1 GPIOBLPEN: IO port B clock enable during Sleep mode
Set and cleared by software.
0: IO port B clock disabled during Sleep mode
1: IO port B clock enabled during Sleep mode
Bit 0 GPIOALPEN: IO port A clock enable during sleep mode
Set and cleared by software.
0: IO port A clock disabled during Sleep mode
1: IO port A clock enabled during Sleep mode

RM0402 Rev 6 147/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.16 RCC AHB2 peripheral clock enable in low power mode register
(RCC_AHB2LPENR)
Address offset: 0x54
Reset value: 0x0000 00C0
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OTGFS RNG
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
LPEN LPEN
rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 OTGFSLPEN: USB OTG FS clock enable during Sleep mode
Set and cleared by software.
0: USB OTG FS clock disabled during Sleep mode
1: USB OTG FS clock enabled during Sleep mode
Bit 6 RNGLPEN: RNG clock enable during Sleep mode
Set and cleared by software.
0: RNG clock disabled during Sleep mode
1: RNG clock enabled during Sleep mode
Bits 5:0 Reserved, must be kept at reset value.

6.3.17 RCC AHB3 peripheral clock enable in low power mode register
(RCC_AHB3LPENR)
Address offset: 0x58
Reset value: 0x0000 0003
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
QSPI FSMC
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
LPEN LPEN
rw rw

148/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bits 31:2 Reserved, must be kept at reset value.

Bit 1 QSPILPEN: QUADSPI memory controller module clock enable during Sleep mode
Set and cleared by software.
0: QUADSPI module clock disabled during Sleep mode
1: QUADSPI module clock enabled during Sleep mode
Bit 0 FSMCLPEN: Flexible memory controller module clock enable during Sleep mode
Set and cleared by software.
0: FSMC clock disabled during Sleep mode
1: FSMC clock enabled during Sleep mode

RM0402 Rev 6 149/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.18 RCC APB1 peripheral clock enable in low power mode register
(RCC_APB1LPENR)
Address offset: 0x60
Reset value: 0x17E6 CDFF
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PWR CAN2 CAN1 I2CFMP1 I2C3 I2C2 I2C1 USART3 USART2
Res. Res. Res. Res. Res. Res. Res.
LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN
rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SPI3 SPI2 WWDG RTCAPB TIM14 TIM13 TIM12 TIM7 TIM6 TIM5 TIM4 TIM3 TIM2
Res. Res. Res.
LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN LPEN
rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:29 Reserved, must be kept at reset value.

Bit 28 PWRLPEN: Power interface clock enable during Sleep mode
Set and cleared by software.
0: Power interface clock disabled during Sleep mode
1: Power interface clock enabled during Sleep mode
Bit 27 Reserved, must be kept at reset value.
Bit 26 CAN2LPEN: CAN2 clock enable during Sleep mode
Set and cleared by software.
0: CAN2 clock disabled during Sleep mode
1: CAN2 clock enabled during Sleep mode
Bit 25 CAN1LPEN: CAN1 clock enable during Sleep mode
Set and cleared by software.
0: CAN1 clock disabled during Sleep mode
1: CAN1 clock enabled during Sleep mode
Bit 24 I2CFMP1LPEN: I2CFMP1 clock enable during Sleep mode
Set and cleared by software.
0: I2CFMP1 clock disabled during Sleep mode
1: I2CFMP1 clock enabled during Sleep mode
Bit 23 I2C3LPEN: I2C3 clock enable during Sleep mode
Set and cleared by software.
0: I2C3 clock disabled during Sleep mode
1: I2C3 clock enabled during Sleep mode
Bit 22 I2C2LPEN: I2C2 clock enable during Sleep mode
Set and cleared by software.
0: I2C2 clock disabled during Sleep mode
1: I2C2 clock enabled during Sleep mode
Bit 21 I2C1LPEN: I2C1 clock enable during Sleep mode
Set and cleared by software.
0: I2C1 clock disabled during Sleep mode
1: I2C1 clock enabled during Sleep mode
Bits 20:19 Reserved, must be kept at reset value.

150/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 18 USART3LPEN: USART3 clock enable during Sleep mode

Set and cleared by software.
0: USART3 clock disabled during Sleep mode
1: USART3 clock enabled during Sleep mode
Bit 17 USART2LPEN: USART2 clock enable during Sleep mode
Set and cleared by software.
0: USART2 clock disabled during Sleep mode
1: USART2 clock enabled during Sleep mode
Bit 16 Reserved, must be kept at reset value.
Bit 15 SPI3LPEN: SPI3 clock enable during Sleep mode
Set and cleared by software.
0: SPI3 clock disabled during Sleep mode
1: SPI3 clock enabled during Sleep mode
Bit 14 SPI2LPEN: SPI2 clock enable during Sleep mode
Set and cleared by software.
0: SPI2 clock disabled during Sleep mode
1: SPI2 clock enabled during Sleep mode
Bits 13:12 Reserved, must be kept at reset value.
Bit 11 WWDGLPEN: Window watchdog clock enable during Sleep mode
Set and cleared by software.
0: Window watchdog clock disabled during sleep mode
1: Window watchdog clock enabled during sleep mode
Bit 10 RTCAPBEN: RTC APB clock enable during Sleep mode
Set and cleared by software.
0: RTC APB clock disabled during sleep mode
1: RTC APB watchdog clock enabled during sleep mode
Bit 9 Reserved, must be kept at reset value.
Bit 8 TIM14LPEN: TIM14 clock enable during Sleep mode
Set and cleared by software.
0: TIM14 clock disabled during Sleep mode
1: TIM14 clock enabled during Sleep mode
Bit 7 TIM13LPEN: TIM13 clock enable during Sleep mode
Set and cleared by software.
0: TIM13 clock disabled during Sleep mode
1: TIM13 clock enabled during Sleep mode
Bit 6 TIM12LPEN: TIM12 clock enable during Sleep mode
Set and cleared by software.
0: TIM12 clock disabled during Sleep mode
1: TIM12 clock enabled during Sleep mode
Bit 5 TIM7LPEN: TIM7 clock enable during Sleep mode
Set and cleared by software.
0: TIM7 clock disabled during Sleep mode
1: TIM7 clock enabled during Sleep mode

RM0402 Rev 6 151/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 4 TIM6LPEN: TIM6 clock enable during Sleep mode

Set and cleared by software.
0: TIM6 clock disabled during Sleep mode
1: TIM6 clock enabled during Sleep mode
Bit 3 TIM5LPEN: TIM5 clock enable during Sleep mode
Set and cleared by software.
0: TIM5 clock disabled during Sleep mode
1: TIM5 clock enabled during Sleep mode
Bit 2 TIM4LPEN: TIM4 clock enable during Sleep mode
Set and cleared by software.
0: TIM4 clock disabled during Sleep mode
1: TIM4 clock enabled during Sleep mode
Bit 1 TIM3LPEN: TIM3 clock enable during Sleep mode
Set and cleared by software.
0: TIM3 clock disabled during Sleep mode
1: TIM3 clock enabled during Sleep mode
Bit 0 TIM2LPEN: TIM2 clock enable during Sleep mode
Set and cleared by software.
0: TIM2 clock disabled during Sleep mode
1: TIM2 clock enabled during Sleep mode

152/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.19 RCC APB2 peripheral clock enabled in low power mode register
(RCC_APB2LPENR)
Address offset: 0x64
Reset value: 0x0117 F933
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DFSDM1 SPI5 TIM11 TIM10 TIM9
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
LPEN LPEN LPEN LPEN LPEN
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SYSC
EXTIT SPI4LP SPI1 SDIO ADC1 USART6 USART1 TIM8 TIM1
FG Res. Res. Res. Res. Res. Res.
EN EN LPEN LPEN LPEN LPEN LPEN LPEN LPEN
LPEN
rw rw rw rw rw rw rw rw rw rw

Bits 31:25 Reserved, must be kept at reset value.

Bit 24 DFSDM1LPEN: DFSDM1 clock enable during Sleep mode
This bit is set and cleared by software
0: DFSDM1 clock disabled during Sleep mode
1: DFSDM1 clock enabled during Sleep mode
Bits 23:21 Reserved, must be kept at reset value.
Bit 20 SPI5LPEN: SPI5 clock enable during Sleep mode
This bit is set and cleared by software
0: SPI5 clock disabled during Sleep mode
1: SPI5 clock enabled during Sleep mode
Bit 19 Reserved, must be kept at reset value.
Bit 18 TIM11LPEN: TIM11 clock enable during Sleep mode
Set and cleared by software.
0: TIM11 clock disabled during Sleep mode
1: TIM11 clock enabled during Sleep mode
Bit 17 TIM10LPEN: TIM10 clock enable during Sleep mode
Set and cleared by software.
0: TIM10 clock disabled during Sleep mode
1: TIM10 clock enabled during Sleep mode
Bit 16 TIM9LPEN: TIM9 clock enable during sleep mode
Set and cleared by software.
0: TIM9 clock disabled during Sleep mode
1: TIM9 clock enabled during Sleep mode
Bit 15 EXTITEN: EXTIT APB and SYSCTRL PFREE clock enable during Sleep mode
Set and cleared by software.
0: EXTIT APB and SYSCTRL PFREE clock disabled during Sleep mode
1: EXTIT APB and SYSCTRL PFREE clock enabled during Sleep mode

RM0402 Rev 6 153/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 14 SYSCFGLPEN: System configuration controller clock enable during Sleep mode
Set and cleared by software.
0: System configuration controller clock disabled during Sleep mode
1: System configuration controller clock enabled during Sleep mode
Bit 13 SPI4LPEN: SPI4 clock enable during sleep mode
Set and reset by software.
0: SPI4 clock disabled during sleep mode
1: SPI4 clock enabled during sleep mode
Bit 12 SPI1LPEN: SPI1 clock enable during Sleep mode
Set and cleared by software.
0: SPI1 clock disabled during Sleep mode
1: SPI1 clock enabled during Sleep mode
Bit 11 SDIOLPEN: SDIO clock enable during Sleep mode
Set and cleared by software.
0: SDIO module clock disabled during Sleep mode
1: SDIO module clock enabled during Sleep mode
Bits 10:9 Reserved, must be kept at reset value.
Bit 8 ADC1LPEN: ADC1 clock enable during Sleep mode
Set and cleared by software.
0: ADC1 clock disabled during Sleep mode
1: ADC1 clock disabled during Sleep mode
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 USART6LPEN: USART6 clock enable during Sleep mode
Set and cleared by software.
0: USART6 clock disabled during Sleep mode
1: USART6 clock enabled during Sleep mode
Bit 4 USART1LPEN: USART1 clock enable during Sleep mode
Set and cleared by software.
0: USART1 clock disabled during Sleep mode
1: USART1 clock enabled during Sleep mode
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 TIM8LPEN: TIM8 clock enable during Sleep mode
Set and cleared by software.
0: TIM8 clock disabled during Sleep mode
1: TIM8 clock enabled during Sleep mode
Bit 0 TIM1LPEN: TIM1 clock enable during Sleep mode
Set and cleared by software.
0: TIM1 clock disabled during Sleep mode
1: TIM1 clock enabled during Sleep mode

154/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.20 RCC Backup domain control register (RCC_BDCR)

Address offset: 0x70
Reset value: 0x0000 0000, reset by Backup domain reset.
Access: 0 ≤ wait state ≤ 3, word, half-word and byte access
Wait states are inserted in case of successive accesses to this register.
The LSEON, LSEBYP, RTCSEL and RTCEN bits in the RCC Backup domain control
register (RCC_BDCR) are in the Backup domain. As a result, after Reset, these bits are
write-protected and the DBP bit in the Section 5.4.1: PWR power control register
(PWR_CR) has to be set before these can be modified. Refer to Section 5.4.2: PWR power
control/status register (PWR_CSR) for further information. These bits are only reset after a
Backup domain Reset (see Section 6.1.3: Backup domain reset). Any internal or external
Reset will not have any effect on these bits.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. BDRST
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RTCEN Res. Res. Res. Res. Res. RTCSEL[1:0] Res. Res. Res. Res. LSEMOD LSEBYP LSERDY LSEON
rw rw rw rw rw r rw

Bits 31:17 Reserved, must be kept at reset value.

Bit 16 BDRST: Backup domain software reset
Set and cleared by software.
0: Reset not activated
1: Resets the entire Backup domain
Bit 15 RTCEN: RTC clock enable
Set and cleared by software.
0: RTC clock disabled
1: RTC clock enabled
Bits 14:10 Reserved, must be kept at reset value.
Bits 9:8 RTCSEL[1:0]: RTC clock source selection
Set by software to select the clock source for the RTC. Once the RTC clock source has been
selected, it cannot be changed anymore unless the Backup domain is reset. The BDRST bit
can be used to reset them.
00: No clock
01: LSE oscillator clock used as the RTC clock
10: LSI oscillator clock used as the RTC clock
11: HSE oscillator clock divided by a programmable prescaler (selection through the
RTCPRE[4:0] bits in the RCC clock configuration register (RCC_CFGR)) used as the RTC
clock
Bits 7:4 Reserved, must be kept at reset value.
Bit 3 LSEMOD: External low-speed oscillator bypass
Set and reset by software to select crystal mode for low speed oscillator. Two power modes
are available.
0: LSE oscillator “low power” mode selection
1: LSE oscillator “high drive” mode selection

RM0402 Rev 6 155/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 2 LSEBYP: External low-speed oscillator bypass

Set and cleared by software to bypass oscillator in debug mode. This bit can be written only
when the LSE clock is disabled.
0: LSE oscillator not bypassed
1: LSE oscillator bypassed
Bit 1 LSERDY: External low-speed oscillator ready
Set and cleared by hardware to indicate when the external 32 kHz oscillator is stable. After
the LSEON bit is cleared, LSERDY goes low after 6 external low-speed oscillator clock
cycles.
0: LSE clock not ready
1: LSE clock ready
Bit 0 LSEON: External low-speed oscillator enable
Set and cleared by software.
0: LSE clock OFF
1: LSE clock ON

6.3.21 RCC clock control & status register (RCC_CSR)

Address offset: 0x74
Reset value: 0x0E00 0000, reset by system reset, except reset flags by power reset only.
Access: 0 ≤ wait state ≤ 3, word, half-word and byte access
Wait states are inserted in case of successive accesses to this register.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
LPWR WWDG IWDG SFT POR PIN BORRS
RMVF Res. Res. Res. Res. Res. Res. Res. Res.
RSTF RSTF RSTF RSTF RSTF RSTF TF
r r r r r r r rt_w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. LSIRDY LSION
r rw

Bit 31 LPWRRSTF: Low-power reset flag

Set by hardware when a Low-power management reset occurs.
Cleared by writing to the RMVF bit.
0: No Low-power management reset occurred
1: Low-power management reset occurred
For further information on Low-power management reset, refer to Low-power management
reset.
Bit 30 WWDGRSTF: Window watchdog reset flag
Set by hardware when a window watchdog reset occurs.
Cleared by writing to the RMVF bit.
0: No window watchdog reset occurred
1: Window watchdog reset occurred
Bit 29 IWDGRSTF: Independent watchdog reset flag
Set by hardware when an independent watchdog reset from VDD domain occurs.
Cleared by writing to the RMVF bit.
0: No watchdog reset occurred
1: Watchdog reset occurred

156/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 28 SFTRSTF: Software reset flag

Set by hardware when a software reset occurs.
Cleared by writing to the RMVF bit.
0: No software reset occurred
1: Software reset occurred
Bit 27 PORRSTF: POR/PDR reset flag
Set by hardware when a POR/PDR reset occurs.
Cleared by writing to the RMVF bit.
0: No POR/PDR reset occurred
1: POR/PDR reset occurred
Bit 26 PINRSTF: PIN reset flag
Set by hardware when a reset from the NRST pin occurs.
Cleared by writing to the RMVF bit.
0: No reset from NRST pin occurred
1: Reset from NRST pin occurred
Bit 25 BORRSTF: BOR reset flag
Cleared by software by writing the RMVF bit.
Set by hardware when a POR/PDR or BOR reset occurs.
0: No POR/PDR or BOR reset occurred
1: POR/PDR or BOR reset occurred
Bit 24 RMVF: Remove reset flag
Set by software to clear the reset flags.
0: No effect
1: Clear the reset flags
Bits 23:2 Reserved, must be kept at reset value.
Bit 1 LSIRDY: Internal low-speed oscillator ready
Set and cleared by hardware to indicate when the internal RC 40 kHz oscillator is stable.
After the LSION bit is cleared, LSIRDY goes low after 3 LSI clock cycles.
0: LSI RC oscillator not ready
1: LSI RC oscillator ready
Bit 0 LSION: Internal low-speed oscillator enable
Set and cleared by software.
0: LSI RC oscillator OFF
1: LSI RC oscillator ON

RM0402 Rev 6 157/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

6.3.22 RCC spread spectrum clock generation register (RCC_SSCGR)

Address offset: 0x80
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.
The spread spectrum clock generation is available only for the main PLL.
The RCC_SSCGR register must be written either before the main PLL is enabled or after
the main PLL disabled.
Note: For full details about PLL spread spectrum clock generation (SSCG) characteristics, refer to
the “Electrical characteristics” section in your device datasheet.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SSCG SPREAD
Res. Res. INCSTEP[14:3]
EN SEL
rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INCSTEP[2:0] MODPER[11:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 SSCGEN: Spread spectrum modulation enable

Set and cleared by software.
0: Spread spectrum modulation DISABLE. (To write after clearing CR[24]=PLLON bit)
1: Spread spectrum modulation ENABLE. (To write before setting CR[24]=PLLON bit)
Bit 30 SPREADSEL: Spread Select
Set and cleared by software.
To write before to set CR[24]=PLLON bit.
0: Center spread
1: Down spread
Bits 29:28 Reserved, must be kept at reset value.
Bits 27:13 INCSTEP[14:0]: Incrementation step
Set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile amplitude.
Bits 12:0 MODPER[11:0]: Modulation period
Set and cleared by software. To write before setting CR[24]=PLLON bit.
Configuration input for modulation profile period.

158/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.23 RCC PLLI2S configuration register (RCC_PLLI2SCFGR)

Address offset: 0x84
Reset value: 0x2400 3010
Access: no wait state, word, half-word and byte access.
This register is used to configure the PLLI2S clock outputs according to the formulas:
• f(VCO clock) = f(PLLI2S clock input) × (PLLI2SN / PLLI2SM)
• f(USB OTG FS, SDIO, RNG clock output) = f(VCO clock) / PLLQ
• f(DFSDM, I2S clock output) = f(VCO clock) / PLLR

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res Res Res Res
Res. PLLI2SR[2:0] PLLI2SQ[3:0] Res. PLLI2SSRC Res. Res.
. . . .
rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. PLLI2SN[8:0] PLLI2SM[5:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.

Bits 30:28 PLLI2SR[2:0]: PLLI2S division factor for I2S clocks
Set and cleared by software to control the I2S clock frequency. These bits should be written
only if the PLLI2S is disabled. The factor must be chosen in accordance with the prescaler
values inside the I2S peripherals, to reach 0.3% error when using standard crystals and 0%
error with audio crystals. For more information about I2S clock frequency and precision, refer
to Section 26.6.4: Clock generator in the I2S chapter.
Caution: The I2Ss requires a frequency lower than or equal to 192 MHz to work correctly.
I2S clock frequency = VCO frequency / PLLR with 2 ≤ PLLR ≤ 7
000: PLLR = 0, wrong configuration
001: PLLR = 1, wrong configuration
010: PLLR = 2
...
111: PLLR = 7
Bits 27:24 PLLI2SQ[3:0]: PLLI2S division factor for USB OTG FS/SDIO/RNG clock
Set and cleared by software to control the USB OTG FS/SDIO/RNG clock frequency. These
bits can be written only when the PLLI2S is disabled.
USB OTG FS/SDIO/RNG clock frequency = VCO frequency / PLLI2SQ with 2 ≤ PLLI2SQ ≤
15
0000: PLLI2SQ = 0, wrong configuration
0001: PLLI2SQ = 1, wrong configuration
0010: PLLI2SQ = 2
0011: PLLI2SQ = 3
0100: PLLI2SQ = 4
0101: PLLI2SQ = 5
...
1111: PLLI2SQ = 15
Bit 23 Reserved, must be kept at reset value.

RM0402 Rev 6 159/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bit 22 PLLI2SSRC: PLLI2S entry clock source

Set and cleared by software to select PLLI2S clock source. This bit can be written only when
PLLI2S is disabled.
0: HSE or HSI depending on PLLSRC of PLLCFGR
1: external AFI clock (CK_I2S_EXT) selected as PLL clock entry
Bits 21:15 Reserved, must be kept at reset value.
Bits 14:6 PLLI2SN[8:0]: PLLI2S multiplication factor for VCO
Set and cleared by software to control the multiplication factor of the VCO. These bits can be
written only when the PLLI2S is disabled. Only half-word and word accesses are allowed to
write these bits.
Caution: The software has to set these bits correctly to ensure that the VCO output frequency
is between 100 and 432 MHz. With VCO input frequency ranges from 1 to 2 MHz
(refer to Figure 14 and divider factor M of the RCC PLL configuration register
(RCC_PLLCFGR))
VCO output frequency = VCO input frequency × PLLI2SN with 50 ≤ PLLI2SN ≤ 432
000000000: PLLI2SN = 0, wrong configuration
000000001: PLLI2SN = 1, wrong configuration
...
001100010: PLLI2SN = 50
...
001100011: PLLI2SN = 99
001100100: PLLI2SN = 100
001100101: PLLI2SN = 101
001100110: PLLI2SN = 102
...
110110000: PLLI2SN = 432
110110000: PLLI2SN = 433, wrong configuration
...
111111111: PLLI2SN = 511, wrong configuration
Note: Between 50 and 99 multiplication factors are possible for VCO input frequency higher
than 1 MHz. However care must be taken to fulfill the minimum VCO output frequency
as specified above.
Bits 5:0 PLLI2SM[5:0]: Division factor for the main PLL (PLL) and audio PLL (PLLI2S) input clock
Set and cleared by software to divide the PLL and PLLI2S input clock before the VCO.
These bits can be written only when the PLL and PLLI2S are disabled.
Caution: The software has to set these bits correctly to ensure that the VCO input frequency
ranges from 1 to 2 MHz.It is recommended to select a frequency of 2 MHz to limit
PLL jitter.
VCO input frequency = PLL input clock frequency / PLLI2SM with 2 ≤ PLLI2SM ≤ 63
000000: PLLI2SM = 0, wrong configuration
000001: PLLI2SM = 1, wrong configuration...
000010: PLLI2SM = 2
000011: PLLI2SM = 3
000100: PLLI2SM = 4
.......
111110: PLLI2SM = 62
111111: PLLI2SM = 63

160/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

6.3.24 RCC Dedicated Clocks Configuration Register (RCC_DCKCFGR)

Address offset: 0x8C
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CKDFSD
Res. Res. I2S2RC[1:0] I2S1RC[1:0] TIMPRE Res. Res. Res. Res. Res. Res. Res. Res.
M1SEL
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKDFSD
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
M1ASEL
rw

Bit 31 CKDFSDMSEL: DFSDM1 Kernel clock selection.

0: APB2 clock used as Kernel clock
1: System clock used as Kernel clock
Bits 30:29 Reserved, must be kept at reset value.
Bits 28:27 I2S2SRC[1:0]: I2S APB2 clocks source selection (I2S1/4/5)
Set and reset by software.
These bits should be written when the PLL and PLLI2S are disabled.
00: I2S APB2 clock frequency = f(PLLI2S_R)
01: I2S APB2 clock frequency = external I2S clock from pads - alternate function input
frequency
10: I2S APB2 clock frequency = f(PLL_R)
11: I2S APB2 clock frequency = HSI/HSE depending on PLLSRC (PLLCFGR(22))
Bits 26:25 I2S1SRC[1:0]: I2S APB1 clocks source selection (I2S2/3)
Set and reset by software to control the frequency of the APB1 I2S clock.
These bits should be written when the PLL and PLLI2S are disabled.
00: I2S APB1 clock frequency = f(PLLI2S_R)
01: I2S APB1 clock frequency = external I2S clock from pads - alternate function input
frequency
10: I2S APB1 clock frequency = f(PLL_R)
11: I2S APB1 clock frequency = HSI/HSE depending on PLLSRC (PLLCFGR(22))
Bit 24 TIMPRE: Timers clocks prescalers selection
Set and reset by software to control the clock frequency of all the timers connected to APB1
and APB2 domain.
0: If the APB prescaler (PPRE1, PPRE2 in the RCC_CFGR register) is configured to a
division factor of 1, TIMxCLK = PCLKx. Otherwise, the timer clock frequencies are set to
twice to the frequency of the APB domain to which the timers are connected:
TIMxCLK = 2xPCLKx.
1:If the APB prescaler (PPRE1, PPRE2 in the RCC_CFGR register) is configured to a
division factor of 1, 2, or 4, TIMxCLK = HCKL. Otherwise, the timer clock frequencies are set
to four times to the frequency of the APB domain to which the timers are connected:
TIMxCLK = 4xPCLKx.

RM0402 Rev 6 161/1163

166
Reset and clock control (RCC) for STM32F412xx RM0402

Bits 23: 16 Reserved, must be kept at reset value.

Bit 15 CKDFSDM1ASEL: DFSDM1 audio clock selection.
0: CK_I2S_APB1 selected as audio clock
1: CK_I2S_APB2 selected as audio clock
Bits 14:0 Reserved, must be kept at reset value.

6.3.25 RCC clocks gated enable register (CKGATENR)

Address offset: 0x90
Reset value: 0x0000 0000
Access: no wait state, word, half-word and byte access.
This register allows to enable or disable the clock gating for the specified IPs.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EVTCL RCC FLITF SRAM SPARE CM4DBG AHB2APB2 AHB2APB1
Res. Res. Res. Res. Res. Res. Res. Res.
_CKEN _CKEN _CKEN _CKEN _CKEN _CKEN _CKEN _CKEN
rw rw rw rw rw rw rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 EVTCL_CKEN
0: the clock gating is enabled
1: the cock gating is disabled, the clock is always enabled
Bit 6 RCC_CKEN: RCC clock enable
0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.
Bit 5 FLITF_CKEN: Flash Interface clock enable
0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.
Bit 4 SRAM_CKEN: SRQAM controller clock enable
0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.
Bit 3 SPARE_CKEN: Spare clock enable
0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.

162/1163 RM0402 Rev 6

RM0402 Reset and clock control (RCC) for STM32F412xx

Bit 2 CM4DBG_CKEN: Cortex M4 ETM clock enable

0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.
Bit 1 AHB2APB2_CKEN: AHB to APB2 Bridge clock enable
0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.
Bit 0 AHB2APB1_CKEN: AHB to APB1 Bridge clock enable
0: the clock gating is enabled
1: the clock gating is disabled, the clock is always enabled.

6.3.26 RCC Dedicated Clocks Configuration Register (RCC_DCKCFGR2)

Address offset: 0x94
Reset value: 0x0000 0000
This register allows to enable or disable the clock gating for the specified IPs.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SDIO CK48M I2CFMP1
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
SEL SEL SEL[1:0]
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

Bits 31:29 Reserved, must be kept at reset value.

Bit 28 CKSDIOSEL: SDIO clock selection.
0: CK_48MHz (see CK48MSEL bit definition)
1: clock system
Bit 27 CK48MSEL: SDIO/USBFS clock selection.
0: f(PLL_Q)
1: f(PLLI2S_Q)
Bits 26:24 Reserved, must be kept at reset value.
Bits 23:22 I2CFMP1SEL[1:0]: I2CFMP1 kernel clock source selection
00: APB clock selected as I2CFMP1 clock
01: System clock selected as I2CFMP1 clock
10: HSI clock selected as I2CFMP1 clock
11: APB clock selected as I2CFMP1 (same as “00”)
Bits 21: 0 Reserved, must be kept at reset value.

RM0402 Rev 6 163/1163

166
 0x34
0x30
0x28
0x24
0x20
0x18
0x14
0x10
0x08
0x04
0x00

0x2C
0x1C
0x0C
Addr.
offset
6.3.27

164/1163
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
name

RCC_CR

RCC_CIR
PLLCFGR

AHB2ENR
AHB1ENR
Register

APB2RSTR
APB1RSTR
AHB3RSTR
AHB2RSTR
AHB1RSTR
RCC_CFGR
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
MCO2[1:0]
Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. PWRRST Res. Res. Res. Res. MCO2PRE[2:0] Res.

PLLR[2:0]
28
Res. Res. Res. Res. Res. Res. Res. Res. PLL I2SRDY 27
Res. Res. Res. CAN2RST Res. Res. Res. Res. PLL I2SON 26
RCC register map

Res. Res. Res. CAN1RST Res. Res. Res. Res. MCO1PRE[2:0] PLL RDY 25

PLLQ[3:0]
Res. Res. DFSDM1RST I2CFMP1RST Res. Res. Res. Res. PLL ON 24
Res. Res. Res. I2C3RST Res. Res. Res. CSSC Res. Res. Res. 23
Res. DMA2EN Res. I2C2RST Res. Res. DMA2RST Res. PLLSRC Res. 22
MCO1[1:0]
Res. DMA1EN Res. I2C1RST Res. Res. DMA1RST PLLI2SRDYC Res. Res. 21
Res. Res. SPI5RST Res. Res. Res. Res. PLLRDYC Res. Res. 20
Reset and clock control (RCC) for STM32F412xx

Res. Res. Res. Res. Res. Res. Res. HSERDYC Res. CSSON 19
Res. Res. TIM11RST USART3RST Res. Res. Res. HSIRDYC Res. HSEBYP 18
Res. Res. TIM10RST USART2RST Res. Res. Res. LSERDYC HSERDY

Reserved
Reserved
Reserved
17

RM0402 Rev 6
PLLP[1:0]
RTCPRE[4:0]

Res. Res. TIM9RST Res. Res. Res. Res. LSIRDYC HSEON 16

Table 24 gives the register map and reset values

Res. Res. Res. SPI3RST Res. Res. Res. Res. Res. 15

Res. Res. SYSCFGRST SPI2RST Res. Res. Res. Res. PPRE2[2:0] 14
Res. Res. SP45RST Res. Res. Res. Res. PLLI2SRDYIE 13
Res. CRCEN SPI1RST Res. Res. Res. CRCRST PLLRDYIE 12
Res. Res. SDIORST WWDGRST Res. Res. Res. HSERDYIE PPRE1[2:0] 11
HSICAL[7:0]

Res. Res. Res. Res. Res. Res. Res. HSIRDYIE 10

Res. Res. Res. Res. Res. Res. Res. LSERDYIE Res.
PLLN[8:0]

9
Res. Res. ADC1RST TIM14RST Res. Res. Res. LSIRDYIE Res. 8
OTGFSEN GPIOHEN Res. TIM13RST Res. OTGFSRST GPIOHRST CSSF
Table 24. RCC register map and reset values for STM32F412xx

7
RNGEN GPIOGEN Res. TIM12RST Res. RNGRST GPIOGRST Res. 6
Res. GPIOFEN USART6RST TIM7RST Res. Res. GPIOFRST PLLI2SRDYF 5
HPRE[3:0]3

Res. GPIOEEN USART1RST TIM6RST Res. Res. GPIOERST PLLRDYF 4

HSITRIM[4:0]

Res. GPIODEN Res. TIM5RST Res. Res. GPIODRST HSERDYF 3

SWS[1:0]
Res. GPIOCEN Res. TIM4RST Res. Res. GPIOCRST HSIRDYF Res. 2
PLLM[5:0]

Res. GPIOBEN TIM8RST TIM3RST QSPIRST Res. GPIOBRST LSERDYF HSIRDY 1

SW[1:0]
Res. GPIOAEN TIM1RST TIM2RST FSMCRST Res. GPIOARST LSIRDYF HSION
RM0402

0
 0x78
0x74
0x70
0x68
0x64
0x60
0x58
0x54
0x50
0x48
0x44
0x40
0x38

0x7C
0x6C
0x5C
0x4C
0x3C
Addr.
offset
RM0402

RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
RCC_
name

APB2ENR
APB1ENR

RCC_CSR
AHB3ENR
Register

RCC_BDCR
APB2LPENR
APB1LPENR
AHB3LPENR
AHB2LPENR
AHB1LPENR
LPWRRSTF Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
WWDGRSTF Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
WDGRSTF Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
SFTRSTF Res. Res. PWRLPEN Res. Res. Res. Res. PWREN Res. 28
PORRSTF Res. Res. Res. Res. Res. Res. Res. Res. Res. 27
PADRSTF Res. Res. CAN2LPEN Res. Res. Res. Res. CAN2EN Res. 26
BORRSTF Res. Res. CAN1LPEN Res. Res. Res. Res. CAN1EN Res. 25
RMVF Res. DFSDM1LPEN I2CFMP1LPEN Res. Res. Res. DFSDM1EN I2CFMP1EN Res. 24
Res. Res. Res. I2C3LPEN Res. Res. Res. Res. I2C3EN Res. 23
Res. Res. Res. I2C2LPEN Res. Res. DMA2LPEN Res. I2C2EN Res. 22
Res. Res. Res. I2C1LPEN Res. Res. DMA1LPEN Res. I2C1EN Res. 21
Res. Res. SPI5LPEN Res. Res. Res. Res. SPI5EN Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. TIM11LPEN USART3LPEN Res. Res. Res. TIM11EN Res. Res. 18
Res. Res. TIM10LPEN USART2LPEN Res. Res. Res. TIM10EN USART2EN Res.

Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved

17

RM0402 Rev 6
Res. BDRST TIM9LPEN Res. Res. Res. SRAM1LPEN TIM9EN Res. Res. 16
Res. RTCEN EXTITEN SPI3LPEN Res. Res. FLITFLPEN Res. SPI3EN Res. 15
Res. Res. SYSCFGLPEN SPI2LPEN Res. Res. Res. SYSCFGEN SPI2EN Res. 14
Res. Res. SPI4LPEN Res. Res. Res. Res. SPI4EN Res. Res. 13
Res. Res. SPI1LPEN Res. Res. Res. CRCLPEN SPI1EN Res. Res. 12
Res. Res. SDIOLPEN WWDGLPEN Res. Res. Res. SDIOEN WWDGEN Res. 11
Res. Res. Res. Res. Res. Res. Res. Res. RTCAPBEN Res. 10
Res. Res. Res. Res. Res. Res. Res. Res. Res. 9
RTCSEL[1:0]
Res. ADC1LPEN TIM14LPEN Res. Res. Res. ADC1EN TIM14EN Res. 8
Res. Res. Res. TIM13LPEN Res. OTGFSLPEN GPIOHLPEN Res. TIM13EN Res. 7
Res. Res. Res. TIM12LPEN Res. RNGLPEN GPIOGLPEN Res. TIM12EN Res. 6
Res. Res. USART6LPEN TIM7LPEN Res. Res. GPIOFLPEN USART6EN TIM7EN Res. 5
Table 24. RCC register map and reset values for STM32F412xx (continued)

Res. Res. USART1LPEN TIM6LPEN Res. Res. GPIOELPEN USART1EN TIM6EN Res. 4
Res. LSEMOD Res. TIM5LPEN Res. Res. GPIODLPEN Res. TIM5EN Res. 3
Res. LSEBYP Res. TIM4LPEN Res. Res. GPIOCLPEN Res. TIM4EN Res. 2
LSIRDY LSERDY TIM8LPEN TIM3LPEN QSPILPEN Res. GPIOBLPEN TIM8EN TIM3EN QSPIEN 1
LSION LSEON TIM1LPEN TIM2LPEN FSMCLPEN Res. GPIOALPEN TIM1EN TIM2EN FSMCEN
Reset and clock control (RCC) for STM32F412xx

165/1163
0

166
 0x94
0x90
0x88
0x84
0x80

0x8C
Addr.
offset

166/1163
RCC_
RCC_
RCC_
name

DCKCFGR
Register

DCKCFGR2
CKGATENR
PLLI2SCFGR
RCC_SSCGR
Res. CKDFSDM1SEL Res. SSCGEN 31
Res. Res. SPREADSEL 30
Res. Res. PLLI2SR[2:0] Res. 29
SDIOSEL Res. 28
I2S2SRC[1:0]
CK48MSEL 27
Res. 26
I2S1SRC[1:0] PLLI2SQ[3:0]
Res. 25
Res. TIMPRE 24
Res. Res. 23
I2CFMP1SEL[1:0]
Res. PLLI2SSRC 22
Res. Res. Res. 21
Res. Res. Res. 20
Reset and clock control (RCC) for STM32F412xx

Res. Res. Res. 19

INCSTEP[14:0]

Res. Res. Res. 18

Res. Res. Res.
Reserved

17

RM0402 Rev 6
Res. Res. Res. 16
Res. CKDFSDM1ASEL Res. 15
Res. Res. 14
Res. Res. 13
Res. Res. 12
Res. Res. 11
Res. Res. 10
Res. Res. 9
PLLI2SN[8:0]]

Refer to Section 2.2 on page 49 for the register boundary addresses.

Res. Res. 8
Res. EVTCL_CKEN Res. 7
Res. RCC_CKEN Res. 6
Res. FLITF_CKEN Res. 5
Table 24. RCC register map and reset values for STM32F412xx (continued)

MODPER[11:0]

Res. SRAM_CKEN Res. 4

Res. SPARE _CKEN Res. 3
Res. CM4DBG_CKEN Res. 2
Res. AHB2APB2_CKEN Res.
PLLI2SM[5:0]

1
Res. AHB2APB1_CKEN Res.
RM0402

0
RM0402 General-purpose I/Os (GPIO)

7 General-purpose I/Os (GPIO)

7.1 GPIO introduction

Each general-purpose I/O port has four 32-bit configuration registers (GPIOx_MODER,
GPIOx_OTYPER, GPIOx_OSPEEDR and GPIOx_PUPDR), two 32-bit data registers
(GPIOx_IDR and GPIOx_ODR), a 32-bit set/reset register (GPIOx_BSRR), a 32-bit locking
register (GPIOx_LCKR) and two 32-bit alternate function selection register (GPIOx_AFRH
and GPIOx_AFRL).

7.2 GPIO main features

• Up to 16 I/Os under control
• Output states: push-pull or open drain + pull-up/down
• Output data from output data register (GPIOx_ODR) or peripheral (alternate function
output)
• Speed selection for each I/O
• Input states: floating, pull-up/down, analog
• Input data to input data register (GPIOx_IDR) or peripheral (alternate function input)
• Bit set and reset register (GPIOx_BSRR) for bitwise write access to GPIOx_ODR
• Locking mechanism (GPIOx_LCKR) provided to freeze the I/O configuration
• Analog function
• Alternate function input/output selection registers (at most 16 AFs per I/O)
• Fast toggle capable of changing every two clock cycles
• Highly flexible pin multiplexing allows the use of I/O pins as GPIOs or as one of several
peripheral functions

7.3 GPIO functional description

Subject to the specific hardware characteristics of each I/O port listed in the datasheet, each
port bit of the general-purpose I/O (GPIO) ports can be individually configured by software in
several modes:
• Input floating
• Input pull-up
• Input-pull-down
• Analog
• Output open-drain with pull-up or pull-down capability
• Output push-pull with pull-up or pull-down capability
• Alternate function push-pull with pull-up or pull-down capability
• Alternate function open-drain with pull-up or pull-down capability
Each I/O port bit is freely programmable, however the I/O port registers have to be
accessed as 32-bit words, half-words or bytes. The purpose of the GPIOx_BSRR register is
to allow atomic read/modify accesses to any of the GPIO registers. In this way, there is no
risk of an IRQ occurring between the read and the modify access.

RM0402 Rev 6 167/1163

186
General-purpose I/Os (GPIO) RM0402

Figure 17 show the basic structure of a 5 V tolerant I/O port bit. Table 25 gives the possible
port bit configurations.

Figure 17. Basic structure of a five-volt tolerant I/O port bit

To on-chip Analog
peripheral
Alternate function input

on/off

Input data register

Read
VDD VDD_FT (1)

TTL Schmitt
Bit set/reset registers

Protection
trigger on/off
Pull diode
up
Input driver I/O pin
Write
Output data register

Output driver VDD on/off Pull Protection

down
diode
P-MOS
Output VSS
VSS
control
Read/write N-MOS

From on-chip VSS

Push-pull,
peripheral Alternate function output open-drain or
disabled
Analog
ai15939b

1. VDD_FT is a potential specific to five-volt tolerant I/Os and different from VDD.

Table 25. Port bit configuration table(1)

MODER(i) OSPEEDR(i) PUPDR(i)
OTYPER(i) I/O configuration
[1:0] [B:A] [1:0]

0 0 0 GP output PP
0 0 1 GP output PP + PU
0 1 0 GP output PP + PD
0 SPEED 1 1 Reserved
01
1 [B:A] 0 0 GP output OD
1 0 1 GP output OD + PU
1 1 0 GP output OD + PD
1 1 1 Reserved (GP output OD)

168/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

Table 25. Port bit configuration table(1) (continued)

MODER(i) OSPEEDR(i) PUPDR(i)
OTYPER(i) I/O configuration
[1:0] [B:A] [1:0]

0 0 0 AF PP
0 0 1 AF PP + PU
0 1 0 AF PP + PD
0 SPEED 1 1 Reserved
10
1 [B:A] 0 0 AF OD
1 0 1 AF OD + PU
1 1 0 AF OD + PD
1 1 1 Reserved
x x x 0 0 Input Floating
x x x 0 1 Input PU
00
x x x 1 0 Input PD
x x x 1 1 Reserved (input floating)
x x x 0 0 Input/output Analog
x x x 0 1
11
x x x 1 0 Reserved
x x x 1 1
1. GP = general-purpose, PP = push-pull, PU = pull-up, PD = pull-down, OD = open-drain, AF = alternate
function.

7.3.1 General-purpose I/O (GPIO)

During and just after reset, the alternate functions are not active and the I/O ports are
configured in input floating mode.
The debug pins are in AF pull-up/pull-down after reset:
• PA15: JTDI in pull-up
• PA14: JTCK/SWCLK in pull-down
• PA13: JTMS/SWDAT in pull-up
• PB4: NJTRST in pull-up
• PB3: JTDO in floating state
When the pin is configured as output, the value written to the output data register
(GPIOx_ODR) is output on the I/O pin. It is possible to use the output driver in push-pull
mode or open-drain mode (only the N-MOS is activated when 0 is output).
The input data register (GPIOx_IDR) captures the data present on the I/O pin at every AHB1
clock cycle.
All GPIO pins have weak internal pull-up and pull-down resistors, which can be activated or
not depending on the value in the GPIOx_PUPDR register.

RM0402 Rev 6 169/1163

186
General-purpose I/Os (GPIO) RM0402

7.3.2 I/O pin multiplexer and mapping

The microcontroller I/O pins are connected to onboard peripherals/modules through a
multiplexer that allows only one peripheral’s alternate function (AF) connected to an I/O pin
at a time. In this way, there can be no conflict between peripherals sharing the same I/O pin.
Each I/O pin has a multiplexer with sixteen alternate function inputs (AF0 to AF15) that can
be configured through the GPIOx_AFRL (for pin 0 to 7) and GPIOx_AFRH (for pin 8 to 15)
registers:
• After reset all I/Os are connected to the system’s alternate function 0 (AF0)
• The peripherals’ alternate functions are mapped from AF1 to AF13
• Cortex®-M4 with FPU EVENTOUT is mapped on AF15
This structure is shown in Figure 18: Selecting an alternate function on STM32F412xx
below.
In addition to this flexible I/O multiplexing architecture, each peripheral has alternate
functions mapped onto different I/O pins to optimize the number of peripherals available in
smaller packages.
To use an I/O in a given configuration, proceed as follows:
• System function
Connect the I/O to AF0 and configure it depending on the function used:
– JTAG/SWD, after each device reset these pins are assigned as dedicated pins
immediately usable by the debugger host (not controlled by the GPIO controller)
– RTC_REFIN: this pin should be configured in Input floating mode
– MCO1 and MCO2: these pins have to be configured in alternate function mode.
Note: You can disable some or all of the JTAG/SWD pins and so release the associated pins for
GPIO usage.
For more details please refer to Section 6.2.10: Clock-out capability.

170/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

Table 26. Flexible SWJ-DP pin assignment

SWJ I/O pin assigned

Available debug ports PA13 / PA14 /

PA15 / PB3 / PB4/
JTMS/ JTCK/
JTDI JTDO NJTRST
SWDIO SWCLK

Full SWJ (JTAG-DP + SW-DP) - Reset state X X X X X

Full SWJ (JTAG-DP + SW-DP) but without
X X X X
NJTRST
JTAG-DP Disabled and SW-DP Enabled X X
JTAG-DP Disabled and SW-DP Disabled Released

• GPIO
Configure the desired I/O as output or input in the GPIOx_MODER register.
• Peripheral alternate function
For the ADC, configure the desired I/O as analog in the GPIOx_MODER register.
For other peripherals:
– Configure the desired I/O as an alternate function in the GPIOx_MODER register
– Select the type, pull-up/pull-down and output speed via the GPIOx_OTYPER,
GPIOx_PUPDR and GPIOx_OSPEEDR registers, respectively
– Connect the I/O to the desired AFx in the GPIOx_AFRL or GPIOx_AFRH register
• EVENTOUT
Configure the I/O pin used to output the Cortex®-M4 with FPU EVENTOUT signal by
connecting it to AF15
Note: Please refer to the “Alternate function mapping” table in the datasheets for the detailed
mapping of the system and peripherals’ alternate function I/O pins.

RM0402 Rev 6 171/1163

186
General-purpose I/Os (GPIO) RM0402

Figure 18. Selecting an alternate function on STM32F412xx

For pins 0 to 7, the GPIOx_AFRL[31:0] register selects the dedicated alternate function

AF0 (system)
AF1 (TIM1/TIM2)
AF2 (TIM3/5)
AF3 (TIM8/9)
AF4 (I2C1..3, I2CFMP1)
AF5 (SPI1/2/3/4)
AF6 (SPI2/3/4, DFSDM) Pin x (x = 0..7)
AF7 (SPI2, USART1..3)
1
AF8 (DFSDM, USART3/6, CAN1)
AF9 (I2C1..3, I2CFMP1, CAN1/2, TIM12..14, QUADSPI)
AF10 (DFSDM, FSMC, QUADSPI, OTG_FS)
AF11
AF12 (FMC, SDIO)
AF13
AF14
AF15 (EVENTOUT)

AFRL[31:0]

For pins 8 to 15, the GPIOx_AFRH[31:0] register selects the dedicated alternate function

AF0 (system)
AF1 (TIM1/TIM2)
AF2 (TIM3/5)
AF3 (TIM8/9)
AF4 (I2C1..3, I2CFMP1)
AF5 (SPI1/2/3/4)
AF6 (SPI2/3/4, DFSDM) Pin x (x = 8..15)
AF7 (SPI2, USART1..3)
AF8 (DFSDM, USART3/6, CAN1) 1

AF9 (I2C1..3, I2CFMP1, CAN1/2, TIM12..14, QUADSPI)

AF10 (DFSDM, FSMC, QUADSPI, OTG_FS)
AF11
AF12 (FMC, SDIO)
AF13
AF14
AF15 (EVENTOUT)

AFRH[31:0]
MSv39420V1

172/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

7.3.3 I/O port control registers

Each of the GPIOs has four 32-bit memory-mapped control registers (GPIOx_MODER,
GPIOx_OTYPER, GPIOx_OSPEEDR, GPIOx_PUPDR) to configure up to 16 I/Os.
The GPIOx_MODER register is used to select the I/O direction (input, output, AF, analog).
The GPIOx_OTYPER and GPIOx_OSPEEDR registers are used to select the output type
(push-pull or open-drain) and speed (the I/O speed pins are directly connected to the
corresponding GPIOx_OSPEEDR register bits whatever the I/O direction). The
GPIOx_PUPDR register is used to select the pull-up/pull-down whatever the I/O direction.

7.3.4 I/O port data registers

Each GPIO has two 16-bit memory-mapped data registers: input and output data registers
(GPIOx_IDR and GPIOx_ODR). GPIOx_ODR stores the data to be output, it is read/write
accessible. The data input through the I/O are stored into the input data register
(GPIOx_IDR), a read-only register.
See Section 7.4.5: GPIO port input data register (GPIOx_IDR) (x = A...H) and Section 7.4.6:
GPIO port output data register (GPIOx_ODR) (x = A...H) for the register descriptions.

7.3.5 I/O data bitwise handling

The bit set reset register (GPIOx_BSRR) is a 32-bit register which allows the application to
set and reset each individual bit in the output data register (GPIOx_ODR). The bit set reset
register has twice the size of GPIOx_ODR.
To each bit in GPIOx_ODR, correspond two control bits in GPIOx_BSRR: BSRR(i) and
BSRR(i+SIZE). When written to 1, bit BSRR(i) sets the corresponding ODR(i) bit. When
written to 1, bit BSRR(i+SIZE) resets the ODR(i) corresponding bit.
Writing any bit to 0 in GPIOx_BSRR does not have any effect on the corresponding bit in
GPIOx_ODR. If there is an attempt to both set and reset a bit in GPIOx_BSRR, the set
action takes priority.
Using the GPIOx_BSRR register to change the values of individual bits in GPIOx_ODR is a
“one-shot” effect that does not lock the GPIOx_ODR bits. The GPIOx_ODR bits can always
be accessed directly. The GPIOx_BSRR register provides a way of performing atomic
bitwise handling.
There is no need for the software to disable interrupts when programming the GPIOx_ODR
at bit level: it is possible to modify one or more bits in a single atomic AHB1 write access.

7.3.6 GPIO locking mechanism

It is possible to freeze the GPIO control registers by applying a specific write sequence to
the GPIOx_LCKR register. The frozen registers are GPIOx_MODER, GPIOx_OTYPER,
GPIOx_OSPEEDR, GPIOx_PUPDR, GPIOx_AFRL and GPIOx_AFRH.
To write the GPIOx_LCKR register, a specific write / read sequence has to be applied. When
the right LOCK sequence is applied to bit 16 in this register, the value of LCKR[15:0] is used
to lock the configuration of the I/Os (during the write sequence the LCKR[15:0] value must
be the same). When the LOCK sequence has been applied to a port bit, the value of the port
bit can no longer be modified until the next MCU or peripheral reset. Each GPIOx_LCKR bit
freezes the corresponding bit in the control registers (GPIOx_MODER, GPIOx_OTYPER,
GPIOx_OSPEEDR, GPIOx_PUPDR, GPIOx_AFRL and GPIOx_AFRH).

RM0402 Rev 6 173/1163

186
General-purpose I/Os (GPIO) RM0402

The LOCK sequence (refer to Section 7.4.8: GPIO port configuration lock register
(GPIOx_LCKR) (x = A...H)) can only be performed using a word (32-bit long) access to the
GPIOx_LCKR register due to the fact that GPIOx_LCKR bit 16 has to be set at the same
time as the [15:0] bits.
For more details please refer to LCKR register description in Section 7.4.8: GPIO port
configuration lock register (GPIOx_LCKR) (x = A...H).

7.3.7 I/O alternate function input/output

Two registers are provided to select one out of the sixteen alternate function inputs/outputs
available for each I/O. With these registers, you can connect an alternate function to some
other pin as required by your application.
This means that a number of possible peripheral functions are multiplexed on each GPIO
using the GPIOx_AFRL and GPIOx_AFRH alternate function registers. The application can
thus select any one of the possible functions for each I/O. The AF selection signal being
common to the alternate function input and alternate function output, a single channel is
selected for the alternate function input/output of one I/O.
To know which functions are multiplexed on each GPIO pin, refer to the datasheets.
Note: The application is allowed to select one of the possible peripheral functions for each I/O at a
time.

7.3.8 External interrupt/wakeup lines

All ports have external interrupt capability. To use external interrupt lines, the port must be
configured in input mode, refer to Section 10.2: External interrupt/event controller (EXTI)
and Section 10.2.3: Wakeup event management.

7.3.9 Input configuration

When the I/O port is programmed as Input:
• the output buffer is disabled
• the Schmitt trigger input is activated
• the pull-up and pull-down resistors are activated depending on the value in the
GPIOx_PUPDR register
• The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle
• A read access to the input data register provides the I/O State
Figure 19 shows the input configuration of the I/O port bit.

174/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

Figure 19. Input floating/pull up/pull down configurations

Input data register

on
Read
VDD VDD
Bit set/reset registers
TTL Schmitt on/off
trigger protection
pull diode
Write up
Output data register input driver I/O pin
on/off
output driver
pull protection
down diode
VSS VSS
Read/write

ai15940b

7.3.10 Output configuration

When the I/O port is programmed as output:
• The output buffer is enabled:
– Open drain mode: A “0” in the Output register activates the N-MOS whereas a “1”
in the Output register leaves the port in Hi-Z (the P-MOS is never activated)
– Push-pull mode: A “0” in the Output register activates the N-MOS whereas a “1” in
the Output register activates the P-MOS
• The Schmitt trigger input is activated
• The weak pull-up and pull-down resistors are activated or not depending on the value
in the GPIOx_PUPDR register
• The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle
• A read access to the input data register gets the I/O state
• A read access to the output data register gets the last written value
Figure 20 shows the output configuration of the I/O port bit.

RM0402 Rev 6 175/1163

186
General-purpose I/Os (GPIO) RM0402

Figure 20. Output configuration

Input data register

on
Read

TTL Schmitt VDD VDD

Bit set/reset registers trigger

on/off
Write protection
Input driver diode

Output data register

pull
up
I/O pin
Output driver VDD on/off

P-MOS pull protection

Output down diode
Read/write control VSS
N-MOS VSS
Push-pull or
VSS Open-drain

ai15941b

7.3.11 Alternate function configuration

When the I/O port is programmed as alternate function:
• The output buffer can be configured as open-drain or push-pull
• The output buffer is driven by the signal coming from the peripheral (transmitter enable
and data)
• The Schmitt trigger input is activated
• The weak pull-up and pull-down resistors are activated or not depending on the value
in the GPIOx_PUPDR register
• The data present on the I/O pin are sampled into the input data register every AHB1
clock cycle
• A read access to the input data register gets the I/O state
Figure 21 shows the Alternate function configuration of the I/O port bit.

Figure 21. Alternate function configuration

To on-chip Alternate function input

peripheral
Input data register

on

Read
VDD VDD
TTL Schmitt on/off
Bit set/reset registers

trigger protection
Pull diode
Input driver up
Write
Output data register

I/O pin
Output driver VDD on/off

Pull protection
P-MOS down diode
Output
control VSS VSS
N-MOS
Read/write
VSS push-pull or
open-drain
From on-chip
peripheral Alternate function output
ai15942b

176/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

7.3.12 Analog configuration

When the I/O port is programmed as analog configuration:
• The output buffer is disabled
• The Schmitt trigger input is deactivated, providing zero consumption for every analog
value of the I/O pin. The output of the Schmitt trigger is forced to a constant value (0).
• The weak pull-up and pull-down resistors are disabled
• Read access to the input data register gets the value “0”
Note: In the analog configuration, the I/O pins cannot be 5 Volt tolerant.
Figure 22 shows the high-impedance, analog-input configuration of the I/O port bit.

Figure 22. High impedance-analog configuration

Analog
To on-chip
peripheral
Input data register

Read off
0
VDD
Bit set/reset registers

TTL Schmitt
trigger protection
Write diode
Output data register

Input driver
I/O pin

protection
diode

Read/write VSS

From on-chip Analog

peripheral
ai15943

7.3.13 Using the OSC32_IN/OSC32_OUT pins as GPIO PC14/PC15

port pins
The LSE oscillator pins OSC32_IN and OSC32_OUT can be used as general-purpose
PC14 and PC15 I/Os, respectively, when the LSE oscillator is off. The PC14 and PC15 I/Os
are only configured as LSE oscillator pins OSC32_IN and OSC32_OUT when the LSE
oscillator is ON. This is done by setting the LSEON bit in the RCC_BDCR register. The LSE
has priority over the GPIO function.
Note: The PC14/PC15 GPIO functionality is lost when the 1.2 V domain is powered off (by the
device entering the standby mode) or when the backup domain is supplied by VBAT (VDD no
more supplied). In this case the I/Os are set in analog input mode.

7.3.14 Using the OSC_IN/OSC_OUT pins as GPIO PH0/PH1 port pins

The HSE oscillator pins OSC_IN/OSC_OUT can be used as general-purpose PH0/PH1
I/Os, respectively, when the HSE oscillator is OFF. (after reset, the HSE oscillator is off). The
PH0/PH1 I/Os are only configured as OSC_IN/OSC_OUT HSE oscillator pins when the
HSE oscillator is ON. This is done by setting the HSEON bit in the RCC_CR register. The
HSE has priority over the GPIO function.

RM0402 Rev 6 177/1163

186
 General-purpose I/Os (GPIO) RM0402

7.3.15 Selection of RTC additional functions

The STM32F4xx feature one GPIO pin RTC_AF1 that can be used for the detection of a
tamper or time stamp event, or RTC_ALARM, or RTC_CALIB RTC outputs.
• The RTC_AF1 (PC13) can be used for the following purposes:
RTC_ALARM output: this output can be RTC Alarm A, RTC Alarm B or RTC Wakeup
depending on the OSEL[1:0] bits in the RTC_CR register
• RTC_CALIB output: this feature is enabled by setting the COE[23] in the RTC_CR
register
• RTC_TAMP1: tamper event detection
• RTC_TS: time stamp event detection
The selection of the corresponding pin is performed through the RTC_TAFCR register as
follows:
• TAMP1INSEL is used to select which pin is used as the RTC_TAMP1 tamper input
• TSINSEL is used to select which pin is used as the RTC_TS time stamp input
• ALARMOUTTYPE is used to select whether the RTC_ALARM is output in push-pull or
open-drain mode
The output mechanism follows the priority order listed in Table 27.

Table 27. RTC additional functions(1)

TSINSEL
Pin Time TAMP1INSEL
Tamper TIMESTAMP ALARMOUTTYP
configuration enabled enabled stamp TAMPER1
enabled pin configuration
and function enabled pin selection
selection

Alarm out Don’t Don’t

1 Don’t care Don’t care Don’t care 0
output OD care care
Alarm out Don’t Don’t
1 Don’t care Don’t care Don’t care 1
output PP care care
Calibration Don’t Don’t
0 1 Don’t care Don’t care Don’t care
out output PP care care
TAMPER1
0 0 1 0 0 Don’t care Don’t care
input floating
TIMESTAMP
and
0 0 1 1 0 0 Don’t care
TAMPER1
input floating
TIMESTAMP
0 0 0 1 Don’t care 0 Don’t care
input floating
Standard
0 0 0 0 Don’t care Don’t care Don’t care
GPIO
1. OD: open drain; PP: push-pull.

178/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

7.4 GPIO registers

This section gives a detailed description of the GPIO registers.
For a summary of register bits, register address offsets and reset values, refer to Table 28.
The GPIO registers can be accessed by byte (8 bits), half-words (16 bits) or words (32 bits).

7.4.1 GPIO port mode register (GPIOx_MODER) (x = A...H)

Address offset: 0x00
Reset values:
• 0x0A800 0000 for port A
• 0x0000 0280 for port B
• 0x0000 0000 for other ports

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

MODER15[1:0] MODER14[1:0] MODER13[1:0] MODER12[1:0] MODER11[1:0] MODER10[1:0] MODER9[1:0] MODER8[1:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

MODER7[1:0] MODER6[1:0] MODER5[1:0] MODER4[1:0] MODER3[1:0] MODER2[1:0] MODER1[1:0] MODER0[1:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 2y:2y+1 MODERy[1:0]: Port x configuration bits (y = 0..15)

These bits are written by software to configure the I/O direction mode.
00: Input (reset state)
01: General purpose output mode
10: Alternate function mode
11: Analog mode

7.4.2 GPIO port output type register (GPIOx_OTYPER)

(x = A...H)
Address offset: 0x04
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OT15 OT14 OT13 OT12 OT11 OT10 OT9 OT8 OT7 OT6 OT5 OT4 OT3 OT2 OT1 OT0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 OTy: Port x configuration bits (y = 0..15)
These bits are written by software to configure the output type of the I/O port.
0: Output push-pull (reset state)
1: Output open-drain

RM0402 Rev 6 179/1163

186
General-purpose I/Os (GPIO) RM0402

7.4.3 GPIO port output speed register (GPIOx_OSPEEDR)

(x = A...H)
Address offset: 0x08
Reset values:
• 0x0C00 0000 for port A
• 0x0000 00C0 for port B
• 0x0000 0000 for other ports

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
OSPEEDR15 OSPEEDR14 OSPEEDR13 OSPEEDR12 OSPEEDR11 OSPEEDR10 OSPEEDR9 OSPEEDR8
[1:0] [1:0] [1:0] [1:0] [1:0] [1:0] [1:0] [1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OSPEEDR1 OSPEEDR0
OSPEEDR7[1:0] OSPEEDR6[1:0] OSPEEDR5[1:0] OSPEEDR4[1:0] OSPEEDR3[1:0] OSPEEDR2[1:0]
[1:0] 1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 2y:2y+1 OSPEEDRy[1:0]: Port x configuration bits (y = 0..15)

These bits are written by software to configure the I/O output speed.
00: Low speed
01: Medium speed
10: Fast speed
11: High speed
Note: Refer to the product datasheets for the values of OSPEEDRy bits versus VDD
range and external load.

7.4.4 GPIO port pull-up/pull-down register (GPIOx_PUPDR)

(x = A...H)
Address offset: 0x0C
Reset values:
• 0x6400 0000 for port A
• 0x0000 0100 for port B
• 0x0000 0000 for other ports

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PUPDR15[1:0] PUPDR14[1:0] PUPDR13[1:0] PUPDR12[1:0] PUPDR11[1:0] PUPDR10[1:0] PUPDR9[1:0] PUPDR8[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PUPDR7[1:0] PUPDR6[1:0] PUPDR5[1:0] PUPDR4[1:0] PUPDR3[1:0] PUPDR2[1:0] PUPDR1[1:0] PUPDR0[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 2y:2y+1 PUPDRy[1:0]: Port x configuration bits (y = 0..15)

These bits are written by software to configure the I/O pull-up or pull-down
00: No pull-up, pull-down
01: Pull-up
10: Pull-down
11: Reserved

180/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

7.4.5 GPIO port input data register (GPIOx_IDR) (x = A...H)

Address offset: 0x10
Reset value: 0x0000 XXXX (where X means undefined)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
IDR15 IDR14 IDR13 IDR12 IDR11 IDR10 IDR9 IDR8 IDR7 IDR6 IDR5 IDR4 IDR3 IDR2 IDR1 IDR0
r r r r r r r r r r r r r r r r

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 IDRy: Port input data (y = 0..15)
These bits are read-only and can be accessed in word mode only. They contain the input
value of the corresponding I/O port.

7.4.6 GPIO port output data register (GPIOx_ODR) (x = A...H)

Address offset: 0x14
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ODR15 ODR14 ODR13 ODR12 ODR11 ODR10 ODR9 ODR8 ODR7 ODR6 ODR5 ODR4 ODR3 ODR2 ODR1 ODR0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 ODRy: Port output data (y = 0..15)
These bits can be read and written by software.
Note: For atomic bit set/reset, the ODR bits can be individually set and reset by writing to the
GPIOx_BSRR register (x = A...H).

7.4.7 GPIO port bit set/reset register (GPIOx_BSRR) (x = A...H)

Address offset: 0x18
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BR15 BR14 BR13 BR12 BR11 BR10 BR9 BR8 BR7 BR6 BR5 BR4 BR3 BR2 BR1 BR0
w w w w w w w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BS15 BS14 BS13 BS12 BS11 BS10 BS9 BS8 BS7 BS6 BS5 BS4 BS3 BS2 BS1 BS0
w w w w w w w w w w w w w w w w

RM0402 Rev 6 181/1163

186
General-purpose I/Os (GPIO) RM0402

Bits 31:16 BRy: Port x reset bit y (y = 0..15)

These bits are write-only and can be accessed in word, half-word or byte mode. A read to
these bits returns the value 0x0000.
0: No action on the corresponding ODRx bit
1: Resets the corresponding ODRx bit
Note: If both BSx and BRx are set, BSx has priority.
Bits 15:0 BSy: Port x set bit y (y= 0..15)
These bits are write-only and can be accessed in word, half-word or byte mode. A read to
these bits returns the value 0x0000.
0: No action on the corresponding ODRx bit
1: Sets the corresponding ODRx bit

7.4.8 GPIO port configuration lock register (GPIOx_LCKR)

(x = A...H)
This register is used to lock the configuration of the port bits when a correct write sequence
is applied to bit 16 (LCKK). The value of bits [15:0] is used to lock the configuration of the
GPIO. During the write sequence, the value of LCKR[15:0] must not change. When the
LOCK sequence has been applied on a port bit, the value of this port bit can no longer be
modified until the next MCU or peripheral reset.
Note: A specific write sequence is used to write to the GPIOx_LCKR register. Only word access
(32-bit long) is allowed during this write sequence.
Each lock bit freezes a specific configuration register (control and alternate function
registers).
Address offset: 0x1C
Reset value: 0x0000 0000
Access: 32-bit word only, read/write register

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. LCKK
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
LCK15 LCK14 LCK13 LCK12 LCK11 LCK10 LCK9 LCK8 LCK7 LCK6 LCK5 LCK4 LCK3 LCK2 LCK1 LCK0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

182/1163 RM0402 Rev 6

RM0402 General-purpose I/Os (GPIO)

Bits 31:17 Reserved, must be kept at reset value.

Bit 16 LCKK[16]: Lock key
This bit can be read any time. It can only be modified using the lock key write sequence.
0: Port configuration lock key not active
1: Port configuration lock key active. The GPIOx_LCKR register is locked until an MCU reset
or a peripheral reset occurs.

LOCK key write sequence:

WR LCKR[16] = ‘1’ + LCKR[15:0]
WR LCKR[16] = ‘0’ + LCKR[15:0]
WR LCKR[16] = ‘1’ + LCKR[15:0]
RD LCKR
RD LCKR[16] = ‘1’ (this read operation is optional but it confirms that the lock is active)
Note: During the LOCK key write sequence, the value of LCK[15:0] must not change.
Any error in the lock sequence aborts the lock.
After the first lock sequence on any bit of the port, any read access on the LCKK bit will
return ‘1’ until the next CPU reset.
Bits 15:0 LCKy: Port x lock bit y (y= 0..15)
These bits are read/write but can only be written when the LCKK bit is ‘0.
0: Port configuration not locked
1: Port configuration locked

7.4.9 GPIO alternate function low register (GPIOx_AFRL) (x = A...H)

Address offset: 0x20
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AFRL7[3:0] AFRL6[3:0] AFRL5[3:0] AFRL4[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AFRL3[3:0] AFRL2[3:0] AFRL1[3:0] AFRL0[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 AFRLy: Alternate function selection for port x bit y (y = 0..7)
These bits are written by software to configure alternate function I/Os
AFRLy selection:
0000: AF0 1000: AF8
0001: AF1 1001: AF9
0010: AF2 1010: AF10
0011: AF3 1011: AF11
0100: AF4 1100: AF12
0101: AF5 1101: AF13
0110: AF6 1110: AF14
0111: AF7 1111: AF15

RM0402 Rev 6 183/1163

186
General-purpose I/Os (GPIO) RM0402

7.4.10 GPIO alternate function high register (GPIOx_AFRH)

(x = A...H)
Address offset: 0x24
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AFRH15[3:0] AFRH14[3:0] AFRH13[3:0] AFRH12[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
AFRH11[3:0] AFRH10[3:0] AFRH9[3:0] AFRH8[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 AFRHy: Alternate function selection for port x bit y (y = 8..15)
These bits are written by software to configure alternate function I/Os
AFRHy selection:
0000: AF0 1000: AF8
0001: AF1 1001: AF9
0010: AF2 1010: AF10
0011: AF3 1011: AF11
0100: AF4 1100: AF12
0101: AF5 1101: AF13
0110: AF6 1110: AF14
0111: AF7 1111: AF15

7.4.11 GPIO register map

The following table gives the GPIO register map and the reset values.

Table 28. GPIO register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
MODER15[1:0]

MODER14[1:0]

MODER13[1:0]

MODER12[1:0]

MODER10[1:0]
MODER11[1:0]

MODER9[1:0]

MODER8[1:0]

MODER7[1:0]

MODER6[1:0]

MODER5[1:0]

MODER4[1:0]

MODER3[1:0]

MODER2[1:0]

MODER1[1:0]

MODER0[1:0]
GPIOA_
0x00 MODER

Reset value 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
MODER15[1:0]

MODER14[1:0]

MODER13[1:0]

MODER12[1:0]

MODER10[1:0]
MODER11[1:0]

MODER9[1:0]

MODER8[1:0]

MODER7[1:0]

MODER6[1:0]

MODER5[1:0]

MODER4[1:0]

MODER3[1:0]

MODER2[1:0]

MODER1[1:0]

MODER0[1:0]

GPIOB_
0x00 MODER

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0
MODER15[1:0]

MODER14[1:0]

MODER13[1:0]

MODER12[1:0]

MODER10[1:0]
MODER11[1:0]

MODER9[1:0]

MODER8[1:0]

MODER7[1:0]

MODER6[1:0]

MODER5[1:0]

MODER4[1:0]

MODER3[1:0]

MODER2[1:0]

MODER1[1:0]

MODER0[1:0]

GPIOx_MODER
(where x =
0x00 C...H)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

184/1163 RM0402 Rev 6

 0x18
0x14
0x10
0x08
0x08
0x08
0x04

0x0C
0x0C
0x0C
Offset
RM0402

C...H)
C...H)
GPIOx_
GPIOx_

GPIOB_
GPIOA_
OTYPER

(where x =
(where x =

OSPEEDR
OSPEEDR
Register

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

GPIOx_IDR
OSPEEDER

GPIOx_ODR

GPIOx_BSRR
GPIOx_PUPDR
GPIOB_PUPDR
GPIOA_PUPDR
(where x = A...H)

(where x = A...H)
(where x = A...H)
(where x = A...H)

0
0
0
0
0
0
0
BR15 Res. Res. Res. 31
PUPDR15[1:0] PUPDR15[1:0] PUPDR15[1:0] OSPEEDR15[1:0] OSPEEDR15[1:0] OSPEEDR15[1:0]

0
0
0
1
0
0
0
BR14 Res. Res. Res. 30

0
0
0
1
0
0
0
BR13 Res. Res. Res. 29
PUPDR14[1:0] PUPDR14[1:0] PUPDR14[1:0] OSPEEDR14[1:0] OSPEEDR14[1:0] OSPEEDR14[1:0]

0
0
0
0
0
0
0
BR12 Res. Res. Res. 28

0
0
0
0
0
1
0
BR11 Res. Res. Res. 27
PUPDR13[1:0] PUPDR13[1:0] PUPDR13[1:0] OSPEEDR13[1:0] OSPEEDR13[1:0] OSPEEDR13[1:0]

0
0
0
1
0
1
0
BR10 Res. Res. Res. 26

0
0
0
0
0
0
0
BR9 Res. Res. Res. 25
PUPDR12[1:0] PUPDR12[1:0] PUPDR12[1:0] OSPEEDR12[1:0] OSPEEDR12[1:0] OSPEEDR12[1:0]

0
0
0
0
0
0
0
BR8 Res. Res. Res. 24

0
0
0
0
0
0
0

BR7 Res. Res. Res. 23

PUPDR11[1:0] PUPDR11[1:0] PUPDR11[1:0] OSPEEDR11[1:0] OSPEEDR11[1:0] OSPEEDR11[1:0]

0
0
0
0
0
0
0

BR6 Res. Res. Res. 22

0
0
0
0
0
0
0

BR5 Res. Res. Res. 21

PUPDR10[1:0] PUPDR10[1:0] PUPDR10[1:0] OSPEEDR10[1:0] OSPEEDR10[1:0] OSPEEDR10[1:0]

0
0
0
0
0
0
0

BR4 Res. Res. Res. 20

0
0
0
0
0
0
0

BR3 Res. Res. Res. 19

PUPDR9[1:0] PUPDR9[1:0] PUPDR9[1:0] OSPEEDR9[1:0] OSPEEDR9[1:0] OSPEEDR9[1:0]

0
0
0
0
0
0
0

BR2 Res. Res. Res. 18

RM0402 Rev 6
0
0
0
0
0
0
0

BR1 Res. Res. Res. 17

PUPDR8[1:0] PUPDR8[1:0] PUPDR8[1:0] OSPEEDR8[1:0] OSPEEDR8[1:0] OSPEEDR8[1:0]

0
0
0
0
0
0
0

BR0 Res. Res. Res. 16

0
0
0
0
0
0
0

0
0

BS15 ODR15 IDR15 OT15 15

PUPDR7[1:0] PUPDR7[1:0] PUPDR7[1:0] OSPEEDR7[1:0] OSPEEDR7[1:0] OSPEEDR7[1:0]

0
0
0
0
0
0
0

0
0

BS14 ODR14 IDR14 OT14 14

0
0
0
0
0
0
0

0
0

BS13 ODR13 IDR13 OT13 13

PUPDR6[1:0] PUPDR6[1:0] PUPDR6[1:0] OSPEEDR6[1:0] OSPEEDR6[1:0] OSPEEDR6[1:0]

0
0
0
0
0
0
0

0
0

BS12 ODR12 IDR12 OT12 12

0
0
0
0
0
0
0

0
0

BS11 ODR11 IDR11 OT11 11

PUPDR5[1:0] PUPDR5[1:0] PUPDR5[1:0] OSPEEDR5[1:0] OSPEEDR5[1:0] OSPEEDR5[1:0]

0
0
0
0
0
0
0

0
0

BS10 ODR10 IDR10 OT10 10

0
0
0
0
0
0
0

0
0

BS9 ODR9 IDR9 OT9 9

PUPDR4[1:0] PUPDR4[1:0] PUPDR4[1:0] OSPEEDR4[1:0] OSPEEDR4[1:0] OSPEEDR4[1:0]
Table 28. GPIO register map and reset values (continued)

0
0
1
0
0
0
0

0
0

BS8 ODR8 IDR8 OT8 8

0
0
0
0
1
0
0

0
0

BS7 ODR7 IDR7 OT7 7

PUPDR3[1:0] PUPDR3[1:0] PUPDR3[1:0] OSPEEDR3[1:0] OSPEEDR3[1:0] OSPEEDR3[1:0]

0
0
0
0
1
0
0

0
0

BS6 ODR6 IDR6 OT6 6

0
0
0
0
0
0
0

0
0

BS5 ODR5 IDR5 OT5 5

PUPDR2[1:0] PUPDR2[1:0] PUPDR2[1:0] OSPEEDR2[1:0] OSPEEDR2[1:0] OSPEEDR2[1:0]

0
0
0
0
0
0
0

0
0

BS4 ODR4 IDR4 OT4 4

0
0
0
0
0
0
0

0
0

BS3 ODR3 IDR3 OT3 3

PUPDR1[1:0] PUPDR1[1:0] PUPDR1[1:0] OSPEEDR1[1:0] OSPEEDR1[1:0] OSPEEDR1[1:0]

0
0
0
0
0
0
0

0
0

BS2 ODR2 IDR2 OT2 2

0
0
0
0
0
0
0

0
0

BS1 ODR1 IDR1 OT1 1

PUPDR0[1:0] PUPDR0[1:0] PUPDR0[1:0] OSPEEDR0[1:0] OSPEEDR0[1:0] OSPEEDR0[1:0]

0
0
0
0
0
0
0

0
0

BS0 ODR0 IDR0 OT0

185/1163
General-purpose I/Os (GPIO)

186
General-purpose I/Os (GPIO) RM0402

Table 28. GPIO register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
LCK15
LCK14
LCK13
LCK12

LCK10
LCK11
LCKK
GPIOx_LCKR

LCK9
LCK8
LCK7
LCK6
LCK5
LCK4
LCK3
LCK2
LCK1
LCK0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
0x1C (where x = A...H)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_AFRL
AFRL7[3:0] AFRL6[3:0] AFRL5[3:0] AFRL4[3:0] AFRL3[3:0] AFRL2[3:0] AFRL1[3:0] AFRL0[3:0]
0x20 (where x = A...H)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
GPIOx_AFRH
AFRH15[3:0] AFRH14[3:0] AFRH13[3:0] AFRH12[3:0] AFRH11[3:0] AFRH10[3:0] AFRH9[3:0] AFRH8[3:0]
0x24 (where x = A...H)
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

186/1163 RM0402 Rev 6

RM0402 System configuration controller (SYSCFG)

8 System configuration controller (SYSCFG)

The system configuration controller is mainly used to remap the memory accessible in the
code area and manage the external interrupt line connection to the GPIOs.

8.1 I/O compensation cell

By default the I/O compensation cell is not used. However when the I/O output buffer speed
is configured in 50 MHz or 100 MHz mode, it is recommended to use the compensation cell
for slew rate control on I/O tf(IO)out)/tr(IO)out commutation to reduce the I/O noise on power
supply.
When the compensation cell is enabled, a READY flag is set to indicate that the
compensation cell is ready and can be used. The I/O compensation cell can be used only
when the supply voltage ranges from 2.4 to 3.6 V.

8.2 SYSCFG registers

8.2.1 SYSCFG memory remap register (SYSCFG_MEMRMP)

This register is used for specific configurations on memory remap:
• Two bits are used to configure the type of memory accessible at address 0x0000 0000.
These bits are used to select the physical remap by software and so, bypass the BOOT
pins.
• After reset these bits take the value selected by the BOOT pins. When booting from
main Flash memory with BOOT0 pin set to 0, this register takes the value 0x00.
In remap mode, the CPU can access the external memory via ICode bus instead of System
bus which boosts up the performance.
Address offset: 0x00
Reset value: 0x0000 000X (X is the memory mode selected by the BOOT pins)
)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. MEM_MODE

rw rw

RM0402 Rev 6 187/1163

194
System configuration controller (SYSCFG) RM0402

Bits 31:2 Reserved, must be kept at reset value.

Bits 1:0 MEM_MODE: Memory mapping selection
Set and cleared by software. This bit controls the memory internal mapping at
address 0x0000 0000. After reset these bits take the value selected by the Boot
pins.
00: Main Flash memory mapped at 0x0000 0000
01: System Flash memory mapped at 0x0000 0000
10: reserved
11: Embedded SRAM mapped at 0x0000 0000
Note: Refer to Figure 2: Memory map for details about the memory mapping at
address 0x0000 0000.

8.2.2 SYSCFG peripheral mode configuration register (SYSCFG_PMC)

Address offset: 0x04
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ADC1D
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
C2

rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

Bits 31:17 Reserved, must be kept at reset value.

Bit 16 ADC1DC2:
0: No effect.
1: Refer to AN4073 on how to use this bit
Note: These bits can be set only if the following conditions are met:
- ADC clock higher or equal to 30 MHz.
- Only one ADC1DC2 bit must be selected if ADC conversions do not start
at the same time and the sampling times differ.
- These bits must not be set when the ADCDC1 bit is set in PWR_CR
register.
Bits 15:0 Reserved, must be kept at reset value.

188/1163 RM0402 Rev 6

RM0402 System configuration controller (SYSCFG)

8.2.3 SYSCFG external interrupt configuration register 1

(SYSCFG_EXTICR1)
Address offset: 0x08
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXTI3[3:0] EXTI2[3:0] EXTI1[3:0] EXTI0[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 0 to 3)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
0111: PH[x] pin (Reserved for EXTI3 and EXTI2 configurations)
Other configurations: reserved

8.2.4 SYSCFG external interrupt configuration register 2

(SYSCFG_EXTICR2)
Address offset: 0x0C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXTI7[3:0] EXTI6[3:0] EXTI5[3:0] EXTI4[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 189/1163

194
System configuration controller (SYSCFG) RM0402

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 4 to 7)
These bits are written by software to select the source input for the EXTIx
external interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
Other configurations: reserved

8.2.5 SYSCFG external interrupt configuration register 3

(SYSCFG_EXTICR3)
Address offset: 0x10
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXTI11[3:0] EXTI10[3:0] EXTI9[3:0] EXTI8[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 8 to 11)
These bits are written by software to select the source input for the EXTIx external
interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin
Other configurations: reserved

190/1163 RM0402 Rev 6

RM0402 System configuration controller (SYSCFG)

8.2.6 SYSCFG external interrupt configuration register 4

(SYSCFG_EXTICR4)
Address offset: 0x14
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EXTI15[3:0] EXTI14[3:0] EXTI13[3:0] EXTI12[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 EXTIx[3:0]: EXTI x configuration (x = 12 to 15)
These bits are written by software to select the source input for the EXTIx external
interrupt.
0000: PA[x] pin
0001: PB[x] pin
0010: PC[x] pin
0011: PD[x] pin
0100: PE[x] pin
0101: PF[x] pin
0110: PG[x] pin

8.2.7 SYSCFG configuration register 2 (SYSCFG_CFGR2)

Address offset: 0x1C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PVDL Res. CLL

rw rw

Bits 31:3 Reserved, must be kept at reset value.

RM0402 Rev 6 191/1163

194
System configuration controller (SYSCFG) RM0402

Bit 2 PVDL: PVD lock

This bit is set by software. It can be cleared only by a system reset. It enables and
locks the PVD connection to TIM1/8 Break input. It also locks (write protection) the
PVDE and PVDS[2:0] bits of PWR_CR register.
0: PVD interrupt not connected to TIM1/8 Break input. PVDE and PVDS[2:0] can be
read and modified
1: PVD interrupt connected to TIM1/8 Break input. PVDE and PVDS[2:0] are read-
only
Bit 1 Reserved, must be kept at reset value.
Bit 0 CLL: core lockup lock
This bit is set and cleared by software. It enables and locks the LOCKUP (Hardfault)
output of the Cortex®-M4 with FPU core with TIM1/8 Break input.
0: Cortex®-M4 with FPU LOCKUP output not connected to TIM1/8 Break input
1: Cortex®-M4 with FPU LOCKUP output connected to TIM1/8 Break input

8.2.8 Compensation cell control register (SYSCFG_CMPCR)

Address offset: 0x20
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. READY Res. Res. Res. Res. Res. Res. Res. CMP_PD

r rw

Bits 31:9 Reserved, must be kept at reset value.

Bit 8 READY: Compensation cell ready flag
0: I/O compensation cell not ready
1: O compensation cell ready
Bits 7:2 Reserved, must be kept at reset value.
Bit 0 CMP_PD: Compensation cell power-down
0: I/O compensation cell power-down mode
1: I/O compensation cell enabled

192/1163 RM0402 Rev 6

RM0402 System configuration controller (SYSCFG)

8.2.9 SYSCFG configuration register (SYSCFG_CFGR)

Address offset: 0x2C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. I2CFMP1_SDA I2CFMP1_SCL

rw rw

Bits 31:2 Reserved, must be kept at reset value.

Bit 1 I2CFMP1_SDA
Set and cleared by software. When this bit is set, it forces FM+ drive capability on
I2CFMP1_SDA pin selected through GPIO port mode register and GPIO alternate
function selection bits.
Bit 0 I2CFMP1_SCL
Set and cleared by software. When this bit is set, it forces FM+ drive capability on
I2CFMP1_SCL pin selected through GPIO port mode register and GPIO alternate
function selection bits.

RM0402 Rev 6 193/1163

194
 0x24
0x20
0x14
0x10
0x08
0x04
0x00

0x1C
0x0C
Offset
8.2.10

194/1163
MEMRMP
SYSCFG_
Register

Reset value
Reset value
Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
SYSCFG_PMC

SYSCFG_CFGR
SYSCFG_CFGR2

SYSCFG_CMPCR
SYSCFG_EXTICR4
SYSCFG_EXTICR3
SYSCFG_EXTICR2
SYSCFG_EXTICR1
Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. Res. Res. 28
Res. Res. Res. Res. Res. Res. Res. Res. Res. 27
Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. Res. Res.
SYSCFG register map

25
Res. Res. Res. Res. Res. Res. Res. Res. Res. 24
System configuration controller (SYSCFG)

Res. Res. Res. Res. Res. Res. Res. Res. Res. 23

Res. Res. Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. Res. Res. 21
Res. Res. Res. Res. Res. Res. Res. Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. Res. Res. Res. Res. Res. Res. Res.

RM0402 Rev 6
18
Res. Res. Res. Res. Res. Res. Res. Res. Res. 17
0

Res. Res. Res. Res. Res. Res. Res. ADC1DC2 Res. 16

0
0
0
0
Res. Res. Res. Res. Res. 15

0
0
0
0

Res. Res. Res. Res. Res. 14

0
0
0
0

Res. Res. Res. Res. Res. 13

EXTI7[3:0]
EXTI3[3:0]

EXTI11[3:0]

EXTI15[3:0]

0
0
0
0

Res. Res. Res. Res. Res. 12

0
0
0
0

Res. Res. Res. Res. Res. 11

Table 29. SYSCFG register map and reset values

0
0
0
0

Res. Res. Res. Res. Res. 10

0
0
0
0

Res. Res. Res. Res. Res. 9

Refer to Section 2.2 on page 49 for the register boundary addresses.

EXTI6[3:0]
EXTI2[3:0]

EXTI10[3:0]

EXTI14[3:0]

0
0
0
0

Res. Res. Res. 0 Res. Res. 8

0
0
0

0
READY
The following table gives the SYSCFG register map and the reset values.

Res. Res. Res. Res. 7

0
0
0
0

Res. Res. Res. Res. Res. 6

0
0
0
0

Res. Res. Res. Res. Res. 5

EXTI9[3:0]
EXTI5[3:0]
EXTI1[3:0]

EXTI13[3:0]

0
0
0
0

Res. Res. Res. Res. Res. 4

0
0
0
0

Res. Res. Res. Res. Res. 3

0
0
0
0

Res. Res. PVDL Res. Res. 2

x

0
0
0
0

0
I2CFMP1_SDA Res. Res. Res. 1
MEM_MODE
EXTI8[3:0]
EXTI4[3:0]
EXTI0[3:0]
x

EXTI12[3:0]

0
0
0
0

0
0
0

I2CFMP1_SCL CMP_PD CLL Res. 0

RM0402
RM0402 Direct memory access controller (DMA)

9 Direct memory access controller (DMA)

9.1 DMA introduction

Direct memory access (DMA) is used in order to provide high-speed data transfer between
peripherals and memory and between memory and memory. Data can be quickly moved by
DMA without any CPU action. This keeps CPU resources free for other operations.
The DMA controller combines a powerful dual AHB master bus architecture with
independent FIFO to optimize the bandwidth of the system, based on a complex bus matrix
architecture.
The two DMA controllers (DMA1, DMA2) have 8 streams each, dedicated to managing
memory access requests from one or more peripherals.
Each stream can have up to 8 channels (requests) in total.
Each DMA controller has an arbiter for handling the priority between DMA requests.

9.2 DMA main features

The main DMA features are:
• Dual AHB master bus architecture, one dedicated to memory accesses and one
dedicated to peripheral accesses
• AHB slave programming interface supporting only 32-bit accesses
• 8 streams for each DMA controller, up to 8 channels (requests) per stream
• Four-word depth 32 first-in, first-out memory buffers (FIFOs) per stream, that can be
used in FIFO mode or direct mode:
– FIFO mode: with threshold level software selectable between 1/4, 1/2 or 3/4 of the
FIFO size
– Direct mode: each DMA request immediately initiates a transfer from/to the
memory. When it is configured in direct mode (FIFO disabled), to transfer data in
memory-to-peripheral mode, the DMA preloads only one data from the memory to
the internal FIFO to ensure an immediate data transfer as soon as a DMA request
is triggered by a peripheral.
• Each stream can be configured to be:
– a regular channel that supports peripheral-to-memory, memory-to-peripheral and
memory-to-memory transfers
– a double buffer channel that also supports double buffering on the memory side
• Priorities between DMA stream requests are software-programmable (four levels
consisting of very high, high, medium, low) or hardware in case of equality (for
example, request 0 has priority over request 1)
• Each stream also supports software trigger for memory-to-memory transfers (only
available for the DMA2 controller)
• Each stream request can be selected among up to 8 possible channel requests. This
selection is software-configurable and allows several peripherals to initiate DMA
requests
• The number of data items to be transferred can be managed either by the DMA
controller or by the peripheral:

RM0402 Rev 6 195/1163

230
Direct memory access controller (DMA) RM0402

– DMA flow controller: the number of data items to be transferred is software-

programmable from 1 to 65535
– Peripheral flow controller: the number of data items to be transferred is unknown
and controlled by the source or the destination peripheral that signals the end of
the transfer by hardware
• Independent source and destination transfer width (byte, half-word, word): when the
data widths of the source and destination are not equal, the DMA automatically
packs/unpacks the necessary transfers to optimize the bandwidth. This feature is only
available in FIFO mode
• Incrementing or non-incrementing addressing for source and destination
• Supports incremental burst transfers of 4, 8 or 16 beats. The size of the burst is
software-configurable, usually equal to half the FIFO size of the peripheral
• Each stream supports circular buffer management
• 5 event flags (DMA half transfer, DMA transfer complete, DMA transfer error, DMA
FIFO error, direct mode error) logically ORed together in a single interrupt request for
each stream

196/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

9.3 DMA functional description

9.3.1 DMA block diagram

The figure below shows the block diagram of a DMA.

Figure 23. DMA block diagram

DMA controller

AHB master
REQ_STR0_CH0
REQ_STR0_CH1 Memory port

REQ_STR0_CH7

STREAM 0

STREAM 1

STREAM 2

STREAM 3

STREAM 4

STREAM 5

STREAM 6

STREAM 7
REQ_STR1_CH0
REQ_STR1_CH1
REQ_STREAM0
REQ_STREAM1
REQ_STR1_CH7 REQ_STREAM2
REQ_STREAM3
FIFO

FIFO

FIFO

FIFO

FIFO

FIFO

FIFO

FIFO
REQ_STREAM4 Arbiter
REQ_STREAM5
REQ_STREAM6
STREAM 0

STREAM 1

STREAM 2

STREAM 3

STREAM 4

STREAM 5

STREAM 6

STREAM 7
REQ_STREAM7

REQ_STR7_CH0
REQ_STR7_CH1

AHB master
REQ_STR7_CH7 Peripheral port

Channel
selection

AHB slave
programming Programming port
interface

ai15945b

9.3.2 DMA overview

The DMA controller performs direct memory transfer: as an AHB master, the DMA controller
can take the control of the AHB bus matrix to initiate AHB transactions.
The DMA controller carries out the following transactions:
• peripheral-to-memory
• memory-to-peripheral
• memory-to-memory
The DMA controller provides two AHB master ports: the AHB memory port, intended to be
connected to memories and the AHB peripheral port, intended to be connected to
peripherals. However, to allow memory-to-memory transfers, the AHB peripheral port must
also have access to the memories.
The AHB slave port is used to program the DMA controller (it supports only 32-bit
accesses).

RM0402 Rev 6 197/1163

230
Direct memory access controller (DMA) RM0402

9.3.3 DMA transactions

A DMA transaction consists of a sequence of a given number of data transfers. The number
of data items to be transferred and their width (8-bit, 16-bit or 32-bit) are software-
programmable.
Each DMA transfer consists of three operations:
• a loading from the peripheral data register or a location in memory, addressed through
the DMA_SxPAR or DMA_SxM0AR register
• a storage of the data loaded to the peripheral data register or a location in memory
addressed through the DMA_SxPAR or DMA_SxM0AR register
• a post-decrement of the DMA_SxNDTR register, containing the number of transactions
that still have to be performed
After an event, the peripheral sends a request signal to the DMA controller. The DMA
controller serves the request depending on the channel priorities. As soon as the DMA
controller accesses the peripheral, an Acknowledge signal is sent to the peripheral by the
DMA controller. The peripheral releases its request as soon as it gets the Acknowledge
signal from the DMA controller. Once the request has been deasserted by the peripheral,
the DMA controller releases the Acknowledge signal. If there are more requests, the
peripheral can initiate the next transaction.

9.3.4 Channel selection

Each stream is associated with a DMA request that can be selected out of 8 possible
channel requests. The selection is controlled by the CHSEL[2:0] bits in the DMA_SxCR
register.
Caution: A same peripheral request can be assigned to two different channels only if the application
ensures that these channels are not requested to be served at the same time. In other
words, if two different channels receive a same asserted peripheral request at the same
time, an unpredictable DMA hardware behavior occurs.

Figure 24. Channel selection

REQ_STRx_CH7

REQ_STRx_CH6

REQ_STRx_CH5
REQ_STREAMx
REQ_STRx_CH4

REQ_STRx_CH3

REQ_STRx_CH2
REQ_STRx_CH1

REQ_STRx_CH0

31 27 25 0

DMA_SxCR CHSEL[2:0]

ai15947b

198/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

The 8 requests from the peripherals (such as TIM, ADC, SPI, I2C) are independently
connected to each channel and their connection depends on the product implementation.
Table 30 and Table 31 give examples of DMA request mappings.

Table 30. DMA1 request mapping

Peripheral
Stream 0 Stream 1 Stream 2 Stream 3 Stream 4 Stream 5 Stream 6 Stream 7
requests
Channel 0 SPI3_RX I2C1_TX SPI3_RX SPI2_RX SPI2_TX SPI3_TX - SPI3_TX

Channel 1 I2C1_RX I2C3_RX TIM7_UP I2CFMP1_RX TIM7_UP I2C1_RX I2C1_TX I2C1_TX

Channel 2 TIM4_CH1 I2CFMP1_TX I2S3_EXT_RX TIM4_CH2 I2S2_EXT_TX I2S3_EXT_TX TIM4_UP TIM4_CH3

TIM2_UP TIM2_CH2 TIM2_UP

Channel 3 I2S3_EXT_RX I2C3_RX I2S2EXT_RX I2C3_TX TIM2_CH1
TIM2_CH3 TIM2_CH4 TIM2_CH4

Channel 4 - USART3_RX - USART3_TX - USART2_RX USART2_TX I2CFMP1_TX

TIM3_UP TIM3_CH1
Channel 5 - - - TIM3_CH2 - TIM3_CH3
TIM3_CH4 TIM3_TRIG

TIM5_CH3 TIM5_CH4 TIM5_CH4

Channel 6 TIM5_CH1 TIM5_CH2 I2C3_TX TIM5_UP USART2_RX
TIM5_UP TIM5_TRIG TIM5_TRIG

Channel 7 I2CFMP1_RX TIM6_UP I2C2_RX I2C2_RX USART3_TX - - I2C2_TX

Table 31. DMA2 request mapping

Peripheral
Stream 0 Stream 1 Stream 2 Stream 3 Stream 4 Stream 5 Stream 6 Stream 7
requests

TIM8_CH1 TIM1_CH1
Channel 0 ADC1 - TIM8_CH2 - ADC1 - TIM1_CH2 -
TIM8_CH3 TIM1_CH3

Channel 1 - - - - - - - -

Channel 2 - - SPI1_TX SPI5_RX SPI5_TX - - -

DFSDM1_ DFSDM1_ DFSDM1_

Channel 3 SPI1_RX SPI1_RX SPI1_TX SPI1_TX QUADSPI
FLT1 FLT1 FLT0

Channel 4 SPI4_RX SPI4_TX USART1_RX SDIO SPI4_RX USART1_RX SDIO USART1_TX

Channel 5 - USART6_RX USART6_RX SPI4_RX SPI4_TX SPI5_TX USART6_TX USART6_TX

TIM1_CH4
Channel 6 TIM1_TRIG TIM1_CH1 TIM1_CH2 TIM1_CH1 TIM1_TRIG TIM1_UP TIM1_CH3 -
TIM1_COM

TIM8_CH4
DFSDM1_
Channel 7 TIM8_UP TIM8_CH1 TIM8_CH2 TIM8_CH3 SPI5_RX SPI5_TX TIM8_TRIG
FLT0
TIM8_COM

RM0402 Rev 6 199/1163

230
Direct memory access controller (DMA) RM0402

9.3.5 Arbiter
An arbiter manages the 8 DMA stream requests based on their priority for each of the two
AHB master ports (memory and peripheral ports) and launches the peripheral/memory
access sequences.
Priorities are managed in two stages:
• Software: each stream priority can be configured in the DMA_SxCR register. There are
four levels:
– Very high priority
– High priority
– Medium priority
– Low priority
• Hardware: If two requests have the same software priority level, the stream with the
lower number takes priority over the stream with the higher number. For example,
stream 2 takes priority over stream 4.

9.3.6 DMA streams

Each of the eight DMA controller streams provides a unidirectional transfer link between a
source and a destination.
Each stream can be configured to perform:
• Regular type transactions: memory-to-peripherals, peripherals-to-memory or memory-
to-memory transfers
• Double-buffer type transactions: double buffer transfers using two memory pointers for
the memory (while the DMA is reading/writing from/to a buffer, the application can
write/read to/from the other buffer).
The amount of data to be transferred (up to 65535) is programmable and related to the
source width of the peripheral that requests the DMA transfer connected to the peripheral
AHB port. The register that contains the amount of data items to be transferred is
decremented after each transaction.

9.3.7 Source, destination and transfer modes

Both source and destination transfers can address peripherals and memories in the entire
4-Gbyte area, at addresses comprised between 0x0000 0000 and 0xFFFF FFFF.
The direction is configured using the DIR[1:0] bits in the DMA_SxCR register and offers
three possibilities: memory-to-peripheral, peripheral-to-memory or memory-to-memory
transfers.
The table below describes the corresponding source and destination addresses.

Table 32. Source and destination address

Bits DIR[1:0] of the
Direction Source address Destination address
DMA_SxCR register

00 Peripheral-to-memory DMA_SxPAR DMA_SxM0AR

01 Memory-to-peripheral DMA_SxM0AR DMA_SxPAR

200/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Table 32. Source and destination address (continued)

Bits DIR[1:0] of the
Direction Source address Destination address
DMA_SxCR register

10 Memory-to-memory DMA_SxPAR DMA_SxM0AR

11 Reserved - -

When the data width (programmed in the PSIZE or MSIZE bits in the DMA_SxCR register)
is a half-word or a word, respectively, the peripheral or memory address written into the
DMA_SxPAR or DMA_SxM0AR/M1AR registers has to be aligned on a word or half-word
address boundary, respectively.

Peripheral-to-memory mode
Figure 25 describes this mode.
When this mode is enabled (by setting the bit EN in the DMA_SxCR register), each time a
peripheral request occurs, the stream initiates a transfer from the source to fill the FIFO.
When the threshold level of the FIFO is reached, the contents of the FIFO are drained and
stored into the destination.
The transfer stops once the DMA_SxNDTR register reaches zero, when the peripheral
requests the end of transfers (in case of a peripheral flow controller) or when the EN bit in
the DMA_SxCR register is cleared by software.
In direct mode (when the DMDIS value in the DMA_SxFCR register is 0), the threshold level
of the FIFO is not used: after each single data transfer from the peripheral to the FIFO, the
corresponding data are immediately drained and stored into the destination.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.

RM0402 Rev 6 201/1163

230
Direct memory access controller (DMA) RM0402

Figure 25. Peripheral-to-memory mode

DMA controller DMA_SxM0AR

DMA_SxM1AR(1)

AHB memory Memory bus

port

Memory
destination
FIFO
Arbiter level
REQ_STREAMx FIFO

AHB peripheral Peripheral bus

port

peripheral
DMA_SxPAR source

Peripheral DMA request

ai15948

1. For double-buffer mode.

Memory-to-peripheral mode
Figure 26 describes this mode.
When this mode is enabled (by setting the EN bit in the DMA_SxCR register), the stream
immediately initiates transfers from the source to entirely fill the FIFO.
Each time a peripheral request occurs, the contents of the FIFO are drained and stored into
the destination. When the level of the FIFO is lower than or equal to the predefined
threshold level, the FIFO is fully reloaded with data from the memory.
The transfer stops once the DMA_SxNDTR register reaches zero, when the peripheral
requests the end of transfers (in case of a peripheral flow controller) or when the EN bit in
the DMA_SxCR register is cleared by software.
In direct mode (when the DMDIS value in the DMA_SxFCR register is 0), the threshold level
of the FIFO is not used. Once the stream is enabled, the DMA preloads the first data to
transfer into an internal FIFO. As soon as the peripheral requests a data transfer, the DMA
transfers the preloaded value into the configured destination. It then reloads again the
empty internal FIFO with the next data to be transfer. The preloaded data size corresponds
to the value of the PSIZE bitfield in the DMA_SxCR register.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.

202/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Figure 26. Memory-to-peripheral mode

DMA controller DMA_SxM0AR

DMA_SxM1AR(1)

Memory bus
AHB memory
port
Memory
source

FIFO
Arbiter
REQ_STREAMx level FIFO

AHB peripheral Peripheral bus

port

Peripheral
DMA_SxPAR destination

Peripheral DMA request

ai15949

1. For double-buffer mode.

Memory-to-memory mode
The DMA channels can also work without being triggered by a request from a peripheral.
This is the memory-to-memory mode, described in Figure 27.
When the stream is enabled by setting the Enable bit (EN) in the DMA_SxCR register, the
stream immediately starts to fill the FIFO up to the threshold level. When the threshold level
is reached, the FIFO contents are drained and stored into the destination.
The transfer stops once the DMA_SxNDTR register reaches zero or when the EN bit in the
DMA_SxCR register is cleared by software.
The stream has access to the AHB source or destination port only if the arbitration of the
corresponding stream is won. This arbitration is performed using the priority defined for
each stream using the PL[1:0] bits in the DMA_SxCR register.
Note: When memory-to-memory mode is used, the circular and direct modes are not allowed.
Only the DMA2 controller is able to perform memory-to-memory transfers.

RM0402 Rev 6 203/1163

230
Direct memory access controller (DMA) RM0402

Figure 27. Memory-to-memory mode

DMA controller DMA_SxM0AR

DMA_SxM1AR(1)

AHB memory Memory bus

port

Memory 2
destination
Arbiter FIFO
Stream enable level FIFO

AHB peripheral Peripheral bus

port

Memory 1
DMA_SxPAR source

ai15950

1. For double-buffer mode.

9.3.8 Pointer incrementation

Peripheral and memory pointers can optionally be automatically post-incremented or kept
constant after each transfer depending on the PINC and MINC bits in the DMA_SxCR
register.
Disabling the increment mode is useful when the peripheral source or destination data is
accessed through a single register.
If the increment mode is enabled, the address of the next transfer is the address of the
previous one incremented by 1 (for bytes), 2 (for half-words) or 4 (for words) depending on
the data width programmed in the PSIZE or MSIZE bits in the DMA_SxCR register.
In order to optimize the packing operation, it is possible to fix the increment offset size for
the peripheral address whatever the size of the data transferred on the AHB peripheral port.
The PINCOS bit in the DMA_SxCR register is used to align the increment offset size with
the data size on the peripheral AHB port, or on a 32-bit address (the address is then
incremented by 4). The PINCOS bit has an impact on the AHB peripheral port only.
If the PINCOS bit is set, the address of the following transfer is the address of the previous
one incremented by 4 (automatically aligned on a 32-bit address), whatever the PSIZE
value. The AHB memory port, however, is not impacted by this operation.

204/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

9.3.9 Circular mode

The circular mode is available to handle circular buffers and continuous data flows (e.g.
ADC scan mode). This feature can be enabled using the CIRC bit in the DMA_SxCR
register.
When the circular mode is activated, the number of data items to be transferred is
automatically reloaded with the initial value programmed during the stream configuration
phase, and the DMA requests continue to be served.
Note: In the circular mode, it is mandatory to respect the following rule in case of a burst mode
configured for memory:
DMA_SxNDTR = Multiple of ((Mburst beat) × (Msize)/(Psize)), where:
– (Mburst beat) = 4, 8 or 16 (depending on the MBURST bits in the DMA_SxCR
register)
– ((Msize)/(Psize)) = 1, 2, 4, 1/2 or 1/4 (Msize and Psize represent the MSIZE and
PSIZE bits in the DMA_SxCR register. They are byte dependent)
– DMA_SxNDTR = Number of data items to transfer on the AHB peripheral port
For example: Mburst beat = 8 (INCR8), MSIZE = 00 (byte) and PSIZE = 01 (half-word), in
this case: DMA_SxNDTR must be a multiple of (8 × 1/2 = 4).
If this formula is not respected, the DMA behavior and data integrity are not guaranteed.
NDTR must also be a multiple of the Peripheral burst size multiplied by the peripheral data
size, otherwise this could result in a bad DMA behavior.

9.3.10 Double-buffer mode

This mode is available for all the DMA1 and DMA2 streams.
The double-buffer mode is enabled by setting the DBM bit in the DMA_SxCR register.
A double-buffer stream works as a regular (single buffer) stream with the difference that it
has two memory pointers. When the double-buffer mode is enabled, the circular mode is
automatically enabled (CIRC bit in DMA_SxCR is not relevant) and at each end of
transaction, the memory pointers are swapped.
In this mode, the DMA controller swaps from one memory target to another at each end of
transaction. This allows the software to process one memory area while the second memory
area is being filled/used by the DMA transfer. The double-buffer stream can work in both
directions (the memory can be either the source or the destination) as described in
Table 33: Source and destination address registers in double-buffer mode (DBM = 1).
Note: In double-buffer mode, it is possible to update the base address for the AHB memory port
on-the-fly (DMA_SxM0AR or DMA_SxM1AR) when the stream is enabled, by respecting the
following conditions:
• When the CT bit is 0 in the DMA_SxCR register, the DMA_SxM1AR register can be
written. Attempting to write to this register while CT = 1 sets an error flag (TEIF) and the
stream is automatically disabled.
• When the CT bit is 1 in the DMA_SxCR register, the DMA_SxM0AR register can be
written. Attempting to write to this register while CT = 0, sets an error flag (TEIF) and
the stream is automatically disabled.
To avoid any error condition, it is advised to change the base address as soon as the TCIF
flag is asserted because, at this point, the targeted memory must have changed from

RM0402 Rev 6 205/1163

230
Direct memory access controller (DMA) RM0402

memory 0 to 1 (or from 1 to 0) depending on the value of CT in the DMA_SxCR register in

accordance with one of the two above conditions.
For all the other modes (except the double-buffer mode), the memory address registers are
write-protected as soon as the stream is enabled.

Table 33. Source and destination address registers in double-buffer mode (DBM = 1)
Bits DIR[1:0] of the
Direction Source address Destination address
DMA_SxCR register

00 Peripheral-to-memory DMA_SxPAR DMA_SxM0AR / DMA_SxM1AR

01 Memory-to-peripheral DMA_SxM0AR / DMA_SxM1AR DMA_SxPAR
10 Not allowed(1)
11 Reserved - -
1. When the double-buffer mode is enabled, the circular mode is automatically enabled. Since the memory-to-memory mode
is not compatible with the circular mode, when the double-buffer mode is enabled, it is not allowed to configure the
memory-to-memory mode.

9.3.11 Programmable data width, packing/unpacking, endianness

The number of data items to be transferred has to be programmed into DMA_SxNDTR
(number of data items to transfer bit, NDT) before enabling the stream (except when the
flow controller is the peripheral, PFCTRL bit in DMA_SxCR is set).
When using the internal FIFO, the data widths of the source and destination data are
programmable through the PSIZE and MSIZE bits in the DMA_SxCR register (can be 8-,
16- or 32-bit).
When PSIZE and MSIZE are not equal:
• The data width of the number of data items to transfer, configured in the DMA_SxNDTR
register is equal to the width of the peripheral bus (configured by the PSIZE bits in the
DMA_SxCR register). For instance, in case of peripheral-to-memory, memory-to-
peripheral or memory-to-memory transfers and if the PSIZE[1:0] bits are configured for
half-word, the number of bytes to be transferred is equal to 2 × NDT.
• The DMA controller only copes with little-endian addressing for both source and
destination. This is described in Table 34: Packing/unpacking and endian behavior (bit
PINC = MINC = 1).
This packing/unpacking procedure may present a risk of data corruption when the operation
is interrupted before the data are completely packed/unpacked. So, to ensure data
coherence, the stream may be configured to generate burst transfers: in this case, each
group of transfers belonging to a burst are indivisible (refer to Section 9.3.12: Single and
burst transfers).
In direct mode (DMDIS = 0 in the DMA_SxFCR register), the packing/unpacking of data is
not possible. In this case, it is not allowed to have different source and destination transfer
data widths: both are equal and defined by the PSIZE bits in the DMA_SxCR register.
MSIZE bits are not relevant.

206/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Table 34. Packing/unpacking and endian behavior (bit PINC = MINC = 1)

Number - Peripheral port address / byte lane
AHB AHB
of data Memory Memory port Peripheral
memory peripheral
items to transfer address / byte transfer
port port
transfer - number lane number PINCOS = 1 PINCOS = 0
width width
(NDT)

1 0x0 / B0[7:0] 1 0x0 / B0[7:0] 0x0 / B0[7:0]

2 0x1 / B1[7:0] 2 0x4 / B1[7:0] 0x1 / B1[7:0]
8 8 4 - 3 0x2 / B2[7:0] 3 0x8 / B2[7:0] 0x2 / B2[7:0]
4 0x3 / B3[7:0] 4 0xC / B3[7:0] 0x3 / B3[7:0]

1 0x0 / B0[7:0] 1 0x0 / B1|B0[15:0] 0x0 / B1|B0[15:0]

2 0x1 / B1[7:0] 2 0x4 / B3|B2[15:0] 0x2 / B3|B2[15:0]
8 16 2 - 3 0x2 / B2[7:0]
4 0x3 / B3[7:0]

1 0x0 / B0[7:0] 1 0x0 / 0x0 /

2 0x1 / B1[7:0] B3|B2|B1|B0[31:0] B3|B2|B1|B0[31:0]
8 32 1 - 3 0x2 / B2[7:0]
4 0x3 / B3[7:0]

1 0x0 / B1|B0[15:0] 1 0x0 / B0[7:0] 0x0 / B0[7:0]

2 0x2 / B3|B2[15:0] 2 0x4 / B1[7:0] 0x1 / B1[7:0]
16 8 4 - 3 0x8 / B2[7:0] 0x2 / B2[7:0]
4 0xC / B3[7:0] 0x3 / B3[7:0]

1 0x0 / B1|B0[15:0] 1 0x0 / B1|B0[15:0] 0x0 / B1|B0[15:0]

16 16 2 - 2 0x2 / B1|B0[15:0] 2 0x4 / B3|B2[15:0] 0x2 / B3|B2[15:0]

1 0x0 / B1|B0[15:0] 1 0x0 / 0x0 /

16 32 1 - 2 0x2 / B3|B2[15:0] B3|B2|B1|B0[31:0] B3|B2|B1|B0[31:0]

1 0x0 / B3|B2|B1|B0[31:0] 1 0x0 / B0[7:0] 0x0 / B0[7:0]

2 0x4 / B1[7:0] 0x1 / B1[7:0]
32 8 4 - 3 0x8 / B2[7:0] 0x2 / B2[7:0]
4 0xC / B3[7:0] 0x3 / B3[7:0]

1 0x0 /B3|B2|B1|B0[31:0] 1 0x0 / B1|B0[15:0] 0x0 / B1|B0[15:0]

32 16 2 - 2 0x4 / B3|B2[15:0] 0x2 / B3|B2[15:0]

1 0x0 /B3|B2|B1|B0 [31:0] 1 0x0 / 0x0 /

32 32 1 - B3|B2|B1|B0 [31:0] B3|B2|B1|B0[31:0]

Note: Peripheral port may be the source or the destination (it can also be the memory source in
the case of memory-to-memory transfer).
PSIZE, MSIZE and NDT[15:0] must be configured so as to ensure that the last transfer is
not incomplete. This can occur when the data width of the peripheral port (PSIZE bits) is
lower than the data width of the memory port (MSIZE bits). This constraint is summarized in
the table below.

Table 35. Restriction on NDT versus PSIZE and MSIZE

PSIZE[1:0] of DMA_SxCR MSIZE[1:0] of DMA_SxCR NDT[15:0] of DMA_SxNDTR

00 (8-bit) 01 (16-bit) Must be a multiple of 2.

00 (8-bit) 10 (32-bit) Must be a multiple of 4.
01 (16-bit) 10 (32-bit) Must be a multiple of 2.

9.3.12 Single and burst transfers

The DMA controller can generate single transfers or incremental burst transfers of 4, 8 or
16 beats.

RM0402 Rev 6 207/1163

230
Direct memory access controller (DMA) RM0402

The size of the burst is configured by software independently for the two AHB ports by using
the MBURST[1:0] and PBURST[1:0] bits in the DMA_SxCR register.
The burst size indicates the number of beats in the burst, not the number of bytes
transferred.
To ensure data coherence, each group of transfers that form a burst are indivisible: AHB
transfers are locked and the arbiter of the AHB bus matrix does not degrant the DMA master
during the sequence of the burst transfer.
Depending on the single or burst configuration, each DMA request initiates a different
number of transfers on the AHB peripheral port:
• When the AHB peripheral port is configured for single transfers, each DMA request
generates a data transfer of a byte, half-word or word depending on the PSIZE[1:0] bits
in the DMA_SxCR register
• When the AHB peripheral port is configured for burst transfers, each DMA request
generates 4,8 or 16 beats of byte, half word or word transfers depending on the
PBURST[1:0] and PSIZE[1:0] bits in the DMA_SxCR register.
The same as above has to be considered for the AHB memory port considering the
MBURST and MSIZE bits.
In direct mode, the stream can only generate single transfers and the MBURST[1:0] and
PBURST[1:0] bits are forced by hardware.
The address pointers (DMA_SxPAR or DMA_SxM0AR registers) must be chosen so as to
ensure that all transfers within a burst block are aligned on the address boundary equal to
the size of the transfer.
The burst configuration has to be selected in order to respect the AHB protocol, where
bursts must not cross the 1 Kbyte address boundary because the minimum address space
that can be allocated to a single slave is 1 Kbyte. This means that the 1-Kbyte address
boundary must not be crossed by a burst block transfer, otherwise an AHB error is
generated, that is not reported by the DMA registers.

9.3.13 FIFO
FIFO structure
The FIFO is used to temporarily store data coming from the source before transmitting them
to the destination.
Each stream has an independent 4-word FIFO and the threshold level is software-
configurable between 1/4, 1/2, 3/4 or full.
To enable the use of the FIFO threshold level, the direct mode must be disabled by setting
the DMDIS bit in the DMA_SxFCR register.

208/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

The structure of the FIFO differs depending on the source and destination data widths, and
is described in the figure below.

Figure 28. FIFO structure

4 words

Empty 1/4 1/2 3/4 Full

byte lane 3 B15 B 11 B7 B3
Source: byte Destination: word
byte lane 2 B14 B10 B6 B2
B15 B14 B13 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B0 W3, W2, W1, W0
byte lane 1 B13 B9 B5 B1

byte lane 0 W3 B12 W2 B8 W1 B4 W0 B0

4 words

Empty 1/4 1/2 3/4 Full

byte lane 3 B15 B 11 B7 B3
Source: byte Destination: half-word
byte lane 2 H7 B14 H5 B10 H3 B6 H1 B2
B15 B14 B13 B12 B11 B10 B9 B8 B7 B6 B5 B4 B3 B2 B1 B0 H7, H6, H5, H4, H3, H2, H1, H0
byte lane 1 B13 B9 B5 B1

byte lane 0 H6 B12 H4 B8 H2 B4 H0 B0

4 words

Empty 1/4 1/2 3/4 Full

byte lane 3
Source: half-word H7 H5 H3 H1 Destination: word
byte lane 2
H7 H6 H5 H4 H3 H2 H1 H0 W3, W2, W1, W0
byte lane 1
H6 H4 H2 H0
byte lane 0 W3 W2 W1 W0

4-words

Empty 1/4 1/2 3/4 Full

byte lane 3 B15 B 11 B7 B3
Source: half-word Destination: byte
byte lane 2 H7 B14 H5 B10 H3 B6 H1 B2
H7 H6 H5 H4 H3 H2 H1 H0 B15 B14 B13 B12 B11 B10 B9 B8
byte lane 1 B13 B9 B5 B1 B7 B6 B5 B4 B3 B2 B1 B0

byte lane 0 H6 B12 H4 B8 H2 B4 H0 B0

ai15951

RM0402 Rev 6 209/1163

230
Direct memory access controller (DMA) RM0402

FIFO threshold and burst configuration

Caution is required when choosing the FIFO threshold (bits FTH[1:0] of the DMA_SxFCR
register) and the size of the memory burst (MBURST[1:0] of the DMA_SxCR register): The
content pointed by the FIFO threshold must exactly match an integer number of memory
burst transfers. If this is not in the case, a FIFO error (flag FEIFx of the DMA_HISR or
DMA_LISR register) is generated when the stream is enabled, then the stream is
automatically disabled. The allowed and forbidden configurations are described in the table
below. The forbidden configurations are highlighted in gray in the table.

Table 36. FIFO threshold configurations

MSIZE FIFO level MBURST = INCR4 MBURST = INCR8 MBURST = INCR16

1/4 1 burst of 4 beats Forbidden

1/2 2 bursts of 4 beats 1 burst of 8 beats Forbidden
Byte
3/4 3 bursts of 4 beats Forbidden
Full 4 bursts of 4 beats 2 bursts of 8 beats 1 burst of 16 beats
1/4 Forbidden
1/2 1 burst of 4 beats Forbidden
Half-word
3/4 Forbidden
Full 2 bursts of 4 beats 1 burst of 8 beats
Forbidden
1/4
1/2 Forbidden
Word Forbidden
3/4
Full 1 burst of 4 beats

In all cases, the burst size multiplied by the data size must not exceed the FIFO size (data
size can be: 1 (byte), 2 (half-word) or 4 (word)).
Incomplete burst transfer at the end of a DMA transfer may happen if one of the following
conditions occurs:
• For the AHB peripheral port configuration: the total number of data items (set in the
DMA_SxNDTR register) is not a multiple of the burst size multiplied by the data size.
• For the AHB memory port configuration: the number of remaining data items in the
FIFO to be transferred to the memory is not a multiple of the burst size multiplied by the
data size.
In such cases, the remaining data to be transferred is managed in single mode by the DMA,
even if a burst transaction is requested during the DMA stream configuration.
Note: When burst transfers are requested on the peripheral AHB port and the FIFO is used
(DMDIS = 1 in the DMA_SxCR register), it is mandatory to respect the following rule to
avoid permanent underrun or overrun conditions, depending on the DMA stream direction:
If (PBURST × PSIZE) = FIFO_SIZE (4 words), FIFO_Threshold = 3/4 is forbidden with
PSIZE = 1, 2 or 4 and PBURST = 4, 8 or 16.
This rule ensures that enough FIFO space at a time is free to serve the request from the
peripheral.

210/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

FIFO flush
The FIFO can be flushed when the stream is disabled by resetting the EN bit in the
DMA_SxCR register and when the stream is configured to manage peripheral-to-memory or
memory-to-memory transfers. If some data are still present in the FIFO when the stream is
disabled, the DMA controller continues transferring the remaining data to the destination
(even though stream is effectively disabled). When this flush is completed, the transfer
complete status bit (TCIFx) in the DMA_LISR or DMA_HISR register is set.
The remaining data counter DMA_SxNDTR keeps the value in this case to indicate how
many data items are currently available in the destination memory.
Note that during the FIFO flush operation, if the number of remaining data items in the FIFO
to be transferred to memory (in bytes) is less than the memory data width (for example 2
bytes in FIFO while MSIZE is configured to word), data is sent with the data width set in the
MSIZE bit in the DMA_SxCR register. This means that memory is written with an undesired
value. The software may read the DMA_SxNDTR register to determine the memory area
that contains the good data (start address and last address).
If the number of remaining data items in the FIFO is lower than a burst size (if the MBURST
bits in DMA_SxCR register are set to configure the stream to manage burst on the AHB
memory port), single transactions are generated to complete the FIFO flush.

Direct mode
By default, the FIFO operates in direct mode (DMDIS bit in the DMA_SxFCR is reset) and
the FIFO threshold level is not used. This mode is useful when the system requires an
immediate and single transfer to or from the memory after each DMA request.
When the DMA is configured in direct mode (FIFO disabled), to transfer data in memory-to-
peripheral mode, the DMA preloads one data from the memory to the internal FIFO to
ensure an immediate data transfer as soon as a DMA request is triggered by a peripheral.
To avoid saturating the FIFO, it is recommended to configure the corresponding stream with
a high priority.
This mode is restricted to transfers where:
• the source and destination transfer widths are equal and both defined by the
PSIZE[1:0] bits in DMA_SxCR (MSIZE[1:0] bits are not relevant)
• burst transfers are not possible (PBURST[1:0] and MBURST[1:0] bits in DMA_SxCR
are don’t care)
Direct mode must not be used when implementing memory-to-memory transfers.

9.3.14 DMA transfer completion

Different events can generate an end of transfer by setting the TCIFx bit in the DMA_LISR
or DMA_HISR status register:
• In DMA flow controller mode:
– The DMA_SxNDTR counter has reached zero in the memory-to-peripheral mode.
– The stream is disabled before the end of transfer (by clearing the EN bit in the
DMA_SxCR register) and (when transfers are peripheral-to-memory or memory-

RM0402 Rev 6 211/1163

230
Direct memory access controller (DMA) RM0402

to-memory) all the remaining data have been flushed from the FIFO into the
memory.
• In Peripheral flow controller mode:
– The last external burst or single request has been generated from the peripheral
and (when the DMA is operating in peripheral-to-memory mode) the remaining
data have been transferred from the FIFO into the memory
– The stream is disabled by software, and (when the DMA is operating in peripheral-
to-memory mode) the remaining data have been transferred from the FIFO into
the memory
Note: The transfer completion is dependent on the remaining data in FIFO to be transferred into
memory only in the case of peripheral-to-memory mode. This condition is not applicable in
memory-to-peripheral mode.
If the stream is configured in non-circular mode, after the end of the transfer (that is when
the number of data to be transferred reaches zero), the DMA is stopped (EN bit in
DMA_SxCR register is cleared by Hardware) and no DMA request is served unless the
software reprograms the stream and re-enables it (by setting the EN bit in the DMA_SxCR
register).

9.3.15 DMA transfer suspension

At any time, a DMA transfer can be suspended to be restarted later on or to be definitively
disabled before the end of the DMA transfer.
There are two cases:
• The stream disables the transfer with no later-on restart from the point where it was
stopped. There is no particular action to do, except to clear the EN bit in the
DMA_SxCR register to disable the stream. The stream may take time to be disabled
(ongoing transfer is completed first). The transfer complete interrupt flag (TCIF in the
DMA_LISR or DMA_HISR register) is set in order to indicate the end of transfer. The
value of the EN bit in DMA_SxCR is now 0 to confirm the stream interruption. The
DMA_SxNDTR register contains the number of remaining data items at the moment
when the stream was stopped so that the software can determine how many data items
have been transferred before the stream was interrupted.
• The stream suspends the transfer before the number of remaining data items to be
transferred in the DMA_SxNDTR register reaches 0. The aim is to restart the transfer
later by re-enabling the stream. In order to restart from the point where the transfer was
stopped, the software has to read the DMA_SxNDTR register after disabling the stream
by writing the EN bit in DMA_SxCR register (and then checking that it is at 0) to know
the number of data items already collected. Then:
– The peripheral and/or memory addresses have to be updated in order to adjust
the address pointers
– The SxNDTR register has to be updated with the remaining number of data items
to be transferred (the value read when the stream was disabled)
– The stream may then be re-enabled to restart the transfer from the point it was
stopped
Note: A transfer complete interrupt flag (TCIF in DMA_LISR or DMA_HISR) is set to indicate the
end of transfer due to the stream interruption.

212/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

9.3.16 Flow controller

The entity that controls the number of data to be transferred is known as the flow controller.
This flow controller is configured independently for each stream using the PFCTRL bit in the
DMA_SxCR register.
The flow controller can be:
• The DMA controller: in this case, the number of data items to be transferred is
programmed by software into the DMA_SxNDTR register before the DMA stream is
enabled.
• The peripheral source or destination: this is the case when the number of data items to
be transferred is unknown. The peripheral indicates by hardware to the DMA controller
when the last data are being transferred. This feature is only supported for peripherals
that are able to signal the end of the transfer.
When the peripheral flow controller is used for a given stream, the value written into the
DMA_SxNDTR has no effect on the DMA transfer. Actually, whatever the value written, it is
forced by hardware to 0xFFFF as soon as the stream is enabled, to respect the following
schemes:
• Anticipated stream interruption: EN bit in DMA_SxCR register is reset to 0 by the
software to stop the stream before the last data hardware signal (single or burst) is sent
by the peripheral. In such a case, the stream is switched off and the FIFO flush is
triggered in the case of a peripheral-to-memory DMA transfer. The TCIFx flag of the
corresponding stream is set in the status register to indicate the DMA completion. To
know the number of data items transferred during the DMA transfer, read the
DMA_SxNDTR register and apply the following formula:
– Number_of_data_transferred = 0xFFFF – DMA_SxNDTR
• Normal stream interruption due to the reception of a last data hardware signal: the
stream is automatically interrupted when the peripheral requests the last transfer
(single or burst) and when this transfer is complete. the TCIFx flag of the corresponding
stream is set in the status register to indicate the DMA transfer completion. To know the
number of data items transferred, read the DMA_SxNDTR register and apply the same
formula as above.
• The DMA_SxNDTR register reaches 0: the TCIFx flag of the corresponding stream is
set in the status register to indicate the forced DMA transfer completion. The stream is
automatically switched off even though the last data hardware signal (single or burst)
has not been yet asserted. The already transferred data is not lost. This means that a
maximum of 65535 data items can be managed by the DMA in a single transaction,
even in peripheral flow control mode.
Note: When configured in memory-to-memory mode, the DMA is always the flow controller and
the PFCTRL bit is forced to 0 by hardware.
The circular mode is forbidden in the peripheral flow controller mode.

RM0402 Rev 6 213/1163

230
Direct memory access controller (DMA) RM0402

9.3.17 Summary of the possible DMA configurations

The table below summarizes the different possible DMA configurations. The forbidden
configurations are highlighted in gray in the table.

Table 37. Possible DMA configurations

DMA transfer Flow Circular Transfer Direct Double-
Source Destination
mode controller mode type mode buffer mode

single Possible
DMA Possible Possible
Peripheral-to- AHB AHB burst Forbidden
memory peripheral port memory port single Possible
Peripheral Forbidden Forbidden
burst Forbidden
single Possible
DMA Possible Possible
Memory-to- AHB AHB burst Forbidden
peripheral memory port peripheral port single Possible
Peripheral Forbidden Forbidden
burst Forbidden

Memory-to- AHB AHB single

DMA only Forbidden Forbidden Forbidden
memory peripheral port memory port burst

9.3.18 Stream configuration procedure

The following sequence must be followed to configure a DMA stream x (where x is the
stream number):
1. If the stream is enabled, disable it by resetting the EN bit in the DMA_SxCR register,
then read this bit in order to confirm that there is no ongoing stream operation. Writing
this bit to 0 is not immediately effective since it is actually written to 0 once all the
current transfers are finished. When the EN bit is read as 0, this means that the stream
is ready to be configured. It is therefore necessary to wait for the EN bit to be cleared
before starting any stream configuration. All the stream dedicated bits set in the status
register (DMA_LISR and DMA_HISR) from the previous data block DMA transfer must
be cleared before the stream can be re-enabled.
2. Set the peripheral port register address in the DMA_SxPAR register. The data is moved
from/ to this address to/ from the peripheral port after the peripheral event.
3. Set the memory address in the DMA_SxMA0R register (and in the DMA_SxMA1R
register in the case of a double-buffer mode). The data is written to or read from this
memory after the peripheral event.
4. Configure the total number of data items to be transferred in the DMA_SxNDTR
register. After each peripheral event or each beat of the burst, this value is
decremented.
5. Select the DMA channel (request) using CHSEL[2:0] in the DMA_SxCR register.
6. If the peripheral is intended to be the flow controller and if it supports this feature, set
the PFCTRL bit in the DMA_SxCR register.
7. Configure the stream priority using the PL[1:0] bits in the DMA_SxCR register.
8. Configure the FIFO usage (enable or disable, threshold in transmission and reception)

214/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

9. Configure the data transfer direction, peripheral and memory incremented/fixed mode,
single or burst transactions, peripheral and memory data widths, circular mode,
double-buffer mode and interrupts after half and/or full transfer, and/or errors in the
DMA_SxCR register.
10. Activate the stream by setting the EN bit in the DMA_SxCR register.
As soon as the stream is enabled, it can serve any DMA request from the peripheral
connected to the stream.
Once half the data have been transferred on the AHB destination port, the half-transfer flag
(HTIF) is set and an interrupt is generated if the half-transfer interrupt enable bit (HTIE) is
set. At the end of the transfer, the transfer complete flag (TCIF) is set and an interrupt is
generated if the transfer complete interrupt enable bit (TCIE) is set.

Warning: To switch off a peripheral connected to a DMA stream

request, it is mandatory to, first, switch off the DMA stream to
which the peripheral is connected, then to wait for EN bit = 0.
Only then can the peripheral be safely disabled.

9.3.19 Error management

The DMA controller can detect the following errors:
• Transfer error: the transfer error interrupt flag (TEIFx) is set when:
– a bus error occurs during a DMA read or a write access
– a write access is requested by software on a memory address register in
double-buffer mode whereas the stream is enabled and the current target memory
is the one impacted by the write into the memory address register (refer to
Section 9.3.10: Double-buffer mode)
• FIFO error: the FIFO error interrupt flag (FEIFx) is set if:
– a FIFO underrun condition is detected
– a FIFO overrun condition is detected (no detection in memory-to-memory mode
because requests and transfers are internally managed by the DMA)
– the stream is enabled while the FIFO threshold level is not compatible with the
size of the memory burst (refer to Table 36: FIFO threshold configurations)
• Direct mode error: the direct mode error interrupt flag (DMEIFx) can only be set in the
peripheral-to-memory mode while operating in direct mode and when the MINC bit in
the DMA_SxCR register is cleared. This flag is set when a DMA request occurs while
the previous data have not yet been fully transferred into the memory (because the
memory bus was not granted). In this case, the flag indicates that two data items were
be transferred successively to the same destination address, which could be an issue if
the destination is not able to manage this situation
In direct mode, the FIFO error flag can also be set under the following conditions:
• In the peripheral-to-memory mode, the FIFO can be saturated (overrun) if the memory
bus is not granted for several peripheral requests.
• In the memory-to-peripheral mode, an underrun condition may occur if the memory bus
has not been granted before a peripheral request occurs.

RM0402 Rev 6 215/1163

230
Direct memory access controller (DMA) RM0402

If the TEIFx or the FEIFx flag is set due to incompatibility between burst size and FIFO
threshold level, the faulty stream is automatically disabled through a hardware clear of its
EN bit in the corresponding stream configuration register (DMA_SxCR).
If the DMEIFx or the FEIFx flag is set due to an overrun or underrun condition, the faulty
stream is not automatically disabled and it is up to the software to disable or not the stream
by resetting the EN bit in the DMA_SxCR register. This is because there is no data loss
when this kind of errors occur.
When the stream's error interrupt flag (TEIF, FEIF, DMEIF) in the DMA_LISR or DMA_HISR
register is set, an interrupt is generated if the corresponding interrupt enable bit (TEIE,
FEIE, DMIE) in the DMA_SxCR or DMA_SxFCR register is set.
Note: When a FIFO overrun or underrun condition occurs, the data is not lost because the
peripheral request is not acknowledged by the stream until the overrun or underrun
condition is cleared. If this acknowledge takes too much time, the peripheral itself may
detect an overrun or underrun condition of its internal buffer and data might be lost.

9.4 DMA interrupts

For each DMA stream, an interrupt can be produced on the following events:
• Half-transfer reached
• Transfer complete
• Transfer error
• FIFO error (overrun, underrun or FIFO level error)
• Direct mode error
Separate interrupt enable control bits are available for flexibility as shown in the table below.

Table 38. DMA interrupt requests

Interrupt event Event flag Enable control bit

Half-transfer HTIF HTIE

Transfer complete TCIF TCIE
Transfer error TEIF TEIE
FIFO overrun/underrun FEIF FEIE
Direct mode error DMEIF DMEIE

Note: Before setting an enable control bit EN = 1, the corresponding event flag must be cleared,
otherwise an interrupt is immediately generated.

216/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

9.5 DMA registers

The DMA registers have to be accessed by words (32 bits).

9.5.1 DMA low interrupt status register (DMA_LISR)

Address offset: 0x000
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. TCIF3 HTIF3 TEIF3 DMEIF3 Res. FEIF3 TCIF2 HTIF2 TEIF2 DMEIF2 Res. FEIF2
r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. TCIF1 HTIF1 TEIF1 DMEIF1 Res. FEIF1 TCIF0 HTIF0 TEIF0 DMEIF0 Res. FEIF0
r r r r r r r r r r

Bits 31:28, 15:12 Reserved, must be kept at reset value.

Bits 27, 21, 11, 5 TCIFx: stream x transfer complete interrupt flag (x = 3 to 0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: no transfer complete event on stream x
1: a transfer complete event occurred on stream x
Bits 26, 20, 10, 4 HTIFx: stream x half transfer interrupt flag (x = 3 to 0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: no half transfer event on stream x
1: a half transfer event occurred on stream x
Bits 25, 19, 9, 3 TEIFx: stream x transfer error interrupt flag (x = 3 to 0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: no transfer error on stream x
1: a transfer error occurred on stream x
Bits 24, 18, 8, 2 DMEIFx: stream x direct mode error interrupt flag (x = 3 to 0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: No direct mode error on stream x
1: a direct mode error occurred on stream x
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 FEIFx: stream x FIFO error interrupt flag (x = 3 to 0)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_LIFCR register.
0: no FIFO error event on stream x
1: a FIFO error event occurred on stream x

RM0402 Rev 6 217/1163

230
Direct memory access controller (DMA) RM0402

9.5.2 DMA high interrupt status register (DMA_HISR)

Address offset: 0x004
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. TCIF7 HTIF7 TEIF7 DMEIF7 Res. FEIF7 TCIF6 HTIF6 TEIF6 DMEIF6 Res. FEIF6

r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. TCIF5 HTIF5 TEIF5 DMEIF5 Res. FEIF5 TCIF4 HTIF4 TEIF4 DMEIF4 Res. FEIF4

r r r r r r r r r r

Bits 31:28, 15:12 Reserved, must be kept at reset value.

Bits 27, 21, 11, 5 TCIFx: stream x transfer complete interrupt flag (x = 7 to 4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: no transfer complete event on stream x
1: a transfer complete event occurred on stream x
Bits 26, 20, 10, 4 HTIFx: stream x half transfer interrupt flag (x = 7 to 4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: no half transfer event on stream x
1: a half transfer event occurred on stream x
Bits 25, 19, 9, 3 TEIFx: stream x transfer error interrupt flag (x = 7 to 4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: no transfer error on stream x
1: a transfer error occurred on stream x
Bits 24, 18, 8, 2 DMEIFx: stream x direct mode error interrupt flag (x = 7 to 4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: no direct mode error on stream x
1: a direct mode error occurred on stream x
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 FEIFx: stream x FIFO error interrupt flag (x = 7 to 4)
This bit is set by hardware. It is cleared by software writing 1 to the corresponding bit in the
DMA_HIFCR register.
0: no FIFO error event on stream x
1: a FIFO error event occurred on stream x

218/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

9.5.3 DMA low interrupt flag clear register (DMA_LIFCR)

Address offset: 0x008
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. CTCIF3 CHTIF3 CTEIF3 CDMEIF3 Res. CFEIF3 CTCIF2 CHTIF2 CTEIF2 CDMEIF2 Res. CFEIF2
w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. CTCIF1 CHTIF1 CTEIF1 CDMEIF1 Res. CFEIF1 CTCIF0 CHTIF0 CTEIF0 CDMEIF0 Res. CFEIF0
w w w w w w w w w w

Bits 31:28, 15:12 Reserved, must be kept at reset value.

Bits 27, 21, 11, 5 CTCIFx: stream x clear transfer complete interrupt flag (x = 3 to 0)
Writing 1 to this bit clears the corresponding TCIFx flag in the DMA_LISR register.
Bits 26, 20, 10, 4 CHTIFx: stream x clear half transfer interrupt flag (x = 3 to 0)
Writing 1 to this bit clears the corresponding HTIFx flag in the DMA_LISR register
Bits 25, 19, 9, 3 CTEIFx: Stream x clear transfer error interrupt flag (x = 3 to 0)
Writing 1 to this bit clears the corresponding TEIFx flag in the DMA_LISR register.
Bits 24, 18, 8, 2 CDMEIFx: stream x clear direct mode error interrupt flag (x = 3 to 0)
Writing 1 to this bit clears the corresponding DMEIFx flag in the DMA_LISR register.
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 CFEIFx: stream x clear FIFO error interrupt flag (x = 3 to 0)
Writing 1 to this bit clears the corresponding CFEIFx flag in the DMA_LISR register.

9.5.4 DMA high interrupt flag clear register (DMA_HIFCR)

Address offset: 0x00C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. CTCIF7 CHTIF7 CTEIF7 CDMEIF7 Res. CFEIF7 CTCIF6 CHTIF6 CTEIF6 CDMEIF6 Res. CFEIF6
w w w w w w w w w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. CTCIF5 CHTIF5 CTEIF5 CDMEIF5 Res. CFEIF5 CTCIF4 CHTIF4 CTEIF4 CDMEIF4 Res. CFEIF4
w w w w w w w w w w

Bits 31:28, 15:12 Reserved, must be kept at reset value.

Bits 27, 21, 11, 5 CTCIFx: stream x clear transfer complete interrupt flag (x = 7 to 4)
Writing 1 to this bit clears the corresponding TCIFx flag in the DMA_HISR register.
Bits 26, 20, 10, 4 CHTIFx: stream x clear half transfer interrupt flag (x = 7 to 4)
Writing 1 to this bit clears the corresponding HTIFx flag in the DMA_HISR register.
Bits 25, 19, 9, 3 CTEIFx: stream x clear transfer error interrupt flag (x = 7 to 4)
Writing 1 to this bit clears the corresponding TEIFx flag in the DMA_HISR register.

RM0402 Rev 6 219/1163

230
Direct memory access controller (DMA) RM0402

Bits 24, 18, 8, 2 CDMEIFx: stream x clear direct mode error interrupt flag (x = 7 to 4)
Writing 1 to this bit clears the corresponding DMEIFx flag in the DMA_HISR register.
Bits 23, 17, 7, 1 Reserved, must be kept at reset value.
Bits 22, 16, 6, 0 CFEIFx: stream x clear FIFO error interrupt flag (x = 7 to 4)
Writing 1 to this bit clears the corresponding CFEIFx flag in the DMA_HISR register.

9.5.5 DMA stream x configuration register (DMA_SxCR)

This register is used to configure the concerned stream.
Address offset: 0x010 + 0x018 * x, (x = 0 to 7)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. CHSEL[2:0] MBURST [1:0] PBURST[1:0] Res. CT DBM PL[1:0]
rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PINCOS MSIZE[1:0] PSIZE[1:0] MINC PINC CIRC DIR[1:0] PFCTRL TCIE HTIE TEIE DMEIE EN
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:28 Reserved, must be kept at reset value.

Bits 27:25 CHSEL[2:0]: channel selection
These bits are set and cleared by software.
000: channel 0 selected
001: channel 1 selected
010: channel 2 selected
011: channel 3 selected
100: channel 4 selected
101: channel 5 selected
110: channel 6 selected
111: channel 7 selected
These bits are protected and can be written only if EN is 0.
Bits 24:23 MBURST[1:0]: memory burst transfer configuration
These bits are set and cleared by software.
00: single transfer
01: INCR4 (incremental burst of 4 beats)
10: INCR8 (incremental burst of 8 beats)
11: INCR16 (incremental burst of 16 beats)
These bits are protected and can be written only if EN = 0.
In direct mode, these bits are forced to 0x0 by hardware as soon as bit EN = 1.
Bits 22:21 PBURST[1:0]: peripheral burst transfer configuration
These bits are set and cleared by software.
00: single transfer
01: INCR4 (incremental burst of 4 beats)
10: INCR8 (incremental burst of 8 beats)
11: INCR16 (incremental burst of 16 beats)
These bits are protected and can be written only if EN = 0.
In direct mode, these bits are forced to 0x0 by hardware.

220/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Bit 20 Reserved, must be kept at reset value.

Bit 19 CT: current target (only in double-buffer mode)
This bit is set and cleared by hardware. It can also be written by software.
0: current target memory is Memory 0 (addressed by the DMA_SxM0AR pointer)
1: current target memory is Memory 1 (addressed by the DMA_SxM1AR pointer)
This bit can be written only if EN = 0 to indicate the target memory area of the first transfer.
Once the stream is enabled, this bit operates as a status flag indicating which memory area
is the current target.
Bit 18 DBM: double-buffer mode
This bit is set and cleared by software.
0: no buffer switching at the end of transfer
1: memory target switched at the end of the DMA transfer
This bit is protected and can be written only if EN = 0.
Bits 17:16 PL[1:0]: priority level
These bits are set and cleared by software.
00: low
01: medium
10: high
11: very high
These bits are protected and can be written only if EN = 0.
Bit 15 PINCOS: peripheral increment offset size
This bit is set and cleared by software
0: The offset size for the peripheral address calculation is linked to the PSIZE
1: The offset size for the peripheral address calculation is fixed to 4 (32-bit alignment).
This bit has no meaning if bit PINC = 0.
This bit is protected and can be written only if EN = 0.
This bit is forced low by hardware when the stream is enabled (EN = 1) if the direct mode is
selected or if PBURST are different from 00.
Bits 14:13 MSIZE[1:0]: memory data size
These bits are set and cleared by software.
00: byte (8-bit)
01: half-word (16-bit)
10: word (32-bit)
11: reserved
These bits are protected and can be written only if EN = 0.
In direct mode, MSIZE is forced by hardware to the same value as PSIZE as soon as
EN = 1.
Bits 12:11 PSIZE[1:0]: peripheral data size
These bits are set and cleared by software.
00: byte (8-bit)
01: half-word (16-bit)
10: word (32-bit)
11: reserved
These bits are protected and can be written only if EN = 0.

RM0402 Rev 6 221/1163

230
Direct memory access controller (DMA) RM0402

Bit 10 MINC: memory increment mode

This bit is set and cleared by software.
0: memory address pointer is fixed
1: memory address pointer is incremented after each data transfer (increment is done
according to MSIZE)
This bit is protected and can be written only if EN = 0.
Bit 9 PINC: peripheral increment mode
This bit is set and cleared by software.
0: peripheral address pointer fixed
1: peripheral address pointer incremented after each data transfer (increment done
according to PSIZE)
This bit is protected and can be written only if EN = 0.
Bit 8 CIRC: circular mode
This bit is set and cleared by software and can be cleared by hardware.
0: circular mode disabled
1: circular mode enabled
When the peripheral is the flow controller (bit PFCTRL = 1) and the stream is enabled
(EN = 1), then this bit is automatically forced by hardware to 0.
It is automatically forced by hardware to 1 if the DBM bit is set, as soon as the stream is
enabled (EN = 1).
Bits 7:6 DIR[1:0]: data transfer direction
These bits are set and cleared by software.
00: peripheral-to-memory
01: memory-to-peripheral
10: memory-to-memory
11: reserved
These bits are protected and can be written only if EN = 0.
Bit 5 PFCTRL: peripheral flow controller
This bit is set and cleared by software.
0: DMA is the flow controller.
1: The peripheral is the flow controller.
This bit is protected and can be written only if EN = 0.
When the memory-to-memory mode is selected (bits DIR[1:0]=10), then this bit is
automatically forced to 0 by hardware.
Bit 4 TCIE: transfer complete interrupt enable
This bit is set and cleared by software.
0: TC interrupt disabled
1: TC interrupt enabled
Bit 3 HTIE: half transfer interrupt enable
This bit is set and cleared by software.
0: HT interrupt disabled
1: HT interrupt enabled
Bit 2 TEIE: transfer error interrupt enable
This bit is set and cleared by software.
0: TE interrupt disabled
1: TE interrupt enabled

222/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Bit 1 DMEIE: direct mode error interrupt enable

This bit is set and cleared by software.
0: DME interrupt disabled
1: DME interrupt enabled
Bit 0 EN: stream enable / flag stream ready when read low
This bit is set and cleared by software.
0: stream disabled
1: stream enabled
This bit may be cleared by hardware:
– on a DMA end of transfer (stream ready to be configured)
– if a transfer error occurs on the AHB master buses
– when the FIFO threshold on memory AHB port is not compatible with the size of the
burst
When this bit is read as 0, the software is allowed to program the configuration and FIFO bits
registers. It is forbidden to write these registers when the EN bit is read as 1.
Note: Before setting EN bit to 1 to start a new transfer, the event flags corresponding to the
stream in DMA_LISR or DMA_HISR register must be cleared.

9.5.6 DMA stream x number of data register (DMA_SxNDTR)

Address offset: 0x014 + 0x018 * x, (x = 0 to 7)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NDT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 NDT[15:0]: number of data items to transfer (0 up to 65535)
This register can be written only when the stream is disabled. When the stream is enabled,
this register is read-only, indicating the remaining data items to be transmitted. This register
decrements after each DMA transfer.
Once the transfer is completed, this register can either stay at zero (when the stream is in
normal mode) or be reloaded automatically with the previously programmed value in the
following cases:
– when the stream is configured in circular mode.
– when the stream is enabled again by setting EN bit to 1.
If the value of this register is zero, no transaction can be served even if the stream is
enabled.

RM0402 Rev 6 223/1163

230
Direct memory access controller (DMA) RM0402

9.5.7 DMA stream x peripheral address register (DMA_SxPAR)

Address offset: 0x018 + 0x018 * x, (x = 0 to 7)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PAR[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PAR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 PAR[31:0]: peripheral address

Base address of the peripheral data register from/to which the data is read/written.
These bits are write-protected and can be written only when bit EN = 0 in DMA_SxCR.

9.5.8 DMA stream x memory 0 address register

(DMA_SxM0AR)
Address offset: 0x01C + 0x018 * x, (x = 0 to 7)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
M0A[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
M0A[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 M0A[31:0]: memory 0 address

Base address of memory area 0 from/to which the data is read/written.
These bits are write-protected. They can be written only if:
– the stream is disabled (EN = 0 in DMA_SxCR) or
– the stream is enabled (EN = 1 in DMA_SxCR) and CT = 1 in DMA_SxCR (in
double-buffer mode).

9.5.9 DMA stream x memory 1 address register

(DMA_SxM1AR)
Address offset: 0x020 + 0x018 * x, (x = 0 to 7)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
M1A[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
M1A[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

224/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Bits 31:0 M1A[31:0]: memory 1 address (used in case of double-buffer mode)

Base address of memory area 1 from/to which the data is read/written.
This register is used only for the double-buffer mode.
These bits are write-protected. They can be written only if:
– the stream is disabled (EN = 0 in DMA_SxCR) or
– the stream is enabled (EN = 1 in DMA_SxCR) and bit CT = 0 in DMA_SxCR .

9.5.10 DMA stream x FIFO control register (DMA_SxFCR)

Address offset: 0x024 + 0x018 * x, (x = 0 to 7)
Reset value: 0x0000 0021

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. FEIE Res. FS[2:0] DMDIS FTH[1:0]
rw r r r rw rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 FEIE: FIFO error interrupt enable
This bit is set and cleared by software.
0: FE interrupt disabled
1: FE interrupt enabled
Bit 6 Reserved, must be kept at reset value.
Bits 5:3 FS[2:0]: FIFO status
These bits are read-only.
000: 0 < fifo_level < 1/4
001: 1/4 ≤ fifo_level < 1/2
010: 1/2 ≤ fifo_level < 3/4
011: 3/4 ≤ fifo_level < full
100: FIFO is empty
101: FIFO is full
others: no meaning
These bits are not relevant in the direct mode (DMDIS = 0).
Bit 2 DMDIS: direct mode disable
This bit is set and cleared by software. It can be set by hardware.
0: direct mode enabled
1: direct mode disabled
This bit is protected and can be written only if EN = 0.
This bit is set by hardware if the memory-to-memory mode is selected (DIR bit in
DMA_SxCR are 10) and the EN = 1 in DMA_SxCR because the direct mode is not allowed
in the memory-to-memory configuration.

RM0402 Rev 6 225/1163

230
Direct memory access controller (DMA) RM0402

Bits 1:0 FTH[1:0]: FIFO threshold selection

These bits are set and cleared by software.
00: 1/4 full FIFO
01: 1/2 full FIFO
10: 3/4 full FIFO
11: full FIFO
These bits are not used in the direct mode when the DMIS = 0.
These bits are protected and can be written only if EN = 0.

226/1163 RM0402 Rev 6

 0x030
0x008
0x024
0x020
0x018
0x014
0x010
0x008
0x004
0x000

0x02C
0x01C
0x00C
9.5.11
RM0402

DMA_LISR

DMA_HISR

Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

DMA_S1CR
DMA_S0CR
DMA_LIFCR

DMA_HIFCR

DMA_S1PAR
DMA_S0PAR

DMA_S0FCR

DMA_S1NDTR
DMA_S0NDTR

DMA_S0M1AR
DMA_S0M0AR
Offset Register name

0
0
0
0
Res Res. Res Res Res Res Res Res Res. 31

0
0
0
0
Res Res. Res Res Res Res Res Res Res. 30

0
0
0
0
Res Res. Res Res Res Res Res Res Res. 29

0
0
0
0
Res Res. Res Res Res Res Res Res Res. 28

0
0
0
0
0
0
0
0
0

Res Res Res CTCIF7 CTCIF3 TCIF7 TCIF3 27

0
0
0
0
0
0
0
0
0

Res CHSEL[2:0] Res Res CHSEL[2:0] CHTIF7 CHTIF3 HTIF7 HTIF3 26

DMA register map

0
0
0
0
0
0
0
0
0

Res Res Res CTEIF7 TEIF3 TEIF7 TEIF3 25

0
0
0
0
0
0
0
0
0

Res Res Res CDMEIF7 CDMEIF3 DMEIF7 DMEIF3 24

MBURST[1:] MBURST[1:0]

0
0
0
0
0
Res Res Res Res Res Res Res. 23

0
0
0
0
0
0
0
0
0

Res Res Res CFEIF7 CFEIF3 FEIF7 FEIF3 22

PBURST[1:0] PBURST[1:0]

0
0
0
0
0
0
0
0
0

Res Res Res CTCIF6 CTCIF2 TCIF6 TCIF2 21

0
0
0
0
0
0
0
0

Res Res. Res Res Res CHTIF6 CHTIF2 HTIF6 HTIF2 20

0
0
0
0
0
0
0
0
0

0
Res CT Res Res CT CTEIF6 CTEIF2 TEIF6 TEIF2 19

0
0
0
0
0
0
0
0
0

0
Res DBM Res Res DBM CDMEIF6 CDMEIF2 DMEIF6 DMEIF2 18

0
0
0
0
0

RM0402 Rev 6
Res Res Res Res Res Res Res. 17
PL[1:0] PL[1:0]

0
0
0
0
0

0
0
0
0
0

Res Res Res CFEIF6 CFEIF2 FEIF6 FEIF2 16

0
0
0
0
0

0
0
0
PINCOS Res PINCOS Res Res Res Res. 15

PA[31:0]
PA[31:0]

M1A[31:0]
M0A[31:0]

0
0
0
0
0

0
0
0
Res Res Res Res Res. 14
MSIZE[1:0] MSIZE[1:0]

0
0
0
0
0

0
0
0
Res Res Res Res Res. 13

0
0
0
0
0

0
0
0
Res Res Res Res Res. 12
PSIZE[1:0] PSIZE[1:0]

0
0
0
0
0

0
0
0
0
0
0
0

Res CTCIF5 CTCIF1 TCIF5 TCIF1 11

Table 39. DMA register map and reset values

0
0
0
0
0

0
0
0
0
0
0
0

MINC Res MINC CHTIF5 CHTIF1 HTIF5 HTIF1 10

0
0
0
0
0

0
0
0
0
0
0
0

PINC Res PINC CTEIF5 CTEIF1 TEIF5 TEIF1 9

0
0
0
0
0

0
0
0
0
0
0
0

CIRC Res CIRC CDMEIF5 CDMEIF1 DMEIF5 DMEIF1 8

0
0
0
0
0

0
0
0

0
FEIE Res Res Res Res. 7
DIR[1:0] DIR[1:0]

NDT[15:0]
NDT[15:0]

0
0
0
0
0

0
0
0
0
0
0
0

Res CFEIF5 CFEIF1 FEIF5 FEIF1 6

0
0
0
0
0

0
0
0
0
0
0
0

1
PFCTRL PFCTRL CTCIF4 CTCIF0 TCIF4 TCIF0 5

0
0
0
0
0

0
0
0
0
0
0
0

0
TCIE TCIE CHTIF4 CHTIF0 HTIF4 HTIF0 4

FS[2:0]

0
0
0
0
0

0
0
0
0
0
0
0

0
HTIE HTIE CTEIF4 CTEIF0 TEIF4 TEIF0 3

0
0
0
0
0

0
0
0
0
0
0
0

0
TEIE DMDIS TEIE CDMEIF4 CDMEIF0 DMEIF4 DMEIF0 2

0
0
0
0
0

0
0
0

0
DMEIE DMEIE Res Res Res Res. 1
FTH[1:0]

0
0
0
0
0

0
0
0
0
0
0
0

1
EN EN CFEIF4 CFEIF0 FEIF4 FEIF0
Direct memory access controller (DMA)

227/1163
0

230
Direct memory access controller (DMA) RM0402

Table 39. DMA register map and reset values (continued)

Offset Register name

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
DMA_S1M0AR M0A[31:0]
0x034
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S1M1AR M1A[31:0]
0x038
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res
DMA_S1FCR FS[2:0]
0x03C

Reset value 0 1 0 0 0 0 1
MBURST[1:0]

PBURST[1:0]
CHSEL[2:0]

MSIZE[1:0]

PSIZE[1:0]

PFCTRL
PINCOS
PL[1:0]

DMEIE
MINC

CIRC
PINC

HTIE
TCIE
DBM

TEIE
[1:0]
Res
Res
Res
Res

Res

DIR

EN
CT
DMA_S2CR
0x040

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
DMA_S2NDTR NDT[15:0]
0x044
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S2PAR PA[31:0]
0x048
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S2M0AR M0A[31:0]
0x04C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S2M1AR M1A[31:0]
0x050
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res
DMA_S2FCR FS[2:0]
0x054

Reset value 0 1 0 0 0 0 1
MBURST[1:0]

PBURST[1:0]
CHSEL[2:0]

MSIZE[1:0]

PSIZE[1:0]

PFCTRL
PINCOS

DIR[1:0]
PL[1:0]

DMEIE
MINC

CIRC
PINC

HTIE
TCIE

TEIE
DBM
Res
Res
Res
Res

Res

EN
CT

DMA_S3CR
0x058

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

DMA_S3NDTR NDT[15:0]
0x05C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S3PAR PA[31:0]
0x060
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S3M0AR M0A[31:0]
0x064
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S3M1AR M1A[31:0]
0x068
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

228/1163 RM0402 Rev 6

RM0402 Direct memory access controller (DMA)

Table 39. DMA register map and reset values (continued)

Offset Register name

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res
DMA_S3FCR FS[2:0]
0x06C

Reset value 0 1 0 0 0 0 1

MBURST[1:0]

PBURST[1:0]
CHSEL[2:0]

MSIZE[1:0]

PSIZE[1:0]

PFCTRL
PINCOS
PL[1:0]

DMEIE
MINC

CIRC
PINC

HTIE
TCIE

TEIE
DBM

[1:0]
Res
Res
Res
Res

Res

DIR

EN
CT
DMA_S4CR
0x070

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
DMA_S4NDTR NDT[15:0]
0x074
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S4PAR PA[31:0]
0x078
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S4M0AR M0A[31:0]
0x07C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S4M1AR M1A[31:0]
0x080
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res
DMA_S4FCR FS[2:0]
0x084

Reset value 0 1 0 0 0 0 1
MBURST[1:0]

PBURST[1:0]
CHSEL[2:0]

MSIZE[1:0]

PSIZE[1:0]

PFCTRL
PINCOS

DIR[1:0]
PL[1:0]

DMEIE
MINC

CIRC
PINC

HTIE
TCIE

TEIE
DBM
Res
Res
Res
Res

Res

EN
CT

DMA_S5CR
0x088

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

DMA_S5NDTR NDT[15:0]
0x08C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S5PAR PA[31:0]
0x090
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S5M0AR M0A[31:0]
0x094
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S5M1AR M1A[31:0]
0x098
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res

DMA_S5FCR FS[2:0]
0x09C

Reset value 0 1 0 0 0 0 1

RM0402 Rev 6 229/1163

230
Direct memory access controller (DMA) RM0402

Table 39. DMA register map and reset values (continued)

Offset Register name

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
MBURST[1:0]

PBURST[1:0]
CHSEL[2:0]

MSIZE[1:0]

PSIZE[1:0]

PFCTRL
PINCOS

DIR[1:0]
PL[1:0]

DMEIE
MINC

CIRC
PINC

TCIE
HTIE
DBM

TEIE
Res
Res
Res
Res

Res

EN
CT
DMA_S6CR
0x0A0

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
DMA_S6NDTR NDT[15:0]
0x0A4
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S6PAR PA[31:0]
0x0A8
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S6M0AR M0A[31:0]
0x0AC
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S6M1AR M1A[31:0]
0x0B0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res
DMA_S6FCR FS[2:0]
0x0B4

Reset value 0 1 0 0 0 0 1
MBURST[1:0]

PBURST[1:0]
CHSEL[2:0]

MSIZE[1:0]

PSIZE[1:0]

PFCTRL
PINCOS

DIR[1:0]
PL[1:0]

DMEIE
MINC

CIRC
PINC

TCIE
HTIE
DBM

TEIE
Res
Res
Res
Res

Res

EN
CT

DMA_S7CR
0x0B8

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

DMA_S7NDTR NDT[15:0]
0x0BC
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S7PAR PA[31:0]
0x0C0
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S7M0AR M0A[31:0]
0x0C4
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DMA_S7M1AR M1A[31:0]
0x0C8
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FTH[1:0]
DMDIS
FEIE
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res
Res

Res

DMA_S7FCR FS[2:0]
0x0CC

Reset value 0 1 0 0 0 0 1

Refer to Section 2.2 for the register boundary addresses.

230/1163 RM0402 Rev 6

RM0402 Interrupts and events

10 Interrupts and events

10.1 Nested vectored interrupt controller (NVIC)

10.1.1 NVIC features

The nested vector interrupt controller NVIC includes the following features:
• 52 maskable interrupt channels (not including the 16 interrupt lines of Cortex®-M4 with
FPU)
• 16 programmable priority levels (4 bits of interrupt priority are used)
• low-latency exception and interrupt handling
• power management control
• implementation of system control registers
The NVIC and the processor core interface are closely coupled, which enables low latency
interrupt processing and efficient processing of late arriving interrupts.
All interrupts including the core exceptions are managed by the NVIC. For more information
on exceptions and NVIC programming, refer to programming manual PM0214.

10.1.2 SysTick calibration value register

The SysTick calibration value is fixed to 10500, which gives a reference time base of 1 ms
with the SysTick clock set to 10.5 MHz (HCLK/8, with HCLK set to 84 MHz).

10.1.3 Interrupt and exception vectors

See Table 40, for the vector table for the STM32F412xx devices.

10.2 External interrupt/event controller (EXTI)

The external interrupt/event controller consists of up to 23 edge detectors for generating
event/interrupt requests. Each input line can be independently configured to select the type
(interrupt or event) and the corresponding trigger event (rising or falling or both). Each line
can also masked independently. A pending register maintains the status line of the interrupt
requests.

RM0402 Rev 6 231/1163

245
Interrupts and events RM0402

Table 40. Vector table for STM32F412xx

Position

Priority

Type of
Acronym Description Address
priority

- - - - Reserved 0x0000 0000

- -3 fixed Reset Reset 0x0000 0004
Non maskable interrupt, Clock Security
- -2 fixed NMI 0x0000 0008
System
- -1 fixed HardFault All class of fault 0x0000 000C
- 0 settable MemManage Memory management 0x0000 0010
- 1 settable BusFault Pre-fetch fault, memory access fault 0x0000 0014
- 2 settable UsageFault Undefined instruction or illegal state 0x0000 0018
0x0000 001C -
- - - - Reserved
0x0000 002B
- 3 settable SVCall System Service call via SWI instruction 0x0000 002C
- 4 settable Debug Monitor Debug Monitor 0x0000 0030
- - - Reserved 0x0000 0034
- 5 settable PendSV Pendable request for system service 0x0000 0038
- 6 settable Systick System tick timer 0x0000 003C
0 7 settable WWDG Window Watchdog interrupt 0x0000 0040
1 8 settable PVD PVD through EXTI line detection interrupt 0x0000 0044
Tamper and TimeStamp interrupts
2 9 settable TAMP_STAMP 0x0000 0048
through the EXTI line
3 10 settable RTC_WKUP RTC Wakeup interrupt through the EXTI line 0x0000 004C
4 11 settable FLASH Flash global interrupt 0x0000 0050
5 12 settable RCC RCC global interrupt 0x0000 0054
6 13 settable EXTI0 EXTI Line0 interrupt 0x0000 0058
7 14 settable EXTI1 EXTI Line1 interrupt 0x0000 005C
8 15 settable EXTI2 EXTI Line2 interrupt 0x0000 0060
9 16 settable EXTI3 EXTI Line3 interrupt 0x0000 0064
10 17 settable EXTI4 EXTI Line4 interrupt 0x0000 0068
11 18 settable DMA1_Stream0 DMA1 Stream0 global interrupt 0x0000 006C
12 19 settable DMA1_Stream1 DMA1 Stream1 global interrupt 0x0000 0070
13 20 settable DMA1_Stream2 DMA1 Stream2 global interrupt 0x0000 0074
14 21 settable DMA1_Stream3 DMA1 Stream3 global interrupt 0x0000 0078

232/1163 RM0402 Rev 6

RM0402 Interrupts and events

Table 40. Vector table for STM32F412xx (continued)

Position

Priority

Type of
Acronym Description Address
priority

15 22 settable DMA1_Stream4 DMA1 Stream4 global interrupt 0x0000 007C

16 23 settable DMA1_Stream5 DMA1 Stream5 global interrupt 0x0000 0080
17 24 settable DMA1_Stream6 DMA1 Stream6 global interrupt 0x0000 0084
18 25 settable ADC ADC1 global interrupt 0x0000 0088
19 26 settable CAN1_TX CAN1 TX interrupt 0x0000 008C
20 27 settable CAN1_RX0 CAN1 RX0 interrupt 0x0000 0090
21 28 settable CAN1_RX1 CAN1 RX1 interrupt 0x0000 0094
22 29 settable CAN1_SCE CAN1 SCE interrupt 0x0000 0098
23 30 settable EXTI9_5 EXTI Line[9:5] interrupts 0x0000 009C
TIM1 Break interrupt
24 31 settable TIM1_BRK_TIM9 0x0000 00A0
and TIM9 global interrupt
TIM1 update interrupt
25 32 settable TIM1_UP_TIM10 0x0000 00A4
and TIM10 global interrupt
TIM1 Trigger & Commutation interrupts
26 33 settable TIM_TRG_COM_TIM11 0x0000 00A8
and TIM11 global interrupt
27 34 settable TIM1_CC TIM1 Capture Compare interrupt 0x0000 00AC
28 35 settable TIM2 TIM2 global interrupt 0x0000 00B0
29 36 settable TIM3 TIM3 global interrupt 0x0000 00B4
30 37 settable TIM4 TIM4 global interrupt 0x0000 00B8
31 38 settable I2C1_EVT I2C1 global event interrupt 0x0000 00BC
32 39 settable I2C1_ERR I2C1 global error interrupt 0x0000 00C0
33 40 settable I2C2_EVT I2C2 global event interrupt 0x0000 00C4
34 41 settable I2C2_ERR I2C2 global error interrupt 0x0000 00C8
35 42 settable SPI1 SPI1 global interrupt 0x0000 00CC
36 43 settable SPI2 SPI2 global interrupt 0x0000 00D0
37 44 settable USART1 USART1 global interrupt 0x0000 00D4
38 45 settable USART2 USART2 global interrupt 0x0000 00D8
39 46 settable USART 3 USART3 global interrupt 0x0000 00DC
40 47 settable EXTI15_10 EXTI Line[15:10] interrupts 0x0000 00E0

RM0402 Rev 6 233/1163

245
Interrupts and events RM0402

Table 40. Vector table for STM32F412xx (continued)

Position

Priority

Type of
Acronym Description Address
priority

EXTI Line 17 interrupt /

EXTI17 /
41 48 settable RTC Alarms (A and B) through EXTI line 0x0000 00E4
RTC Alarm
interrupt
EXTI Line 18 interrupt / USB On-The-Go FS
42 49 settable EXTI18 / OTG_FS_WKUP 0x0000 00E8
Wakeup through EXTI line interrupt
TIM8 Break interrupt
43 50 settable TIM8_BRK_TIM12 0x0000 00EC
TIM12 global interrupt
TIM8 Update interrupt
44 51 settable TIM8_UP_TIM13 0x0000 00F0
TIM13 global interrupt
TIM8 Trigger & Commutation interrupt
45 52 settable TIM8_TRG_COM_TIM14 0x0000 00F4
TIM14 global interrupt
46 53 settable TIM8_CC TIM8 Cap/Com interrupt 0x0000 00F8
47 54 settable DMA1_Stream7 DMA1 global interrupt Channel 7 0x0000 00FC
48 55 settable FSMC FSMC global interrupt 0x0000 0100
49 56 settable SDIO SDIO global interrupt 0x0000 0104
50 57 settable TIM5 TIM5 global interrupt 0x0000 0108
51 58 settable SPI3 SPI3 global interrupt 0x0000 010C
54 61 settable TIM6 TIM6 global interrupt 0x0000 0118
55 62 settable TIM7 TIM7 global interrupt 0x0000 011C
56 63 settable DMA2_Stream0 DMA2 Stream0 global interrupt 0x0000 0120
57 64 settable DMA2_Stream1 DMA2 Stream1 global interrupt 0x0000 0124
58 65 settable DMA2_Stream2 DMA2 Stream2 global interrupt 0x0000 0128
59 66 settable DMA2_Stream3 DMA2 Stream3 global interrupt 0x0000 012C
60 67 settable DMA2_Stream4 DMA2 Stream4 global interrupt 0x0000 0130
61 68 settable DFSDM1_FLT0 SD filter0 global interrupt 0x0000 0134
62 69 settable DFSDM1_FLT1 SD filter1 global interrupt 0x0000 0138
63 70 settable CAN2_TX CAN2 TX interrupt 0x0000 013C
64 71 settable CAN2_RX0 BXCAN2 RX0 interrupt 0x0000 0140
65 72 settable CAN2_RX1 BXCAN2 RX1 interrupt 0x0000 0144
66 73 settable CAN2_SCE CAN2 SCE interrupt 0x0000 0148
67 74 settable OTG_FS USB On The Go FS global interrupt 0x0000 014C
68 75 settable DMA2_Stream5 DMA2 Stream5 global interrupt 0x0000 0150

234/1163 RM0402 Rev 6

RM0402 Interrupts and events

Table 40. Vector table for STM32F412xx (continued)

Position

Priority

Type of
Acronym Description Address
priority

69 76 settable DMA2_Stream6 DMA2 Stream6 global interrupt 0x0000 0154

70 77 settable DMA2_Stream7 DMA2 Stream7 global interrupt 0x0000 0158
71 78 settable USART6 USART6 global interrupt 0x0000 015C
72 79 settable I2C3_EV I2C3 event interrupt 0x0000 0160
73 80 settable I2C3_ER I2C3 error interrupt 0x0000 0164
80 87 settable RNG RNG global interrupt 0x0000 0180
81 88 settable FPU FPU global interrupt 0x0000 0184
84 91 settable SPI4 SPI4 global interrupt 0x0000 0190
85 92 settable SPI5 SPI5 global interrupt 0x0000 0194

92 99 settable Quad-SPI Quad-SPI global interrupt 0x0000 01B0

95 102 settable I2CFMP1 event I2CFMP1 event interrupt 0x0000 01BC
96 103 settable I2CFMP1 error I2CFMP1 error interrupt 0x0000 01C0

10.2.1 EXTI main features

The main features of the EXTI controller are the following:
• independent trigger and mask on each interrupt/event line
• dedicated status bit for each interrupt line
• generation of up to 23 software event/interrupt requests
• detection of external signals with a pulse width lower than the APB2 clock period. Refer
to the electrical characteristics section of the STM32F4xx datasheets for details on this
parameter.

RM0402 Rev 6 235/1163

245
Interrupts and events RM0402

10.2.2 EXTI block diagram

Figure 29 shows the block diagram.

Figure 29. External interrupt/event controller block diagram

AMBA APB bus

PCLK2 Peripheral interface

23 23 23 23 23

Pending Interrupt Software Rising Falling

request mask interrupt trigger trigger
register register event selection selection
register register register

To NVIC interrupt 23 23 23 23
controller
23

Pulse Edge detect Input

23 generator 23 23 circuit line

Event
mask
register
MS32662V1

10.2.3 Wakeup event management

The STM32F4xx are able to handle external or internal events in order to wake up the core
(WFE). The wakeup event can be generated either by:
• enabling an interrupt in the peripheral control register but not in the NVIC, and enabling
the SEVONPEND bit in the Cortex®-M4 with FPU System Control register. When the
MCU resumes from WFE, the peripheral interrupt pending bit and the peripheral NVIC
IRQ channel pending bit (in the NVIC interrupt clear pending register) have to be
cleared.
• or configuring an external or internal EXTI line in event mode. When the CPU resumes
from WFE, it is not necessary to clear the peripheral interrupt pending bit or the NVIC
IRQ channel pending bit as the pending bit corresponding to the event line is not set.
To use an external line as a wakeup event, refer to Section 10.2.4: Functional description.

236/1163 RM0402 Rev 6

RM0402 Interrupts and events

10.2.4 Functional description

To generate the interrupt, the interrupt line should be configured and enabled. This is done
by programming the two trigger registers with the desired edge detection and by enabling
the interrupt request by writing a ‘1’ to the corresponding bit in the interrupt mask register.
When the selected edge occurs on the external interrupt line, an interrupt request is
generated. The pending bit corresponding to the interrupt line is also set. This request is
reset by writing a ‘1’ in the pending register.
To generate the event, the event line should be configured and enabled. This is done by
programming the two trigger registers with the desired edge detection and by enabling the
event request by writing a ‘1’ to the corresponding bit in the event mask register. When the
selected edge occurs on the event line, an event pulse is generated. The pending bit
corresponding to the event line is not set.
An interrupt/event request can also be generated by software by writing a ‘1’ in the software
interrupt/event register.

Hardware interrupt selection

To configure the 23 lines as interrupt sources, use the following procedure:
• Configure the mask bits of the 23 interrupt lines (EXTI_IMR)
• Configure the Trigger selection bits of the interrupt lines (EXTI_RTSR and EXTI_FTSR)
• Configure the enable and mask bits that control the NVIC IRQ channel mapped to the
external interrupt controller (EXTI) so that an interrupt coming from one of the 23 lines
can be correctly acknowledged.

Hardware event selection

To configure the 23 lines as event sources, use the following procedure:
• Configure the mask bits of the 23 event lines (EXTI_EMR)
• Configure the Trigger selection bits of the event lines (EXTI_RTSR and EXTI_FTSR)

Software interrupt/event selection

The 23 lines can be configured as software interrupt/event lines. The following is the
procedure to generate a software interrupt.
• Configure the mask bits of the 23 interrupt/event lines (EXTI_IMR, EXTI_EMR)
• Set the required bit in the software interrupt register (EXTI_SWIER)

RM0402 Rev 6 237/1163

245
Interrupts and events RM0402

10.2.5 External interrupt/event line mapping

Up to STM32F412xx are connected to the 16 external interrupt/event lines in the following
manner:

Figure 30. External interrupt/event GPIO mapping

EXTI0[3:0] bits in the SYSCFG_EXTICR1 register

PA0
PB0
PC0
PD0 EXTI0
PE0
PF0
PG0
PH0

EXTI1[3:0] bits in the SYSCFG_EXTICR1 register

PA1
PB1
PC1
PD1 EXTI1
PE1
PF1
PG1
PH1
...

EXTI15[3:0] bits in the SYSCFG_EXTICR4 register

PA15
PB15
PC15 EXTI15
PD15
PE15
PF15
PG15

MSv39617V1

The five other EXTI lines are connected as follows:

• EXTI line 16 is connected to the PVD output
• EXTI line 17 is connected to the RTC Alarm event
• EXTI line 18 is connected to the USB OTG FS Wakeup event
• EXTI line 21 is connected to the RTC Tamper and TimeStamp events
• EXTI line 22 is connected to the RTC Wakeup event

238/1163 RM0402 Rev 6

RM0402 Interrupts and events

10.3 EXTI registers

Refer to Section 1.2: List of abbreviations for registers for a list of abbreviations used in
register descriptions.

10.3.1 Interrupt mask register (EXTI_IMR)

Address offset: 0x00
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. MR22 MR21 Res. Res. MR18 MR17 MR16
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

MR15 MR14 MR13 MR12 MR11 MR10 MR9 MR8 MR7 MR6 MR5 MR4 MR3 MR2 MR1 MR0

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bits 22:21 MR[22:21]: Interrupt mask on line x
0: Interrupt request from line x is masked
1: Interrupt request from line x is not masked
Bits 20:19 Reserved, must be kept at reset value.
Bits 18:0 MR[18:0]: Interrupt mask on line x
0: Interrupt request from line x is masked
1: Interrupt request from line x is not masked

10.3.2 Event mask register (EXTI_EMR)

Address offset: 0x04
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. MR22 MR21 Res. Res. MR18 MR17 MR16
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

MR15 MR14 MR13 MR12 MR11 MR10 MR9 MR8 MR7 MR6 MR5 MR4 MR3 MR2 MR1 MR0

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

RM0402 Rev 6 239/1163

245
Interrupts and events RM0402

Bits 22:21 MR[22:21]: Event mask on line x

0: Interrupt request from line x is masked
1: Interrupt request from line x is not masked
Bits 20:19 Reserved, must be kept at reset value.
Bits 18:0 MR[18:0]: Event mask on line x
0: Interrupt request from line x is masked
1: Interrupt request from line x is not masked

240/1163 RM0402 Rev 6

RM0402 Interrupts and events

10.3.3 Rising trigger selection register (EXTI_RTSR)

Address offset: 0x08
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. TR22 TR21 Res. Res. TR18 TR17 TR16
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

TR15 TR14 TR13 TR12 TR11 TR10 TR9 TR8 TR7 TR6 TR5 TR4 TR3 TR2 TR1 TR0

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bits 22:21 TR[22:21]: Rising trigger event configuration bit of line x
0: Rising trigger disabled (for Event and Interrupt) for input line
1: Rising trigger enabled (for Event and Interrupt) for input line
Bits 20:19 Reserved, must be kept at reset value.
Bits 18:0 TR[18:0]: Rising trigger event configuration bit of line x
0: Rising trigger disabled (for Event and Interrupt) for input line
1: Rising trigger enabled (for Event and Interrupt) for input line

Note: The external wakeup lines are edge triggered, no glitch must be generated on these lines.
If a rising edge occurs on the external interrupt line while writing to the EXTI_RTSR register,
the pending bit is be set.
Rising and falling edge triggers can be set for the same interrupt line. In this configuration,
both generate a trigger condition.

RM0402 Rev 6 241/1163

245
Interrupts and events RM0402

10.3.4 Falling trigger selection register (EXTI_FTSR)

Address offset: 0x0C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. TR22 TR21 Res. Res. TR18 TR17 TR16
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

TR15 TR14 TR13 TR12 TR11 TR10 TR9 TR8 TR7 TR6 TR5 TR4 TR3 TR2 TR1 TR0

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bits 22:21 TR[22:21]: Falling trigger event configuration bit of line x
0: Falling trigger disabled (for Event and Interrupt) for input line
1: Falling trigger enabled (for Event and Interrupt) for input line.
Bits 20:19 Reserved, must be kept at reset value.
Bits 18:0 TR[18:0]: Falling trigger event configuration bit of line x
0: Falling trigger disabled (for Event and Interrupt) for input line
1: Falling trigger enabled (for Event and Interrupt) for input line.

Note: The external wakeup lines are edge triggered, no glitch must be generated on these lines.
If a falling edge occurs on the external interrupt line while writing to the EXTI_FTSR register,
the pending bit is not set.
Rising and falling edge triggers can be set for the same interrupt line. In this configuration,
both generate a trigger condition.

242/1163 RM0402 Rev 6

RM0402 Interrupts and events

10.3.5 Software interrupt event register (EXTI_SWIER)

Address offset: 0x10
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SWIER SWIER SWIER SWIER SWIER
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
22 21 18 17 16
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER SWIER
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bits 22:21 SWIER[22:21]: Software Interrupt on line x
If interrupt are enabled on line x in the EXTI_IMR register, writing '1' to SWIERx bit when it is
set at '0' sets the corresponding pending bit in the EXTI_PR register, thus resulting in an
interrupt request generation.
This bit is cleared by clearing the corresponding bit in EXTI_PR (by writing a 1 to the bit).
Bits 20:19 Reserved, must be kept at reset value.
Bits 18:0 SWIER[18:0]: Software Interrupt on line x
If interrupt are enabled on line x in the EXTI_IMR register, writing '1' to SWIERx bit when it is
set at '0' sets the corresponding pending bit in the EXTI_PR register, thus resulting in an
interrupt request generation.
This bit is cleared by clearing the corresponding bit in EXTI_PR (by writing a 1 to the bit).

RM0402 Rev 6 243/1163

245
Interrupts and events RM0402

10.3.6 Pending register (EXTI_PR)

Address offset: 0x14
Reset value: undefined

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. PR22 PR21 Res. Res. PR18 PR17 PR16
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

PR15 PR14 PR13 PR12 PR11 PR10 PR9 PR8 PR7 PR6 PR5 PR4 PR3 PR2 PR1 PR0

rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bits 31:23 Reserved, must be kept at reset value.

Bits 22:21 PR[22:21]: Pending bit
0: No trigger request occurred
1: selected trigger request occurred
This bit is set when the selected edge event arrives on the external interrupt line.
This bit is cleared by programming it to ‘1’.
Bits 20:19 Reserved, must be kept at reset value.
Bits 18:0 PR[18:0]: Pending bit
0: No trigger request occurred
1: selected trigger request occurred
This bit is set when the selected edge event arrives on the external interrupt line.
This bit is cleared by programming it to ‘1’.

244/1163 RM0402 Rev 6

RM0402 Interrupts and events

10.3.7 EXTI register map

Table 41 gives the EXTI register map and the reset values.

Table 41. External interrupt/event controller register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
MR
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
EXTI_IMR MR[18:0]
0x00 [22:21]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

MR
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
EXTI_EMR MR[18:0]
0x04 [22:21]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

TR
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
EXTI_RTSR TR[18:0]
0x08 [22:21]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

TR
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.

EXTI_FTSR TR[18:0]
0x0C [22:21]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

SWIER
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.

EXTI_SWIER SWIER[18:0]
0x10 [22:21]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

PR
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.

EXTI_PR PR[18:0]
0x14 [22:21]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

RM0402 Rev 6 245/1163

245
Flexible static memory controller (FSMC) RM0402

11 Flexible static memory controller (FSMC)

11.1 Introduction
The flexible static memory controller (FSMC) includes one memory controller:
• The NOR/PSRAM memory controller

11.2 FSMC main features

The FSMC functional block makes the interface with: synchronous and asynchronous static
memories. Its main purposes are:
• to translate AHB transactions into the appropriate external device protocol
• to meet the access time requirements of the external memory devices
All external memories share the addresses, data and control signals with the controller.
Each external device is accessed by means of a unique chip select. The FSMC performs
only one access at a time to an external device.
The main features of the FSMC controller are the following:
• Interface with static-memory mapped devices including:
– Static random access memory (SRAM)
– NOR Flash memory/OneNAND Flash memory
– PSRAM (4 memory banks)
• Interface with parallel LCD modules, supporting Intel 8080 and Motorola 6800 modes.
• Burst mode support for faster access to synchronous devices such as NOR Flash
memory, PSRAM)
• Programmable continuous clock output for asynchronous and synchronous accesses
• 8-,16-bit wide data bus
• Independent chip select control for each memory bank
• Independent configuration for each memory bank
• Write enable and byte lane select outputs for use with PSRAM, SRAM devices
• External asynchronous wait control
• Write FIFO with 16 x32-bit depth
The Write FIFO is common to all memory controllers and consists of:
• a Write Data FIFO which stores the AHB data to be written to the memory (up to 32
bits) plus one bit for the AHB transfer (burst or not sequential mode)
• a Write Address FIFO which stores the AHB address (up to 28 bits) plus the AHB data
size (up to 2 bits). When operating in burst mode, only the start address is stored
except when crossing a page boundary (for PSRAM). In this case, the AHB burst is
broken into two FIFO entries.
The Write FIFO can be disabled by setting the WFDIS bit in the FSMC_BCR1 register.
At startup the FSMC pins must be configured by the user application. The FSMC I/O pins
which are not used by the application can be used for other purposes.

246/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

The FSMC registers that define the external device type and associated characteristics are
usually set at boot time and do not change until the next reset or power-up. However, the
settings can be changed at any time.

11.3 FMC block diagram

The FSMC consists of the following main blocks:
• The AHB interface (including the FSMC configuration registers)
• The NOR Flash/PSRAM/SRAM controller
The block diagram is shown in the figure below.

Figure 31. FSMC block diagram

FSMC interrupts to NVIC

NOR/PSRAM
FSMC_NL (or NADV)
signals
From clock FSMC_CLK
controller
HCLK NOR / PSRAM / SRAM
FSMC_NBL[1:0]
shared signals

FSMC_A[25:0]
Shared signals
FSMC_D[15:0]
NOR/PSRAM
Configuration
memory FSMC_NE[4:1]
registers
controller FSMC_NOE NOR / PSRAM / SRAM
FSMC_NWE shared signals
FSMC_NWAIT

MSv39279V2

RM0402 Rev 6 247/1163

287
Flexible static memory controller (FSMC) RM0402

11.4 AHB interface

The AHB slave interface allows internal CPUs and other bus master peripherals to access
the external memories.
AHB transactions are translated into the external device protocol. In particular, if the
selected external memory is 16- or 8-bit wide, 32-bit wide transactions on the AHB are split
into consecutive 16- or 8-bit accesses. The FSMC chip select (FSMC_NEx) does not toggle
between the consecutive accesses except in case of Access mode D when the Extended
mode is enabled.
The FSMC generates an AHB error in the following conditions:
• When reading or writing to an FSMC bank (Bank 1 to 4) which is not enabled.
• When reading or writing to the NOR Flash bank while the FACCEN bit is reset in the
FSMC_BCRx register.
The effect of an AHB error depends on the AHB master which has attempted the R/W
access:
• If the access has been attempted by the Cortex®-M4 with FPU CPU, a hard fault
interrupt is generated.
• If the access has been performed by a DMA controller, a DMA transfer error is
generated and the corresponding DMA channel is automatically disabled.
The AHB clock (HCLK) is the reference clock for the FSMC.

11.4.1 Supported memories and transactions

General transaction rules
The requested AHB transaction data size can be 8-, 16- or 32-bit wide whereas the
accessed external device has a fixed data width. This may lead to inconsistent transfers.
Therefore, some simple transaction rules must be followed:
• AHB transaction size and memory data size are equal
There is no issue in this case.
• AHB transaction size is greater than the memory size:
In this case, the FSMC splits the AHB transaction into smaller consecutive memory
accesses to meet the external data width. The FSMC chip select (FSMC_NEx) does
not toggle between the consecutive accesses.
• AHB transaction size is smaller than the memory size:
The transfer may or not be consistent depending on the type of external device:
– Accesses to devices that have the byte select feature (SRAM, ROM, PSRAM)
In this case, the FSMC allows read/write transactions and accesses the right data
through its byte lanes NBL[1:0].
Bytes to be written are addressed by NBL[1:0].
All memory bytes are read (NBL[1:0] are driven low during read transaction) and
the useless ones are discarded.
– Accesses to devices that do not have the byte select feature (NOR)
This situation occurs when a byte access is requested to a 16-bit wide Flash
memory. Since the device cannot be accessed in Byte mode (only 16-bit words
can be read/written from/to the Flash memory), Write transactions and Read

248/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

transactions are allowed (the controller reads the entire 16-bit memory word and
uses only the required byte).

Wrap support for NOR Flash/PSRAM

Wrap burst mode for synchronous memories is not supported. The memories must be
configured in Linear burst mode of undefined length.

Configuration registers
The FSMC can be configured through a set of registers. Refer to Section 11.6.6, for a
detailed description of the NOR Flash/PSRAM controller registers.

11.5 External device address mapping

From the FSMC point of view, the external memory is divided into fixed-size banks of
256 Mbytes each (see Figure 32):
• Bank 1 used to address up to 4 NOR Flash memory or PSRAM devices. This bank is
split into 4 NOR/PSRAM subbanks with 4 dedicated chip selects, as follows:
– Bank 1 - NOR/PSRAM 1
– Bank 1 - NOR/PSRAM 2
– Bank 1 - NOR/PSRAM 3
– Bank 1 - NOR/PSRAM 4
• Bank 3 used to address NAND Flash memory devices.The MPU memory attribute for
this space must be reconfigured by software to Device.
For each bank the type of memory to be used can be configured by the user application
through the Configuration register.

Figure 32. FSMC memory banks

Address Bank Supported memory type
0x6000 0000
Bank 1 NOR/PSRAM/SRAM
4 x 64 MB
0x6FFF FFFF
MS34476V1

11.5.1 NOR/PSRAM address mapping

HADDR[27:26] bits are used to select one of the four memory banks as shown in Table 42.

Table 42. NOR/PSRAM bank selection

HADDR[27:26](1) Selected bank

00 Bank 1 - NOR/PSRAM 1
01 Bank 1 - NOR/PSRAM 2
10 Bank 1 - NOR/PSRAM 3
11 Bank 1 - NOR/PSRAM 4

RM0402 Rev 6 249/1163

287
Flexible static memory controller (FSMC) RM0402

1. HADDR are internal AHB address lines that are translated to external memory.

The HADDR[25:0] bits contain the external memory address. Since HADDR is a byte
address whereas the memory is addressed at word level, the address actually issued to the
memory varies according to the memory data width, as shown in the following table.

Table 43. NOR/PSRAM External memory address

(1)
Memory width Data address issued to the memory Maximum memory capacity (bits)

8-bit HADDR[25:0] 64 Mbytes x 8 = 512 Mbit

16-bit HADDR[25:1] >> 1 64 Mbytes/2 x 16 = 512 Mbit
1. In case of a 16-bit external memory width, the FSMC internally uses HADDR[25:1] to generate the address
for external memory FSMC_A[24:0].
Whatever the external memory width, FSMC_A[0] should be connected to external memory address A[0].

11.6 NOR Flash/PSRAM controller

The FSMC generates the appropriate signal timings to drive the following types of
memories:
• Asynchronous SRAM and ROM
– 8 bits
– 16 bits
• PSRAM (CellularRAM™)
– Asynchronous mode
– Burst mode for synchronous accesses
– Multiplexed or non-multiplexed
• NOR Flash memory
– Asynchronous mode
– Burst mode for synchronous accesses
– Multiplexed or non-multiplexed
The FSMC outputs a unique chip select signal, NE[4:1], per bank. All the other signals
(addresses, data and control) are shared.
The FSMC supports a wide range of devices through a programmable timings among
which:
• Programmable wait states (up to 15)
• Programmable bus turnaround cycles (up to 15)
• Programmable output enable and write enable delays (up to 15)
• Independent read and write timings and protocol to support the widest variety of
memories and timings
• Programmable continuous clock (FSMC_CLK) output.
The FSMC Clock (FSMC_CLK) is a submultiple of the HCLK clock. It can be delivered to the
selected external device either during synchronous accesses only or during asynchronous

250/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

and synchronous accesses depending on the CCKEN bit configuration in the FSMC_BCR1
register:
• If the CCLKEN bit is reset, the FSMC generates the clock (CLK) only during
synchronous accesses (Read/write transactions).
• If the CCLKEN bit is set, the FSMC generates a continuous clock during asynchronous
and synchronous accesses. To generate the FSMC_CLK continuous clock, Bank 1
must be configured in Synchronous mode (see Section 11.6.6: NOR/PSRAM controller
registers). Since the same clock is used for all synchronous memories, when a
continuous output clock is generated and synchronous accesses are performed, the
AHB data size has to be the same as the memory data width (MWID) otherwise the
FSMC_CLK frequency is changed depending on AHB data transaction (refer to
Section 11.6.5: Synchronous transactions for FSMC_CLK divider ratio formula).
The size of each bank is fixed and equal to 64 Mbytes. Each bank is configured through
dedicated registers (see Section 11.6.6: NOR/PSRAM controller registers).
The programmable memory parameters include access times (see Table 44) and support
for wait management (for PSRAM and NOR Flash accessed in Burst mode).

Table 44. Programmable NOR/PSRAM access parameters

Parameter Function Access mode Unit Min. Max.

Address Duration of the address AHB clock cycle

Asynchronous 0 15
setup setup phase (HCLK)
Duration of the address hold Asynchronous, AHB clock cycle
Address hold 1 15
phase muxed I/Os (HCLK)
Duration of the data setup AHB clock cycle
Data setup Asynchronous 1 256
phase (HCLK)
Asynchronous and
Duration of the bus AHB clock cycle
Bust turn synchronous read 0 15
turnaround phase (HCLK)
/ write
Number of AHB clock cycles
Clock divide AHB clock cycle
(HCLK) to build one memory Synchronous 2 16
ratio (HCLK)
clock cycle (CLK)
Number of clock cycles to
Memory clock
Data latency issue to the memory before Synchronous 2 17
cycle (CLK)
the first data of the burst

11.6.1 External memory interface signals

Table 45, Table 46 and Table 47 list the signals that are typically used to interface with NOR
Flash memory, SRAM and PSRAM.
Note: The prefix “N” identifies the signals that are active low.

RM0402 Rev 6 251/1163

287
Flexible static memory controller (FSMC) RM0402

NOR Flash memory, non-multiplexed I/Os

Table 45. Non-multiplexed I/O NOR Flash memory

FSMC signal name I/O Function

CLK O Clock (for synchronous access)

A[25:0] O Address bus
D[15:0] I/O Bidirectional data bus
NE[x] O Chip select, x = 1..4
NOE O Output enable
NWE O Write enable
Latch enable (this signal is called address
NL(=NADV) O
valid, NADV, by some NOR Flash devices)
NWAIT I NOR Flash wait input signal to the FSMC

The maximum capacity is 512 Mbits (26 address lines).

NOR Flash memory, 16-bit multiplexed I/Os

Table 46. 16-bit multiplexed I/O NOR Flash memory

FSMC signal name I/O Function

CLK O Clock (for synchronous access)

A[25:16] O Address bus
16-bit multiplexed, bidirectional address/data bus (the 16-bit address
AD[15:0] I/O
A[15:0] and data D[15:0] are multiplexed on the databus)
NE[x] O Chip select, x = 1..4
NOE O Output enable
NWE O Write enable
Latch enable (this signal is called address valid, NADV, by some NOR
NL(=NADV) O
Flash devices)
NWAIT I NOR Flash wait input signal to the FSMC

The maximum capacity is 512 Mbits.

PSRAM/SRAM, non-multiplexed I/Os

Table 47. Non-multiplexed I/Os PSRAM/SRAM

FSMC signal
I/O Function
name

CLK O Clock (only for PSRAM synchronous access)

A[25:0] O Address bus
D[15:0] I/O Data bidirectional bus

252/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Table 47. Non-multiplexed I/Os PSRAM/SRAM (continued)

FSMC signal
I/O Function
name

NE[x] O Chip select, x = 1..4 (called NCE by PSRAM (CellularRAM™ i.e. CRAM))
NOE O Output enable
NWE O Write enable
NL(= NADV) O Address valid only for PSRAM input (memory signal name: NADV)
NWAIT I PSRAM wait input signal to the FSMC
NBL[1:0] O Byte lane output. Byte 0 and Byte 1 control (upper and lower byte enable)

The maximum capacity is 512 Mbits.

PSRAM, 16-bit multiplexed I/Os

Table 48. 16-Bit multiplexed I/O PSRAM

FSMC signal
I/O Function
name

CLK O Clock (for synchronous access)

A[25:16] O Address bus
16-bit multiplexed, bidirectional address/data bus (the 16-bit address
AD[15:0] I/O
A[15:0] and data D[15:0] are multiplexed on the databus)
NE[x] O Chip select, x = 1..4 (called NCE by PSRAM (CellularRAM™ i.e. CRAM))
NOE O Output enable
NWE O Write enable
NL(= NADV) O Address valid PSRAM input (memory signal name: NADV)
NWAIT I PSRAM wait input signal to the FSMC
NBL[1:0] O Byte lane output. Byte 0 and Byte 1 control (upper and lower byte enable)

The maximum capacity is 512 Mbits (26 address lines).

11.6.2 Supported memories and transactions

Table 49 below shows an example of the supported devices, access modes and
transactions when the memory data bus is 16-bit wide for NOR Flash memory, PSRAM and
SRAM. The transactions not allowed (or not supported) by the FSMC are shown in gray in
this example.

RM0402 Rev 6 253/1163

287
Flexible static memory controller (FSMC) RM0402

Table 49. NOR Flash/PSRAM: example of supported memories

and transactions
AHB Allowed/
Memory
Device Mode R/W data not Comments
data size
size allowed

Asynchronous R 8 16 Y -
Asynchronous W 8 16 N -
Asynchronous R 16 16 Y -
Asynchronous W 16 16 Y -
NOR Flash Asynchronous R 32 16 Y Split into 2 FSMC accesses
(muxed I/Os
and nonmuxed Asynchronous W 32 16 Y Split into 2 FSMC accesses
I/Os) Asynchronous
R - 16 N Mode is not supported
page
Synchronous R 8 16 N -
Synchronous R 16 16 Y -
Synchronous R 32 16 Y -
Asynchronous R 8 16 Y -
Asynchronous W 8 16 Y Use of byte lanes NBL[1:0]
Asynchronous R 16 16 Y -
Asynchronous W 16 16 Y -
Asynchronous R 32 16 Y Split into 2 FSMC accesses
PSRAM
(multiplexed Asynchronous W 32 16 Y Split into 2 FSMC accesses
I/Os and non- Asynchronous
multiplexed R - 16 N Mode is not supported
page
I/Os)
Synchronous R 8 16 N -
Synchronous R 16 16 Y -
Synchronous R 32 16 Y -
Synchronous W 8 16 Y Use of byte lanes NBL[1:0]
Synchronous W 16/32 16 Y -
Asynchronous R 8 / 16 16 Y -
Asynchronous W 8 / 16 16 Y Use of byte lanes NBL[1:0]
SRAM and
ROM Asynchronous R 32 16 Y Split into 2 FSMC accesses
Split into 2 FSMC accesses
Asynchronous W 32 16 Y
Use of byte lanes NBL[1:0]

254/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

11.6.3 General timing rules

Signals synchronization
• All controller output signals change on the rising edge of the internal clock (HCLK)
• In Synchronous mode (read or write), all output signals change on the rising edge of
HCLK. Whatever the CLKDIV value, all outputs change as follows:
– NOEL/NWEL/ NEL/NADVL/ NADVH /NBLL/ Address valid outputs change on the
falling edge of FSMC_CLK clock.
– NOEH/ NWEH / NEH/ NOEH/NBLH/ Address invalid outputs change on the rising
edge of FSMC_CLK clock.

11.6.4 NOR Flash/PSRAM controller asynchronous transactions

Asynchronous static memories (NOR Flash, PSRAM, SRAM)
• Signals are synchronized by the internal clock HCLK. This clock is not issued to the
memory
• The FSMC always samples the data before de-asserting the NOE signal. This
guarantees that the memory data hold timing constraint is met (minimum Chip Enable
high to data transition is usually 0 ns)
• If the Extended mode is enabled (EXTMOD bit is set in the FSMC_BCRx register), up
to four extended modes (A, B, C and D) are available. It is possible to mix A, B, C and
D modes for read and write operations. For example, read operation can be performed
in mode A and write in mode B.
• If the Extended mode is disabled (EXTMOD bit is reset in the FSMC_BCRx register),
the FSMC can operate in mode 1 or mode 2 as follows:
– Mode 1 is the default mode when SRAM/PSRAM memory type is selected (MTYP
= 0x0 or 0x01 in the FSMC_BCRx register)
– Mode 2 is the default mode when NOR memory type is selected (MTYP = 0x10 in
the FSMC_BCRx register).

RM0402 Rev 6 255/1163

287
Flexible static memory controller (FSMC) RM0402

Mode 1 - SRAM/PSRAM (CRAM)

The next figures show the read and write transactions for the supported modes followed by
the required configuration of FSMC_BCRx, and FSMC_BTRx/FSMC_BWTRx registers.

Figure 33. Mode 1 read access waveforms

Memory transaction

A[25:0]

NBL[1:0]

NEx

NOE

NWE
High

data driven
D[15:0] by memory

ADDSET DATAST
HCLK cycles HCLK cycles

MS34477V1

Figure 34. Mode 1 write access waveforms

Memory transaction

A[25:0]

NBL[1:0]

NEx

NOE

1HCLK
NWE

D[15:0] data driven by FSMC

ADDSET (DATAST + 1)
HCLK cycles HCLK cycles

MSv40113V1

256/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

The one HCLK cycle at the end of the write transaction helps guarantee the address and
data hold time after the NWE rising edge. Due to the presence of this HCLK cycle, the
DATAST value must be greater than zero (DATAST > 0).

Table 50. FSMC_BCRx bitfields (mode 1)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x0 (no effect in Asynchronous mode)
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT Set to 1 if the memory supports this feature. Otherwise keep at 0.
14 EXTMOD 0x0
13 WAITEN 0x0 (no effect in Asynchronous mode)
12 WREN As needed
11 Reserved 0x0
10 Reserved 0x0
9 WAITPOL Meaningful only if bit 15 is 1
8 BURSTEN 0x0
7 Reserved 0x1
6 FACCEN Don’t care
5:4 MWID As needed
3:2 MTYP As needed, exclude 0x2 (NOR Flash memory)
1 MUXE 0x0
0 MBKEN 0x1

Table 51. FSMC_BTRx bitfields (mode 1)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD Don’t care
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST+1 HCLK cycles for
15:8 DATAST
write accesses, DATAST HCLK cycles for read accesses).
7:4 ADDHLD Don’t care
Duration of the first access phase (ADDSET HCLK cycles).
3:0 ADDSET
Minimum value for ADDSET is 0.

RM0402 Rev 6 257/1163

287
Flexible static memory controller (FSMC) RM0402

Mode A - SRAM/PSRAM (CRAM) OE toggling

Figure 35. Mode A read access waveforms

Memory transaction

A[25:0]

NBL[1:0]

NEx

NOE

NWE
High

D[15:0] data driven

by memory

ADDSET DATAST
HCLK cycles HCLK cycles

MS34479V1

1. NBL[1:0] are driven low during the read access

Figure 36. Mode A write access waveforms

Memory transaction

A[25:0]

NBL[1:0]

NEx

NOE

1HCLK
NWE

D[15:0] data driven by FSMC

ADDSET (DATAST + 1)
HCLK cycles HCLK cycles
MS34480V1

258/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

The differences compared with Mode 1 are the toggling of NOE and the independent read
and write timings.

Table 52. FSMC_BCRx bitfields (mode A)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x0 (no effect in Asynchronous mode)
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT Set to 1 if the memory supports this feature. Otherwise keep at 0.
14 EXTMOD 0x1
13 WAITEN 0x0 (no effect in Asynchronous mode)
12 WREN As needed
11 WAITCFG Don’t care
10 Reserved 0x0
9 WAITPOL Meaningful only if bit 15 is 1
8 BURSTEN 0x0
7 Reserved 0x1
6 FACCEN Don’t care
5:4 MWID As needed
3:2 MTYP As needed, exclude 0x2 (NOR Flash memory)
1 MUXEN 0x0
0 MBKEN 0x1

Table 53. FSMC_BTRx bitfields (mode A)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x0
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST HCLK cycles) for read
15:8 DATAST
accesses.
7:4 ADDHLD Don’t care
Duration of the first access phase (ADDSET HCLK cycles) for read
3:0 ADDSET accesses.
Minimum value for ADDSET is 0.

RM0402 Rev 6 259/1163

287
Flexible static memory controller (FSMC) RM0402

Table 54. FSMC_BWTRx bitfields (mode A)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x0
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST HCLK cycles) for write
15:8 DATAST
accesses.
7:4 ADDHLD Don’t care
Duration of the first access phase (ADDSET HCLK cycles) for write
3:0 ADDSET accesses.
Minimum value for ADDSET is 0.

Mode 2/B - NOR Flash

Figure 37. Mode 2 and mode B read access waveforms

Memory transaction

A[25:0]

NADV

NEx

NOE

NWE
High

D[15:0] data driven

by memory

ADDSET DATAST
HCLK cycles HCLK cycles
MS34481V2

260/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Figure 38. Mode 2 write access waveforms

Memory transaction

A[25:0]

NADV

NEx

NOE

1HCLK
NWE

D[15:0] data driven by FMC

ADDSET (DATAST + 1)
HCLK cycles HCLK cycles
MS34482V2

Figure 39. Mode B write access waveforms

Memory transaction

A[25:0]

NADV

NEx

NOE

1HCLK
NWE

D[15:0] data driven by FSMC

ADDSET (DATAST + 1)
HCLK cycles HCLK cycles
MSv40114V1

The differences with mode 1 are the toggling of NWE and the independent read and write
timings when extended mode is set (mode B).

RM0402 Rev 6 261/1163

287
Flexible static memory controller (FSMC) RM0402

Table 55. FSMC_BCRx bitfields (mode 2/B)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x0 (no effect in Asynchronous mode)
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT Set to 1 if the memory supports this feature. Otherwise keep at 0.
14 EXTMOD 0x1 for mode B, 0x0 for mode 2
13 WAITEN 0x0 (no effect in Asynchronous mode)
12 WREN As needed
11 WAITCFG Don’t care
10 Reserved 0x0
9 WAITPOL Meaningful only if bit 15 is 1
8 BURSTEN 0x0
7 Reserved 0x1
6 FACCEN 0x1
5:4 MWID As needed
3:2 MTYP 0x2 (NOR Flash memory)
1 MUXEN 0x0
0 MBKEN 0x1

Table 56. FSMC_BTRx bitfields (mode 2/B)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x1 if Extended mode is set
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the access second phase (DATAST HCLK cycles) for
15:8 DATAST
read accesses.
7:4 ADDHLD Don’t care
Duration of the access first phase (ADDSET HCLK cycles) for read
3:0 ADDSET
accesses. Minimum value for ADDSET is 0.

262/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Table 57. FSMC_BWTRx bitfields (mode 2/B)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x1 if Extended mode is set
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the access second phase (DATAST HCLK cycles) for
15:8 DATAST
write accesses.
7:4 ADDHLD Don’t care
Duration of the access first phase (ADDSET HCLK cycles) for write
3:0 ADDSET
accesses. Minimum value for ADDSET is 0.

Note: The FSMC_BWTRx register is valid only if the Extended mode is set (mode B), otherwise its
content is don’t care.

Mode C - NOR Flash - OE toggling

Figure 40. Mode C read access waveforms

Memory transaction

A[25:0]

NADV

NEx

NOE

NWE
High

D[15:0] data driven

by memory

ADDSET DATAST
HCLK cycles HCLK cycles
MS34484V1

RM0402 Rev 6 263/1163

287
Flexible static memory controller (FSMC) RM0402

Figure 41. Mode C write access waveforms

Memory transaction

A[25:0]

NADV

NEx

NOE

1HCLK
NWE

D[15:0] data driven by FSMC

ADDSET (DATAST + 1)
HCLK cycles HCLK cycles
MS34485V1

The differences compared with mode 1 are the toggling of NOE and the independent read
and write timings.

Table 58. FSMC_BCRx bitfields (mode C)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x0 (no effect in Asynchronous mode)
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT Set to 1 if the memory supports this feature. Otherwise keep at 0.
14 EXTMOD 0x1
13 WAITEN 0x0 (no effect in Asynchronous mode)
12 WREN As needed
11 WAITCFG Don’t care
10 Reserved 0x0
9 WAITPOL Meaningful only if bit 15 is 1
8 BURSTEN 0x0
7 Reserved 0x1
6 FACCEN 0x1
5:4 MWID As needed

264/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Table 58. FSMC_BCRx bitfields (mode C) (continued)

Bit number Bit name Value to set

3:2 MTYP 0x02 (NOR Flash memory)

1 MUXEN 0x0
0 MBKEN 0x1

Table 59. FSMC_BTRx bitfields (mode C)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x2
27:24 DATLAT 0x0
23:20 CLKDIV 0x0
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST HCLK cycles) for
15:8 DATAST
read accesses.
7:4 ADDHLD Don’t care
Duration of the first access phase (ADDSET HCLK cycles) for read
3:0 ADDSET
accesses. Minimum value for ADDSET is 0.

Table 60. FSMC_BWTRx bitfields (mode C)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x2
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST HCLK cycles) for
15:8 DATAST
write accesses.
7:4 ADDHLD Don’t care
Duration of the first access phase (ADDSET HCLK cycles) for write
3:0 ADDSET
accesses. Minimum value for ADDSET is 0.

RM0402 Rev 6 265/1163

287
Flexible static memory controller (FSMC) RM0402

Mode D - asynchronous access with extended address

Figure 42. Mode D read access waveforms

Memory transaction

A[25:0]

NADV

NEx

NOE

NWE
High

D[15:0] data driven

by memory

ADDSET DATAST
HCLK cycles HCLK cycles
ADDHLD
HCLK cycles MS34486V1

Figure 43. Mode D write access waveforms

266/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

The differences with mode 1 are the toggling of NOE that goes on toggling after NADV
changes and the independent read and write timings.

Table 61. FSMC_BCRx bitfields (mode D)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x0 (no effect in Asynchronous mode)
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT Set to 1 if the memory supports this feature. Otherwise keep at 0.
14 EXTMOD 0x1
13 WAITEN 0x0 (no effect in Asynchronous mode)
12 WREN As needed
11 WAITCFG Don’t care
10 Reserved 0x0
9 WAITPOL Meaningful only if bit 15 is 1
8 BURSTEN 0x0
7 Reserved 0x1
6 FACCEN Set according to memory support
5:4 MWID As needed
3:2 MTYP As needed
1 MUXEN 0x0
0 MBKEN 0x1

Table 62. FSMC_BTRx bitfields (mode D)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x3
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST HCLK cycles) for read
15:8 DATAST
accesses.
Duration of the middle phase of the read access (ADDHLD HCLK
7:4 ADDHLD
cycles)
Duration of the first access phase (ADDSET HCLK cycles) for read
3:0 ADDSET
accesses. Minimum value for ADDSET is 1.

RM0402 Rev 6 267/1163

287
Flexible static memory controller (FSMC) RM0402

Table 63. FSMC_BWTRx bitfields (mode D)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x3
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST + 1 HCLK cycles) for
15:8 DATAST
write accesses.
Duration of the middle phase of the write access (ADDHLD HCLK
7:4 ADDHLD
cycles)
Duration of the first access phase (ADDSET HCLK cycles) for write
3:0 ADDSET
accesses. Minimum value for ADDSET is 1.

Muxed mode - multiplexed asynchronous access to NOR Flash memory

Figure 44. Muxed read access waveforms

268/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Figure 45. Muxed write access waveforms

Memory transaction

A[25:16]

NADV

NEx

NOE

1HCLK
NWE

AD[15:0] Lower address data driven by FSMC

ADDSET ADDHLD (DATAST + 1)

HCLK cycles HCLK cycles HCLK cycles
MSv40112V2

The difference with mode D is the drive of the lower address byte(s) on the data bus.

Table 64. FSMC_BCRx bitfields (Muxed mode)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x0 (no effect in Asynchronous mode)
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT Set to 1 if the memory supports this feature. Otherwise keep at 0.
14 EXTMOD 0x0
13 WAITEN 0x0 (no effect in Asynchronous mode)
12 WREN As needed
11 WAITCFG Don’t care
10 Reserved 0x0
9 WAITPOL Meaningful only if bit 15 is 1
8 BURSTEN 0x0
7 Reserved 0x1
6 FACCEN 0x1
5:4 MWID As needed

RM0402 Rev 6 269/1163

287
Flexible static memory controller (FSMC) RM0402

Table 64. FSMC_BCRx bitfields (Muxed mode) (continued)

Bit number Bit name Value to set

3:2 MTYP 0x2 (NOR Flash memory) or 0x1(PSRAM)

1 MUXEN 0x1
0 MBKEN 0x1

Table 65. FSMC_BTRx bitfields (Muxed mode)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x0
27:24 DATLAT Don’t care
23:20 CLKDIV Don’t care
19:16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
Duration of the second access phase (DATAST HCLK cycles for read
15:8 DATAST
accesses and DATAST+1 HCLK cycles for write accesses).
7:4 ADDHLD Duration of the middle phase of the access (ADDHLD HCLK cycles).
Duration of the first access phase (ADDSET HCLK cycles). Minimum
3:0 ADDSET
value for ADDSET is 1.

WAIT management in asynchronous accesses

If the asynchronous memory asserts the WAIT signal to indicate that it is not yet ready to
accept or to provide data, the ASYNCWAIT bit has to be set in FSMC_BCRx register.
If the WAIT signal is active (high or low depending on the WAITPOL bit), the second access
phase (Data setup phase), programmed by the DATAST bits, is extended until WAIT
becomes inactive. Unlike the data setup phase, the first access phases (Address setup and
Address hold phases), programmed by the ADDSET and ADDHLD bits, are not WAIT
sensitive and so they are not prolonged.
The data setup phase must be programmed so that WAIT can be detected 4 HCLK cycles
before the end of the memory transaction. The following cases must be considered:

270/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

1. The memory asserts the WAIT signal aligned to NOE/NWE which toggles:
DATAST ≥ ( 4 × HCLK ) + max_wait_assertion_time

2. The memory asserts the WAIT signal aligned to NEx (or NOE/NWE not toggling):
if
max_wait_assertion_time > address_phase + hold_phase

then:

DATAST ≥ ( 4 × HCLK ) + ( max_wait_assertion_time – address_phase – hold_phase )

otherwise
DATAST ≥ 4 × HCLK

where max_wait_assertion_time is the maximum time taken by the memory to assert

the WAIT signal once NEx/NOE/NWE is low.
Figure 46 and Figure 47 show the number of HCLK clock cycles that are added to the
memory access phase after WAIT is released by the asynchronous memory (independently
of the above cases).

Figure 46. Asynchronous wait during a read access waveforms

Memory transaction

A[25:0]

address phase data setup phase

NEx

NWAIT don’t care don’t care

NOE

D[15:0] data driven by memory

4HCLK

MS30463V2

1. NWAIT polarity depends on WAITPOL bit setting in FSMC_BCRx register.

RM0402 Rev 6 271/1163

287
Flexible static memory controller (FSMC) RM0402

Figure 47. Asynchronous wait during a write access waveforms

Memory transaction

A[25:0]

address phase data setup phase

NEx

NWAIT don’t care don’t care

1HCLK

NWE

D[15:0] data driven by FSMC

3HCLK

MS30464V2

1. NWAIT polarity depends on WAITPOL bit setting in FSMC_BCRx register.

11.6.5 Synchronous transactions

The memory clock, FSMC_CLK, is a submultiple of HCLK. It depends on the value of
CLKDIV and the MWID/ AHB data size, following the formula given below:

FSMC_CLK divider ratio = max (CLKDIV + 1,MWID ( AHB data size ))

Whatever MWID size: 16 or 8-bit, the FSMC_CLK divider ratio is always defined by the
programmed CLKDIV value.
Example:
• If CLKDIV=1, MWID = 16 bits, AHB data size=8 bits, FSMC_CLK=HCLK/2.
NOR Flash memories specify a minimum time from NADV assertion to CLK high. To meet
this constraint, the FSMC does not issue the clock to the memory during the first internal
clock cycle of the synchronous access (before NADV assertion). This guarantees that the
rising edge of the memory clock occurs in the middle of the NADV low pulse.

Data latency versus NOR memory latency

The data latency is the number of cycles to wait before sampling the data. The DATLAT
value must be consistent with the latency value specified in the NOR Flash configuration
register. The FSMC does not include the clock cycle when NADV is low in the data latency
count.

272/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Caution: Some NOR Flash memories include the NADV Low cycle in the data latency count, so that
the exact relation between the NOR Flash latency and the FSMC DATLAT parameter can be
either:
• NOR Flash latency = (DATLAT + 2) CLK clock cycles
• or NOR Flash latency = (DATLAT + 3) CLK clock cycles
Some recent memories assert NWAIT during the latency phase. In such cases DATLAT can
be set to its minimum value. As a result, the FSMC samples the data and waits long enough
to evaluate if the data are valid. Thus the FSMC detects when the memory exits latency and
real data are processed.
Other memories do not assert NWAIT during latency. In this case the latency must be set
correctly for both the FSMC and the memory, otherwise invalid data are mistaken for good
data, or valid data are lost in the initial phase of the memory access.

Single-burst transfer
When the selected bank is configured in Burst mode for synchronous accesses, if for
example an AHB single-burst transaction is requested on 16-bit memories, the FSMC
performs a burst transaction of length 1 (if the AHB transfer is 16 bits), or length 2 (if the
AHB transfer is 32 bits) and de-assert the chip select signal when the last data is strobed.
Such transfers are not the most efficient in terms of cycles compared to asynchronous read
operations. Nevertheless, a random asynchronous access would first require to re-program
the memory access mode, which would altogether last longer.

Cross boundary page for CellularRAM™ 1.5

CellularRAM™ 1.5 does not allow burst access to cross the page boundary. The FSMC
controller allows to split automatically the burst access when the memory page size is
reached by configuring the CPSIZE bits in the FSMC_BCR1 register following the memory
page size.

Wait management
For synchronous NOR Flash memories, NWAIT is evaluated after the programmed latency
period, which corresponds to (DATLAT+2) CLK clock cycles.
If NWAIT is active (low level when WAITPOL = 0, high level when WAITPOL = 1), wait
states are inserted until NWAIT is inactive (high level when WAITPOL = 0, low level when
WAITPOL = 1).
When NWAIT is inactive, the data is considered valid either immediately (bit WAITCFG = 1)
or on the next clock edge (bit WAITCFG = 0).
During wait-state insertion via the NWAIT signal, the controller continues to send clock
pulses to the memory, keeping the chip select and output enable signals valid. It does not
consider the data as valid.
In Burst mode, there are two timing configurations for the NOR Flash NWAIT signal:
• The Flash memory asserts the NWAIT signal one data cycle before the wait state
(default after reset).
• The Flash memory asserts the NWAIT signal during the wait state
The FSMC supports both NOR Flash wait state configurations, for each chip select, thanks
to the WAITCFG bit in the FSMC_BCRx registers (x = 0..3).

RM0402 Rev 6 273/1163

287
Flexible static memory controller (FSMC) RM0402

Figure 48. Wait configuration waveforms

Memory transaction = burst of 4 half words

HCLK

CLK

A[25:16] addr[25:16]

NADV

NWAIT
(WAITCFG = 0)

NWAIT
(WAITCFG = 1)
inserted wait state

A/D[15:0] addr[15:0] data data data

ai15798c

274/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Figure 49. Synchronous multiplexed read mode waveforms - NOR, PSRAM (CRAM)

Memory transaction = burst of 4 half words

HCLK

CLK

A[25:16] addr[25:16]

NEx

NOE

High
NWE

NADV

NWAIT
(WAITCFG=
0)
(DATLAT + 2) inserted wait state
CLK cycles
A/D[15:0] Addr[15:0] data data data data

1 clock 1 clock
cycle cycle
Data strobes Data strobes
ai17723f

1. Byte lane outputs (NBL are not shown; for NOR access, they are held high, and, for PSRAM (CRAM)
access, they are held low.

Table 66. FSMC_BCRx bitfields (Synchronous multiplexed read mode)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW No effect on synchronous read
18:16 CPSIZE 0x0 (no effect in Asynchronous mode)
15 ASYNCWAIT 0x0
14 EXTMOD 0x0
To be set to 1 if the memory supports this feature, to be kept at 0
13 WAITEN
otherwise
12 WREN No effect on synchronous read
11 WAITCFG To be set according to memory
10 Reserved 0x0

RM0402 Rev 6 275/1163

287
Flexible static memory controller (FSMC) RM0402

Table 66. FSMC_BCRx bitfields (Synchronous multiplexed read mode) (continued)

Bit number Bit name Value to set

9 WAITPOL To be set according to memory

8 BURSTEN 0x1
7 Reserved 0x1
6 FACCEN Set according to memory support (NOR Flash memory)
5-4 MWID As needed
3-2 MTYP 0x1 or 0x2
1 MUXEN As needed
0 MBKEN 0x1

Table 67. FSMC_BTRx bitfields (Synchronous multiplexed read mode)

Bit number Bit name Value to set

31:30 Reserved 0x0

29:28 ACCMOD 0x0
27-24 DATLAT Data latency
27-24 DATLAT Data latency
0x0 to get CLK = HCLK
23-20 CLKDIV 0x1 to get CLK = 2 × HCLK
..
19-16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
15-8 DATAST Don’t care
7-4 ADDHLD Don’t care
3-0 ADDSET Don’t care

276/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Figure 50. Synchronous multiplexed write mode waveforms - PSRAM (CRAM)

Memory transaction = burst of 2 half words

HCLK

CLK

A[25:16] addr[25:16]

NEx

Hi-Z
NOE

NWE

NADV

NWAIT
(WAITCFG = 0)

(DATLAT + 2) inserted wait state

CLK cycles
A/D[15:0] Addr[15:0] data data

1 clock 1 clock ai14731f

1. The memory must issue NWAIT signal one cycle in advance, accordingly WAITCFG must be programmed to 0.
2. Byte Lane (NBL) outputs are not shown, they are held low while NEx is active.

Table 68. FSMC_BCRx bitfields (Synchronous multiplexed write mode)

Bit number Bit name Value to set

31:22 Reserved 0x000

21 WFDIS As needed
20 CCLKEN As needed
19 CBURSTRW 0x1
18:16 CPSIZE As needed (0x1 for CRAM 1.5)
15 ASYNCWAIT 0x0
14 EXTMOD 0x0
To be set to 1 if the memory supports this feature, to be kept at 0
13 WAITEN
otherwise.

RM0402 Rev 6 277/1163

287
Flexible static memory controller (FSMC) RM0402

Table 68. FSMC_BCRx bitfields (Synchronous multiplexed write mode) (continued)

Bit number Bit name Value to set

12 WREN 0x1
11 WAITCFG 0x0
10 Reserved 0x0
9 WAITPOL to be set according to memory
8 BURSTEN no effect on synchronous write
7 Reserved 0x1
6 FACCEN Set according to memory support
5-4 MWID As needed
3-2 MTYP 0x1
1 MUXEN As needed
0 MBKEN 0x1

Table 69. FSMC_BTRx bitfields (Synchronous multiplexed write mode)

Bit number Bit name Value to set

31-30 Reserved 0x0

29:28 ACCMOD 0x0
27-24 DATLAT Data latency
0x0 to get CLK = HCLK
23-20 CLKDIV
0x1 to get CLK = 2 × HCLK
19-16 BUSTURN Time between NEx high to NEx low (BUSTURN HCLK).
15-8 DATAST Don’t care
7-4 ADDHLD Don’t care
3-0 ADDSET Don’t care

278/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

11.6.6 NOR/PSRAM controller registers

SRAM/NOR-Flash chip-select control register for bank x
(FSMC_BCRx) (x = 1 to 4)
Address offset: 8 * (x – 1), (x = 1 to 4)
Reset value: Bank 1: 0x0000 30DB
Reset value: Bank 2: 0x0000 30D2
Reset value: Bank 3: 0x0000 30D2
Reset value: Bank 4: 0x0000 30D2
This register contains the control information of each memory bank, used for SRAMs,
PSRAM and NOR Flash memories.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CCLK CBURST
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. WFDIS CPSIZE[2:0]
EN RW

rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ASYNC EXT WAIT WAIT WAIT BURST FACC MUX MBK
WREN Res. Res. MWID[1:0] MTYP[1:0]
WAIT MOD EN CFG POL EN EN EN EN

rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:22 Reserved, must be kept at reset value.

Bit 21 WFDIS: Write FIFO disable
This bit disables the Write FIFO used by the FSMC controller.
0: Write FIFO enabled (Default after reset)
1: Write FIFO disabled
Note: The WFDIS bit of the FSMC_BCR2..4 registers is don’t care. It is only enabled through the
FSMC_BCR1 register.
Bit 20 CCLKEN: Continuous clock enable
This bit enables the FSMC_CLK clock output to external memory devices.
0: The FSMC_CLK is only generated during the synchronous memory access (read/write
transaction). The FSMC_CLK clock ratio is specified by the programmed CLKDIV value in the
FSMC_BCRx register (default after reset).
1: The FSMC_CLK is generated continuously during asynchronous and synchronous access. The
FSMC_CLK clock is activated when the CCLKEN is set.
Note: The CCLKEN bit of the FSMC_BCR2..4 registers is don’t care. It is only enabled through the
FSMC_BCR1 register. Bank 1 must be configured in Synchronous mode to generate the
FSMC_CLK continuous clock.
Note: If CCLKEN bit is set, the FSMC_CLK clock ratio is specified by CLKDIV value in the
FSMC_BTR1 register. CLKDIV in FSMC_BWTR1 is don’t care.
Note: If the Synchronous mode is used and CCLKEN bit is set, the synchronous memories
connected to other banks than Bank 1 are clocked by the same clock (the CLKDIV value in the
FSMC_BTR2..4 and FSMC_BWTR2..4 registers for other banks has no effect.)

RM0402 Rev 6 279/1163

287
Flexible static memory controller (FSMC) RM0402

Bit 19 CBURSTRW: Write burst enable

For PSRAM (CRAM) operating in Burst mode, the bit enables synchronous accesses during write
operations. The enable bit for synchronous read accesses is the BURSTEN bit in the FSMC_BCRx
register.
0: Write operations are always performed in Asynchronous mode.
1: Write operations are performed in Synchronous mode.
Bits 18:16 CPSIZE[2:0]: CRAM page size
These are used for CellularRAM™ 1.5 which does not allow burst access to cross the address
boundaries between pages. When these bits are configured, the FSMC controller splits
automatically the burst access when the memory page size is reached (refer to memory datasheet
for page size).
000: No burst split when crossing page boundary (default after reset)
001: 128 bytes
010: 256 bytes
011: 512 bytes
100: 1024 bytes
Others: reserved
Bit 15 ASYNCWAIT: Wait signal during asynchronous transfers
This bit enables/disables the FSMC to use the wait signal even during an asynchronous protocol.
0: NWAIT signal is not taken in to account when running an asynchronous protocol (default after
reset).
1: NWAIT signal is taken in to account when running an asynchronous protocol.
Bit 14 EXTMOD: Extended mode enable
This bit enables the FSMC to program the write timings for non multiplexed asynchronous accesses
inside the FSMC_BWTR register, thus resulting in different timings for read and write operations.
0: values inside FSMC_BWTR register are not taken into account (default after reset)
1: values inside FSMC_BWTR register are taken into account
Note: When the Extended mode is disabled, the FSMC can operate in mode 1 or mode 2 as follows:
– Mode 1 is the default mode when the SRAM/PSRAM memory type is selected
(MTYP = 0x0 or 0x01)
– Mode 2 is the default mode when the NOR memory type is selected (MTYP = 0x10).
Bit 13 WAITEN: Wait enable bit
This bit enables/disables wait-state insertion via the NWAIT signal when accessing the memory in
Synchronous mode.
0: NWAIT signal is disabled (its level not taken into account, no wait state inserted after the
programmed Flash latency period).
1: NWAIT signal is enabled (its level is taken into account after the programmed latency period to
insert wait states if asserted) (default after reset).
Bit 12 WREN: Write enable bit
This bit indicates whether write operations are enabled/disabled in the bank by the FSMC.
0: Write operations are disabled in the bank by the FSMC, an AHB error is reported.
1: Write operations are enabled for the bank by the FSMC (default after reset).
Bit 11 WAITCFG: Wait timing configuration
The NWAIT signal indicates whether the data from the memory are valid or if a wait state must be
inserted when accessing the memory in Synchronous mode. This configuration bit determines if
NWAIT is asserted by the memory one clock cycle before the wait state or during the wait state:
0: NWAIT signal is active one data cycle before wait state (default after reset).
1: NWAIT signal is active during wait state (not used for PSRAM).
Bit 10 Reserved, must be kept at reset value.

280/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Bit 9 WAITPOL: Wait signal polarity bit

Defines the polarity of the wait signal from memory used for either in Synchronous or Asynchronous
mode.
0: NWAIT active low (default after reset)
1: NWAIT active high
Bit 8 BURSTEN: Burst enable bit
This bit enables/disables synchronous accesses during read operations. It is valid only for
synchronous memories operating in Burst mode.
0: Burst mode disabled (default after reset). Read accesses are performed in Asynchronous mode.
1: Burst mode enable. Read accesses are performed in Synchronous mode.
Bit 7 Reserved, must be kept at reset value.
Bit 6 FACCEN: Flash access enable
Enables NOR Flash memory access operations.
0: Corresponding NOR Flash memory access is disabled.
1: Corresponding NOR Flash memory access is enabled (default after reset).
Bits 5:4 MWID[1:0]: Memory data bus width
Defines the external memory device width, valid for all type of memories.
00: 8 bits
01: 16 bits (default after reset)
10: reserved
11: reserved
Bits 3:2 MTYP[1:0]: Memory type
Defines the type of external memory attached to the corresponding memory bank.
00: SRAM (default after reset for Bank 2...4)
01: PSRAM (CRAM)
10: NOR Flash/OneNAND Flash (default after reset for Bank 1)
11: reserved
Bit 1 MUXEN: Address/data multiplexing enable bit
When this bit is set, the address and data values are multiplexed on the data bus, valid only with
NOR and PSRAM memories:
0: Address/data non multiplexed
1: Address/data multiplexed on databus (default after reset)
Bit 0 MBKEN: Memory bank enable bit
Enables the memory bank. After reset Bank1 is enabled, all others are disabled. Accessing a
disabled bank causes an ERROR on AHB bus.
0: Corresponding memory bank is disabled.
1: Corresponding memory bank is enabled.

SRAM/NOR-Flash chip-select timing register for bank x (FSMC_BTRx)

Address offset: 0x04 + 8 * (x – 1), (x = 1 to 4)
Reset value: 0x0FFF FFFF
This register contains the control information of each memory bank, used for SRAMs,
PSRAM and NOR Flash memories.If the EXTMOD bit is set in the FSMC_BCRx register,
then this register is partitioned for write and read access, that is, 2 registers are available:

RM0402 Rev 6 281/1163

287
Flexible static memory controller (FSMC) RM0402

one to configure read accesses (this register) and one to configure write accesses
(FSMC_BWTRx registers).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. ACCMOD[1:0] DATLAT[3:0] CLKDIV[3:0] BUSTURN[3:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:30 Reserved, must be kept at reset value.

Bits 29:28 ACCMOD[1:0]: Access mode
Specifies the asynchronous access modes as shown in the timing diagrams. These bits are
taken into account only when the EXTMOD bit in the FSMC_BCRx register is 1.
00: Access mode A
01: Access mode B
10: Access mode C
11: Access mode D
Bits 27:24 DATLAT[3:0]: (see note below bit descriptions): Data latency for synchronous memory
For synchronous access with read/write Burst mode enabled (BURSTEN / CBURSTRW bits
set), defines the number of memory clock cycles (+2) to issue to the memory before
reading/writing the first data:
This timing parameter is not expressed in HCLK periods, but in FSMC_CLK periods.
For asynchronous access, this value is don't care.
0000: Data latency of 2 CLK clock cycles for first burst access
1111: Data latency of 17 CLK clock cycles for first burst access (default value after reset)
Bits 23:20 CLKDIV[3:0]: Clock divide ratio (for FSMC_CLK signal)
Defines the period of FSMC_CLK clock output signal, expressed in number of HCLK cycles:
0000: FSMC_CLK period= 1x HCLK period
0001: FSMC_CLK period = 2 × HCLK periods
0010: FSMC_CLK period = 3 × HCLK periods
1111: FSMC_CLK period = 16 × HCLK periods (default value after reset)
In asynchronous NOR Flash, SRAM or PSRAM accesses, this value is don’t care.
Note: Refer to Section 11.6.5: Synchronous transactions for FSMC_CLK divider ratio formula)

282/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Bits 19:16 BUSTURN[3:0]: Bus turnaround phase duration

These bits are written by software to add a delay at the end of a write-to-read (and read-to-
write) transaction. This delay allows to match the minimum time between consecutive
transactions (tEHEL from NEx high to NEx low) and the maximum time needed by the
memory to free the data bus after a read access (tEHQZ). The programmed bus turnaround
delay is inserted between an asynchronous read (muxed or mode D) or write transaction and
any other asynchronous /synchronous read or write to or from a static bank. The bank can be
the same or different in case of read, in case of write the bank can be different except for
muxed or mode D.
In some cases, whatever the programmed BUSTURN values, the bus turnaround delay is
fixed
as follows:
• The bus turnaround delay is not inserted between two consecutive asynchronous write
transfers to the same static memory bank except for muxed mode and mode D.
• There is a bus turnaround delay of 1 HCLK clock cycle between:
–Two consecutive asynchronous read transfers to the same static memory bank except for
muxed mode and mode D.
–An asynchronous read to an asynchronous or synchronous write to any static bank or
dynamic bank except for muxed mode and mode D.
–An asynchronous (modes 1, 2, A, B or C) read and a read from another static bank.
• There is a bus turnaround delay of 2 HCLK clock cycle between:
–Two consecutive synchronous writes (burst or single) to the same bank.
–A synchronous write (burst or single) access and an asynchronous write or read transfer
to or from static memory bank (the bank can be the same or different for the case of
read.
–Two consecutive synchronous reads (burst or single) followed by any
synchronous/asynchronous read or write from/to another static memory bank.
• There is a bus turnaround delay of 3 HCLK clock cycle between:
–Two consecutive synchronous writes (burst or single) to different static bank.
–A synchronous write (burst or single) access and a synchronous read from the same or a
different bank.

0000: BUSTURN phase duration = 0 HCLK clock cycle added

...
1111: BUSTURN phase duration = 15 x HCLK clock cycles added (default value after reset)
Bits 15:8 DATAST[7:0]: Data-phase duration
These bits are written by software to define the duration of the data phase (refer to Figure 33
to Figure 45), used in asynchronous accesses:
0000 0000: Reserved
0000 0001: DATAST phase duration = 1 × HCLK clock cycles
0000 0010: DATAST phase duration = 2 × HCLK clock cycles
...
1111 1111: DATAST phase duration = 255 × HCLK clock cycles (default value after reset)
For each memory type and access mode data-phase duration, refer to the respective figure
(Figure 33 to Figure 45).
Example: Mode 1, write access, DATAST=1: Data-phase duration= DATAST+1 = 2 HCLK
clock cycles.
Note: In synchronous accesses, this value is don’t care.

RM0402 Rev 6 283/1163

287
Flexible static memory controller (FSMC) RM0402

Bits 7:4 ADDHLD[3:0]: Address-hold phase duration

These bits are written by software to define the duration of the address hold phase (refer to
Figure 33 to Figure 45), used in mode D or multiplexed accesses:
0000: Reserved
0001: ADDHLD phase duration =1 × HCLK clock cycle
0010: ADDHLD phase duration = 2 × HCLK clock cycle
...
1111: ADDHLD phase duration = 15 × HCLK clock cycles (default value after reset)
For each access mode address-hold phase duration, refer to the respective figure (Figure 33
to Figure 45).
Note: In synchronous accesses, this value is not used, the address hold phase is always 1
memory clock period duration.
Bits 3:0 ADDSET[3:0]: Address setup phase duration
These bits are written by software to define the duration of the address setup phase (refer to
Figure 33 to Figure 45), used in SRAMs, ROMs, asynchronous NOR Flash and PSRAM:
0000: ADDSET phase duration = 0 × HCLK clock cycle
...
1111: ADDSET phase duration = 15 × HCLK clock cycles (default value after reset)
For each access mode address setup phase duration, refer to the respective figure
(Figure 33 to Figure 45).
Note: In synchronous accesses, this value is don’t care.
In Muxed mode or mode D, the minimum value for ADDSET is 1.
In mode 1 and PSRAM memory, the minimum value for ADDSET is 1.

Note: PSRAMs (CRAMs) have a variable latency due to internal refresh. Therefore these
memories issue the NWAIT signal during the whole latency phase to prolong the latency as
needed.
With PSRAMs (CRAMs) the filled DATLAT must be set to 0, so that the FSMC exits its
latency phase soon and starts sampling NWAIT from memory, then starts to read or write
when the memory is ready.
This method can be used also with the latest generation of synchronous Flash memories
that issue the NWAIT signal, unlike older Flash memories (check the datasheet of the
specific Flash memory being used).

SRAM/NOR-Flash write timing registers x (FSMC_BWTRx)

Address offset: 0x104 + 8 * (x – 1), (x = 1 to 4)
Reset value: 0x0FFF FFFF
This register contains the control information of each memory bank. It is used for SRAMs,
PSRAMs and NOR Flash memories. When the EXTMOD bit is set in the FSMC_BCRx
register, then this register is active for write access.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. ACCMOD[1:0] Res. Res. Res. Res. Res. Res. Res. Res. BUSTURN[3:0]

rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:30 Reserved, must be kept at reset value.

284/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Bits 29:28 ACCMOD[1:0]: Access mode.

Specifies the asynchronous access modes as shown in the next timing diagrams.These bits are
taken into account only when the EXTMOD bit in the FSMC_BCRx register is 1.
00: Access mode A
01: Access mode B
10: Access mode C
11: Access mode D
Bits 27:20 Reserved, must be kept at reset value.
Bits 19:16 BUSTURN[3:0]: Bus turnaround phase duration
The programmed bus turnaround delay is inserted between an asynchronous write transfer and
any other asynchronous /synchronous read or write transfer to or from a static bank. The bank can
be the same or different in case of read, in case of write the bank can be different expect for muxed
or mode D.
In some cases, whatever the programmed BUSTURN values, the bus turnaround delay is fixed as
follows:
• The bus turnaround delay is not inserted between two consecutive asynchronous write transfers
to the same static memory bank except for muxed and D modes.
• There is a bus turnaround delay of 2 HCLK clock cycle between:
–Two consecutive synchronous writes (burst or single) to the same bank.
–A synchronous write (burst or single) transfer and an asynchronous write or read transfer to or
from static memory bank.
• There is a bus turnaround delay of 3 HCLK clock cycle between:
–Two consecutive synchronous writes (burst or single) to different static bank.
–A synchronous write (burst or single) transfer and a synchronous read from the same or a
different bank.

0000: BUSTURN phase duration = 0 HCLK clock cycle added

...
1111: BUSTURN phase duration = 15 HCLK clock cycles added (default value after reset)
Bits 15:8 DATAST[7:0]: Data-phase duration.
These bits are written by software to define the duration of the data phase (refer to Figure 33 to
Figure 45), used in asynchronous SRAM, PSRAM and NOR Flash memory accesses:
0000 0000: Reserved
0000 0001: DATAST phase duration = 1 × HCLK clock cycles
0000 0010: DATAST phase duration = 2 × HCLK clock cycles
...
1111 1111: DATAST phase duration = 255 × HCLK clock cycles (default value after reset)
Bits 7:4 ADDHLD[3:0]: Address-hold phase duration.
These bits are written by software to define the duration of the address hold phase (refer to
Figure 42 to Figure 45), used in asynchronous multiplexed accesses:
0000: Reserved
0001: ADDHLD phase duration = 1 × HCLK clock cycle
0010: ADDHLD phase duration = 2 × HCLK clock cycle
...
1111: ADDHLD phase duration = 15 × HCLK clock cycles (default value after reset)
Note: In synchronous NOR Flash accesses, this value is not used, the address hold phase is always
1 Flash clock period duration.

RM0402 Rev 6 285/1163

287
Flexible static memory controller (FSMC) RM0402

Bits 3:0 ADDSET[3:0]: Address setup phase duration.

These bits are written by software to define the duration of the address setup phase in HCLK cycles
(refer to Figure 33 to Figure 45), used in asynchronous accesses:
0000: ADDSET phase duration = 0 × HCLK clock cycle
...
1111: ADDSET phase duration = 15 × HCLK clock cycles (default value after reset)
Note: In synchronous accesses, this value is not used, the address setup phase is always 1 Flash
clock period duration. In muxed mode, the minimum ADDSET value is 1.

11.6.7 FSMC register map

Table 70. FSMC register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
ASYNCWAIT
CBURSTRW

BURSTEN
WAITCFG

WAITPOL
EXTMOD
CCLKEN

FACCEN
WAITEN

MUXEN
MBKEN
WFDIS

WREN
CPSIZE MWID MTYP
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
FSMC_BCR1
0x00 [2:0] [1:0] [1:0]

Reset value 0 0 0 0 0 0 0 0 1 1 0 0 0 1 0 1 1 0 1 1
ASYNCWAIT
CBURSTRW

BURSTEN
WAITCFG

WAITPOL
EXTMOD

FACCEN
WAITEN

MUXEN
MBKEN
WREN
CPSIZE MWID MTYP
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
FSMC_BCR2
0x08 [2:0] [1:0] [1:0]

Reset value 0 0 0 0 0 0 1 1 0 0 0 1 0 1 0 0 1 0
ASYNCWAIT
CBURSTRW

BURSTEN
WAITCFG

WAITPOL
EXTMOD

FACCEN
WAITEN

MUXEN
MBKEN
WREN

CPSIZE MWID MTYP

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
FSMC_BCR3
0x10 [2:0] [1:0] [1:0]

Reset value 0 0 0 0 0 0 1 1 0 0 0 1 0 1 0 0 1 0
ASYNCWAIT
CBURSTRW

BURSTEN
WAITCFG

WAITPOL
EXTMOD

FACCEN
WAITEN

MUXEN
MBKEN
WREN

CPSIZE MWID MTYP

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.

FSMC_BCR4
0x18 [2:0] [1:0] [1:0]

Reset value 0 0 0 0 0 0 1 1 0 0 0 1 0 1 0 0 1 0
ACCMOD[1:0]

BUSTURN[3:0
Res.
Res.

FSMC_BTR1 DATLAT[3:0] CLKDIV[3:0] DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]

0x04 ]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ACCMOD[1:0]

BUSTURN[3:0
Res.
Res.

FSMC_BTR2 DATLAT[3:0] CLKDIV[3:0] DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]

0x0C ]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ACCMOD[1:0]

BUSTURN[3:0
Res.
Res.

FSMC_BTR3 DATLAT[3:0] CLKDIV[3:0] DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]

0x14 ]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

286/1163 RM0402 Rev 6

RM0402 Flexible static memory controller (FSMC)

Table 70. FSMC register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
ACCMOD[1:0]
BUSTURN[3:0
Res.
Res.
FSMC_BTR4 DATLAT[3:0] CLKDIV[3:0] DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]
0x1C ]

Reset value 0
ACCMOD[1:0] 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

BUSTURN[3:0
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
FSMC_BWTR1 DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]
0x104 ]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ACCMOD[1:0]

BUSTURN[3:0
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
FSMC_BWTR2 DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]
0x10C ]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ACCMOD[1:0]

BUSTURN
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

FSMC_BWTR3 DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]

0x114 [3:0]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
ACCMOD[1:0]

BUSTURN
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

FSMC_BWTR4 DATAST[7:0] ADDHLD[3:0] ADDSET[3:0]

0x11C [3:0]

Reset value 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Refer to Section 2.2 on page 49 for the register boundary addresses.

RM0402 Rev 6 287/1163

287
Quad-SPI interface (QUADSPI) RM0402

12 Quad-SPI interface (QUADSPI)

12.1 Introduction
The QUADSPI is a specialized communication interface targeting single, dual or quad SPI
Flash memories. It can operate in any of the three following modes:
• indirect mode: all the operations are performed using the QUADSPI registers
• status polling mode: the external Flash memory status register is periodically read and
an interrupt can be generated in case of flag setting
• memory-mapped mode: the external Flash memory is mapped to the device address
space and is seen by the system as if it was an internal memory
Both throughput and capacity can be increased two-fold using dual-flash mode, where two
Quad-SPI Flash memories are accessed simultaneously.

12.2 QUADSPI main features

• Three functional modes: indirect, status-polling, and memory-mapped
• Dual-flash mode, where 8 bits can be sent/received simultaneously by accessing two
Flash memories in parallel.
• SDR and DDR support
• Fully programmable opcode for both indirect and memory mapped mode
• Fully programmable frame format for both indirect and memory mapped mode
• Integrated FIFO for reception and transmission
• 8, 16, and 32-bit data accesses are allowed
• DMA channel for indirect mode operations
• Interrupt generation on FIFO threshold, timeout, operation complete, and access error

12.3 QUADSPI functional description

12.3.1 QUADSPI block diagram

Figure 51. QUADSPI block diagram when dual-flash mode is disabled

QUADSPI
Registers / Clock
control management
SPI FLASH
CLK
CLK
BK1_IO0/SO
AHB Q0/SI
BK1_IO1/SI
Q1/SO
BK1_IO2
FIFO Shift register Q2/WP
BK1_IO3
Q3/HOLD
BK1_nCS
CS
MS35315V1

288/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

Figure 52. QUADSPI block diagram when dual-flash mode is enabled

QUADSPI
Registers / Clock
control management
SPI FLASH 1
CLK
CLK
BK1_IO0/SO
AHB Q0/SI
BK1_IO1/SI
Q1/SO
BK1_IO2
FIFO Q2/WP
BK1_IO3
Q3/HOLD
BK1_nCS
CS

Shift register SPI FLASH 2

CLK
BK2_IO0/SO
Q0/SI
BK2_IO1/SI
Q1/SO
BK2_IO2
Q2/WP
BK2_IO3
Q3/HOLD
BK2_nCS
CS
MS35316V1

12.3.2 QUADSPI pins

Table 71 lists the QUADSPI pins, six for interfacing with a single Flash memory, or 10 to 11
for interfacing with two Flash memories (FLASH 1 and FLASH 2) in dual-flash mode.

Table 71. QUADSPI pins

Signal name Signal type Description

CLK Digital output Clock to FLASH 1 and FLASH 2

Bidirectional IO in dual/quad modes or serial output
BK1_IO0/SO Digital input/output
in single mode, for FLASH 1
Bidirectional IO in dual/quad modes or serial input
BK1_IO1/SI Digital input/output
in single mode, for FLASH 1
BK1_IO2 Digital input/output Bidirectional IO in quad mode, for FLASH 1
BK1_IO3 Digital input/output Bidirectional IO in quad mode, for FLASH 1
Bidirectional IO in dual/quad modes or serial output
BK2_IO0/SO Digital input/output
in single mode, for FLASH 2
Bidirectional IO in dual/quad modes or serial input
BK2_IO1/SI Digital input/output
in single mode, for FLASH 2
BK2_IO2 Digital input/output Bidirectional IO in quad mode, for FLASH 2
BK2_IO3 Digital input/output Bidirectional IO in quad mode, for FLASH 2
Chip select (active low) for FLASH 1. Can also be
BK1_nCS Digital output used for FLASH 2 if QUADSPI is always used in
dual-flash mode.
Chip select (active low) for FLASH 2. Can also be
BK2_nCS Digital output used for FLASH 1 if QUADSPI is always used in
dual-flash mode.

RM0402 Rev 6 289/1163

316
Quad-SPI interface (QUADSPI) RM0402

12.3.3 QUADSPI command sequence

The QUADSPI communicates with the Flash memory using commands. Each command
can include 5 phases: instruction, address, alternate byte, dummy, data. Any of these
phases can be configured to be skipped, but at least one of the instruction, address,
alternate byte, or data phase must be present.
nCS falls before the start of each command and rises again after each command finishes.

Figure 53. An example of a read command in quad mode

Instruction Address Alt. Dummy Data

nCS
SCLK
IO0 7 6 5 4 3 2 1 0 4 0 4 0 4 0 4 0 4 0 4 0

IO1 5 1 5 1 5 1 5 1 5 1 5 1

IO2 6 2 6 2 6 2 6 2 6 2 6 2

IO3 7 3 7 3 7 3 7 3 7 3 7 3
A23-16 A15-8 A7-0 M7-0 Byte 1 Byte 2

IO switch from
output to input MS35317V1

Instruction phase
During this phase, an 8-bit instruction, configured in INSTRUCTION field of
QUADSPI_CCR[7:0] register, is sent to the Flash memory, specifying the type of operation
to be performed.
Though most Flash memories can receive instructions only one bit at a time from the
IO0/SO signal (single SPI mode), the instruction phase can optionally send 2 bits at a time
(over IO0/IO1 in dual SPI mode) or 4 bits at a time (over IO0/IO1/IO2/IO3 in quad SPI
mode). This can be configured using the IMODE[1:0] field of QUADSPI_CCR[9:8] register.
When IMODE = 00, the instruction phase is skipped, and the command sequence starts
with the address phase, if present.

Address phase
In the address phase, 1-4 bytes are sent to the Flash memory to indicate the address of the
operation. The number of address bytes to be sent is configured in the ADSIZE[1:0] field of
QUADSPI_CCR[13:12] register. In indirect and automatic-polling modes, the address bytes
to be sent are specified in the ADDRESS[31:0] field of QUADSPI_AR register, while in
memory-mapped mode the address is given directly via the AHB (from the Cortex® or from
a DMA).
The address phase can send 1 bit at a time (over SO in single SPI mode), 2 bits at a time
(over IO0/IO1 in dual SPI mode), or 4 bits at a time (over IO0/IO1/IO2/IO3 in quad SPI
mode). This can be configured using the ADMODE[1:0] field of QUADSPI_CCR[11:10]
register.
When ADMODE = 00, the address phase is skipped, and the command sequence proceeds
directly to the next phase, if any.

290/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

Alternate-bytes phase
In the alternate-bytes phase, 1-4 bytes are sent to the Flash memory, generally to control
the mode of operation. The number of alternate bytes to be sent is configured in the
ABSIZE[1:0] field of QUADSPI_CCR[17:16] register. The bytes to be sent are specified in
the QUADSPI_ABR register.
The alternate-bytes phase can send 1 bit at a time (over SO in single SPI mode), 2 bits at a
time (over IO0/IO1 in dual SPI mode), or 4 bits at a time (over IO0/IO1/IO2/IO3 in quad SPI
mode). This can be configured using the ABMODE[1:0] field of QUADSPI_CCR[15:14]
register.
When ABMODE = 00, the alternate-bytes phase is skipped, and the command sequence
proceeds directly to the next phase, if any.
There may be times when only a single nibble needs to be sent during the alternate-byte
phase rather than a full byte, such as when dual-mode is used and only two cycles are used
for the alternate bytes. In this case, firmware can use quad-mode (ABMODE = 11) and send
a byte with bits 7 and 3 of ALTERNATE set to ‘1’ (keeping the IO3 line high), and bits 6 and
2 set to ‘0’ (keeping the IO2 line low). In this case the upper two bits of the nibble to be sent
are placed in bits 4:3 of ALTERNATE while the lower two bits are placed in bits 1 and 0. For
example, if the nibble 2 (0010) is to be sent over IO0/IO1, then ALTERNATE should be set
to 0x8A (1000_1010).

Dummy-cycles phase
In the dummy-cycles phase, 1-31 cycles are given without any data being sent or received,
in order to allow the Flash memory the time to prepare for the data phase when higher clock
frequencies are used. The number of cycles given during this phase is specified in the
DCYC[4:0] field of QUADSPI_CCR[22:18] register. In both SDR and DDR modes, the
duration is specified as a number of full CLK cycles.
When DCYC is zero, the dummy-cycles phase is skipped, and the command sequence
proceeds directly to the data phase, if present.
The operating mode of the dummy-cycles phase is determined by DMODE.
In order to assure enough “turn-around” time for changing the data signals from output
mode to input mode, there must be at least one dummy cycle when using dual or quad
mode to receive data from the Flash memory.

Data phase
During the data phase, any number of bytes can be sent to, or received from the Flash
memory.
In indirect and automatic-polling modes, the number of bytes to be sent/received is specified
in the QUADSPI_DLR register.
In indirect write mode the data to be sent to the Flash memory must be written to the
QUADSPI_DR register, while in indirect read mode the data received from the Flash
memory is obtained by reading from the QUADSPI_DR register.
In memory-mapped mode, the data which is read is sent back directly over the AHB to the
Cortex or to a DMA.
The data phase can send/receive 1 bit at a time (over SO/SI in single SPI mode), 2 bits at a
time (over IO0/IO1 in dual SPI mode), or 4 bits at a time (over IO0/IO1/IO2/IO3 in quad SPI

RM0402 Rev 6 291/1163

316
Quad-SPI interface (QUADSPI) RM0402

mode). This can be configured using the ABMODE[1:0] field of QUADSPI_CCR[15:14]

register.
When DMODE = 00, the data phase is skipped, and the command sequence finishes
immediately by raising nCS. This configuration must only be used in only indirect write
mode.

12.3.4 QUADSPI signal interface protocol modes

Single SPI mode
Legacy SPI mode allows just a single bit to be sent/received serially. In this mode, data is
sent to the Flash memory over the SO signal (whose I/O shared with IO0). Data received
from the Flash memory arrives via SI (whose I/O shared with IO1).
The different phases can each be configured separately to use this single bit mode by
setting the IMODE/ADMODE/ABMODE/DMODE fields (in QUADSPI_CCR) to 01.
In each phase which is configured in single mode:
• IO0 (SO) is in output mode
• IO1 (SI) is in input mode (high impedance)
• IO2 is in output mode and forced to ‘0’
• IO3 is in output mode and forced to ‘1’ (to deactivate the “hold” function)
This is the case even for the dummy phase if DMODE = 01.

Dual SPI mode

In dual SPI mode, two bits are sent/received simultaneously over the IO0/IO1 signals.
The different phases can each be configured separately to use dual SPI mode by setting the
IMODE/ADMODE/ABMODE/DMODE fields of QUADSPI_CCR register to 10.
In each phase which is configured in dual mode:
• IO0/IO1 are at high-impedance (input) during the data phase for read operations, and
outputs in all other cases
• IO2 is in output mode and forced to ‘0’
• IO3 is in output mode and forced to ‘1’
In the dummy phase when DMODE = 01, IO0/IO1 are always high-impedance.

Quad SPI mode

In quad SPI mode, four bits are sent/received simultaneously over the IO0/IO1/IO2/IO3
signals.
The different phases can each be configured separately to use quad SPI mode by setting
the IMODE/ADMODE/ABMODE/DMODE fields of QUADSPI_CCR register to 11.
In each phase which is configured in quad mode, IO0/IO1/IO2/IO3 are all are at high-
impedance (input) during the data phase for read operations, and outputs in all other cases.
In the dummy phase when DMODE = 11, IO0/IO1/IO2/IO3 are all high-impedance.
IO2 and IO3 are used only in Quad SPI mode. If none of the phases are configured to use
Quad SPI mode, then the pins corresponding to IO2 and IO3 can be used for other functions
even while QUADSPI is active.

292/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

SDR mode
By default, the DDRM bit (QUADSPI_CCR[31]) is 0 and the QUADSPI operates in single
data rate (SDR) mode.
In SDR mode, when the QUADSPI is driving the IO0/SO, IO1, IO2, IO3 signals, these
signals transition only with the falling edge of CLK.
When receiving data in SDR mode, the QUADSPI assumes that the Flash memories also
send the data using CLK’s falling edge. By default (when SSHIFT = 0), the signals are
sampled using the following (rising) edge of CLK.

DDR mode
When the DDRM bit (QUADSPI_CCR[31]) is set to 1, the QUADSPI operates in double data
rate (DDR) mode.
In DDR mode, when the QUADSPI is driving the IO0/SO, IO1, IO2, IO3 signals in the
address/alternate-byte/data phases, a bit is sent on each of the falling and rising edges of
CLK.
The instruction phase is not affected by DDRM. The instruction is always sent using CLK’s
falling edge.
When receiving data in DDR mode, the QUADSPI assumes that the Flash memories also
send the data using both rising and falling CLK edges. When DDRM = 1, firmware must
clear SSHIFT bit (bit 4 of QUADSPI_CR). Thus, the signals are sampled one half of a CLK
cycle later (on the following, opposite edge).

Figure 54. An example of a DDR command in quad mode

Instruction Address Alt. Dummy Data

nCS
SCLK
IO0 7 6 5 4 3 2 1 0 4 0 4 0 4 0 4 0 4 0 4 0

IO1 5 4 5 4 5 4 5 4 5 4 5 4

IO2 6 2 6 2 6 2 6 2 6 2 6 2

IO3 7 3 7 3 7 3 7 3 7 3 7 3
A23-16A15-8 A7-0 M7-0 Byte1Byte2

IO switch from
output to input MS35318V1

Dual-flash mode
When the DFM bit (bit 6 of QUADSPI_CR) is 1, the QUADSPI is in dual-flash mode, where
two external quad SPI Flash memories (FLASH 1 and FLASH 2) are used in order to
send/receive 8 bits (or 16 bits in DDR mode) every cycle, effectively doubling the throughput
as well as the capacity.
Each of the Flash memories use the same CLK and optionally the same nCS signals, but
each have separate IO0, IO1, IO2, and IO3 signals.
Dual-flash mode can be used in conjunction with single-bit, dual-bit, and quad-bit modes, as
well as with either SDR or DDR mode.

RM0402 Rev 6 293/1163

316
Quad-SPI interface (QUADSPI) RM0402

The Flash memory size, as specified in FSIZE[4:0] (QUADSPI_DCR[20:16]), should reflect

the total Flash memory capacity, which is double the size of one individual component.
If address X is even, then the byte which the QUADSPI gives for address X is the byte at the
address X/2 of FLASH 1, and the byte which the QUADSPI gives for address X+1 is the
byte at the address X/2 of FLASH 2. In other words, bytes at even addresses are all stored
in FLASH 1 and bytes at odd addresses are all stored in FLASH 2.
When reading the Flash memories status registers in dual-flash mode, twice as many bytes
should be read compared to doing the same read in single-flash mode. This means that if
each Flash memory gives 8 valid bits after the instruction for fetching the status register,
then the QUADSPI must be configured with a data length of 2 bytes (16 bits), and the
QUADSPI receives one byte from each Flash memory. If each Flash memory gives a status
of 16 bits, then the QUADSPI must be configured to read 4 bytes to get all the status bits of
both Flash memories in dual-flash mode. The least-significant byte of the result (in the data
register) is the least-significant byte of FLASH 1 status register, while the next byte is the
least-significant byte of FLASH 2 status register. Then, the third byte of the data register is
FLASH 1 second byte, while the forth byte is FLASH 2 second byte (in the case that the
Flash memories have 16-bit status registers).
An even number of bytes must always be accessed in dual-flash mode. For this reason, bit
0 of the data length field (QUADSPI_DLR[0]) is stuck at 1 when DRM = 1.
In dual-flash mode, the behavior of FLASH 1 interface signals are basically the same as in
normal mode. FLASH 2 interface signals have exactly the same waveforms as FLASH 1
during the instruction, address, alternate-byte, and dummy-cycles phases. In other words,
each Flash memory always receives the same instruction and the same address. Then,
during the data phase, the BK1_IOx and BK2_IOx buses are both transferring data in
parallel, but the data that are sent to (or received from) FLASH 1 are distinct from those of
FLASH 2.

12.3.5 QUADSPI indirect mode

When in indirect mode, commands are started by writing to QUADSPI registers and data is
transferred by writing or reading the data register, in the same way as for other
communication peripherals.
When FMODE = 00 (QUADSPI_CCR[27:26]), the QUADSPI is in indirect write mode,
where bytes are sent to the Flash memory during the data phase. Data are provided by
writing to the data register (QUADSPI_DR).
When FMODE = 01, the QUADSPI is in indirect read mode, where bytes are received from
the Flash memory during the data phase. Data are recovered by reading QUADSPI_DR.
The number of bytes to be read/written is specified in the data length register
(QUADSPI_DLR). If QUADSPI_DLR = 0xFFFF_FFFF (all 1’s), then the data length is
considered undefined and the QUADSPI simply continues to transfer data until the end of
Flash memory (as defined by FSIZE) is reached. If no bytes are to be transferred, DMODE
(QUADSPI_CCR[25:24]) should be set to 00.
If QUADSPI_DLR = 0xFFFF_FFFF and FSIZE = 0x1F (max value indicating a 4GB Flash
memory), then in this special case the transfers continue indefinitely, stopping only after an
abort request or after the QUADSPI is disabled. After the last memory address is read (at
address 0xFFFF_FFFF), reading continues with address = 0x0000_0000.
When the programmed number of bytes to be transmitted or received is reached, TCF is set
and an interrupt is generated if TCIE = 1. In the case of undefined number of data, the TCF

294/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

is set when the limit of the external SPI memory is reached according to the Flash memory
size defined in the QUADSPI_CR.

Triggering the start of a command

Essentially, a command starts as soon as firmware gives the last information that is
necessary for this command. Depending on the QUADSPI configuration, there are three
different ways to trigger the start of a command in indirect mode. The commands starts
immediately after:
1. a write is performed to INSTRUCTION[7:0] (QUADSPI_CCR), if no address is
necessary (when ADMODE = 00) and if no data needs to be provided by the firmware
(when FMODE = 01 or DMODE = 00)
2. a write is performed to ADDRESS[31:0] (QUADSPI_AR), if an address is necessary
(when ADMODE != 00) and if no data needs to be provided by the firmware (when
FMODE = 01 or DMODE = 00)
3. a write is performed to DATA[31:0] (QUADSPI_DR), if an address is necessary (when
ADMODE != 00) and if data needs to be provided by the firmware (when FMODE = 00
and DMODE != 00)
Writes to the alternate byte register (QUADSPI_ABR) never trigger the communication start.
If alternate bytes are required, they must be programmed before.
As soon as a command is started, the BUSY bit (bit 5 of QUADSPI_SR) is automatically set.

FIFO and data management

In indirect mode, data go through a 32-byte FIFO which is internal to the QUADSPI.
FLEVEL[5:0] (QUADSPI_SR[13:8]) indicates how many bytes are currently being held in
the FIFO.
In indirect write mode (FMODE = 00), firmware adds data to the FIFO when it writes
QUADSPI_DR. Word writes add 4 bytes to the FIFO, halfword writes add 2 bytes, and byte
writes add only 1 byte. If firmware adds too many bytes to the FIFO (more than is indicated
by DL[31:0]), the extra bytes are flushed from the FIFO at the end of the write operation
(when TCF is set).
Byte/halfword accesses to QUADSPI_DR must be done only to the least significant
byte/halfword of the 32-bit register.
FTHRES[3:0] is used to define a FIFO threshold. When the threshold is reached, the FTF
(FIFO threshold flag) is set. In indirect read mode, FTF is set when the number of valid
bytes to be read from the FIFO is above the threshold. FTF is also set if there are data in the
FIFO after the last byte is read from the Flash memory, regardless of the FTHRES setting.
In indirect write mode, FTF is set when the number of empty bytes in the FIFO is above the
threshold.
If FTIE = 1, there is an interrupt when FTF is set. If DMAEN = 1, a DMA transfer is initiated
when FTF is set. FTF is cleared by HW as soon as the threshold condition is no longer true
(after enough data is transferred by the CPU or DMA).
In indirect read mode when the FIFO becomes full, the QUADSPI temporarily stops reading
bytes from the Flash memory to avoid an overrun. Note that the reading of the Flash
memory does not restart until 4 bytes become vacant in the FIFO (when FLEVEL ≤ 11).
Thus, when FTHRES ≥ 13, the application must take care to read enough bytes to assure
that the QUADSPI starts retrieving data from the Flash memory again. Otherwise, the FTF
flag stays at '0' as long as 11 < FLEVEL < FTHRES.

RM0402 Rev 6 295/1163

316
Quad-SPI interface (QUADSPI) RM0402

12.3.6 QUADSPI status flag polling mode

In automatic-polling mode, the QUADSPI periodically starts a command to read a defined
number of status bytes (up to 4). The received bytes can be masked to isolate some status
bits and an interrupt can be generated when the selected bits have a defined value.
The accesses to the Flash memory begin in the same way as in indirect read mode: if no
address is required (AMODE = 00), accesses begin as soon as the QUADSPI_CCR is
written. Otherwise, if an address is required, the first access begins when QUADSPI_AR is
written. BUSY goes high at this point and stays high even between the periodic accesses.
The contents of MASK[31:0] (QUADSPI_PSMAR) are used to mask the data from the Flash
memory in automatic-polling mode. If the MASK[n] = 0, then bit n of the result is masked
and not considered. If MASK[n] = 1, and the content of bit[n] is the same as MATCH[n]
(QUADSPI_PSMAR), then there is a match for bit n.
If the polling match mode bit (PMM, bit 23 of QUADSPI_CR) is 0, then “AND” match mode is
activated. This means status match flag (SMF) is set only when there is a match on all of the
unmasked bits.
If PMM = 1, then “OR” match mode is activated. This means SMF is set if there is a match
on any of the unmasked bits.
An interrupt is called when SMF is set if SMIE = 1.
If the automatic-polling-mode-stop (APMS) bit is set, operation stops and BUSY goes to 0
as soon as a match is detected. Otherwise, BUSY stays at ‘1’ and the periodic accesses
continue until there is an abort or the QUADSPI is disabled (EN = 0).
The data register (QUADSPI_DR) contains the latest received status bytes (the FIFO is
deactivated). The content of the data register is not affected by the masking used in the
matching logic. The FTF status bit is set as soon as a new reading of the status is complete,
and FTF is cleared as soon as the data is read.

12.3.7 QUADSPI memory-mapped mode

When configured in memory-mapped mode, the external SPI device is seen as an internal
memory.
It is forbidden to access QUADSPI Flash bank area before having properly configured and
enabled the QUADSPI peripheral.
No more than 256MB can addressed even if the Flash memory capacity is larger.
If an access is made to an address outside of the range defined by FSIZE but still within the
256MB range, then a bus error is given. The effect of this error depends on the bus master
that attempted the access:
• If it is the Cortex® CPU, bus fault exception is generated when enabled (or a hard fault
exception when bus fault is disabled)
• If it is a DMA, a DMA transfer error is generated and the corresponding DMA channel is
automatically disabled.
Byte, halfword, and word access types are all supported.
Support for execute in place (XIP) operation is implemented, where the QUADSPI
anticipates the next access and load in advance the byte at the following address. If the
subsequent access is indeed made at a continuous address, the access is completed faster
since the value is already prefetched.

296/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

By default, the QUADSPI never stops its prefetch operation, keeping the previous read
operation active with nCS maintained low, even if no access to the Flash memory occurs for
a long time. Since Flash memories tend to consume more when nCS is held low, the
application might want to activate the timeout counter (TCEN = 1, bit 3 of QUADSPI_CR) so
that nCS is released after a period of TIMEOUT[15:0] (QUADSPI_LPTR) cycles have
elapsed without any access since when the FIFO becomes full with prefetch data.
BUSY goes high as soon as the first memory-mapped access occurs. Because of the
prefetch operations, BUSY does not fall until there is a timeout, there is an abort, or the
peripheral is disabled.

12.3.8 QUADSPI Flash memory configuration

The device configuration register (QUADSPI_DCR) can be used to specify the
characteristics of the external SPI Flash memory.
The FSIZE[4:0] field defines the size of external memory using the following formula:
Number of bytes in Flash memory = 2[FSIZE+1]
FSIZE+1 is effectively the number of address bits required to address the Flash memory.
The Flash memory capacity can be up to 4GB (addressed using 32 bits) in indirect mode,
but the addressable space in memory-mapped mode is limited to 256MB.
If DFM = 1, FSIZE indicates the total capacity of the two Flash memories together.
When the QUADSPI executes two commands, one immediately after the other, it raises the
chip select signal (nCS) high between the two commands for only one CLK cycle by default.
If the Flash memory requires more time between commands, the chip select high time
(CSHT) field can be used to specify the minimum number of CLK cycles (up to 8) that nCS
must remain high.
The clock mode (CKMODE) bit indicates the CLK signal logic level in between commands
(when nCS = 1).

12.3.9 QUADSPI delayed data sampling

By default, the QUADSPI samples the data driven by the Flash memory one half of a CLK
cycle after the Flash memory drives the signal.
In case of external signal delays, it may be beneficial to sample the data later. Using the
SSHIFT bit (bit 4 of QUADSPI_CR), the sampling of the data can be shifted by half of a CLK
cycle.
Clock shifting is not supported in DDR mode: the SSHIFT bit must be clear when DDRM bit
is set.

12.3.10 QUADSPI configuration

The QUADSPI configuration is done in two phases:
• QUADSPI IP configuration
• QUADSPI Flash memory configuration
Once configured and enabled, the QUADSPI can be used in one of its three operating
modes: indirect mode, status-polling mode, or memory-mapped mode.
QUADSPI configuration

RM0402 Rev 6 297/1163

316
Quad-SPI interface (QUADSPI) RM0402

The QUADSPI is configured using the QUADSPI_CR. The user shall configure the clock
prescaler division factor and the sample shifting settings for the incoming data.
DDR mode can be set through the DDRM bit. When setting QUADSPI interface in DDR
mode, the internal divider of kernel clock must be set with a division ratio of 2 or more. Once
enabled, the address and the alternate bytes are sent on both clock edges and the data are
sent/received on both clock edges. Regardless of the DDRM bit setting, instructions are
always sent in SDR mode.
The DMA requests are enabled setting the DMAEN bit. In case of interrupt usage, their
respective enable bit can be also set during this phase.
FIFO level for either DMA request generation or interrupt generation is programmed in the
FTHRES bits.
If timeout counter is needed, the TCEN bit can be set and the timeout value programmed in
the QUADSPI_LPTR register.
Dual-flash mode can be activated by setting DFM to 1.

QUADSPI Flash memory configuration

The parameters related to the targeted external Flash memory are configured through the
QUADSPI_DCR register.The user shall program the Flash memory size in the FSIZE bits,
the Chip Select minimum high time in the CSHT bits, and the functional mode (Mode 0 or
Mode 3) in the MODE bit.

12.3.11 QUADSPI usage

The operating mode is selected using FMODE[1:0] (QUADSPI_CCR[27:26]).

Indirect mode procedure

When FMODE is programmed to 00, indirect write mode is selected and data can be sent to
the Flash memory. With FMODE = 01, indirect read mode is selected where data can be
read from the Flash memory.
When the QUADSPI is used in indirect mode, the frames are constructed in the following
way:
1. Specify a number of data bytes to read or write in the QUADSPI_DLR.
2. Specify the frame format, mode and instruction code in the QUADSPI_CCR.
3. Specify optional alternate byte to be sent right after the address phase in the
QUADSPI_ABR.
4. Specify the operating mode in the QUADSPI_CR. If FMODE = 00 (indirect write mode)
and DMAEN = 1, then QUADSPI_AR should be specified before QUADSPI_CR,
because otherwise QUADSPI_DR might be written by the DMA before QUADSPI_AR
is updated (if the DMA controller has already been enabled)
5. Specify the targeted address in the QUADSPI_AR.
6. Read/Write the data from/to the FIFO through the QUADSPI_DR.

298/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

When writing the control register (QUADSPI_CR) the user specifies the following settings:
• The enable bit (EN) set to ‘1’
• The DMA enable bit (DMAEN) for transferring data to/from RAM
• Timeout counter enable bit (TCEN)
• Sample shift setting (SSHIFT)
• FIFO threshold level (FTRHES) to indicate when the FTF flag should be set
• Interrupt enables
• Automatic polling mode parameters: match mode and stop mode (valid when
FMODE = 11)
• Clock prescaler
When writing the communication configuration register (QUADSPI_CCR) the user specifies
the following parameters:
• The instruction byte through the INSTRUCTION bits
• The way the instruction has to be sent through the IMODE bits (1/2/4 lines)
• The way the address has to be sent through the ADMODE bits (None/1/2/4 lines)
• The address size (8/16/24/32-bit) through the ADSIZE bits
• The way the alternate bytes have to be sent through the ABMODE (None/1/2/4 lines)
• The alternate bytes number (1/2/3/4) through the ABSIZE bits
• The presence or not of dummy bytes through the DBMODE bit
• The number of dummy bytes through the DCYC bits
• The way the data have to be sent/received (None/1/2/4 lines) through the DMODE bits
If neither the address register (QUADSPI_AR) nor the data register (QUADSPI_DR) need to
be updated for a particular command, then the command sequence starts as soon as
QUADSPI_CCR is written. This is the case when both ADMODE and DMODE are 00, or if
just ADMODE = 00 when in indirect read mode (FMODE = 01).
When an address is required (ADMODE is not 00) and the data register does not need to be
written (when FMODE = 01 or DMODE = 00), the command sequence starts as soon as the
address is updated with a write to QUADSPI_AR.
In case of data transmission (FMODE = 00 and DMODE! = 00), the communication start is
triggered by a write in the FIFO through QUADSPI_DR.

Status flag polling mode

The status flag polling mode is enabled setting the FMODE field (QUADSPI_CCR[27:26]) to
10. In this mode, the programmed frame is sent and the data retrieved periodically.
The maximum amount of data read in each frame is 4 bytes. If more data is requested in
QUADSPI_DLR, it is ignored and only 4 bytes are read.
The periodicity is specified in the QUADSPI_PISR register.
Once the status data is retrieved, it can internally be processed i order to:
• set the status match flag and generate an interrupt if enabled
• stop automatically the periodic retrieving of the status bytes
The received value can be masked with the value stored in the QUADSPI_PSMKR and
ORed or ANDed with the value stored in the QUADSPI_PSMAR.

RM0402 Rev 6 299/1163

316
Quad-SPI interface (QUADSPI) RM0402

In case of match, the status match flag is set and an interrupt is generated if enabled, and
the QUADSPI can be automatically stopped if the AMPS bit is set.
In any case, the latest retrieved value is available in the QUADSPI_DR.

Memory-mapped mode
In memory-mapped mode, the external Flash memory is seen as internal memory but with
some latency during accesses. Only read operations are allowed to the external Flash
memory in this mode.
Memory-mapped mode is entered by setting the FMODE to 11 in the QUADSPI_CCR
register.
The programmed instruction and frame is sent when a master is accessing the memory
mapped space.
The FIFO is used as a prefetch buffer to anticipate linear reads. Any access to
QUADSPI_DR in this mode returns zero.
The data length register (QUADSPI_DLR) has no meaning in memory-mapped mode.

12.3.12 Sending the instruction only once

Some Flash memories (e.g. Winbound) might provide a mode where an instruction must be
sent only with the first command sequence, while subsequent commands start directly with
the address. One can take advantage of such a feature using the SIOO bit
(QUADSPI_CCR[28]).
SIOO is valid for all functional modes (indirect, automatic polling, and memory-mapped). If
the SIOO bit is set, the instruction is sent only for the first command following a write to
QUADSPI_CCR. Subsequent command sequences skip the instruction phase, until there is
a write to QUADSPI_CCR.
SIOO has no effect when IMODE = 00 (no instruction).

12.3.13 QUADSPI error management

An error can be generated in the following case:
• In indirect mode or status flag polling mode when a wrong address is programmed in
the QUADSPI_AR (according to the Flash memory size defined by FSIZE[4:0] in the
QUADSPI_DCR): this sets the TEF and an interrupt is generated if enabled.
• Also in indirect mode, if the address plus the data length exceeds the Flash memory
size, TEF is set as soon as the access is triggered.
• In memory-mapped mode, when an out of range access is done by a master or when
the QUADSPI is disabled: this generates a bus error as a response to the faulty bus
master request.
• When a master is accessing the memory mapped space while the memory mapped
mode is disabled: this generates a bus error as a response to the faulty bus master
request.

12.3.14 QUADSPI busy bit and abort functionality

Once the QUADSPI starts an operation with the Flash memory, the BUSY bit is
automatically set in the QUADSPI_SR.

300/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

In indirect mode, the BUSY bit is reset once the QUADSPI has completed the requested
command sequence and the FIFO is empty.
In automatic-polling mode, BUSY goes low only after the last periodic access is complete,
due to a match when APMS = 1, or due to an abort.
After the first access in memory-mapped mode, BUSY goes low only on a timeout event or
on an abort.
Any operation can be aborted by setting the ABORT bit in the QUADSPI_CR. Once the
abort is completed, the BUSY bit and the ABORT bit are automatically reset, and the FIFO
is flushed.
Note: Some Flash memories might misbehave if a write operation to a status registers is aborted.

12.3.15 nCS behavior

By default, nCS is high, deselecting the external Flash memory. nCS falls before an
operation begins and rises as soon as it finishes.
When CKMODE = 0 (“mode0”, where CLK stays low when no operation is in progress) nCS
falls one CLK cycle before an operation first rising CLK edge, and nCS rises one CLK cycle
after the operation final rising CLK edge, as shown in Figure 55.

Figure 55. nCS when CKMODE = 0 (T = CLK period)

T T

nCS

SCLK

MS35319V1

When CKMODE=1 (“mode3”, where CLK goes high when no operation is in progress) and
DDRM=0 (SDR mode), nCS still falls one CLK cycle before an operation first rising CLK
edge, and nCS rises one CLK cycle after the operation final rising CLK edge, as shown in
Figure 56.

Figure 56. nCS when CKMODE = 1 in SDR mode (T = CLK period)

T T

nCS

SCLK

MS35320V1

When CKMODE = 1 (“mode3”) and DDRM = 1 (DDR mode), nCS falls one CLK cycle
before an operation first rising CLK edge, and nCS rises one CLK cycle after the operation
final active rising CLK edge, as shown in Figure 57. Because DDR operations must finish
with a falling edge, CLK is low when nCS rises, and CLK rises back up one half of a CLK
cycle afterwards.

RM0402 Rev 6 301/1163

316
Quad-SPI interface (QUADSPI) RM0402

Figure 57. nCS when CKMODE = 1 in DDR mode (T = CLK period)

T T T/2

nCS

SCLK

MS35321V1

When the FIFO stays full in a read operation or if the FIFO stays empty in a write operation,
the operation stalls and CLK stays low until firmware services the FIFO. If an abort occurs
when an operation is stalled, nCS rises just after the abort is requested and then CLK rises
one half of a CLK cycle later, as shown in Figure 58.

Figure 58. nCS when CKMODE = 1 with an abort (T = CLK period)

T Clock stalled T/2

nCS

SCLK

Abort

MS35322V1

When not in dual-flash mode (DFM = 0) and FSEL = 0 (default value), only FLASH 1 is
accessed and thus BK2_nCS stays high, if FSEL = 1, only FLASH 2 is accessed and
BK1_nCS stays high. In dual-flash mode, BK2_nCS behaves exactly the same as
BK1_nCS. Thus, if there is a FLASH 2 and if the application is dual-flash mode only, then
BK1_nCS signal can be used for FLASH 2 as well, and the pin devoted to BK2_nCS can be
used for other functions.

12.4 QUADSPI interrupts

An interrupt can be produced on the following events:
• Timeout
• Status match
• FIFO threshold
• Transfer complete
• Transfer error
Separate interrupt enable bits are available for flexibility.

302/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

Table 72. QUADSPI interrupt requests

Interrupt event Event flag Enable control bit

Timeout TOF TOIE

Status match SMF SMIE
FIFO threshold FTF FTIE
Transfer complete TCF TCIE
Transfer error TEF TEIE

RM0402 Rev 6 303/1163

316
Quad-SPI interface (QUADSPI) RM0402

12.5 QUADSPI registers

12.5.1 QUADSPI control register (QUADSPI_CR)

Address offset: 0x0000
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PRESCALER[7:0] PMM APMS Res. TOIE SMIE FTIE TCIE TEIE

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. FTHRES[4:0] FSEL DFM Res. SSHIFT TCEN DMAEN ABORT EN

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:24 PRESCALER[7:0]: Clock prescaler

This field defines the scaler factor for generating CLK based on the AHB clock
(value+1).
0: FCLK = FAHB, AHB clock used directly as QUADSPI CLK (prescaler bypassed)
1: FCLK = FAHB/2
2: FCLK = FAHB/3
...
255: FCLK = FAHB/256
For odd clock division factors, CLK’s duty cycle is not 50%. The clock signal remains
low one cycle longer than it stays high.
This field can be modified only when BUSY = 0.
When setting QUADSPI interface in DDR mode, the prescaler must be set with a
division ratio of 2 or more.
Bit 23 PMM: Polling match mode
This bit indicates which method should be used for determining a “match” during
automatic polling mode.
0: AND match mode. SMF is set if all the unmasked bits received from the Flash
memory match the corresponding bits in the match register.
1: OR match mode. SMF is set if any one of the unmasked bits received from the Flash
memory matches its corresponding bit in the match register.
This bit can be modified only when BUSY = 0.
Bit 22 APMS: Automatic poll mode stop
This bit determines if automatic polling is stopped after a match.
0: Automatic polling mode is stopped only by abort or by disabling the QUADSPI.
1: Automatic polling mode stops as soon as there is a match.
This bit can be modified only when BUSY = 0.
Bit 21 Reserved, must be kept at reset value.
Bit 20 TOIE: TimeOut interrupt enable
This bit enables the TimeOut interrupt.
0: Interrupt disable
1: Interrupt enabled

304/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

Bit 19 SMIE: Status match interrupt enable

This bit enables the status match interrupt.
0: Interrupt disable
1: Interrupt enabled
Bit 18 FTIE: FIFO threshold interrupt enable
This bit enables the FIFO threshold interrupt.
0: Interrupt disabled
1: Interrupt enabled
Bit 17 TCIE: Transfer complete interrupt enable
This bit enables the transfer complete interrupt.
0: Interrupt disabled
1: Interrupt enabled
Bit 16 TEIE: Transfer error interrupt enable
This bit enables the transfer error interrupt.
0: Interrupt disable
1: Interrupt enabled
Bits 15:13 Reserved, must be kept at reset value.
Bits 12:8 FTHRES[4:0] FIFO threshold level
Defines, in indirect mode, the threshold number of bytes in the FIFO that causes the
FIFO threshold flag (bit FTF in register QUADSPI_SR) to be set.
0: In indirect write mode (FMODE = 00) FTF is set if there are 1 or more free bytes
location left in the FIFO or indirect read mode (FMODE = 01) FTF is set if there are 1 or
more valid bytes that can be read from the FIFO
1: In indirect write mode (FMODE = 00) FTF is set if there are 2 or more free bytes
location left in the FIFO or indirect read mode (FMODE = 01) FTF is set if there are 2 or
more valid bytes that can be read from the FIFO
...
31: In indirect write mode (FMODE = 00) FTF is set if there are 32 free bytes location
left in the FIFO or indirect read mode (FMODE = 01) FTF is set if there are 32 valid
bytes that can be read from the FIFO
If DMAEN = 1, then the DMA controller for the corresponding channel must be disabled
before changing the FTHRES value.
Bit 7 FSEL: Flash memory selection
This bit selects the Flash memory to be addressed in single flash mode (when DFM =
0).
0: FLASH 1 selected
1: FLASH 2 selected
This bit can be modified only when BUSY = 0.
This bit is ignored when DFM = 1.
Bit 6 DFM: Dual-flash mode
This bit activates dual-flash mode, where two external Flash memories are used
simultaneously to double throughput and capacity.
0: Dual-flash mode disabled
1: Dual-flash mode enabled
This bit can be modified only when BUSY = 0.
Bit 5 Reserved, must be kept at reset value.

RM0402 Rev 6 305/1163

316
Quad-SPI interface (QUADSPI) RM0402

Bit 4 SSHIFT: Sample shift

By default, the QUADSPI samples data 1/2 of a CLK cycle after the data is driven by the
Flash memory. This bit allows the data to be sampled later in order to account for
external signal delays.
0: No shift
1: 1/2 cycle shift
Firmware must assure that SSHIFT = 0 when in DDR mode (when DDRM = 1).
This field can be modified only when BUSY = 0.
Bit 3 TCEN: Timeout counter enable
This bit is valid only when memory-mapped mode (FMODE = 11) is selected. Activating
this bit causes the chip select (nCS) to be released (and thus reduces consumption) if
there has not been an access after a certain amount of time, where this time is defined
by TIMEOUT[15:0] (QUADSPI_LPTR).
Enable the timeout counter.
By default, the QUADSPI never stops its prefetch operation, keeping the previous read
operation active with nCS maintained low, even if no access to the Flash memory
occurs for a long time. Since Flash memories tend to consume more when nCS is held
low, the application might want to activate the timeout counter (TCEN = 1, bit 3 of
QUADSPI_CR) so that nCS is released after a period of TIMEOUT[15:0]
(QUADSPI_LPTR) cycles have elapsed without an access since when the FIFO
becomes full with prefetch data.
0: Timeout counter is disabled, and thus the chip select (nCS) remains active
indefinitely after an access in memory-mapped mode.
1: Timeout counter is enabled, and thus the chip select is released in memory-mapped
mode after TIMEOUT[15:0] cycles of Flash memory inactivity.
This bit can be modified only when BUSY = 0.
Bit 2 DMAEN: DMA enable
In indirect mode, DMA can be used to input or output data via the QUADSPI_DR
register. DMA transfers are initiated when the FIFO threshold flag, FTF, is set.
0: DMA is disabled for indirect mode
1: DMA is enabled for indirect mode
Bit 1 ABORT: Abort request
This bit aborts the on-going command sequence. It is automatically reset once the abort
is complete.
This bit stops the current transfer.
In polling mode or memory-mapped mode, this bit also reset the APM bit or the DM bit.
0: No abort requested
1: Abort requested
Bit 0 EN: Enable
Enable the QUADSPI.
0: QUADSPI is disabled
1: QUADSPI is enabled

306/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

12.5.2 QUADSPI device configuration register (QUADSPI_DCR)

Address offset: 0x0004
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. FSIZE[4:0]

rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CK
Res. Res. Res. Res. Res. CSHT[2:0] Res. Res. Res. Res. Res. Res. Res.
MODE
rw rw rw rw

Bits 31:21 Reserved, must be kept at reset value.

Bits 20:16 FSIZE[4:0]: Flash memory size
This field defines the size of external memory using the following formula:
Number of bytes in Flash memory = 2[FSIZE+1]
FSIZE+1 is effectively the number of address bits required to address the Flash
memory. The Flash memory capacity can be up to 4GB (addressed using 32 bits) in
indirect mode, but the addressable space in memory-mapped mode is limited to
256MB.
If DFM = 1, FSIZE indicates the total capacity of the two Flash memories together.
This field can be modified only when BUSY = 0.
Bits 15:11 Reserved, must be kept at reset value.
Bits 10:8 CSHT[2:0]: Chip select high time
CSHT+1 defines the minimum number of CLK cycles which the chip select (nCS) must
remain high between commands issued to the Flash memory.
0: nCS stays high for at least 1 cycle between Flash memory commands
1: nCS stays high for at least 2 cycles between Flash memory commands
...
7: nCS stays high for at least 8 cycles between Flash memory commands
This field can be modified only when BUSY = 0.
Bits 7:1 Reserved, must be kept at reset value.
Bit 0 CKMODE: Mode 0 / mode 3
This bit indicates the level that CLK takes between commands (when nCS = 1).
0: CLK must stay low while nCS is high (chip select released). This is referred to as
mode 0.
1: CLK must stay high while nCS is high (chip select released). This is referred to as
mode 3.
This field can be modified only when BUSY = 0.

RM0402 Rev 6 307/1163

316
Quad-SPI interface (QUADSPI) RM0402

12.5.3 QUADSPI status register (QUADSPI_SR)

Address offset: 0x0008
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. FLEVEL[5:0] Res. Res. BUSY TOF SMF FTF TCF TEF

r r r r r r r r r r r r

Bits 31:14 Reserved, must be kept at reset value.

Bits 13:8 FLEVEL[5:0]: FIFO level
This field gives the number of valid bytes which are being held in the FIFO. FLEVEL = 0
when the FIFO is empty, and 32 when it is full. In memory-mapped mode and in
automatic status polling mode, FLEVEL is zero.
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 BUSY: Busy
This bit is set when an operation is on going. This bit clears automatically when the
operation with the Flash memory is finished and the FIFO is empty.
Bit 4 TOF: Timeout flag
This bit is set when timeout occurs. It is cleared by writing 1 to CTOF.
Bit 3 SMF: Status match flag
This bit is set in automatic polling mode when the unmasked received data matches the
corresponding bits in the match register (QUADSPI_PSMAR). It is cleared by writing 1
to CSMF.
Bit 2 FTF: FIFO threshold flag
In indirect mode, this bit is set when the FIFO threshold is reached, or if there is any
data left in the FIFO after reads from the Flash memory are complete. It is cleared
automatically as soon as threshold condition is no longer true.
In automatic polling mode this bit is set every time the status register is read, and the bit
is cleared when the data register is read.
Bit 1 TCF: Transfer complete flag
This bit is set in indirect mode when the programmed number of data is transferred or in
any mode when the transfer is aborted.It is cleared by writing 1 to CTCF.
Bit 0 TEF: Transfer error flag
This bit is set in indirect mode when an invalid address is being accessed in indirect
mode. It is cleared by writing 1 to CTEF.

308/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

12.5.4 QUADSPI flag clear register (QUADSPI_FCR)

Address offset: 0x000C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. CTOF CSMF Res. CTCF CTEF

w w w w

Bits 31:5 Reserved, must be kept at reset value.

Bit 4 CTOF: Clear timeout flag
Writing 1 clears the TOF flag in the QUADSPI_SR register
Bit 3 CSMF: Clear status match flag
Writing 1 clears the SMF flag in the QUADSPI_SR register
Bit 2 Reserved, must be kept at reset value.
Bit 1 CTCF: Clear transfer complete flag
Writing 1 clears the TCF flag in the QUADSPI_SR register
Bit 0 CTEF: Clear transfer error flag
Writing 1 clears the TEF flag in the QUADSPI_SR register

12.5.5 QUADSPI data length register (QUADSPI_DLR)

Address offset: 0x0010
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DL[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DL[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 309/1163

316
Quad-SPI interface (QUADSPI) RM0402

Bits 31:0 DL[31:0]: Data length

Number of data to be retrieved (value+1) in indirect and status-polling modes. A value
no greater than 3 (indicating 4 bytes) should be used for status-polling mode.
All 1s in indirect mode means undefined length, where QUADSPI continues until the
end of memory, as defined by FSIZE.
0x0000_0000: 1 byte is to be transferred
0x0000_0001: 2 bytes are to be transferred
0x0000_0002: 3 bytes are to be transferred
0x0000_0003: 4 bytes are to be transferred
...
0xFFFF_FFFD: 4,294,967,294 (4G-2) bytes are to be transferred
0xFFFF_FFFE: 4,294,967,295 (4G-1) bytes are to be transferred
0xFFFF_FFFF: undefined length -- all bytes until the end of Flash memory (as defined
by FSIZE) are to be transferred. Continue reading indefinitely if FSIZE = 0x1F.
DL[0] is stuck at ‘1’ in dual-flash mode (DFM = 1) even when ‘0’ is written to this bit, thus
assuring that each access transfers an even number of bytes.
This field has no effect when in memory-mapped mode (FMODE = 10).
This field can be written only when BUSY = 0.

12.5.6 QUADSPI communication configuration register (QUADSPI_CCR)

Address offset: 0x0014
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DDRM DHHC Res. SIOO FMODE[1:0] DMODE[1:0] Res. DCYC[4:0] ABSIZE[1:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

ABMODE[1:0] ADSIZE[1:0] ADMODE[1:0] IMODE[1:0] INSTRUCTION[7:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 DDRM: Double data rate mode

This bit sets the DDR mode for the address, alternate byte and data phase:
0: DDR Mode disabled
1: DDR Mode enabled
This field can be written only when BUSY = 0.
Bit 30 DHHC: DDR hold
Delay the data output by 1/4 of the QUADSPI output clock cycle in DDR mode:
0: Delay the data output using analog delay
1: Delay the data output by 1/4 of a QUADSPI output clock cycle.
This feature is only active in DDR mode.
This field can be written only when BUSY = 0.
Note: PRESCALER>0 is mandatory when DHHC=1.
Bit 29 Reserved, must be kept at reset value.

310/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

Bit 28 SIOO: Send instruction only once mode

See Section 12.3.12: Sending the instruction only once on page 300. This bit has no
effect when IMODE = 00.
0: Send instruction on every transaction
1: Send instruction only for the first command
This field can be written only when BUSY = 0.
Bits 27:26 FMODE[1:0]: Functional mode
This field defines the QUADSPI functional mode of operation.
00: Indirect write mode
01: Indirect read mode
10: Automatic polling mode
11: Memory-mapped mode
If DMAEN = 1 already, then the DMA controller for the corresponding channel must be
disabled before changing the FMODE value.
This field can be written only when BUSY = 0.
Bits 25:24 DMODE[1:0]: Data mode
This field defines the data phase’s mode of operation:
00: No data
01: Data on a single line
10: Data on two lines
11: Data on four lines
This field also determines the dummy phase mode of operation.
This field can be written only when BUSY = 0.
Bit 23 Reserved, must be kept at reset value.
Bits 22:18 DCYC[4:0]: Number of dummy cycles
This field defines the duration of the dummy phase. In both SDR and DDR modes, it
specifies a number of CLK cycles (0-31).
This field can be written only when BUSY = 0.
Bits 17:16 ABSIZE[1:0]: Alternate bytes size
This bit defines alternate bytes size:
00: 8-bit alternate byte
01: 16-bit alternate bytes
10: 24-bit alternate bytes
11: 32-bit alternate bytes
This field can be written only when BUSY = 0.
Bits 15:14 ABMODE[1:0]: Alternate bytes mode
This field defines the alternate-bytes phase mode of operation:
00: No alternate bytes
01: Alternate bytes on a single line
10: Alternate bytes on two lines
11: Alternate bytes on four lines
This field can be written only when BUSY = 0.

RM0402 Rev 6 311/1163

316
Quad-SPI interface (QUADSPI) RM0402

Bits 13:12 ADSIZE[1:0]: Address size

This bit defines address size:
00: 8-bit address
01: 16-bit address
10: 24-bit address
11: 32-bit address
This field can be written only when BUSY = 0.
Bits 11:10 ADMODE[1:0]: Address mode
This field defines the address phase mode of operation:
00: No address
01: Address on a single line
10: Address on two lines
11: Address on four lines
This field can be written only when BUSY = 0.
Bits 9:8 IMODE[1:0]: Instruction mode
This field defines the instruction phase mode of operation:
00: No instruction
01: Instruction on a single line
10: Instruction on two lines
11: Instruction on four lines
This field can be written only when BUSY = 0.
Bits 7:0 INSTRUCTION[7:0]: Instruction
Instruction to be send to the external SPI device.
This field can be written only when BUSY = 0.

12.5.7 QUADSPI address register (QUADSPI_AR)

Address offset: 0x0018
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

ADDRESS[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

ADDRESS[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 ADDRESS[31:0]: Address

Address to be send to the external Flash memory
Writes to this field are ignored when BUSY = 1 or when FMODE = 11 (memory-mapped
mode).
In dual flash mode, ADDRESS[0] is automatically stuck to ‘0’ as the address should
always be even

312/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

12.5.8 QUADSPI alternate bytes registers (QUADSPI_ABR)

Address offset: 0x001C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

ALTERNATE[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

ALTERNATE[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 ALTERNATE[31:0]: Alternate Bytes

Optional data to be send to the external SPI device right after the address.
This field can be written only when BUSY = 0.

12.5.9 QUADSPI data register (QUADSPI_DR)

Address offset: 0x0020
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DATA[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DATA[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 DATA[31:0]: Data

Data to be sent/received to/from the external SPI device.
In indirect write mode, data written to this register is stored on the FIFO before it is sent
to the Flash memory during the data phase. If the FIFO is too full, a write operation is
stalled until the FIFO has enough space to accept the amount of data being written.
In indirect read mode, reading this register gives (via the FIFO) the data which was
received from the Flash memory. If the FIFO does not have as many bytes as requested
by the read operation and if BUSY=1, the read operation is stalled until enough data is
present or until the transfer is complete, whichever happens first.
In automatic polling mode, this register contains the last data read from the Flash
memory (without masking).
Word, halfword, and byte accesses to this register are supported. In indirect write mode,
a byte write adds 1 byte to the FIFO, a halfword write 2, and a word write 4. Similarly, in
indirect read mode, a byte read removes 1 byte from the FIFO, a halfword read 2, and a
word read 4. Accesses in indirect mode must be aligned to the bottom of this register: a
byte read must read DATA[7:0] and a halfword read must read DATA[15:0].

RM0402 Rev 6 313/1163

316
Quad-SPI interface (QUADSPI) RM0402

12.5.10 QUADSPI polling status mask register (QUADSPI_PSMKR)

Address offset: 0x0024
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

MASK[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

MASK[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 MASK[31:0]: Status mask

Mask to be applied to the status bytes received in polling mode.
For bit n:
0: Bit n of the data received in automatic polling mode is masked and its value is not
considered in the matching logic
1: Bit n of the data received in automatic polling mode is unmasked and its value is
considered in the matching logic
This field can be written only when BUSY = 0.

12.5.11 QUADSPI polling status match register (QUADSPI_PSMAR)

Address offset: 0x0028
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

MATCH[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

MATCH[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 MATCH[31:0]: Status match

Value to be compared with the masked status register to get a match.
This field can be written only when BUSY = 0.

314/1163 RM0402 Rev 6

RM0402 Quad-SPI interface (QUADSPI)

12.5.12 QUADSPI polling interval register (QUADSPI_PIR)

Address offset: 0x002C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

INTERVAL[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 INTERVAL[15:0]: Polling interval
Number of CLK cycles between to read during automatic polling phases.
This field can be written only when BUSY = 0.

12.5.13 QUADSPI low-power timeout register (QUADSPI_LPTR)

Address offset: 0x0030
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

TIMEOUT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 TIMEOUT[15:0]: Timeout period
After each access in memory-mapped mode, the QUADSPI prefetches the subsequent
bytes and holds these bytes in the FIFO. This field indicates how many CLK cycles the
QUADSPI waits after the FIFO becomes full until it raises nCS, putting the Flash
memory in a lower-consumption state.
This field can be written only when BUSY = 0.

RM0402 Rev 6 315/1163

316
Quad-SPI interface (QUADSPI) RM0402

12.5.14 QUADSPI register map

Table 73. QUADSPI register map and reset values

Register
Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name

DMAEN
SSHIFT

ABORT
APMS
FTHRES

TCEN
FSEL
SMIE
PMM

TOIE

TCIE
TEIE

DFM
FTIE
Res.

Res.
Res.
Res.

Res.
QUADSPI_CR PRESCALER[7:0]

EN
[4:0]
0x0000

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CKMODE
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
QUADSPI_DCR FSIZE[4:0] CSHT
0x0004

Reset value 0 0 0 0 0 0 0 0 0

SMF
BUS
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.

TOF

TCF
TEF
FTF
QUADSPI_SR FLEVEL[5:0]
0x0008
Reset value 0 0 0 0 0 0 0 0 0 0 0 0

CSMF
CTOF

CTCF
CTEF
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
QUADSPI_FCR
0x000C
Reset value 0 0 0 0

QUADSPI_DLR DL[31:0]
0x0010
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ADMODE[1:0]
ABMODE[1:0]
DMODE[1:0]
FMODE[1:0]

ADSIZE[1:0]
ABSIZE[1:0]

IMODE[1:0]
DDRM
DHHC

SIOO
Res.

Res.

QUADSPI_CCR DCYC[4:0] INSTRUCTION[7:0]

0x0014

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

QUADSPI_AR ADDRESS[31:0]
0x0018
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
QUADSPI_ABR ALTERNATE[31:0]
0x001C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
QUADSPI_DR DATA[31:0]
0x0020
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
QUADSPI_
MASK[31:0]
0x0024 PSMKR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
QUADSPI_
MATCH[31:0]
0x0028 PSMAR
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

QUADSPI_PIR INTERVAL[15:0]
0x002C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

QUADSPI_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

TIMEOUT[15:0]
LPTR
0x0030
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

316/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13 Analog-to-digital converter (ADC)

13.1 ADC introduction

The 12-bit ADC is a successive approximation analog-to-digital converter. It has up to 19
multiplexed channels allowing it to measure signals from 16 external sources, two internal
sources, and the VBAT channel. The A/D conversion of the channels can be performed in
single, continuous, scan or discontinuous mode. The result of the ADC is stored into a left-
or right-aligned 16-bit data register.
The analog watchdog feature allows the application to detect if the input voltage goes
beyond the user-defined, higher or lower thresholds.

13.2 ADC main features

• 12-bit, 10-bit, 8-bit or 6-bit configurable resolution
• Interrupt generation at the end of conversion, end of injected conversion, and in case of
analog watchdog or overrun events
• Single and continuous conversion modes
• Scan mode for automatic conversion of channel 0 to channel ‘n’
• Data alignment with in-built data coherency
• Channel-wise programmable sampling time
• External trigger option with configurable polarity for both regular and injected
conversions
• Discontinuous mode
• Configurable delay between conversions in Dual/Triple interleaved mode
• ADC supply requirements: 2.4 V to 3.6 V at full speed and down to 1.8 V at slower
speed
• ADC input range: VREF– ≤ VIN ≤ VREF+
• DMA request generation during regular channel conversion
Figure 59 shows the block diagram of the ADC.
Note: VREF–, if available (depending on package), must be tied to VSSA.

13.3 ADC functional description

Figure 59 shows a single ADC block diagram and Table 74 gives the ADC pin description.

RM0402 Rev 6 317/1163

347
Analog-to-digital converter (ADC) RM0402

Figure 59. Single ADC block diagram

Flags Interrupt
enable bits
DMA overrun
OVR OVRIE
End of conversion
EOC EOCIE
End of injected conversion ADC Interrupt to NVIC
JEOC JEOCIE
Analog watchdog event
AWD AWDIE

Analog watchdog
Compare result

Higher threshold (12 bits)

Lower threshold (12 bits)

Address/data bus
Injected data registers
V REF+ (4 x 16 bits)
V REF-
Regular data register
V DDA (16 bits)
V SSA
Analog DMA request
mux

ADCx_IN0

ADCx_IN1
GPIO up to 4 ADCCLK
Injected
ports channels Analog to digital
up to 16 Regular converter
ADCx_IN15
channels

Temp. sensor
V REFINT
V BAT

From ADC prescaler

JEXTSEL[3:0] bits EXTSEL[3:0] bits

TIM1_CH4 TIM1_CH1
TIM1_TRGO JEXTEN EXTEN TIM1_CH2
TIM2_CH1 [1:0] bits [1:0] bits TIM1_CH3
TIM2_TRGO TIM2_CH2
TIM3_CH2 TIM2_CH3
TIM3_CH4 TIM2_CH4
TIM4_CH1 TIM2_TRGO
TIM4_CH2 TIM3_CH1
TIM4_CH3 TIM3_TRGO
TIM4_TRGO TIM4_CH4
Start trigger Start trigger
TIM5_CH4 TIM5_CH1
(injected group) (regular group)
TIM5_TRGO TIM5_CH2
TIM8_CH2 TIM5_CH3
TIM8_CH3 TIM8_CH1
TIM8_CH4 TIM8_TRGO
EXTI_15

EXTI_11

ai16046

318/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Table 74. ADC pins

Name Signal type Remarks

Input, analog reference The higher/positive reference voltage for the ADC,
VREF+
positive 1.8 V ≤ VREF+ ≤ VDDA
Analog power supply equal to VDD and
VDDA Input, analog supply 2.4 V ≤VDDA ≤VDD (3.6 V) for full speed
1.8 V ≤VDDA ≤VDD (3.6 V) for reduced speed
Input, analog reference The lower/negative reference voltage for the ADC,
VREF–
negative VREF– = VSSA
Input, analog supply
VSSA Ground for analog power supply equal to VSS
ground
ADCx_IN[15:0] Analog input signals 16 analog input channels

13.3.1 ADC on-off control

The ADC is powered on by setting the ADON bit in the ADC_CR2 register. When the ADON
bit is set for the first time, it wakes up the ADC from the Power-down mode.
The conversion starts when either the SWSTART or the JSWSTART bit is set.
The user can stop conversion and put the ADC in power down mode by clearing the ADON
bit. In this mode the ADC consumes almost no power (only a few µA).

13.3.2 ADC clock

The ADC features two clock schemes:
• Clock for the analog circuitry: ADCCLK
This clock is generated from the APB2 clock divided by a programmable prescaler that
allows the ADC to work at fPCLK2/2, /4, /6 or /8. Refer to the datasheets for the
maximum value of ADCCLK.
• Clock for the digital interface (used for registers read/write access)
This clock is equal to the APB2 clock. The digital interface clock can be
enabled/disabled individually for each ADC through the RCC APB2 peripheral clock
enable register (RCC_APB2ENR).

13.3.3 Channel selection

There are 16 multiplexed channels. It is possible to organize the conversions in two groups:
regular and injected. A group consists of a sequence of conversions that can be done on
any channel and in any order. For instance, it is possible to implement the conversion
sequence in the following order: ADC_IN3, ADC_IN8, ADC_IN2, ADC_IN2, ADC_IN0,
ADC_IN2, ADC_IN2, ADC_IN15.
• A regular group is composed of up to 16 conversions. The regular channels and their
order in the conversion sequence must be selected in the ADC_SQRx registers. The
total number of conversions in the regular group must be written in the L[3:0] bits in the
ADC_SQR1 register.
• An injected group is composed of up to 4 conversions. The injected channels and
their order in the conversion sequence must be selected in the ADC_JSQR register.

RM0402 Rev 6 319/1163

347
Analog-to-digital converter (ADC) RM0402

The total number of conversions in the injected group must be written in the L[1:0] bits
in the ADC_JSQR register.
If the ADC_SQRx or ADC_JSQR registers are modified during a conversion, the current
conversion is reset and a new start pulse is sent to the ADC to convert the newly chosen
group.

Temperature sensor, VREFINT and VBAT internal channels

• The temperature sensor is internally connected to ADC1_IN18 channel which is shared
with VBAT. Only one conversion, temperature sensor or VBAT, must be selected at a
time. When the temperature sensor and VBAT conversion are set simultaneously, only
the VBAT conversion is performed.
The internal reference voltage VREFINT is connected to ADC1_IN17.
The VBAT channel is connected to ADC1_IN18 channel. It can also be converted as an
injected or regular channel.

13.3.4 Single conversion mode

In Single conversion mode the ADC does one conversion. This mode is started with the
CONT bit at 0 by either:
• setting the SWSTART bit in the ADC_CR2 register (for a regular channel only)
• setting the JSWSTART bit (for an injected channel)
• external trigger (for a regular or injected channel)
Once the conversion of the selected channel is complete:
• If a regular channel was converted:
– The converted data are stored into the 16-bit ADC_DR register
– The EOC (end of conversion) flag is set
– An interrupt is generated if the EOCIE bit is set
• If an injected channel was converted:
– The converted data are stored into the 16-bit ADC_JDR1 register
– The JEOC (end of conversion injected) flag is set
– An interrupt is generated if the JEOCIE bit is set
Then the ADC stops.

13.3.5 Continuous conversion mode

In continuous conversion mode, the ADC starts a new conversion as soon as it finishes one.
This mode is started with the CONT bit at 1 either by external trigger or by setting the
SWSTRT bit in the ADC_CR2 register (for regular channels only).
After each conversion:
• If a regular group of channels was converted:
– The last converted data are stored into the 16-bit ADC_DR register
– The EOC (end of conversion) flag is set
– An interrupt is generated if the EOCIE bit is set

320/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Note: Injected channels cannot be converted continuously. The only exception is when an injected
channel is configured to be converted automatically after regular channels in continuous
mode (using JAUTO bit), refer to Auto-injection section).

13.3.6 Timing diagram

As shown in Figure 60, the ADC needs a stabilization time of tSTAB before it starts
converting accurately. After the start of the ADC conversion and after 15 clock cycles, the
EOC flag is set and the 16-bit ADC data register contains the result of the conversion.

Figure 60. Timing diagram

ADC_CLK

ADON

SWSTART/
JSWSTART

Start 1st conversion Start next conversion

ADC conversion Next ADC conversion

ADC
Conversion time
tSTAB
(total conv. time)
EOC

Software clears the EOC bit

ai16047b

13.3.7 Analog watchdog

The AWD analog watchdog status bit is set if the analog voltage converted by the ADC is
below a lower threshold or above a higher threshold. These thresholds are programmed in
the 12 least significant bits of the ADC_HTR and ADC_LTR 16-bit registers. An interrupt can
be enabled by using the AWDIE bit in the ADC_CR1 register.
The threshold value is independent of the alignment selected by the ALIGN bit in the
ADC_CR2 register. The analog voltage is compared to the lower and higher thresholds
before alignment.
Table 75 shows how the ADC_CR1 register should be configured to enable the analog
watchdog on one or more channels.

RM0402 Rev 6 321/1163

347
Analog-to-digital converter (ADC) RM0402

Figure 61. Analog watchdog’s guarded area

Analog voltage
Higher threshold HTR
Guarded area
Lower threshold LTR
ai16048

Table 75. Analog watchdog channel selection

ADC_CR1 register control bits (x = don’t care)
Channels guarded by the analog
watchdog
AWDSGL bit AWDEN bit JAWDEN bit

None x 0 0
All injected channels 0 0 1
All regular channels 0 1 0
All regular and injected channels 0 1 1
Single(1) injected channel 1 0 1
Single(1) regular channel 1 1 0
(1)
Single regular or injected channel 1 1 1
1. Selected by the AWDCH[4:0] bits

13.3.8 Scan mode

This mode is used to scan a group of analog channels.
The Scan mode is selected by setting the SCAN bit in the ADC_CR1 register. Once this bit
has been set, the ADC scans all the channels selected in the ADC_SQRx registers (for
regular channels) or in the ADC_JSQR register (for injected channels). A single conversion
is performed for each channel of the group. After each end of conversion, the next channel
in the group is converted automatically. If the CONT bit is set, regular channel conversion
does not stop at the last selected channel in the group but continues again from the first
selected channel.
If the DMA bit is set, the direct memory access (DMA) controller is used to transfer the data
converted from the regular group of channels (stored in the ADC_DR register) to SRAM
after each regular channel conversion.
The EOC bit is set in the ADC_SR register:
• At the end of each regular group sequence if the EOCS bit is cleared to 0
• At the end of each regular channel conversion if the EOCS bit is set to 1
The data converted from an injected channel are always stored into the ADC_JDRx
registers.

322/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13.3.9 Injected channel management

Triggered injection
To use triggered injection, the JAUTO bit must be cleared in the ADC_CR1 register.
1. Start the conversion of a group of regular channels either by external trigger or by
setting the SWSTART bit in the ADC_CR2 register.
2. If an external injected trigger occurs or if the JSWSTART bit is set during the
conversion of a regular group of channels, the current conversion is reset and the
injected channel sequence switches to Scan-once mode.
3. Then, the regular conversion of the regular group of channels is resumed from the last
interrupted regular conversion.
If a regular event occurs during an injected conversion, the injected conversion is not
interrupted but the regular sequence is executed at the end of the injected sequence.
Figure 62 shows the corresponding timing diagram.
Note: When using triggered injection, one must ensure that the interval between trigger events is
longer than the injection sequence. For instance, if the sequence length is 30 ADC clock
cycles (that is two conversions with a sampling time of 3 clock periods), the minimum
interval between triggers must be 31 ADC clock cycles.

Auto-injection
If the JAUTO bit is set, then the channels in the injected group are automatically converted
after the regular group of channels. This can be used to convert a sequence of up to 20
conversions programmed in the ADC_SQRx and ADC_JSQR registers.
In this mode, external trigger on injected channels must be disabled.
If the CONT bit is also set in addition to the JAUTO bit, regular channels followed by injected
channels are continuously converted.
Note: It is not possible to use both the auto-injected and discontinuous modes simultaneously.

Figure 62. Injected conversion latency

ADCCLK

Injection event

Reset ADC

max latency (1)

SOC

ai16049

1. The maximum latency value can be found in the electrical characteristics of the STM32F412xx datasheets.

RM0402 Rev 6 323/1163

347
Analog-to-digital converter (ADC) RM0402

13.3.10 Discontinuous mode

Regular group
This mode is enabled by setting the DISCEN bit in the ADC_CR1 register. It can be used to
convert a short sequence of n conversions (n ≤ 8) that is part of the sequence of
conversions selected in the ADC_SQRx registers. The value of n is specified by writing to
the DISCNUM[2:0] bits in the ADC_CR1 register.
When an external trigger occurs, it starts the next n conversions selected in the ADC_SQRx
registers until all the conversions in the sequence are done. The total sequence length is
defined by the L[3:0] bits in the ADC_SQR1 register.
Example:
• n = 3, channels to be converted = 0, 1, 2, 3, 6, 7, 9, 10
• 1st trigger: sequence converted 0, 1, 2. An EOC event is generated at each
conversion.
• 2nd trigger: sequence converted 3, 6, 7. An EOC event is generated at each
conversion
• 3rd trigger: sequence converted 9, 10.An EOC event is generated at each conversion
• 4th trigger: sequence converted 0, 1, 2. An EOC event is generated at each conversion
Note: When a regular group is converted in discontinuous mode, no rollover occurs.
When all subgroups are converted, the next trigger starts the conversion of the first
subgroup. In the example above, the 4th trigger reconverts the channels 0, 1 and 2 in the
1st subgroup.

Injected group
This mode is enabled by setting the JDISCEN bit in the ADC_CR1 register. It can be used to
convert the sequence selected in the ADC_JSQR register, channel by channel, after an
external trigger event.
When an external trigger occurs, it starts the next channel conversions selected in the
ADC_JSQR registers until all the conversions in the sequence are done. The total sequence
length is defined by the JL[1:0] bits in the ADC_JSQR register.
Example:
n = 1, channels to be converted = 1, 2, 3
1st trigger: channel 1 converted
2nd trigger: channel 2 converted
3rd trigger: channel 3 converted and JEOC event generated
4th trigger: channel 1
Note: When all injected channels are converted, the next trigger starts the conversion of the first
injected channel. In the example above, the 4th trigger reconverts the 1st injected channel
1.
It is not possible to use both the auto-injected and discontinuous modes simultaneously.
Discontinuous mode must not be set for regular and injected groups at the same time.
Discontinuous mode must be enabled only for the conversion of one group.

324/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13.4 Data alignment

The ALIGN bit in the ADC_CR2 register selects the alignment of the data stored after
conversion. Data can be right- or left-aligned as shown in Figure 63 and Figure 64.
The converted data value from the injected group of channels is decreased by the user-
defined offset written in the ADC_JOFRx registers so the result can be a negative value.
The SEXT bit represents the extended sign value.
For channels in a regular group, no offset is subtracted so only twelve bits are significant.

Figure 63. Right alignment of 12-bit data

Injected group

SEXT SEXT SEXT SEXT D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0

Regular group

0 0 0 0 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0

ai16050

Figure 64. Left alignment of 12-bit data

Injected group

SEXT D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 0 0 0

Regular group

D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 0 0 0 0

ai16051

Special case: when left-aligned, the data are aligned on a half-word basis except when the
resolution is set to 6-bit. in that case, the data are aligned on a byte basis as shown in
Figure 65.

Figure 65. Left alignment of 6-bit data

Injected group

SEXT SEXT SEXT SEXT SEXT SEXT SEXT SEXT SEXT D5 D4 D3 D2 D1 D0 0

Regular group

0 0 0 0 0 0 0 0 D5 D4 D3 D2 D1 D0 0 0

ai16052

RM0402 Rev 6 325/1163

347
Analog-to-digital converter (ADC) RM0402

13.5 Channel-wise programmable sampling time

The ADC samples the input voltage for a number of ADCCLK cycles that can be modified
using the SMP[2:0] bits in the ADC_SMPR1 and ADC_SMPR2 registers. Each channel can
be sampled with a different sampling time.
The total conversion time is calculated as follows:
Tconv = Sampling time + 12 cycles
Example:
With ADCCLK = 30 MHz and sampling time = 3 cycles:
Tconv = 3 + 12 = 15 cycles = 0.5 µs with APB2 at 60 MHz

13.6 Conversion on external trigger and trigger polarity

Conversion can be triggered by an external event (e.g. timer capture, EXTI line). If the
EXTEN[1:0] control bits (for a regular conversion) or JEXTEN[1:0] bits (for an injected
conversion) are different from “0b00”, then external events are able to trigger a conversion
with the selected polarity. Table 76 provides the correspondence between the EXTEN[1:0]
and JEXTEN[1:0] values and the trigger polarity.

Table 76. Configuring the trigger polarity

Source EXTEN[1:0] / JEXTEN[1:0]

Trigger detection disabled 00

Detection on the rising edge 01
Detection on the falling edge 10
Detection on both the rising and falling edges 11

Note: The polarity of the external trigger can be changed on the fly.
The EXTSEL[3:0] and JEXTSEL[3:0] control bits are used to select which out of 16 possible
events can trigger conversion for the regular and injected groups.
Table 77 gives the possible external trigger for regular conversion.

326/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Table 77. External trigger for regular channels

Source Type EXTSEL[3:0]

TIM1_CH1 event 0000

TIM1_CH2 event 0001
TIM1_CH3 event 0010
TIM2_CH2 event 0011
TIM2_CH3 event 0100
TIM2_CH4 event 0101
TIM2_TRGO event 0110
TIM3_CH1 event Internal signal from on-chip timers 0111
TIM3_TRGO event 1000
TIM4_CH4 event 1001
TIM5_CH1 event 1010
TIM5_CH2 event 1011
TIM5_CH3 event 1100
TIM8_CH1 event 1101
TIM8_TRGO event 1110
EXTI line11 External pin 1111

RM0402 Rev 6 327/1163

347
Analog-to-digital converter (ADC) RM0402

Table 78 gives the possible external trigger for injected conversion.

Table 78. External trigger for injected channels
Source Connection type JEXTSEL[3:0]

TIM1_CH4 event 0000

TIM1_TRGO event 0001
TIM2_CH1 event 0010
TIM2_TRGO event 0011
TIM3_CH2 event 0100
TIM3_CH4 event 0101
TIM4_CH1 event 0110
TIM4_CH2 event Internal signal from on-chip timers 0111
TIM4_CH3 event 1000
TIM4_TRGO event 1001
TIM5_CH4 event 1010
TIM5_TRGO event 1011
TIM8_CH2 event 1100
TIM8_CH3 event 1101
TIM8_CH4 event 1110
EXTI line15 External pin 1111

Software source trigger events can be generated by setting SWSTART (for regular
conversion) or JSWSTART (for injected conversion) in ADC_CR2.
A regular group conversion can be interrupted by an injected trigger.
Note: The trigger selection can be changed on the fly. However, when the selection changes,
there is a time frame of 1 APB clock cycle during which the trigger detection is disabled.
This is to avoid spurious detection during transitions.

13.7 Fast conversion mode

It is possible to perform faster conversion by reducing the ADC resolution. The RES bits are
used to select the number of bits available in the data register. The minimum conversion
time for each resolution is then as follows:
• 12 bits: 3 + 12 = 15 ADCCLK cycles
• 10 bits: 3 + 10 = 13 ADCCLK cycles
• 8 bits: 3 + 8 = 11 ADCCLK cycles
• 6 bits: 3 + 6 = 9 ADCCLK cycles

328/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13.8 Data management

13.8.1 Using the DMA

Since converted regular channel values are stored into a unique data register, it is useful to
use DMA for conversion of more than one regular channel. This avoids the loss of the data
already stored in the ADC_DR register.
When the DMA mode is enabled (DMA bit set to 1 in the ADC_CR2 register), after each
conversion of a regular channel, a DMA request is generated. This allows the transfer of the
converted data from the ADC_DR register to the destination location selected by the
software.
Despite this, if data are lost (overrun), the OVR bit in the ADC_SR register is set and an
interrupt is generated (if the OVRIE enable bit is set). DMA transfers are then disabled and
DMA requests are no longer accepted. In this case, if a DMA request is made, the regular
conversion in progress is aborted and further regular triggers are ignored. It is then
necessary to clear the OVR flag and the DMAEN bit in the used DMA stream, and to re-
initialize both the DMA and the ADC to have the wanted converted channel data transferred
to the right memory location. Only then can the conversion be resumed and the data
transfer, enabled again. Injected channel conversions are not impacted by overrun errors.
When OVR = 1 in DMA mode, the DMA requests are blocked after the last valid data have
been transferred, which means that all the data transferred to the RAM can be considered
as valid.
At the end of the last DMA transfer (number of transfers configured in the DMA controller’s
DMA_SxNDTR register):
• No new DMA request is issued to the DMA controller if the DDS bit is cleared to 0 in the
ADC_CR2 register (this avoids generating an overrun error). However the DMA bit is
not cleared by hardware. It must be written to 0, then to 1 to start a new transfer.
• Requests can continue to be generated if the DDS bit is set to 1. This allows
configuring the DMA in double-buffer circular mode.
To recover the ADC from OVR state when the DMA is used, follow the steps below:
1. Reinitialize the DMA (adjust destination address and NDTR counter)
2. Clear the ADC OVR bit in ADC_SR register
3. Trigger the ADC to start the conversion.

13.8.2 Managing a sequence of conversions without using the DMA

If the conversions are slow enough, the conversion sequence can be handled by the
software. In this case the EOCS bit must be set in the ADC_CR2 register for the EOC status
bit to be set at the end of each conversion, and not only at the end of the sequence. When
EOCS = 1, overrun detection is automatically enabled. Thus, each time a conversion is
complete, EOC is set and the ADC_DR register can be read. The overrun management is
the same as when the DMA is used.
To recover the ADC from OVR state when the EOCS is set, follow the steps below:
1. Clear the ADC OVR bit in ADC_SR register
2. Trigger the ADC to start the conversion.

RM0402 Rev 6 329/1163

347
Analog-to-digital converter (ADC) RM0402

13.8.3 Conversions without DMA and without overrun detection

It may be useful to let the ADC convert one or more channels without reading the data each
time (if there is an analog watchdog for instance). For that, the DMA must be disabled
(DMA = 0) and the EOC bit must be set at the end of a sequence only (EOCS = 0). In this
configuration, overrun detection is disabled.

13.9 Temperature sensor

The temperature sensor can be used to measure the ambient temperature (TA) of the
device.
Figure 66 shows the block diagram of the temperature sensor.
When not in use, the sensor can be put in power down mode.
Note: The TSVREFE bit must be set to enable the conversion of both internal channels: the
ADC1_IN18 (temperature sensor) and the ADC1_IN17 (VREFINT).

Main features
• Supported temperature range: –40 to 125 °C
• Precision: ±1.5 °C

Figure 66. Temperature sensor and VREFINT channel block diagram

TSVREFE control bit

Temperature V SENSE
sensor ADC1_IN18

Address/data bus
converted data
ADC1

VREFINT
Internal
power block ADC1_IN17

MS35936V1

1. VSENSE is input to ADC1_IN18.

330/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Reading the temperature

To use the sensor:
3. Select ADC1_IN18 input channel.
4. Select a sampling time greater than the minimum sampling time specified in the
datasheet.
5. Set the TSVREFE bit in the ADC_CCR register to wake up the temperature sensor
from power down mode
6. Start the ADC conversion by setting the SWSTART bit (or by external trigger)
7. Read the resulting VSENSE data in the ADC data register
8. Calculate the temperature using the following formula:
Temperature (in °C) = {(VSENSE – V25) / Avg_Slope} + 25
Where:
– V25 = VSENSE value for 25° C
– Avg_Slope = average slope of the temperature vs. VSENSE curve (given in mV/°C
or µV/°C)
Refer to the datasheet electrical characteristics section for the actual values of V25 and
Avg_Slope.
Note: The sensor has a startup time after waking from power down mode before it can output
VSENSE at the correct level. The ADC also has a startup time after power-on, so to minimize
the delay, the ADON and TSVREFE bits should be set at the same time.
The temperature sensor output voltage changes linearly with temperature. The offset of this
linear function depends on each chip due to process variation (up to 45 °C from one chip to
another).
The internal temperature sensor is more suited for applications that detect temperature
variations instead of absolute temperatures. If accurate temperature reading is required, an
external temperature sensor should be used.

13.10 Battery charge monitoring

The VBATE bit in the ADC_CCR register is used to switch to the battery voltage. As the
VBAT voltage could be higher than VDDA, to ensure the correct operation of the ADC, the
VBAT pin is internally connected to a bridge divider.
When the VBATE is set, the bridge is automatically enabled to connect:
• VBAT/4 to the ADC1_IN18 input channel
Note: The VBAT and temperature sensor are connected to the same ADC internal channel
(ADC1_IN18). Only one conversion, either temperature sensor or VBAT, must be selected
at a time. When both conversion are enabled simultaneously, only the VBAT conversion is
performed.

RM0402 Rev 6 331/1163

347
Analog-to-digital converter (ADC) RM0402

13.11 ADC interrupts

An interrupt can be produced on the end of conversion for regular and injected groups,
when the analog watchdog status bit is set and when the overrun status bit is set. Separate
interrupt enable bits are available for flexibility.
Two other flags are present in the ADC_SR register, but there is no interrupt associated with
them:
• JSTRT (Start of conversion for channels of an injected group)
• STRT (Start of conversion for channels of a regular group)

Table 79. ADC interrupts

Interrupt event Event flag Enable control bit

End of conversion of a regular group EOC EOCIE

End of conversion of an injected group JEOC JEOCIE
Analog watchdog status bit is set AWD AWDIE
Overrun OVR OVRIE

332/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13.12 ADC registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers must be written at word level (32 bits). Read accesses can be done
by bytes (8 bits), half-words (16 bits) or words (32 bits).

13.12.1 ADC status register (ADC_SR)

Address offset: 0x00
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. OVR STRT JSTRT JEOC EOC AWD
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0

Bits 31:6 Reserved, must be kept at reset value.

Bit 5 OVR: Overrun
This bit is set by hardware when data are lost (either in single mode or in dual/triple mode). It
is cleared by software. Overrun detection is enabled only when DMA = 1 or EOCS = 1.
0: No overrun occurred
1: Overrun has occurred
Bit 4 STRT: Regular channel start flag
This bit is set by hardware when regular channel conversion starts. It is cleared by software.
0: No regular channel conversion started
1: Regular channel conversion has started
Bit 3 JSTRT: Injected channel start flag
This bit is set by hardware when injected group conversion starts. It is cleared by software.
0: No injected group conversion started
1: Injected group conversion has started
Bit 2 JEOC: Injected channel end of conversion
This bit is set by hardware at the end of the conversion of all injected channels in the group.
It is cleared by software.
0: Conversion is not complete
1: Conversion complete
Bit 1 EOC: Regular channel end of conversion
This bit is set by hardware at the end of the conversion of a regular group of channels. It is
cleared by software or by reading the ADC_DR register.
0: Conversion not complete (EOCS=0), or sequence of conversions not complete (EOCS=1)
1: Conversion complete (EOCS=0), or sequence of conversions complete (EOCS=1)
Bit 0 AWD: Analog watchdog flag
This bit is set by hardware when the converted voltage crosses the values programmed in
the ADC_LTR and ADC_HTR registers. It is cleared by software.
0: No analog watchdog event occurred
1: Analog watchdog event occurred

RM0402 Rev 6 333/1163

347
Analog-to-digital converter (ADC) RM0402

13.12.2 ADC control register 1 (ADC_CR1)

Address offset: 0x04
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. OVRIE RES AWDEN JAWDEN Res. Res. Res. Res. Res. Res.
rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DISCNUM[2:0] JDISCEN DISCEN JAUTO AWDSGL SCAN JEOCIE AWDIE EOCIE AWDCH[4:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:27 Reserved, must be kept at reset value.

Bit 26 OVRIE: Overrun interrupt enable
This bit is set and cleared by software to enable/disable the Overrun interrupt.
0: Overrun interrupt disabled
1: Overrun interrupt enabled. An interrupt is generated when the OVR bit is set.
Bits 25:24 RES[1:0]: Resolution
These bits are written by software to select the resolution of the conversion.
00: 12-bit (minimum 15 ADCCLK cycles)
01: 10-bit (minimum 13 ADCCLK cycles)
10: 8-bit (minimum 11 ADCCLK cycles)
11: 6-bit (minimum 9 ADCCLK cycles)
Bit 23 AWDEN: Analog watchdog enable on regular channels
This bit is set and cleared by software.
0: Analog watchdog disabled on regular channels
1: Analog watchdog enabled on regular channels
Bit 22 JAWDEN: Analog watchdog enable on injected channels
This bit is set and cleared by software.
0: Analog watchdog disabled on injected channels
1: Analog watchdog enabled on injected channels
Bits 21:16 Reserved, must be kept at reset value.
Bits 15:13 DISCNUM[2:0]: Discontinuous mode channel count
These bits are written by software to define the number of regular channels to be converted
in discontinuous mode, after receiving an external trigger.
000: 1 channel
001: 2 channels
...
111: 8 channels
Bit 12 JDISCEN: Discontinuous mode on injected channels
This bit is set and cleared by software to enable/disable discontinuous mode on the injected
channels of a group.
0: Discontinuous mode on injected channels disabled
1: Discontinuous mode on injected channels enabled

334/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Bit 11 DISCEN: Discontinuous mode on regular channels

This bit is set and cleared by software to enable/disable Discontinuous mode on regular
channels.
0: Discontinuous mode on regular channels disabled
1: Discontinuous mode on regular channels enabled
Bit 10 JAUTO: Automatic injected group conversion
This bit is set and cleared by software to enable/disable automatic injected group conversion
after regular group conversion.
0: Automatic injected group conversion disabled
1: Automatic injected group conversion enabled
Bit 9 AWDSGL: Enable the watchdog on a single channel in scan mode
This bit is set and cleared by software to enable/disable the analog watchdog on the channel
identified by the AWDCH[4:0] bits.
0: Analog watchdog enabled on all channels
1: Analog watchdog enabled on a single channel
Bit 8 SCAN: Scan mode
This bit is set and cleared by software to enable/disable the Scan mode. In Scan mode, the
inputs selected through the ADC_SQRx or ADC_JSQRx registers are converted.
0: Scan mode disabled
1: Scan mode enabled
Note: An EOC interrupt is generated if the EOCIE bit is set:
– At the end of each regular group sequence if the EOCS bit is cleared to 0
– At the end of each regular channel conversion if the EOCS bit is set to 1
Note: A JEOC interrupt is generated only on the end of conversion of the last channel if the
JEOCIE bit is set.
Bit 7 JEOCIE: Interrupt enable for injected channels
This bit is set and cleared by software to enable/disable the end of conversion interrupt for
injected channels.
0: JEOC interrupt disabled
1: JEOC interrupt enabled. An interrupt is generated when the JEOC bit is set.
Bit 6 AWDIE: Analog watchdog interrupt enable
This bit is set and cleared by software to enable/disable the analog watchdog interrupt.
0: Analog watchdog interrupt disabled
1: Analog watchdog interrupt enabled
Bit 5 EOCIE: Interrupt enable for EOC
This bit is set and cleared by software to enable/disable the end of conversion interrupt.
0: EOC interrupt disabled
1: EOC interrupt enabled. An interrupt is generated when the EOC bit is set.
Bits 4:0 AWDCH[4:0]: Analog watchdog channel select bits
These bits are set and cleared by software. They select the input channel to be guarded by
the analog watchdog.
Note: 00000: ADC analog input Channel0
00001: ADC analog input Channel1
...
01111: ADC analog input Channel15
10000: ADC analog input Channel16
10001: ADC analog input Channel17
10010: ADC analog input Channel18
Other values reserved

RM0402 Rev 6 335/1163

347
Analog-to-digital converter (ADC) RM0402

13.12.3 ADC control register 2 (ADC_CR2)

Address offset: 0x08
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. SWSTART EXTEN EXTSEL[3:0] Res. JSWSTART JEXTEN JEXTSEL[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. ALIGN EOCS DDS DMA Res. Res. Res. Res. Res. Res. CONT ADON
rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.

Bit 30 SWSTART: Start conversion of regular channels
This bit is set by software to start conversion and cleared by hardware as soon as the
conversion starts.
0: Reset state
1: Starts conversion of regular channels
Note: This bit can be set only when ADON = 1 otherwise no conversion is launched.
Bits 29:28 EXTEN: External trigger enable for regular channels
These bits are set and cleared by software to select the external trigger polarity and enable
the trigger of a regular group.
00: Trigger detection disabled
01: Trigger detection on the rising edge
10: Trigger detection on the falling edge
11: Trigger detection on both the rising and falling edges
Bits 27:24 EXTSEL[3:0]: External event select for regular group
These bits select the external event used to trigger the start of conversion of a regular group:
0000: Timer 1 CC1 event
0001: Timer 1 CC2 event
0010: Timer 1 CC3 event
0011: Timer 2 CC2 event
0100: Timer 2 CC3 event
0101: Timer 2 CC4 event
0110: Timer 2 TRGO event
0111: Timer 3 CC1 event
1000: Timer 3 TRGO event
1001: Timer 4 CC4 event
1010: Timer 5 CC1 event
1011: Timer 5 CC2 event
1100: Timer 5 CC3 event
1101: Timer 8 CC1 event
1110: Timer 8 TRGO event
1111: EXTI line 11
Bit 23 Reserved, must be kept at reset value.

336/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Bit 22 JSWSTART: Start conversion of injected channels

This bit is set by software and cleared by hardware as soon as the conversion starts.
0: Reset state
1: Starts conversion of injected channels
This bit can be set only when ADON = 1 otherwise no conversion is launched.
Bits 21:20 JEXTEN: External trigger enable for injected channels
These bits are set and cleared by software to select the external trigger polarity and enable
the trigger of an injected group.
00: Trigger detection disabled
01: Trigger detection on the rising edge
10: Trigger detection on the falling edge
11: Trigger detection on both the rising and falling edges
Bits 19:16 JEXTSEL[3:0]: External event select for injected group
These bits select the external event used to trigger the start of conversion of an injected group.
0000: Timer 1 CC4 event
0001: Timer 1 TRGO event
0010: Timer 2 CC1 event
0011: Timer 2 TRGO event
0100: Timer 3 CC2 event
0101: Timer 3 CC4 event
0110: Timer 4 CC1 event
0111: Timer 4 CC2 event
1000: Timer 4 CC3 event
1001: Timer 4 TRGO event
1010: Timer 5 CC4 event
1011: Timer 5 TRGO event
1100: Timer 8 CC2 event
1101: Timer 8 CC3 event
1110: Timer 8 CC4 event
1111: EXTI line15
Bits 15:12 Reserved, must be kept at reset value.
Bit 11 ALIGN: Data alignment
This bit is set and cleared by software. Refer to Figure 63 and Figure 64.
0: Right alignment
1: Left alignment
Bit 10 EOCS: End of conversion selection
This bit is set and cleared by software.
0: The EOC bit is set at the end of each sequence of regular conversions. Overrun detection
is enabled only if DMA=1.
1: The EOC bit is set at the end of each regular conversion. Overrun detection is enabled.
Bit 9 DDS: DMA disable selection (for single ADC mode)
This bit is set and cleared by software.
0: No new DMA request is issued after the last transfer (as configured in the DMA controller)
1: DMA requests are issued as long as data are converted and DMA=1

RM0402 Rev 6 337/1163

347
Analog-to-digital converter (ADC) RM0402

Bit 8 DMA: Direct memory access mode (for single ADC mode)
This bit is set and cleared by software. Refer to the DMA controller chapter for more details.
0: DMA mode disabled
1: DMA mode enabled
Bits 7:2 Reserved, must be kept at reset value.
Bit 1 CONT: Continuous conversion
This bit is set and cleared by software. If it is set, conversion takes place continuously until it
is cleared.
0: Single conversion mode
1: Continuous conversion mode
Bit 0 ADON: A/D Converter ON / OFF
This bit is set and cleared by software.
0: Disable ADC conversion and go to power down mode
1: Enable ADC

13.12.4 ADC sample time register 1 (ADC_SMPR1)

Address offset: 0x0C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. SMP18[2:0] SMP17[2:0] SMP16[2:0] SMP15[2:1]
rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMP15_0 SMP14[2:0] SMP13[2:0] SMP12[2:0] SMP11[2:0] SMP10[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31: 27 Reserved, must be kept at reset value.

Bits 26:0 SMPx[2:0]: Channel x sampling time selection
These bits are written by software to select the sampling time individually for each channel.
During sampling cycles, the channel selection bits must remain unchanged.
Note: 000: 3 cycles
001: 15 cycles
010: 28 cycles
011: 56 cycles
100: 84 cycles
101: 112 cycles
110: 144 cycles
111: 480 cycles

338/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13.12.5 ADC sample time register 2 (ADC_SMPR2)

Address offset: 0x10
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. SMP9[2:0] SMP8[2:0] SMP7[2:0] SMP6[2:0] SMP5[2:1]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMP5_0 SMP4[2:0] SMP3[2:0] SMP2[2:0] SMP1[2:0] SMP0[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:30 Reserved, must be kept at reset value.

Bits 29:0 SMPx[2:0]: Channel x sampling time selection
These bits are written by software to select the sampling time individually for each channel.
During sample cycles, the channel selection bits must remain unchanged.
Note: 000: 3 cycles
001: 15 cycles
010: 28 cycles
011: 56 cycles
100: 84 cycles
101: 112 cycles
110: 144 cycles
111: 480 cycles

13.12.6 ADC injected channel data offset register x (ADC_JOFRx) (x=1..4)

Address offset: 0x14-0x20
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. JOFFSETx[11:0]
rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:12 Reserved, must be kept at reset value.

Bits 11:0 JOFFSETx[11:0]: Data offset for injected channel x
These bits are written by software to define the offset to be subtracted from the raw
converted data when converting injected channels. The conversion result can be read from
in the ADC_JDRx registers.

13.12.7 ADC watchdog higher threshold register (ADC_HTR)

Address offset: 0x24
Reset value: 0x0000 0FFF

RM0402 Rev 6 339/1163

347
Analog-to-digital converter (ADC) RM0402

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. HT[11:0]
rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:12 Reserved, must be kept at reset value.

Bits 11:0 HT[11:0]: Analog watchdog higher threshold
These bits are written by software to define the higher threshold for the analog watchdog.

Note: The software can write to these registers when an ADC conversion is ongoing. The
programmed value will be effective when the next conversion is complete. Writing to this
register is performed with a write delay that can create uncertainty on the effective time at
which the new value is programmed.

13.12.8 ADC watchdog lower threshold register (ADC_LTR)

Address offset: 0x28
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. LT[11:0]
rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:12 Reserved, must be kept at reset value.

Bits 11:0 LT[11:0]: Analog watchdog lower threshold
These bits are written by software to define the lower threshold for the analog watchdog.

Note: The software can write to these registers when an ADC conversion is ongoing. The
programmed value will be effective when the next conversion is complete. Writing to this
register is performed with a write delay that can create uncertainty on the effective time at
which the new value is programmed.

13.12.9 ADC regular sequence register 1 (ADC_SQR1)

Address offset: 0x2C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. L[3:0] SQ16[4:1]
rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SQ16_0 SQ15[4:0] SQ14[4:0] SQ13[4:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

340/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Bits 31:24 Reserved, must be kept at reset value.

Bits 23:20 L[3:0]: Regular channel sequence length
These bits are written by software to define the total number of conversions in the regular
channel conversion sequence.
0000: 1 conversion
0001: 2 conversions
...
1111: 16 conversions
Bits 19:15 SQ16[4:0]: 16th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 16th in
the conversion sequence.
Bits 14:10 SQ15[4:0]: 15th conversion in regular sequence
Bits 9:5 SQ14[4:0]: 14th conversion in regular sequence
Bits 4:0 SQ13[4:0]: 13th conversion in regular sequence

13.12.10 ADC regular sequence register 2 (ADC_SQR2)

Address offset: 0x30
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. SQ12[4:0] SQ11[4:0] SQ10[4:1]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SQ10_0 SQ9[4:0] SQ8[4:0] SQ7[4:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:30 Reserved, must be kept at reset value.

Bits 29:26 SQ12[4:0]: 12th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 12th in
the sequence to be converted.
Bits 24:20 SQ11[4:0]: 11th conversion in regular sequence
Bits 19:15 SQ10[4:0]: 10th conversion in regular sequence
Bits 14:10 SQ9[4:0]: 9th conversion in regular sequence
Bits 9:5 SQ8[4:0]: 8th conversion in regular sequence
Bits 4:0 SQ7[4:0]: 7th conversion in regular sequence

RM0402 Rev 6 341/1163

347
Analog-to-digital converter (ADC) RM0402

13.12.11 ADC regular sequence register 3 (ADC_SQR3)

Address offset: 0x34
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. SQ6[4:0] SQ5[4:0] SQ4[4:1]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SQ4_0 SQ3[4:0] SQ2[4:0] SQ1[4:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:30 Reserved, must be kept at reset value.

Bits 29:25 SQ6[4:0]: 6th conversion in regular sequence
These bits are written by software with the channel number (0..18) assigned as the 6th in the
sequence to be converted.
Bits 24:20 SQ5[4:0]: 5th conversion in regular sequence
Bits 19:15 SQ4[4:0]: 4th conversion in regular sequence
Bits 14:10 SQ3[4:0]: 3rd conversion in regular sequence
Bits 9:5 SQ2[4:0]: 2nd conversion in regular sequence
Bits 4:0 SQ1[4:0]: 1st conversion in regular sequence

342/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

13.12.12 ADC injected sequence register (ADC_JSQR)

Address offset: 0x38
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. JL[1:0] JSQ4[4:1]
rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JSQ4[0] JSQ3[4:0] JSQ2[4:0] JSQ1[4:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:22 Reserved, must be kept at reset value.

Bits 21:20 JL[1:0]: Injected sequence length
These bits are written by software to define the total number of conversions in the injected
channel conversion sequence.
00: 1 conversion
01: 2 conversions
10: 3 conversions
11: 4 conversions
Bits 19:15 JSQ4[4:0]: 4th conversion in injected sequence (when JL[1:0]=3, see note below)
These bits are written by software with the channel number (0..18) assigned as the 4th in the
sequence to be converted.
Bits 14:10 JSQ3[4:0]: 3rd conversion in injected sequence (when JL[1:0]=3, see note below)
Bits 9:5 JSQ2[4:0]: 2nd conversion in injected sequence (when JL[1:0]=3, see note below)
Bits 4:0 JSQ1[4:0]: 1st conversion in injected sequence (when JL[1:0]=3, see note below)

Note: When JL[1:0]=3 (4 injected conversions in the sequencer), the ADC converts the channels
in the following order: JSQ1[4:0], JSQ2[4:0], JSQ3[4:0], and JSQ4[4:0].
When JL=2 (3 injected conversions in the sequencer), the ADC converts the channels in the
following order: JSQ2[4:0], JSQ3[4:0], and JSQ4[4:0].
When JL=1 (2 injected conversions in the sequencer), the ADC converts the channels in
starting from JSQ3[4:0], and then JSQ4[4:0].
When JL=0 (1 injected conversion in the sequencer), the ADC converts only JSQ4[4:0]
channel.

13.12.13 ADC injected data register x (ADC_JDRx) (x= 1..4)

Address offset: 0x3C - 0x48
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JDATA[15:0]
r r r r r r r r r r r r r r r r

RM0402 Rev 6 343/1163

347
Analog-to-digital converter (ADC) RM0402

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 JDATA[15:0]: Injected data
These bits are read-only. They contain the conversion result from injected channel x. The
data are left -or right-aligned as shown in Figure 63 and Figure 64.

13.12.14 ADC regular data register (ADC_DR)

Address offset: 0x4C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA[15:0]
r r r r r r r r r r r r r r r r

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 DATA[15:0]: Regular data
These bits are read-only. They contain the conversion result from the regular
channels. The data are left- or right-aligned as shown in Figure 63 and
Figure 64.

13.12.15 ADC Common status register (ADC_CSR)

Address offset: 0x00 (this offset address is relative to ADC1 base address + 0x300)
Reset value: 0x0000 0000
This register provides an image of the status bits of ADC1. Nevertheless it is read-only and
does not allow to clear the different status bits. Instead each status bit must be cleared by
writing it to 0 in the corresponding ADC_SR register.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. OVR1 STRT1 JSTRT1 JEOC 1 EOC1 AWD1
r r r r r r

Bits 31:6 Reserved, must be kept at reset value.

Bit 5 OVR1: Overrun flag of ADC1
This bit is a copy of the OVR bit in the ADC1_SR register.
Bit 4 STRT1: Regular channel Start flag of ADC1
This bit is a copy of the STRT bit in the ADC1_SR register.
Bit 3 JSTRT1: Injected channel Start flag of ADC1
This bit is a copy of the JSTRT bit in the ADC1_SR register.

344/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Bit 2 JEOC1: Injected channel end of conversion of ADC1

This bit is a copy of the JEOC bit in the ADC1_SR register.
Bit 1 EOC1: End of conversion of ADC1
This bit is a copy of the EOC bit in the ADC1_SR register.
Bit 0 AWD1: Analog watchdog flag of ADC1
This bit is a copy of the AWD bit in the ADC1_SR register.

13.12.16 ADC common control register (ADC_CCR)

Address offset: 0x04 (this offset address is relative to ADC1 base address + 0x300)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. TSVREFE VBATE Res. Res. Res. Res. ADCPRE
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

Bits 31:24 Reserved, must be kept at reset value.

Bit 23 TSVREFE: Temperature sensor and VREFINT enable
This bit is set and cleared by software to enable/disable the temperature sensor and the
VREFINT channel.
0: Temperature sensor and VREFINT channel disabled
1: Temperature sensor and VREFINT channel enabled
Note: VBATE must be disabled when TSVREFE is set. If both bits are set, only the VBAT
conversion is performed.
Bit 22 VBATE: VBAT enable
This bit is set and cleared by software to enable/disable the VBAT channel.
0: VBAT channel disabled
1: VBAT channel enabled
Bits 21:18 Reserved, must be kept at reset value.
Bits 17:16 ADCPRE: ADC prescaler
Set and cleared by software to select the frequency of the clock to the ADC. The clock is
common for all the ADCs.
Note: 00: PCLK2 divided by 2
01: PCLK2 divided by 4
10: PCLK2 divided by 6
11: PCLK2 divided by 8
Bits 15:0 Reserved, must be kept at reset value.

RM0402 Rev 6 345/1163

347
Analog-to-digital converter (ADC) RM0402

13.13 ADC register map

The following table summarizes the ADC registers.

Table 80. ADC global register map

Offset Register

0x000 - 0x04C ADC1

0x050 - 0x2FC Reserved
0x300 - 0x308 Common registers

Table 81. ADC register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
JSTRT
JEOC
STRT

AWD
OVR

EOC
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_SR
0x00
Reset value 0 0 0 0 0 0

AWD SGL
JDISCEN
JAWDEN
RES[1:0]

DISCEN
AWDEN

JEOCIE
JAUTO

AWDIE
OVRIE

EOCIE
SCAN
DISC
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
ADC_CR1 Res. AWDCH[4:0]
0x04 NUM [2:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
JEXTEN[1:0]
JSWSTART
EXTEN[1:0]
SWSTART

ALIGN

ADON
CONT
EOCS
JEXTSEL

DMA
DDS
Res.

Res.

Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
ADC_CR2 EXTSEL [3:0]
0x08 [3:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SMPR1 Sample time bits SMPx_x
0x0C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ADC_SMPR2 Sample time bits SMPx_x
0x10
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_JOFR1 JOFFSET1[11:0]
0x14
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_JOFR2 JOFFSET2[11:0]
0x18
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_JOFR3 JOFFSET3[11:0]
0x1C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_JOFR4 JOFFSET4[11:0]
0x20
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_HTR HT[11:0]
0x24
Reset value 1 1 1 1 1 1 1 1 1 1 1 1
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_LTR LT[11:0]
0x28
Reset value 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

ADC_SQR1 L[3:0] Regular channel sequence SQx_x bits

0x2C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.

ADC_SQR2 Regular channel sequence SQx_x bits

0x30
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

346/1163 RM0402 Rev 6

RM0402 Analog-to-digital converter (ADC)

Table 81. ADC register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
ADC_SQR3 Regular channel sequence SQx_x bits
0x34
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_JSQR JL[1:0] Injected channel sequence JSQx_x bits
0x38
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_JDR1 JDATA[15:0]
0x3C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_JDR2 JDATA[15:0]
0x40
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_JDR3 JDATA[15:0]
0x44
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_JDR4 JDATA[15:0]
0x48
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_DR Regular DATA[15:0]
0x4C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Table 82. ADC register map and reset values (common ADC registers)

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
JSTRT
JEOC
STRT

AWD
OVR

EOC
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_CSR
0x00
Reset value 0 0 0 0 0 0
ADCPRE[1:0]
TSVREFE.
VBATE
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
ADC_CCR
0x04

Reset value 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

RM0402 Rev 6 347/1163

347
Digital filter for sigma delta modulators (DFSDM) RM0402

14 Digital filter for sigma delta modulators (DFSDM)

14.1 Introduction
Digital filter for sigma delta modulators (DFSDM) is a high-performance module dedicated to
interface external Σ∆ modulators. It is featuring up to 4 external digital serial interfaces
(channels) and up to 2 digital filters with flexible Sigma Delta stream digital processing
options to offer up to 24-bit final ADC resolution. DFSDM also features optional parallel data
stream input from device memory.
An external Σ∆ modulator provides digital data stream of converted analog values from the
external Σ∆ modulator analog input. This digital data stream is sent into a DFSDM input
channel through a serial interface. DFSDM supports several standards to connect various
Σ∆ modulator outputs: SPI interface and Manchester coded 1-wire interface (both with
adjustable parameters). DFSDM module supports the connection of up to 4 multiplexed
input digital serial channels which are shared with up to 2 DFSDM modules. DFSDM
module also supports alternative parallel data inputs from up to 4 internal 16-bit data
channels (from device memory).
DFSDM is converting an input data stream into a final digital data word which represents an
analog input value on a Σ∆ modulator analog input. The conversion is based on a
configurable digital process: the digital filtering and decimation of the input serial data
stream.
The conversion speed and resolution are adjustable according to configurable parameters
for digital processing: filter type, filter order, length of filter, integrator length. The maximum
output data resolution is up to 24 bits. There are two conversion modes: single conversion
mode and continuous mode. The data can be automatically stored in a system RAM buffer
through DMA, thus reducing the software overhead.
A flexible timer triggering system can be used to control the start of conversion of DFSDM.
This timing control is capable of triggering simultaneous conversions or inserting a
programmable delay between conversions.
DFSDM features an analog watchdog function. Analog watchdog can be assigned to any of
the input channel data stream or to final output data. Analog watchdog has its own digital
filtering of input data stream to reach the required speed and resolution of watched data.
To detect short-circuit in control applications, there is a short-circuit detector. This block
watches each input channel data stream for occurrence of stable data for a defined time
duration (several 0’s or 1’s in an input data stream).
An extremes detector block watches final output data and stores maximum and minimum
values from the output data values. The extremes values stored can be restarted by
software.
Two power modes are supported: normal mode and stop mode.

348/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.2 DFSDM main features

• Up to 4 multiplexed input digital serial channels:
– configurable SPI interface to connect various Σ∆ modulators
– configurable Manchester coded 1 wire interface support
– clock output for Σ∆ modulator(s)
• Alternative inputs from up to 4 internal digital parallel channels:
– inputs with up to 16 bit resolution
– internal sources: memory (CPU/DMA write) data streams
• Adjustable digital signal processing:
– Sincx filter: filter order/type (1..5), oversampling ratio (up to 1..1024)
– integrator: oversampling ratio (1..256)
• Up to 24-bit output data resolution:
– right bit-shifter on final data (0..31 bits)
• Signed output data format
• Automatic data offset correction (offset stored in register by user)
• Continuous or single conversion
• Start-of-conversion synchronization with:
– software trigger
– internal timers
– external events
– start-of-conversion synchronously with first DFSDM filter (DFSDM_FLT0)
• Analog watchdog feature:
– low value and high value data threshold registers
– own configurable Sincx digital filter (order = 1..3, oversampling ratio = 1..32)
– input from output data register or from one or more input digital serial channels
– continuous monitoring independently from standard conversion
• Short-circuit detector to detect saturated analog input values (bottom and top ranges):
– up to 8-bit counter to detect 1..256 consecutive 0’s or 1’s on input data stream
– monitoring continuously each channel (4 serial channel transceiver outputs)
• Break generation on analog watchdog event or short-circuit detector event
• Extremes detector:
– store minimum and maximum values of output data values
– refreshed by software
• DMA may be used to read the conversion data
• Interrupts: end of conversion, overrun, analog watchdog, short-circuit, channel clock
absence
• “regular” or “injected” conversions:
– “regular” conversions can be requested at any time or even in continuous mode
without having any impact on the timing of “injected” conversions
– “injected” conversions for precise timing and with high conversion priority

RM0402 Rev 6 349/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

14.3 DFSDM implementation

This section describes the configuration implemented in DFSDMx.

Table 83. DFSDM1 implementation

DFSDM features DFSDM1

Number of channels 4
Number of filters 2
Input from internal ADC -
Supported trigger sources 11(1)
Pulses skipper -
ID registers support -
1. Refer to Table 86: DFSDM triggers connection for available trigger sources.

350/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.4 DFSDM functional description

14.4.1 DFSDM block diagram

Figure 67. Single DFSDM block diagram

APB bus

Sample 1 Sample 0 16
Parallel input data
register 0
Sample 1 Sample 0
16
Parallel input data
register 3

Channel multiplexer
EXTRG[1:0]

Data 0
Filter Oversampling Oversampling
Clock 0 order ratio ratio
16
CKOUT Clock Mode Sincx filter 0 Integrator unit 0
control control
DATIN0
CKIN0 Serial transceiver 0 Data 1
Filter Oversampling Oversampling
Clock Mode Clock 1 order ratio ratio
control Data in
control 16
DATIN3 Clock in Sincx filter 1 Integrator unit 1
CKIN3 Serial transceiver 3

Right bit-shift Right bit-shift

count count
4 watchdog filters
4 watchdog comparators Calibration data Calibration data
correction unit correction unit
Config
Status

Interrupt, DFSDM data 0 DFSDM data 1

1's, 0's counter break
threshold Interrupt,
Short circuit High threshold break Data output
1's, 0
detector 0's counter Filter 0 Low threshold
threshold config
Analog watchdog 0
Short circuit
detector 3
Interrupt,
High threshold break
Filter 1 Low threshold
config
Analog watchdog 1
APB bus

Control unit

Configuration Maximum value Maximum value

Interrupts and events:
registers Minimum value Minimum value
1) end of conversion
2) analog watchdog Extremes Extremes
DMA, interrupt, break 3) short circuit detection detector 0 detector 1
control, clock control 4) overrun
MSv40119V4

RM0402 Rev 6 351/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

1. This example shows 2 DFSDM filters and 4 input channels.

14.4.2 DFSDM pins and internal signals

Table 84. DFSDM external pins

Name Signal Type Remarks

VDD Power supply Digital power supply.

VSS Power supply Digital ground power supply.
CKIN[3:0] Clock input Clock signal provided from external Σ∆ modulator. FT input.
DATIN[3:0] Data input Data signal provided from external Σ∆ modulator. FT input.
Clock output to provide clock signal into external Σ∆
CKOUT Clock output
modulator.
External trigger Input trigger from two EXTI signals to start analog
EXTRG[1:0]
signal conversion (from GPIOs: EXTI11, EXTI15).

Table 85. DFSDM internal signals

Name Signal Type Remarks

Internal/
Input trigger from internal/external trigger sources to start
dfsdm_jtrg[10:0] external trigger
analog conversion, see Table 86 for details.
signal
break signal Break signals event generation from Analog watchdog or
dfsdm_break[3:0]
output short-circuit detector
DMA request DMA request signal from each DFSDM_FLTx (x=0..1):
dfsdm_dma[1:0]
signal end of injected conversion event.
Interrupt
dfsdm_it[1:0] Interrupt signal for each DFSDM_FLTx (x=0..1)
request signal

Table 86. DFSDM triggers connection

Trigger name Trigger source

dfsdm_jtrg0 TIM1_TRGO
dfsdm_jtrg1 TIM3_TRGO
dfsdm_jtrg2 TIM8_TRGO
dfsdm_jtrg3 TIM10_OC1
dfsdm_jtrg4 N/A
dfsdm_jtrg5 TIM4_TRGO
dfsdm_jtrg6 N/A
dfsdm_jtrg7 TIM6_TRGO
dfsdm_jtrg8 N/A
dfsdm_jtrg9 EXTI11
dfsdm_jtrg10 EXTI15

352/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Table 87. DFSDM break connection

Break name Break destination

dfsdm_break[0] TIM1 break

dfsdm_break[1] -
dfsdm_break[2] TIM8 break
dfsdm_break[3] -

14.4.3 DFSDM reset and clocks

DFSDM on-off control
The DFSDM interface is globally enabled by setting DFSDMEN=1 in the
DFSDM_CH0CFGR1 register. Once DFSDM is globally enabled, all input channels (y=0..3)
and digital filters DFSDM_FLTx (x=0..1) start to work if their enable bits are set (channel
enable bit CHEN in DFSDM_CHyCFGR1 and DFSDM_FLTx enable bit DFEN in
DFSDM_FLTxCR1).
Digital filter x DFSDM_FLTx (x=0..1) is enabled by setting DFEN=1 in the
DFSDM_FLTxCR1 register. Once DFSDM_FLTx is enabled (DFEN=1), both Sincx digital
filter unit and integrator unit are reinitialized.
By clearing DFEN, any conversion which may be in progress is immediately stopped and
DFSDM_FLTx is put into stop mode. All register settings remain unchanged except
DFSDM_FLTxAWSR and DFSDM_FLTxISR (which are reset).
Channel y (y=0..3) is enabled by setting CHEN=1 in the DFSDM_CHyCFGR1 register.
Once the channel is enabled, it receives serial data from the external Σ∆ modulator or
parallel internal data sources (CPU/DMA wire from memory).
DFSDM must be globally disabled (by DFSDMEN=0 in DFSDM_CH0CFGR1) before
stopping the system clock to enter in the STOP mode of the device.

DFSDM clocks
The internal DFSDM clock fDFSDMCLK, which is used to drive the channel transceivers,
digital processing blocks (digital filter, integrator) and next additional blocks (analog
watchdog, short-circuit detector, extremes detector, control block) is generated by the RCC
block and is derived from the system clock SYSCLK or peripheral clock PCLK2 (see
Section 6.3.24: RCC Dedicated Clocks Configuration Register (RCC_DCKCFGR)). The
DFSDM clock is automatically stopped in stop mode (if DFEN = 0 for all DFSDM_FLTx,
x=0..1).
The DFSDM serial channel transceivers can receive an external serial clock to sample an
external serial data stream. The internal DFSDM clock must be at least 4 times faster than
the external serial clock if standard SPI coding is used, and 6 times faster than the external
serial clock if Manchester coding is used.
DFSDM can provide one external output clock signal to drive external Σ∆ modulator(s) clock
input(s). It is provided on CKOUT pin. This output clock signal must be in the range
specified in given device datasheet and is derived from DFSDM clock or from audio clock
(see CKOUTSRC bit in DFSDM_CH0CFGR1 register) by programmable divider in the
range 2 - 256 (CKOUTDIV in DFSDM_CH0CFGR1 register). Audio clock source is SAI1

RM0402 Rev 6 353/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

clock selected by SAI1SEL[1:0] field in RCC configuration (see Section 6.3.24: RCC
Dedicated Clocks Configuration Register (RCC_DCKCFGR)).

14.4.4 Serial channel transceivers

There are 4 multiplexed serial data channels which can be selected for conversion by each
filter or Analog watchdog or Short-circuit detector. Those serial transceivers receive data
stream from external Σ∆ modulator. Data stream can be sent in SPI format or Manchester
coded format (see SITP[1:0] bits in DFSDM_CHyCFGR1 register).
The channel is enabled for operation by setting CHEN=1 in DFSDM_CHyCFGR1 register.

Channel inputs selection

Serial inputs (data and clock signals) from DATINy and CKINy pins can be redirected from
the following channel pins. This serial input channel redirection is set by CHINSEL bit in
DFSDM_CHyCFGR1 register.
Channel redirection can be used to collect audio data from PDM (pulse density modulation)
stereo microphone type. PDM stereo microphone has one data and one clock signal. Data
signal provides information for both left and right audio channel (rising clock edge samples
for left channel and falling clock edge samples for right channel).
Configuration of serial channels for PDM microphone input:
• PDM microphone signals (data, clock) will be connected to DFSDM input serial channel
y (DATINy, CKOUT) pins.
• Channel y will be configured: CHINSEL = 0 (input from given channel pins: DATINy,
CKINy).
• Channel (y-1) (modulo 4) will be configured: CHINSEL = 1 (input from the following
channel ((y-1)+1) pins: DATINy, CKINy).
• Channel y: SITP[1:0] = 0 (rising edge to strobe data) => left audio channel on channel
y.
• Channel (y-1): SITP[1:0] = 1 (falling edge to strobe data) => right audio channel on
channel y-1.
• Two DFSDM filters will be assigned to channel y and channel (y-1) (to filter left and
right channels from PDM microphone).

354/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Figure 68. Input channel pins redirection

(. . .)

CH(ymax)
Decode
DATIN(ymax)
CKIN(ymax)

. . .
. . .
. . .
FLT(xmax)
.
CHy .
Decode .
DATINy
CKINy FLT(x+1)

FLTx
CH(y-1) .
Decode .
DATIN(y-1)
.
CKIN(y-1)
FLT0
. . .
. . .
. . .

CH0
Decode
DATAIN0
CKIN0

(. . .)
CHINSEL

RCH

MSv41632V1

Output clock generation

A clock signal can be provided on CKOUT pin to drive external Σ∆ modulator clock inputs.
The frequency of this CKOUT signal is derived from DFSDM clock or from audio clock (see
CKOUTSRC bit in DFSDM_CH0CFGR1 register) divided by a predivider (see CKOUTDIV
bits in DFSDM_CH0CFGR1 register). If the output clock is stopped, then CKOUT signal is
set to low state (output clock can be stopped by CKOUTDIV=0 in DFSDM_CHyCFGR1
register or by DFSDMEN=0 in DFSDM_CH0CFGR1 register). The output clock stopping is
performed:
• 4 system clocks after DFSDMEN is cleared (if CKOUTSRC=0)
• 1 system clock and 3 audio clocks after DFSDMEN is cleared (if CKOUTSRC=1)
Before changing CKOUTSRC the software has to wait for CKOUT being stopped to avoid
glitch on CKOUT pin. The output clock signal frequency must be in the range 0 - 20 MHz.

RM0402 Rev 6 355/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

SPI data input format operation

In SPI format, the data stream is sent in serial format through data and clock signals. Data
signal is always provided from DATINy pin. A clock signal can be provided externally from
CKINy pin or internally from a signal derived from the CKOUT signal source.
In case of external clock source selection (SPICKSEL[1:0]=0) data signal (on DATINy pin) is
sampled on rising or falling clock edge (of CKINy pin) according SITP[1:0] bits setting (in
DFSDM_CHyCFGR1 register).
Internal clock sources - see SPICKSEL[1:0] in DFSDM_CHyCFGR1 register:
• CKOUT signal:
– For connection to external Σ∆ modulator which uses directly its clock input (from
CKOUT) to generate its output serial communication clock.
– Sampling point: on rising/falling edge according SITP[1:0] setting.
• CKOUT/2 signal (generated on CKOUT rising edge):
– For connection to external Σ∆ modulator which divides its clock input (from
CKOUT) by 2 to generate its output serial communication clock (and this output
clock change is active on each clock input rising edge).
– Sampling point: on each second CKOUT falling edge.
• CKOUT/2 signal (generated on CKOUT falling edge):
– For connection to external Σ∆ modulator which divides its clock input (from
CKOUT) by 2 to generate its output serial communication clock (and this output
clock change is active on each clock input falling edge).
– Sampling point: on each second CKOUT rising edge.
Note: An internal clock source can only be used when the external Σ∆ modulator uses CKOUT
signal as a clock input (to have synchronous clock and data operation).
Internal clock source usage can save CKINy pin connection (CKINy pins can be used for
other purpose).
The clock source signal frequency must be in the range 0 - 20 MHz for SPI coding and less
than fDFSDMCLK/4.

Manchester coded data input format operation

In Manchester coded format, the data stream is sent in serial format through DATINy pin
only. Decoded data and clock signal are recovered from serial stream after Manchester
decoding. There are two possible settings of Manchester codings (see SITP[1:0] bits in
DFSDM_CHyCFGR1 register):
• signal rising edge = log 0; signal falling edge = log 1
• signal rising edge = log 1; signal falling edge = log 0
The recovered clock signal frequency for Manchester coding must be in the range
0 - 10 MHz and less than fDFSDMCLK/6.
To correctly receive Manchester coded data, the CKOUTDIV divider (in
DFSDM_CH0CFGR1 register) must be set with respect to expected Manchester data rate
according formula:

( ( CKOUTDIV + 1 ) × T SYSCLK ) < T Manchester clock < ( 2 × CKOUTDIV × T SYSCLK )

356/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Figure 69. Channel transceiver timing diagrams

CKINy (SPICKSEL=0)
SPI timing : SPICKSEL = 0

twl twh tr tf
tsu th

SITP = 00
DATINy

tsu th

SITP = 01

SPICKSEL=3
CKOUT

SPICKSEL=2
SPI timing : SPICKSEL = 1, 2, 3

SPICKSEL=1

twl twh tr tf
tsu th

SITP = 0
DATINy

tsu th

SITP = 1

SITP = 2
DATINy
Manchester timing

SITP = 3

recovered clock

recovered data 0 0 1 1 0
MS30766V3

RM0402 Rev 6 357/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Clock absence detection

Channels serial clock inputs can be checked for clock absence/presence to ensure the
correct operation of conversion and error reporting. Clock absence detection can be
enabled or disabled on each input channel y by bit CKABEN in DFSDM_CHyCFGR1
register. If enabled, then this clock absence detection is performed continuously on a given
channel. A clock absence flag is set (CKABF[y] = 1) and an interrupt can be invoked (if
CKABIE=1) in case of an input clock error (see CKABF[3:0] in DFSDM_FLT0ISR register
and CKABEN in DFSDM_CHyCFGR1). After a clock absence flag clearing (by CLRCKABF
in DFSDM_FLT0ICR register), the clock absence flag is refreshed. Clock absence status bit
CKABF[y] is set also by hardware when corresponding channel y is disabled (if CHEN[y] = 0
then CKABF[y] is held in set state).
When a clock absence event has occurred, the data conversion (and/or analog watchdog
and short-circuit detector) provides incorrect data. The user should manage this event and
discard given data while a clock absence is reported.
The clock absence feature is available only when the system clock is used for the CKOUT
signal (CKOUTSRC=0 in DFSDM_CH0CFGR1 register).
When the transceiver is not yet synchronized, the clock absence flag is set and cannot be
cleared by CLRCKABF[y] bit (in DFSDM_FLT0ICR register). The software sequence
concerning clock absence detection feature should be:
• Enable given channel by CHEN = 1
• Try to clear the clock absence flag (by CLRCKABF = 1) until the clock absence flag is
really cleared (CKABF = 0). At this time, the transceiver is synchronized (signal clock is
valid) and is able to receive data.
• Enable the clock absence feature CKABEN = 1 and the associated interrupt CKABIE =
1 to detect if the SPI clock is lost or Manchester data edges are missing.
If SPI data format is used, then the clock absence detection is based on the comparison of
an external input clock with an output clock generation (CKOUT signal). The external input
clock signal into the input channel must be changed at least once per 8 signal periods of
CKOUT signal (which is controlled by CKOUTDIV field in DFSDM_CH0CFGR1 register).

Figure 70. Clock absence timing diagram for SPI

max. 8 periods

CKOUT 2 0 1 2 3 4 5 6 7 0
SPI clock presence

restart counting
CKINy
timing

last clock change

CKABF[y]

error reported
MS30767V2

If Manchester data format is used, then the clock absence means that the clock recovery is
unable to perform from Manchester coded signal. For a correct clock recovery, it is first
necessary to receive data with 1 to 0 or 0 to 1 transition (see Figure 72 for Manchester
synchronization).

358/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

The detection of a clock absence in Manchester coding (after a first successful

synchronization) is based on changes comparison of coded serial data input signal with
output clock generation (CKOUT signal). There must be a voltage level change on DATINy
pin during 2 periods of CKOUT signal (which is controlled by CKOUTDIV bits in
DFSDM_CH0CFGR1 register). This condition also defines the minimum data rate to be able
to correctly recover the Manchester coded data and clock signals.
The maximum data rate of Manchester coded data must be less than the CKOUT signal.
So to correctly receive Manchester coded data, the CKOUTDIV divider must be set
according the formula:

( ( CKOUTDIV + 1 ) × T SYSCLK ) < T Manchester clock < ( 2 × CKOUTDIV × T SYSCLK )

A clock absence flag is set (CKABF[y] = 1) and an interrupt can be invoked (if CKABIE=1) in
case of an input clock recovery error (see CKABF[3:0] in DFSDM_FLT0ISR register and
CKABEN in DFSDM_CHyCFGR1). After a clock absence flag clearing (by CLRCKABF in
DFSDM_FLT0ICR register), the clock absence flag is refreshed.

Figure 71. Clock absence timing diagram for Manchester coding

max. 2 periods

CKOUT 0 0 0 1 0

restart counting

SITP = 2
Manchester clock presence

DATINy

last data change

timing

SITP = 3

recovered clock

recovered data 0 0 1 ? ?

CKABF[y]

error reported
MS30768V2

RM0402 Rev 6 359/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Manchester/SPI code synchronization

The Manchester coded stream must be synchronized the first time after enabling the
channel (CHEN=1 in DFSDM_CHyCFGR1 register). The synchronization ends when a data
transition from 0 to 1 or from 1 to 0 (to be able to detect valid data edge) is received. The
end of the synchronization can be checked by polling CKABF[y]=0 for a given channel after
it has been cleared by CLRCKABF[y] in DFSDM_FLT0ICR, following the software sequence
detailed hereafter:
CKABF[y] flag is cleared by setting CLRCKABF[y] bit. If channel y is not yet synchronized
the hardware immediately set the CKABF[y] flag. Software is then reading back the
CKABF[y] flag and if it is set then perform again clearing of this flag by setting
CLRCKABF[y] bit. This software sequence (polling of CKABF[y] flag) continues until
CKABF[y] flag is set (signalizing that Manchester stream is synchronized). To be able to
synchronize/receive Manchester coded data the CKOUTDIV divider (in
DFSDM_CH0CFGR1 register) must be set with respect to expected Manchester data rate
according the formula below.

( ( CKOUTDIV + 1 ) × T SYSCLK ) < T Manchester clock < ( 2 × CKOUTDIV × T SYSCLK )

SPI coded stream is synchronized after first detection of clock input signal (valid
rising/falling edge).
Note: When the transceiver is not yet synchronized, the clock absence flag is set and cannot be
cleared by CLRCKABF[y] bit (in DFSDM_FLT0ICR register).

360/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Figure 72. First conversion for Manchester coding (Manchester synchronization)

SITP = 2
DATINy
Manchester timing

SITP = 3

recovered clock

data from
modulator 0 0 1 1 0

CHEN
real start of first conversion
first conversion
start trigger first data bit toggle - end of Manchester synchronization

recovered data ? ? 1 1 0

CKABF[y]

clearing of CKABF[y] flag by software polling

MS30769V2

External serial clock frequency measurement

The measuring of a channel serial clock input frequency provides a real data rate from an
external Σ∆ modulator, which is important for application purposes.
An external serial clock input frequency can be measured by a timer counting DFSDM
clocks (fDFSDMCLK) during one conversion duration. The counting starts at the first input data
clock after a conversion trigger (regular or injected) and finishes by last input data clock
before conversion ends (end of conversion flag is set). Each conversion duration (time
between first serial sample and last serial sample) is updated in counter CNVCNT[27:0] in
register DFSDM_FLTxCNVTIMR when the conversion finishes (JEOCF=1 or REOCF=1).
The user can then compute the data rate according to the digital filter settings (FORD,
FOSR, IOSR, FAST). The external serial frequency measurement is stopped only if the filter
is bypassed (FOSR=0, only integrator is active, CNVCNT[27:0]=0 in
DFSDM_FLTxCNVTIMR register).
In case of parallel data input (Section 14.4.6: Parallel data inputs) the measured frequency
is the average input data rate during one conversion.

RM0402 Rev 6 361/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Note: When conversion is interrupted (e.g. by disabling/enabling the selected channel) the
interruption time is also counted in CNVCNT[27:0]. Therefore it is recommended to not
interrupt the conversion for correct conversion duration result.
Conversion times:
injected conversion or regular conversion with FAST = 0 (or first conversion if
FAST=1):
for Sincx filters (x=1..5):
t = CNVCNT/fDFSDMCLK = [FOSR * (IOSR-1 + FORD) + FORD] / fCKIN
for FastSinc filter:
t = CNVCNT/fDFSDMCLK = [FOSR * (IOSR-1 + 4) + 2] / fCKIN
regular conversion with FAST = 1 (except first conversion):
for Sincx and FastSinc filters:
t = CNVCNT/fDFSDMCLK = [FOSR * IOSR] / fCKIN
in case if FOSR = FOSR[9:0]+1 = 1 (filter bypassed, active only integrator):
t = IOSR / fCKIN (... but CNVCNT=0)
where:
• fCKIN is the channel input clock frequency (on given channel CKINy pin) or input data
rate (in case of parallel data input)
• FOSR is the filter oversampling ratio: FOSR = FOSR[9:0]+1 (see DFSDM_FLTxFCR
register)
• IOSR is the integrator oversampling ratio: IOSR = IOSR[7:0]+1 (see DFSDM_FLTxFCR
register)
• FORD is the filter order: FORD = FORD[2:0] (see DFSDM_FLTxFCR register)

Channel offset setting

Each channel has its own offset setting (in register) which is finally subtracted from each
conversion result (injected or regular) from a given channel. Offset correction is performed
after the data right bit shift. The offset is stored as a 24-bit signed value in OFFSET[23:0]
field in DFSDM_CHyCFGR2 register.

Data right bit shift

To have the result aligned to a 24-bit value, each channel defines a number of right bit shifts
which will be applied on each conversion result (injected or regular) from a given channel.
The data bit shift number is stored in DTRBS[4:0] bits in DFSDM_CHyCFGR2 register.
The right bit-shift is rounding the result to nearest integer value. The sign of shifted result is
maintained, in order to have valid 24-bit signed format of result data.

362/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.4.5 Configuring the input serial interface

The following parameters must be configured for the input serial interface:
• Output clock predivider. There is a programmable predivider to generate the output
clock from DFSDM clock (2 - 256). It is defined by CKOUTDIV[7:0] bits in
DFSDM_CH0CFGR1 register.
• Serial interface type and input clock phase. Selection of SPI or Manchester coding
and sampling edge of input clock. It is defined by SITP [1:0] bits in
DFSDM_CHyCFGR1 register.
• Input clock source. External source from CKINy pin or internal from CKOUT pin. It is
defined by SPICKSEL[1:0] field in DFSDM_CHyCFGR1 register.
• Final data right bit-shift. Defines the final data right bit shift to have the result aligned
to a 24-bit value. It is defined by DTRBS[4:0] in DFSDM_CHyCFGR2 register.
• Channel offset per channel. Defines the analog offset of a given serial channel (offset
of connected external Σ∆ modulator). It is defined by OFFSET[23:0] bits in
DFSDM_CHyCFGR2 register.
• short-circuit detector and clock absence per channel enable. To enable or disable
the short-circuit detector (by SCDEN bit) and the clock absence monitoring (by
CKABEN bit) on a given serial channel in register DFSDM_CHyCFGR1.
• Analog watchdog filter and short-circuit detector threshold settings. To configure
channel analog watchdog filter parameters and channel short-circuit detector
parameters. Configurations are defined in DFSDM_CHyAWSCDR register.

14.4.6 Parallel data inputs

Each input channel provides a register for 16-bit parallel data input (besides serial data
input). Each 16-bit parallel input can be sourced from internal data sources only:
• direct CPU/DMA writing.
The selection for using serial or parallel data input for a given channel is done by field
DATMPX[1:0] of DFSDM_CHyCFGR1 register. In DATMPX[1:0] is also defined the parallel
data source: direct write by CPU/DMA.
Each channel contains a 32-bit data input register DFSDM_CHyDATINR in which it can be
written a 16-bit data. Data are in 16-bit signed format. Those data can be used as input to
the digital filter which is accepting 16-bit parallel data.
If serial data input is selected (DATMPX[1:0] = 0), the DFSDM_CHyDATINR register is write
protected.

Input from memory (direct CPU/DMA write)

The direct data write into DFSDM_CHyDATINR register by CPU or DMA (DATMPX[1:0]=2)
can be used as data input in order to process digital data streams from memory or
peripherals.
Data can be written by CPU or DMA into DFSDM_CHyDATINR register:
1. CPU data write:
Input data are written directly by CPU into DFSDM_CHyDATINR register.
2. DMA data write:
The DMA should be configured in memory-to-memory transfer mode to transfer data
from memory buffer into DFSDM_CHyDATINR register. The destination memory

RM0402 Rev 6 363/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

address is the address of DFSDM_CHyDATINR register. Data are transferred at DMA

transfer speed from memory to DFSDM parallel input.
This DMA transfer is different from DMA used to read DFSDM conversion results. Both
DMA can be used at the same time - first DMA (configured as memory-to-memory
transfer) for input data writings and second DMA (configured as peripheral-to-memory
transfer) for data results reading.
The accesses to DFSDM_CHyDATINR can be either 16-bit or 32-bit wide, allowing to load
respectively one or two samples in one write operation. 32-bit input data register
(DFSDM_CHyDATINR) can be filled with one or two 16-bit data samples, depending on the
data packing operation mode defined in field DATPACK[1:0] of DFSDM_CHyCFGR1
register:
1. Standard mode (DATPACK[1:0]=0):
Only one sample is stored in field INDAT0[15:0] of DFSDM_CHyDATINR register which
is used as input data for channel y. The upper 16 bits (INDAT1[15:0]) are ignored and
write protected. The digital filter must perform one input sampling (from INDAT0[15:0])
to empty data register after it has been filled by CPU/DMA. This mode is used together
with 16-bit CPU/DMA access to DFSDM_CHyDATINR register to load one sample per
write operation.
2. Interleaved mode (DATPACK[1:0]=1):
DFSDM_CHyDATINR register is used as a two sample buffer. The first sample is
stored in INDAT0[15:0] and the second sample is stored in INDAT1[15:0]. The digital
filter must perform two input samplings from channel y to empty DFSDM_CHyDATINR
register. This mode is used together with 32-bit CPU/DMA access to
DFSDM_CHyDATINR register to load two samples per write operation.
3. Dual mode (DATPACK[1:0]=2):
Two samples are written into DFSDM_CHyDATINR register. The data INDAT0[15:0] is
for channel y, the data in INDAT1[15:0] is for channel y+1. The data in INDAT1[15:0] is
automatically copied INDAT0[15:0] of the following (y+1) channel data register
DFSDM_CH[y+1]DATINR). The digital filters must perform two samplings - one from
channel y and one from channel (y+1) - in order to empty DFSDM_CHyDATINR
registers.
Dual mode setting (DATPACK[1:0]=2) is available only on even channel numbers (y =
0, 2). If odd channel (y = 1, 3) is set to Dual mode then both INDAT0[15:0] and
INDAT1[15:0] parts are write protected for this channel. If even channel is set to Dual
mode then the following odd channel must be set into Standard mode
(DATPACK[1:0]=0) for correct cooperation with even channels.
See Figure 73 for DFSDM_CHyDATINR registers data modes and assignments of data
samples to channels.

Figure 73. DFSDM_CHyDATINR registers operation modes and assignment

Standard mode Interleaved mode Dual mode

31 16 15 0 31 16 15 0 31 16 15 0
Unused Ch0 (sample 0) Ch0 (sample 1) Ch0 (sample 0) Ch1 (sample 0) Ch0 (sample 0) y=0

Unused Ch1 (sample 0) Ch1 (sample 1) Ch1 (sample 0) Unused Ch1 (sample 0) y=1

Unused Ch2 (sample 0) Ch2 (sample 1) Ch2 (sample 0) Ch3 (sample 0) Ch2 (sample 0) y=2

Unused Ch3 (sample 0) Ch3 (sample 1) Ch3 (sample 0) Unused Ch3 (sample 0) y=3
MSv40123V1

364/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

The write into DFSDM_CHyDATINR register to load one or two samples must be performed
after the selected input channel (channel y) is enabled for data collection (starting
conversion for channel y). Otherwise written data are lost for next processing.
For example: for single conversion and interleaved mode, do not start writing pair of data
samples into DFSDM_CHyDATINR before the single conversion is started (any data
present in the DFSDM_CHyDATINR before starting a conversion is discarded).

14.4.7 Channel selection

There are 4 multiplexed channels which can be selected for conversion using the injected
channel group and/or using the regular channel.
The injected channel group is a selection of any or all of the 4 channels. JCHG[3:0] in the
DFSDM_FLTxJCHGR register selects the channels of the injected group, where JCHG[y]=1
means that channel y is selected.
Injected conversions can operate in scan mode (JSCAN=1) or single mode (JSCAN=0). In
scan mode, each of the selected channels is converted, one after another. The lowest
channel (channel 0, if selected) is converted first, followed immediately by the next higher
channel until all the channels selected by JCHG[3:0] have been converted. In single mode
(JSCAN=0), only one channel from the selected channels is converted, and the channel
selection is moved to the next channel. Writing to JCHG[3:0] if JSCAN=0 resets the channel
selection to the lowest selected channel.
Injected conversions can be launched by software or by a trigger. They are never
interrupted by regular conversions.
The regular channel is a selection of just one of the 4 channels. RCH[1:0] in the
DFSDM_FLTxCR1 register indicates the selected channel.
Regular conversions can be launched only by software (not by a trigger). A sequence of
continuous regular conversions is temporarily interrupted when an injected conversion is
requested.
Performing a conversion on a disabled channel (CHEN=0 in DFSDM_CHyCFGR1 register)
causes that the conversion will never end - because no input data is provided (with no clock
signal). In this case, it is necessary to enable a given channel (CHEN=1 in
DFSDM_CHyCFGR1 register) or to stop the conversion by DFEN=0 in DFSDM_FLTxCR1
register.

14.4.8 Digital filter configuration

DFSDM contains a Sincx type digital filter implementation. This Sincx filter performs an input
digital data stream filtering, which results in decreasing the output data rate (decimation)
and increasing the output data resolution. The Sincx digital filter is configurable in order to

RM0402 Rev 6 365/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

reach the required output data rates and required output data resolution. The configurable
parameters are:
• Filter order/type: (see FORD[2:0] bits in DFSDM_FLTxFCR register):
– FastSinc
– Sinc1
– Sinc2
– Sinc3
– Sinc4
– Sinc5
• Filter oversampling/decimation ratio (see FOSR[9:0] bits in DFSDM_FLTxFCR
register):
– FOSR = 1-1024 - for FastSinc filter and Sincx filter x = FORD = 1..3
– FOSR = 1-215 - for Sincx filter x = FORD = 4
– FOSR = 1-73 - for Sincx filter x = FORD = 5
The filter has the following transfer function (impulse response in H domain):
x
 1 – z – FOSR
• Sincx filter type: H ( z ) =  ----------------------------
-
 1 – z–1 

2
 1 – z – FOSR
• FastSinc filter type: - ⋅ ( 1 + z – ( 2 ⋅ FOSR ) )
H ( z ) =  ----------------------------
 1 – z–1 

Figure 74. Example: Sinc3 filter response

Gain (dB)

Normalized frequency (fIN/fDATA )

MS30770V1

366/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Table 88. Filter maximum output resolution (peak data values from filter output)
for some FOSR values
FOSR Sinc1 Sinc2 FastSinc Sinc3 Sinc4 Sinc5

x +/- x +/- x2 +/- 2x2 +/- x3 +/- x4 +/- x5

4 +/- 4 +/- 16 +/- 32 +/- 64 +/- 256 +/- 1024
8 +/- 8 +/- 64 +/- 128 +/- 512 +/- 4096 -
32 +/- 32 +/- 1024 +/- 2048 +/- 32768 +/- 1048576 +/- 33554432
64 +/- 64 +/- 4096 +/- 8192 +/- 262144 +/- 16777216 +/- 1073741824
128 +/- 128 +/- 16384 +/- 32768 +/- 2097152 +/- 268435456
256 +/- 256 +/- 65536 +/- 131072 +/- 16777216 Result can overflow on full scale
1024 +/- 1024 +/- 1048576 +/- 2097152 +/- 1073741824 input (> 32-bit signed integer)

For more information about Sinc filter type properties and usage, it is recommended to study
the theory about digital filters (more resources can be downloaded from internet).

14.4.9 Integrator unit

The integrator performs additional decimation and a resolution increase of data coming from
the digital filter. The integrator simply performs the sum of data from a digital filter for a given
number of data samples from a filter.
The integrator oversampling ratio parameter defines how many data counts will be summed
to one data output from the integrator. IOSR can be set in the range 1-256 (see IOSR[7:0]
bits description in DFSDM_FLTxFCR register).

Table 89. Integrator maximum output resolution (peak data values from integrator
output) for some IOSR values and FOSR = 256 and Sinc3 filter type (largest data)
IOSR Sinc1 Sinc2 FastSinc Sinc3 Sinc4 Sinc5

x +/- FOSR. x +/- FOSR2. x +/- 2.FOSR2. x +/- FOSR3. x +/- FOSR4. x +/- FOSR5. x
4 - - - +/- 67 108 864 - -
32 - - - +/- 536 870 912 - -
+/- 2 147 483
128 - - - - -
648
256 - - - +/- 232 - -

14.4.10 Analog watchdog

The analog watchdog purpose is to trigger an external signal (break or interrupt) when an
analog signal reaches or crosses given maximum and minimum threshold values. An
interrupt/event/break generation can then be invoked.
Each analog watchdog will supervise serial data receiver outputs (after the analog watchdog
filter on each channel) or data output register (current injected or regular conversion result)
according to AWFSEL bit setting (in DFSDM_FLTxCR1 register). The input channels to be
monitored or not by the analog watchdog x will be selected by AWDCH[3:0] in
DFSDM_FLTxCR2 register.

RM0402 Rev 6 367/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Analog watchdog conversions on input channels are independent from standard

conversions. In this case, the analog watchdog uses its own filters and signal processing on
each input channel independently from the main injected or regular conversions. Analog
watchdog conversions are performed in a continuous mode on the selected input channels
in order to watch channels also when main injected or regular conversions are paused
(RCIP = 0, JCIP = 0).
There are high and low threshold registers which are compared with given data values (set
by AWHT[23:0] bits in DFSDM_FLTxAWHTR register and by AWLT[23:0] bits in
DFSDM_FLTxAWLTR register).
There are 2 options for comparing the threshold registers with the data values
• Option1: in this case, the input data are taken from final output data register
(AWFSEL=0). This option is characterized by:
– high input data resolution (up to 24-bits)
– slow response time - inappropriate for fast response applications like overcurrent
detection
– for the comparison the final data are taken after bit shifting and offset data
correction
– final data are available only after main regular or injected conversions are
performed
– can be used in case of parallel input data source (DATMPX[1:0] ≠ 0 in
DFSDM_CHyCFGR1 register)
• Option2: in this case, the input data are taken from any serial data receivers output
(AWFSEL=1). This option is characterized by:
– input serial data are processed by dedicated analog watchdog Sincx channel
filters with configurable oversampling ratio (1..32) and filter order (1..3) (see
AWFOSR[4:0] and AWFORD[1:0] bits setting in DFSDM_CHyAWSCDR register)
– lower resolution (up to 16-bit)
– fast response time - appropriate for applications which require a fast response like
overcurrent/overvoltage detection)
– data are available in continuous mode independently from main regular or injected
conversions activity
In case of input channels monitoring (AWFSEL=1), the data for comparison to threshold is
taken from channels selected by AWDCH[3:0] field (DFSDM_FLTxCR2 register). Each of
the selected channels filter result is compared to one threshold value pair (AWHT[23:0] /
AWLT[23:0]). In this case, only higher 16 bits (AWHT[23:8] / AWLT[23:8]) define the 16-bit
threshold compared with the analog watchdog filter output because data coming from the
analog watchdog filter is up to a 16-bit resolution. Bits AWHT[7:0] / AWLT[7:0] are not taken
into comparison in this case (AWFSEL=1).
Parameters of the analog watchdog filter configuration for each input channel are set in
DFSDM_CHyAWSCDR register (filter order AWFORD[1:0] and filter oversampling ratio
AWFOSR[4:0]).
Each input channel has its own comparator which compares the analog watchdog data
(from analog watchdog filter) with analog watchdog threshold values (AWHT/AWLT). When
several channels are selected (field AWDCH[3:0] field of DFSDM_FLTxCR2 register),
several comparison requests may be received simultaneously. In this case, the channel
request with the lowest number is managed first and then continuing to higher selected
channels. For each channel, the result can be recorded in a separate flag (fields

368/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

AWHTF[3:0], AWLTF[3:0] of DFSDM_FLTxAWSR register). Each channel request is

executed in 8 DFSDM clock cycles. So, the bandwidth from each channel is limited to 8
DFSDM clock cycles (if AWDCH[3:0] = 0x0F). Because the maximum input channel
sampling clock frequency is the DFSDM clock frequency divided by 4, the configuration
AWFOSR = 0 (analog watchdog filter is bypassed) cannot be used for analog watchdog
feature at this input clock speed. Therefore user must properly configure the number of
watched channels and analog watchdog filter parameters with respect to input sampling
clock speed and DFSDM frequency.
Analog watchdog filter data for given channel y is available for reading by firmware on field
WDATA[15:0] in DFSDM_CHyWDATR register. That analog watchdog filter data is
converted continuously (if CHEN=1 in DFSDM_CHyCFGR1 register) with the data rate
given by the analog watchdog filter setting and the channel input clock frequency.
The analog watchdog filter conversion works like a regular Fast Continuous Conversion
without the intergator. The number of serial samples needed for one result from analog
watchdog filter output (at channel input clock frequency fCKIN):
first conversion:
for Sincx filters (x=1..5): number of samples = [FOSR * FORD + FORD + 1]
for FastSinc filter: number of samples = [FOSR * 4 + 2 + 1]
next conversions:
for Sincx and FastSinc filters: number of samples = [FOSR * IOSR]
where:
FOSR ....... filter oversampling ratio: FOSR = AWFOSR[4:0]+1 (see DFSDM_CHyAWSCDR
register)
FORD ....... the filter order: FORD = AWFORD[1:0] (see DFSDM_CHyAWSCDR register)
In case of output data register monitoring (AWFSEL=0), the comparison is done after a right
bit shift and an offset correction of final data (see OFFSET[23:0] and DTRBS[4:0] fields in
DFSDM_CHyCFGR2 register). A comparison is performed after each injected or regular
end of conversion for the channels selected by AWDCH[3:0] field (in DFSDM_FLTxCR2
register).
The status of an analog watchdog event is signalized in DFSDM_FLTxAWSR register where
a given event is latched. AWHTF[y]=1 flag signalizes crossing AWHT[23:0] value on
channel y. AWLTF[y]=1 flag signalizes crossing AWLT[23:0] value on channel y. Latched
events in DFSDM_FLTxAWSR register are cleared by writing ‘1’ into the corresponding
clearing bit CLRAWHTF[y] or CLRAWLTF[y] in DFSDM_FLTxAWCFR register.
The global status of an analog watchdog is signalized by the AWDF flag bit in
DFSDM_FLTxISR register (it is used for the fast detection of an interrupt source). AWDF=1
signalizes that at least one watchdog occurred (AWHTF[y]=1 or AWLTF[y]=1 for at least one
channel). AWDF bit is cleared when all AWHTF[3:0] and AWLTF[3:0] are cleared.
An analog watchdog event can be assigned to break output signal. There are four break
outputs to be assigned to a high or low threshold crossing event (dfsdm_break[3:0]). The
break signal assignment to a given analog watchdog event is done by BKAWH[3:0] and
BKAWL[3:0] fields in DFSDM_FLTxAWHTR and DFSDM_FLTxAWLTR register.

RM0402 Rev 6 369/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

14.4.11 Short-circuit detector

The purpose of a short-circuit detector is to signalize with a very fast response time if an
analog signal reached saturated values (out of full scale ranges) and remained on this value
given time. This behavior can detect short-circuit or open circuit errors (e.g. overcurrent or
overvoltage). An interrupt/event/break generation can be invoked.
Input data into a short-circuit detector is taken from channel transceiver outputs.
There is an upcounting counter on each input channel which is counting consecutive 0’s or
1’s on serial data receiver outputs. A counter is restarted if there is a change in the data
stream received - 1 to 0 or 0 to 1 change of data signal. If this counter reaches a short-circuit
threshold register value (SCDT[7:0] bits in DFSDM_CHyAWSCDR register), then a short-
circuit event is invoked. Each input channel has its short-circuit detector. Any channel can
be selected to be continuously monitored by setting the SCDEN bit (in DFSDM_CHyCFGR1
register) and it has its own short-circuit detector settings (threshold value in SCDT[7:0] bits,
status bit SCDF[3:0], status clearing bits CLRSCDF[3:0]). Status flag SCDF[y] is cleared
also by hardware when corresponding channel y is disabled (CHEN[y] = 0).
On each channel, a short-circuit detector event can be assigned to break output signal
dfsdm_break[3:0]. There are four break outputs to be assigned to a short-circuit detector
event. The break signal assignment to a given channel short-circuit detector event is done
by BKSCD[3:0] field in DFSDM_CHyAWSCDR register.
Short circuit detector cannot be used in case of parallel input data channel selection
(DATMPX[1:0] ≠ 0 in DFSDM_CHyCFGR1 register).
Four break outputs are totally available (shared with the analog watchdog function).

14.4.12 Extreme detector

The purpose of an extremes detector is to collect the minimum and maximum values of final
output data words (peak to peak values).
If the output data word is higher than the value stored in the extremes detector maximum
register (EXMAX[23:0] bits in DFSDM_FLTxEXMAX register), then this register is updated
with the current output data word value and the channel from which the data is stored is in
EXMAXCH[1:0] bits (in DFSDM_FLTxEXMAX register) .
If the output data word is lower than the value stored in the extremes detector minimum
register (EXMIN[23:0] bits in DFSDM_FLTxEXMIN register), then this register is updated
with the current output data word value and the channel from which the data is stored is in
EXMINCH[1:0] bits (in DFSDM_FLTxEXMIN register).
The minimum and maximum register values can be refreshed by software (by reading given
DFSDM_FLTxEXMAX or DFSDM_FLTxEXMIN register). After refresh, the extremes
detector minimum data register DFSDM_FLTxEXMIN is filled with 0x7FFFFF (maximum
positive value) and the extremes detector maximum register DFSDM_FLTxEXMAX is filled
with 0x800000 (minimum negative value).
The extremes detector performs a comparison after a right bit shift and an offset data
correction. For each extremes detector, the input channels to be considered into computing
the extremes value are selected in EXCH[3:0] bits (in DFSDM_FLTxCR2 register).

370/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.4.13 Data unit block

The data unit block is the last block of the whole processing path: External Σ∆ modulators -
Serial transceivers - Sinc filter - Integrator - Data unit block.
The output data rate depends on the serial data stream rate, and filter and integrator
settings. The maximum output data rate is:
f CKIN
Datarate samples ⁄ s = ------------------------------------------------------------------------------------------------------
- ...FAST = 0, Sincx filter
F OSR ⋅ ( I OSR – 1 + F ORD ) + ( F ORD + 1 )

f CKIN
Datarate samples ⁄ s = -------------------------------------------------------------------------------
- ...FAST = 0, FastSinc filter
F OSR ⋅ ( I OSR – 1 + 4 ) + ( 2 + 1 )

or
f CKIN
Datarate samples ⁄ s = ------------------------------- ...FAST = 1
F OSR ⋅ I OSR

Maximum output data rate in case of parallel data input:

f DATAIN_RATE
Datarate samples ⁄ s = ------------------------------------------------------------------------------------------------------
- ...FAST = 0, Sincx filter
F OSR ⋅ ( I OSR – 1 + F ORD ) + ( F ORD + 1 )

or
f DATAIN_RATE
Datarate samples ⁄ s = -------------------------------------------------------------------------------
- ...FAST = 0, FastSinc filter
F OSR ⋅ ( I OSR – 1 + 4 ) + ( 2 + 1 )

or
f DATAIN_RATE
Datarate samples ⁄ s = -----------------------------------
- ...FAST=1 or any filter bypass case ( F OSR = 1 )
F OSR ⋅ I OSR

where: f DATAIN_RATE ...input data rate from CPU/DMA

The right bit-shift of final data is performed in this module because the final data width is 24-
bit and data coming from the processing path can be up to 32 bits. This right bit-shift is
configurable in the range 0-31 bits for each selected input channel (see DTRBS[4:0] bits in
DFSDM_CHyCFGR2 register). The right bit-shift is rounding the result to nearest integer
value. The sign of shifted result is maintained - to have valid 24-bit signed format of result
data.
In the next step, an offset correction of the result is performed. The offset correction value
(OFFSET[23:0] stored in register DFSDM_CHyCFGR2) is subtracted from the output data
for a given channel. Data in the OFFSET[23:0] field is set by software by the appropriate
calibration routine.
Due to the fact that all operations in digital processing are performed on 32-bit signed
registers, the following conditions must be fulfilled not to overflow the result:
FOSR FORD . IOSR <= 231 ... for Sincx filters, x = 1..5)
2 . FOSR 2 . IOSR <= 231 ... for FastSinc filter)

RM0402 Rev 6 371/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Note: In case of filter and integrator bypass (IOSR[7:0]=0, FOSR[9:0]=0), the input data rate
(fDATAIN_RATE) must be limited to be able to read all output data:
fDATAIN_RATE ≤ fAPB
where fAPB is the bus frequency to which the DFSDM peripheral is connected.

14.4.14 Signed data format

Each DFSDM input serial channel can be connected to one external Σ∆ modulator. An
external Σ∆ modulator can have 2 differential inputs (positive and negative) which can be
used for a differential or single-ended signal measurement.
A Σ∆ modulator output is always assumed in a signed format (a data stream of zeros and
ones from a Σ∆ modulator represents values -1 and +1).
Signed data format in registers: Data is in a signed format in registers for final output data,
analog watchdog, extremes detector, offset correction. The msb of output data word
represents the sign of value (two’s complement format).

14.4.15 Launching conversions

Injected conversions can be launched using the following methods:
• Software: writing ‘1’ to JSWSTART in the DFSDM_FLTxCR1 register.
• Trigger: JEXTSEL[2:0] selects the trigger signal while JEXTEN activates the trigger
and selects the active edge at the same time (see the DFSDM_FLTxCR1 register).
• Synchronous with DFSDM_FLT0 if JSYNC=1: for DFSDM_FLTx (x>0), an injected
conversion is automatically launched when in DFSDM_FLT0; the injected conversion is
started by software (JSWSTART=1 in DFSDM_FLT0CR2 register). Each injected
conversion in DFSDM_FLTx (x>0) is always executed according to its local
configuration settings (JSCAN, JCHG, etc.).
If the scan conversion is enabled (bit JSCAN=1) then, each time an injected conversion is
triggered, all of the selected channels in the injected group (JCHG[3:0] bits in
DFSDM_FLTxJCHGR register) are converted sequentially, starting with the lowest channel
(channel 0, if selected).
If the scan conversion is disabled (bit JSCAN=0) then, each time an injected conversion is
triggered, only one of the selected channels in the injected group (JCHG[3:0] bits in
DFSDM_FLTxJCHGR register) is converted and the channel selection is then moved to the
next selected channel. Writing to the JCHG[3:0] bits when JSCAN=0 sets the channel
selection to the lowest selected injected channel.
Only one injected conversion can be ongoing at a given time. Thus, any request to launch
an injected conversion is ignored if another request for an injected conversion has already
been issued but not yet completed.
Regular conversions can be launched using the following methods:
• Software: by writing ‘1’ to RSWSTART in the DFSDM_FLTxCR1 register.
• Synchronous with DFSDM_FLT0 if RSYNC=1: for DFSDM_FLTx (x>0), a regular
conversion is automatically launched when in DFSDM_FLT0; a regular conversion is
started by software (RSWSTART=1 in DFSDM_FLT0CR2 register). Each regular
conversion in DFSDM_FLTx (x>0) is always executed according to its local
configuration settings (RCONT, RCH, etc.).
Only one regular conversion can be pending or ongoing at a given time. Thus, any request
to launch a regular conversion is ignored if another request for a regular conversion has

372/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

already been issued but not yet completed. A regular conversion can be pending if it was
interrupted by an injected conversion or if it was started while an injected conversion was in
progress. This pending regular conversion is then delayed and is performed when all
injected conversion are finished. Any delayed regular conversion is signalized by RPEND bit
in DFSDM_FLTxRDATAR register.

14.4.16 Continuous and fast continuous modes

Setting RCONT in the DFSDM_FLTxCR1 register causes regular conversions to execute in
continuous mode. RCONT=1 means that the channel selected by RCH[1:0] is converted
repeatedly after ‘1’ is written to RSWSTART.
The regular conversions executing in continuous mode can be stopped by writing ‘0’ to
RCONT. After clearing RCONT, the on-going conversion is stopped immediately.
In continuous mode, the data rate can be increased by setting the FAST bit in the
DFSDM_FLTxCR1 register. In this case, the filter does not need to be refilled by new fresh
data if converting continuously from one channel because data inside the filter is valid from
previously sampled continuous data. The speed increase depends on the chosen filter
order. The first conversion in fast mode (FAST=1) after starting a continuous conversion by
RSWSTART=1 takes still full time (as when FAST=0), then each subsequent conversion is
finished in shorter intervals.
Conversion time in continuous mode:
if FAST = 0 (or first conversion if FAST=1):
for Sincx filters:
t = CNVCNT/fDFSDMCLK = [FOSR * (IOSR-1 + FORD) + FORD] / fCKIN
for FastSinc filter:
t = CNVCNT/fDFSDMCLK = [FOSR * (IOSR-1 + 4) + 2] / fCKIN
if FAST = 1 (except first conversion):
for Sincx and FastSinc filters:
t = CNVCNT/fDFSDMCLK = [FOSR * IOSR] / fCKIN
in case FOSR = FOSR[9:0]+1 = 1 (filter bypassed, only integrator active):
t = IOSR / fCKIN (... but CNVCNT=0)
Continuous mode is not available for injected conversions. Injected conversions can be
started by timer trigger to emulate the continuous mode with precise timing.
If a regular continuous conversion is in progress (RCONT=1) and if a write access to
DFSDM_FLTxCR1 register requesting regular continuous conversion (RCONT=1) is
performed, then regular continuous conversion is restarted from the next conversion cycle
(like new regular continuous conversion is applied for new channel selection - even if there
is no change in DFSDM_FLTxCR1 register).

14.4.17 Request precedence

An injected conversion has a higher precedence than a regular conversion. A regular
conversion which is already in progress is immediately interrupted by the request of an
injected conversion; this regular conversion is restarted after the injected conversion
finishes.

RM0402 Rev 6 373/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

An injected conversion cannot be launched if another injected conversion is pending or

already in progress: any request to launch an injected conversion (either by JSWSTART or
by a trigger) is ignored as long as bit JCIP is ‘1’ (in the DFSDM_FLTxISR register).
Similarly, a regular conversion cannot be launched if another regular conversion is pending
or already in progress: any request to launch a regular conversion (using RSWSTART) is
ignored as long as bit RCIP is ‘1’ (in the DFSDM_FLTxISR register).
However, if an injected conversion is requested while a regular conversion is already in
progress, the regular conversion is immediately stopped and an injected conversion is
launched. The regular conversion is then restarted and this delayed restart is signalized in
bit RPEND.
Injected conversions have precedence over regular conversions in that a injected
conversion can temporarily interrupt a sequence of continuous regular conversions. When
the sequence of injected conversions finishes, the continuous regular conversions start
again if RCONT is still set (and RPEND bit will signalize the delayed start on the first regular
conversion result).
Precedence also matters when actions are initiated by the same write to DFSDM, or if
multiple actions are pending at the end of another action. For example, suppose that, while
an injected conversion is in process (JCIP=1), a single write operation to DFSDM_FLTxCR1
writes ‘1’ to RSWSTART, requesting a regular conversion. When the injected sequence
finishes, the precedence dictates that the regular conversion is performed next and its
delayed start is signalized in RPEND bit.

14.4.18 Power optimization in run mode

In order to reduce the consumption, the DFSDM filter and integrator are automatically put
into idle when not used by conversions (RCIP=0, JCIP=0).

14.5 DFSDM interrupts

In order to increase the CPU performance, a set of interrupts related to the CPU event
occurrence has been implemented:
• End of injected conversion interrupt:
– enabled by JEOCIE bit in DFSDM_FLTxCR2 register
– indicated in JEOCF bit in DFSDM_FLTxISR register
– cleared by reading DFSDM_FLTxJDATAR register (injected data)
– indication of which channel end of conversion occurred, reported in JDATACH[1:0]
bits in DFSDM_FLTxJDATAR register
• End of regular conversion interrupt:
– enabled by REOCIE bit in DFSDM_FLTxCR2 register
– indicated in REOCF bit in DFSDM_FLTxISR register
– cleared by reading DFSDM_FLTxRDATAR register (regular data)
– indication of which channel end of conversion occurred, reported in
RDATACH[1:0] bits in DFSDM_FLTxRDATAR register
• Data overrun interrupt for injected conversions:
– occurred when injected converted data were not read from DFSDM_FLTxJDATAR
register (by CPU or DMA) and were overwritten by a new injected conversion

374/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

– enabled by JOVRIE bit in DFSDM_FLTxCR2 register

– indicated in JOVRF bit in DFSDM_FLTxISR register
– cleared by writing ‘1’ into CLRJOVRF bit in DFSDM_FLTxICR register
• Data overrun interrupt for regular conversions:
– occurred when regular converted data were not read from DFSDM_FLTxRDATAR
register (by CPU or DMA) and were overwritten by a new regular conversion
– enabled by ROVRIE bit in DFSDM_FLTxCR2 register
– indicated in ROVRF bit in DFSDM_FLTxISR register
– cleared by writing ‘1’ into CLRROVRF bit in DFSDM_FLTxICR register
• Analog watchdog interrupt:
– occurred when converted data (output data or data from analog watchdog filter -
according to AWFSEL bit setting in DFSDM_FLTxCR1 register) crosses
over/under high/low thresholds in DFSDM_FLTxAWHTR / DFSDM_FLTxAWLTR
registers
– enabled by AWDIE bit in DFSDM_FLTxCR2 register (on selected channels
AWDCH[3:0])
– indicated in AWDF bit in DFSDM_FLTxISR register
– separate indication of high or low analog watchdog threshold error by AWHTF[3:0]
and AWLTF[3:0] fields in DFSDM_FLTxAWSR register
– cleared by writing ‘1’ into corresponding CLRAWHTF[3:0] or CLRAWLTF[3:0] bits
in DFSDM_FLTxAWCFR register
• Short-circuit detector interrupt:
– occurred when the number of stable data crosses over thresholds in
DFSDM_CHyAWSCDR register
– enabled by SCDIE bit in DFSDM_FLTxCR2 register (on channel selected by
SCDEN bi tin DFSDM_CHyCFGR1 register)
– indicated in SCDF[3:0] bits in DFSDM_FLTxISR register (which also reports the
channel on which the short-circuit detector event occurred)
– cleared by writing ‘1’ into the corresponding CLRSCDF[3:0] bit in
DFSDM_FLTxICR register
• Channel clock absence interrupt:
– occurred when there is clock absence on CKINy pin (see Clock absence detection
in Section 14.4.4: Serial channel transceivers)
– enabled by CKABIE bit in DFSDM_FLTxCR2 register (on channels selected by
CKABEN bit in DFSDM_CHyCFGR1 register)
– indicated in CKABF[y] bit in DFSDM_FLTxISR register
– cleared by writing ‘1’ into CLRCKABF[y] bit in DFSDM_FLTxICR register

Table 90. DFSDM interrupt requests

Event/Interrupt clearing Interrupt enable
Interrupt event Event flag
method control bit

End of injected conversion JEOCF reading DFSDM_FLTxJDATAR JEOCIE

End of regular conversion REOCF reading DFSDM_FLTxRDATAR REOCIE
Injected data overrun JOVRF writing CLRJOVRF = 1 JOVRIE

RM0402 Rev 6 375/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Table 90. DFSDM interrupt requests (continued)

Event/Interrupt clearing Interrupt enable
Interrupt event Event flag
method control bit

Regular data overrun ROVRF writing CLRROVRF = 1 ROVRIE

AWDF,
writing CLRAWHTF[3:0] = 1 AWDIE,
Analog watchdog AWHTF[3:0],
writing CLRAWLTF[3:0] = 1 (AWDCH[3:0])
AWLTF[3:0]
SCDIE,
short-circuit detector SCDF[3:0] writing CLRSCDF[3:0] = 1
(SCDEN)
CKABIE,
Channel clock absence CKABF[3:0] writing CLRCKABF[3:0] = 1
(CKABEN)

14.6 DFSDM DMA transfer

To decrease the CPU intervention, conversions can be transferred into memory using a
DMA transfer. A DMA transfer for injected conversions is enabled by setting bit JDMAEN=1
in DFSDM_FLTxCR1 register. A DMA transfer for regular conversions is enabled by setting
bit RDMAEN=1 in DFSDM_FLTxCR1 register.
Note: With a DMA transfer, the interrupt flag is automatically cleared at the end of the injected or
regular conversion (JEOCF or REOCF bit in DFSDM_FLTxISR register) because DMA is
reading DFSDM_FLTxJDATAR or DFSDM_FLTxRDATAR register.

14.7 DFSDM channel y registers (y=0..3)

Word access (32-bit) must be used for registers write access except DFSDM_CHyDATINR
register. Write access to DFSDM_CHyDATINR register can be either word access (32-bit)
or half-word access (16-bit).

14.7.1 DFSDM channel y configuration register (DFSDM_CHyCFGR1)

This register specifies the parameters used by channel y.
Address offset: 0x00 + 0x20 * y, (y = 0 to 3)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DFSDM CKOUT
Res. Res. Res. Res. Res. Res. CKOUTDIV[7:0]
EN SRC
rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

CHIN CKAB
DATPACK[1:0] DATMPX[1:0] Res. Res. Res. CHEN SCDEN Res. SPICKSEL[1:0] SITP[1:0]
SEL EN

rw rw rw rw rw rw rw rw rw rw rw rw

376/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bit 31 DFSDMEN: Global enable for DFSDM interface

0: DFSDM interface disabled
1: DFSDM interface enabled
If DFSDM interface is enabled, then it is started to operate according to enabled y channels and
enabled x filters settings (CHEN bit in DFSDM_CHyCFGR1 and DFEN bit in DFSDM_FLTxCR1).
Data cleared by setting DFSDMEN=0:
–all registers DFSDM_FLTxISR are set to reset state (x = 0..1)
–all registers DFSDM_FLTxAWSR are set to reset state (x = 0..1)
Note: DFSDMEN is present only in DFSDM_CH0CFGR1 register (channel y=0)
Bit 30 CKOUTSRC: Output serial clock source selection
0: Source for output clock is from system clock
1: Source for output clock is from audio clock
–I2S clock selected by CKDFSDM1ASEL bit (Section 6.3.24: RCC Dedicated Clocks Configuration
Register (RCC_DCKCFGR))
This value can be modified only when DFSDMEN=0 (in DFSDM_CH0CFGR1 register).
Note: CKOUTSRC is present only in DFSDM_CH0CFGR1 register (channel y=0)
Bits 29:24 Reserved, must be kept at reset value.
Bits 23:16 CKOUTDIV[7:0]: Output serial clock divider
0: Output clock generation is disabled (CKOUT signal is set to low state)
1- 255: Defines the division of system clock for the serial clock output for CKOUT signal in range 2 -
256 (Divider = CKOUTDIV+1).
CKOUTDIV also defines the threshold for a clock absence detection.
This value can only be modified when DFSDMEN=0 (in DFSDM_CH0CFGR1 register).
If DFSDMEN=0 (in DFSDM_CH0CFGR1 register) then CKOUT signal is set to low state (setting is
performed one DFSDM clock cycle after DFSDMEN=0).
Note: CKOUTDIV is present only in DFSDM_CH0CFGR1 register (channel y=0)
Bits 15:14 DATPACK[1:0]: Data packing mode in DFSDM_CHyDATINR register.
0: Standard: input data in DFSDM_CHyDATINR register are stored only in INDAT0[15:0]. To empty
DFSDM_CHyDATINR register one sample must be read by the DFSDM filter from channel y.
1: Interleaved: input data in DFSDM_CHyDATINR register are stored as two samples:
–first sample in INDAT0[15:0] (assigned to channel y)
–second sample INDAT1[15:0] (assigned to channel y)
To empty DFSDM_CHyDATINR register, two samples must be read by the digital filter from
channel y (INDAT0[15:0] part is read as first sample and then INDAT1[15:0] part is read as next
sample).
2: Dual: input data in DFSDM_CHyDATINR register are stored as two samples:
–first sample INDAT0[15:0] (assigned to channel y)
–second sample INDAT1[15:0] (assigned to channel y+1)
To empty DFSDM_CHyDATINR register first sample must be read by the digital filter from channel
y and second sample must be read by another digital filter from channel y+1. Dual mode is
available only on even channel numbers (y = 0, 2), for odd channel numbers (y = 1, 3)
DFSDM_CHyDATINR is write protected. If an even channel is set to dual mode then the following
odd channel must be set into standard mode (DATPACK[1:0]=0) for correct cooperation with even
channel.
3: Reserved
This value can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).

RM0402 Rev 6 377/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bits 13:12 DATMPX[1:0]: Input data multiplexer for channel y

0: Data to channel y are taken from external serial inputs as 1-bit values. DFSDM_CHyDATINR
register is write protected.
1: Reserved
2: Data to channel y are taken from internal DFSDM_CHyDATINR register by direct CPU/DMA write.
There can be written one or two 16-bit data samples according DATPACK[1:0] bit field setting.
3: Reserved
This value can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).

Bits 11:9 Reserved, must be kept at reset value.

Bit 8 CHINSEL: Channel inputs selection
0: Channel inputs are taken from pins of the same channel y.
1: Channel inputs are taken from pins of the following channel (channel (y+1) modulo 8).
This value can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).
Bit 7 CHEN: Channel y enable
0: Channel y disabled
1: Channel y enabled
If channel y is enabled, then serial data receiving is started according to the given channel setting.
Bit 6 CKABEN: Clock absence detector enable on channel y
0: Clock absence detector disabled on channel y
1: Clock absence detector enabled on channel y
Bit 5 SCDEN: Short-circuit detector enable on channel y
0: Input channel y will not be guarded by the short-circuit detector
1: Input channel y will be continuously guarded by the short-circuit detector
Bit 4 Reserved, must be kept at reset value.
Bits 3:2 SPICKSEL[1:0]: SPI clock select for channel y
0: clock coming from external CKINy input - sampling point according SITP[1:0]
1: clock coming from internal CKOUT output - sampling point according SITP[1:0]
2: clock coming from internal CKOUT - sampling point on each second CKOUT falling edge.
For connection to external Σ∆ modulator which divides its clock input (from CKOUT) by 2 to
generate its output serial communication clock (and this output clock change is active on each
clock input rising edge).
3: clock coming from internal CKOUT output - sampling point on each second CKOUT rising edge.
For connection to external Σ∆ modulator which divides its clock input (from CKOUT) by 2 to
generate its output serial communication clock (and this output clock change is active on each
clock input falling edge).
This value can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).
Bits 1:0 SITP[1:0]: Serial interface type for channel y
00: SPI with rising edge to strobe data
01: SPI with falling edge to strobe data
10: Manchester coded input on DATINy pin: rising edge = logic 0, falling edge = logic 1
11: Manchester coded input on DATINy pin: rising edge = logic 1, falling edge = logic 0
This value can only be modified when CHEN=0 (in DFSDM_CHyCFGR1 register).

378/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.7.2 DFSDM channel y configuration register (DFSDM_CHyCFGR2)

This register specifies the parameters used by channel y.
Address offset: 0x04 + 0x20 * y, (y = 0 to 3)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

OFFSET[23:8]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

OFFSET[7:0] DTRBS[4:0] Res. Res. Res.

rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:8 OFFSET[23:0]: 24-bit calibration offset for channel y

For channel y, OFFSET is applied to the results of each conversion from this channel.
This value is set by software.
Bits 7:3 DTRBS[4:0]: Data right bit-shift for channel y
0-31: Defines the shift of the data result coming from the integrator - how many bit shifts to the right
will be performed to have final results. Bit-shift is performed before offset correction. The data shift is
rounding the result to nearest integer value. The sign of shifted result is maintained (to have valid
24-bit signed format of result data).
This value can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).
Bits 2:0 Reserved, must be kept at reset value.

14.7.3 DFSDM channel y analog watchdog and short-circuit detector register

(DFSDM_CHyAWSCDR)
Short-circuit detector and analog watchdog settings for channel y.
Address offset: 0x08 + 0x20 * y, (y = 0 to 3)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. AWFORD[1:0] Res. AWFOSR[4:0]

rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

BKSCD[3:0] Res. Res. Res. Res. SCDT[7:0]

rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 379/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bits 31:24 Reserved, must be kept at reset value.

Bits 23:22 AWFORD[1:0]: Analog watchdog Sinc filter order on channel y
0: FastSinc filter type
1: Sinc1 filter type
2: Sinc2 filter type
3: Sinc3 filter type x
 1 – z – FOSR
Sincx filter type transfer function: H ( z ) =  ----------------------------
-
 1 – z –1 

2
 1 – z – FOSR
FastSinc filter type transfer function: - ⋅ ( 1 + z – ( 2 ⋅ FOSR ) )
H ( z ) =  ----------------------------
 1 – z–1 
This bit can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).
Bit 21 Reserved, must be kept at reset value.
Bits 20:16 AWFOSR[4:0]: Analog watchdog filter oversampling ratio (decimation rate) on channel y
0 - 31: Defines the length of the Sinc type filter in the range 1 - 32 (AWFOSR + 1). This number is
also the decimation ratio of the analog data rate.
This bit can be modified only when CHEN=0 (in DFSDM_CHyCFGR1 register).
Note: If AWFOSR = 0 then the filter has no effect (filter bypass).
Bits 15:12 BKSCD[3:0]: Break signal assignment for short-circuit detector on channel y
BKSCD[i] = 0: Break i signal not assigned to short-circuit detector on channel y
BKSCD[i] = 1: Break i signal assigned to short-circuit detector on channel y
Bits 11:8 Reserved, must be kept at reset value.
Bits 7:0 SCDT[7:0]: short-circuit detector threshold for channel y
These bits are written by software to define the threshold counter for the short-circuit detector. If this
value is reached, then a short-circuit detector event occurs on a given channel.

14.7.4 DFSDM channel y watchdog filter data register

(DFSDM_CHyWDATR)
This register contains the data resulting from the analog watchdog filter associated to the
input channel y.
Address offset: 0x0C + 0x20 * y, (y = 0 to 3)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

WDATA[15:0]

r r r r r r r r r r r r r r r r

380/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 WDATA[15:0]: Input channel y watchdog data
Data converted by the analog watchdog filter for input channel y. This data is continuously converted
(no trigger) for this channel, with a limited resolution (OSR=1..32/sinc order = 1..3).

14.7.5 DFSDM channel y data input register (DFSDM_CHyDATINR)

This register contains 16-bit input data to be processed by DFSDM filter module. Write
access can be either word access (32-bit) or half-word access (16-bit).
Address offset: 0x10 + 0x20 * y, (y = 0 to 3)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

INDAT1[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

INDAT0[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 INDAT1[15:0]: Input data for channel y or channel y+1

Input parallel channel data to be processed by the digital filter if DATMPX[1:0]=1 or DATMPX[1:0]=2.
Data can be written by CPU/DMA (if DATMPX[1:0]=2).
If DATPACK[1:0]=0 (standard mode)
INDAT0[15:0] is write protected (not used for input sample).
If DATPACK[1:0]=1 (interleaved mode)
Second channel y data sample is stored into INDAT1[15:0]. First channel y data sample is stored
into INDAT0[15:0]. Both samples are read sequentially by DFSDM_FLTx filter as two channel y
data samples.
If DATPACK[1:0]=2 (dual mode).
For even y channels: sample in INDAT1[15:0] is automatically copied into INDAT0[15:0] of
channel (y+1).
For odd y channels: INDAT1[15:0] is write protected.
See Section 14.4.6: Parallel data inputs for more details.
INDAT0[15:1] is in the16-bit signed format.
Bits 15:0 INDAT0[15:0]: Input data for channel y
Input parallel channel data to be processed by the digital filter if DATMPX[1:0]=1 or DATMPX[1:0]=2.
Data can be written by CPU/DMA (if DATMPX[1:0]=2).
If DATPACK[1:0]=0 (standard mode)
Channel y data sample is stored into INDAT0[15:0].
If DATPACK[1:0]=1 (interleaved mode)
First channel y data sample is stored into INDAT0[15:0]. Second channel y data sample is stored
into INDAT1[15:0]. Both samples are read sequentially by DFSDM_FLTx filter as two channel y
data samples.
If DATPACK[1:0]=2 (dual mode).
For even y channels: Channel y data sample is stored into INDAT0[15:0].
For odd y channels: INDAT0[15:0] is write protected.
See Section 14.4.6: Parallel data inputs for more details.
INDAT0[15:0] is in the16-bit signed format.

RM0402 Rev 6 381/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

14.8 DFSDM filter x module registers (x=0..1)

Word access (32-bit) must be used for registers write access except DFSDM_CHyDATINR
register.

14.8.1 DFSDM filter x control register 1 (DFSDM_FLTxCR1)

Address offset: 0x100 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AWF RDMA RCON RSW
Res. FAST Res. Res. Res. RCH[1:0] Res. Res. Res. RSYNC Res.
SEL EN T START
rw rw rw rw rw rw rw rt_w1
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
JDMA JSW
Res. JEXTEN[1:0] Res. Res. JEXTSEL[2:0] Res. Res. JSCAN JSYNC Res. DFEN
EN START
rw rw rw rw rw rw rw rw rt_w1 rw

Bit 31 Reserved, must be kept at reset value.

Bit 30 AWFSEL: Analog watchdog fast mode select
0: Analog watchdog on data output value (after the digital filter). The comparison is done after offset
correction and shift
1: Analog watchdog on channel transceivers value (after watchdog filter)
Bit 29 FAST: Fast conversion mode selection for regular conversions
0: Fast conversion mode disabled
1: Fast conversion mode enabled
When converting a regular conversion in continuous mode, having enabled the fast mode causes
each conversion (except the first) to execute faster than in standard mode. This bit has no effect on
conversions which are not continuous.
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
if FAST=0 (or first conversion in continuous mode if FAST=1):
t = [FOSR * (IOSR-1 + FORD) + FORD] / fCKIN..... for Sincx filters
t = [FOSR * (IOSR-1 + 4) + 2] / fCKIN..... for FastSinc filter
if FAST=1 in continuous mode (except first conversion):
t = [FOSR * IOSR] / fCKIN
in case if FOSR = FOSR[9:0]+1 = 1 (filter bypassed, active only integrator):
t = IOSR / fCKIN (... but CNVCNT=0)
where: fCKIN is the channel input clock frequency (on given channel CKINy pin) or input data rate in
case of parallel data input.
Bits 28:26 Reserved, must be kept at reset value.
Bits 25:24 RCH[1:0]: Regular channel selection
0: Channel 0 is selected as the regular channel
1: Channel 1 is selected as the regular channel
...
3: Chanel 3 is selected as the regular channel
Writing this bit when RCIP=1 takes effect when the next regular conversion begins. This is
especially useful in continuous mode (when RCONT=1). It also affects regular conversions which
are pending (due to ongoing injected conversion).
Bits 23:22 Reserved, must be kept at reset value.

382/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bit 21 RDMAEN: DMA channel enabled to read data for the regular conversion
0: The DMA channel is not enabled to read regular data
1: The DMA channel is enabled to read regular data
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
Bit 20 Reserved, must be kept at reset value.
Bit 19 RSYNC: Launch regular conversion synchronously with DFSDM_FLT0
0: Do not launch a regular conversion synchronously with DFSDM_FLT0
1: Launch a regular conversion in this DFSDM_FLTx at the very moment when a regular conversion
is launched in DFSDM_FLT0
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
Bit 18 RCONT: Continuous mode selection for regular conversions
0: The regular channel is converted just once for each conversion request
1: The regular channel is converted repeatedly after each conversion request
Writing ‘0’ to this bit while a continuous regular conversion is already in progress stops the
continuous mode immediately.

Bit 17 RSWSTART: Software start of a conversion on the regular channel

0: Writing ‘0’ has no effect
1: Writing ‘1’ makes a request to start a conversion on the regular channel and causes RCIP to
become ‘1’. If RCIP=1 already, writing to RSWSTART has no effect. Writing ‘1’ has no effect if
RSYNC=1.
This bit is always read as ‘0’.
Bits 16:15 Reserved, must be kept at reset value.
Bits 14:13 JEXTEN[1:0]: Trigger enable and trigger edge selection for injected conversions
00: Trigger detection is disabled
01: Each rising edge on the selected trigger makes a request to launch an injected conversion
10: Each falling edge on the selected trigger makes a request to launch an injected conversion
11: Both rising edges and falling edges on the selected trigger make requests to launch injected
conversions
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
Bits 12:11 Reserved, must be kept at reset value.
Bits 10:8 JEXTSEL[2:0]: Trigger signal selection for launching injected conversions
0x0-0x7: Trigger inputs selected by the following table.
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
DFSDM_FLT0 DFSDM_FLT1
0x00 dfsdm_jtrg0 dfsdm_jtrg0
0x01 dfsdm_jtrg1 dfsdm_jtrg1
0x02 dfsdm_jtrg2 dfsdm_jtrg2
0x03 dfsdm_jtrg3 dfsdm_jtrg3
0x04 dfsdm_jtrg5 dfsdm_jtrg5
0x05 dfsdm_jtrg7 dfsdm_jtrg7
0x06 dfsdm_jtrg9 dfsdm_jtrg9
0x07 dfsdm_jtrg10 dfsdm_jtrg10
Refer to Table 86: DFSDM triggers connection.
Bits 7:6 Reserved, must be kept at reset value.

RM0402 Rev 6 383/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bit 5 JDMAEN: DMA channel enabled to read data for the injected channel group
0: The DMA channel is not enabled to read injected data
1: The DMA channel is enabled to read injected data
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
Bit 4 JSCAN: Scanning conversion mode for injected conversions
0: One channel conversion is performed from the injected channel group and next the selected
channel from this group is selected.
1: The series of conversions for the injected group channels is executed, starting over with the
lowest selected channel.
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
Writing JCHG if JSCAN=0 resets the channel selection to the lowest selected channel.
Bit 3 JSYNC: Launch an injected conversion synchronously with the DFSDM_FLT0 JSWSTART trigger
0: Do not launch an injected conversion synchronously with DFSDM_FLT0
1: Launch an injected conversion in this DFSDM_FLTx at the very moment when an injected
conversion is launched in DFSDM_FLT0 by its JSWSTART trigger
This bit can be modified only when DFEN=0 (DFSDM_FLTxCR1).
Bit 2 Reserved, must be kept at reset value.
Bit 1 JSWSTART: Start a conversion of the injected group of channels
0: Writing ‘0’ has no effect.
1: Writing ‘1’ makes a request to convert the channels in the injected conversion group, causing
JCIP to become ‘1’ at the same time. If JCIP=1 already, then writing to JSWSTART has no effect.
Writing ‘1’ has no effect if JSYNC=1.
This bit is always read as ‘0’.
Bit 0 DFEN: DFSDM_FLTx enable
0: DFSDM_FLTx is disabled. All conversions of given DFSDM_FLTx are stopped immediately and
all DFSDM_FLTx functions are stopped.
1: DFSDM_FLTx is enabled. If DFSDM_FLTx is enabled, then DFSDM_FLTx starts operating
according to its setting.
Data which are cleared by setting DFEN=0:
–register DFSDM_FLTxISR is set to the reset state
–register DFSDM_FLTxAWSR is set to the reset state

14.8.2 DFSDM filter x control register 2 (DFSDM_FLTxCR2)

Address offset: 0x104 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. AWDCH[3:0]

rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CKAB ROVR JOVR REOC JEOC
Res. Res. Res. Res. EXCH[3:0] Res. SCDIE AWDIE
IE IE IE IE IE
rw rw rw rw rw rw rw rw rw rw rw

384/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bits 31:20 Reserved, must be kept at reset value.

Bits 19:16 AWDCH[3:0]: Analog watchdog channel selection
These bits select the input channel to be guarded continuously by the analog watchdog.
AWDCH[y] = 0: Analog watchdog is disabled on channel y
AWDCH[y] = 1: Analog watchdog is enabled on channel y
Bits 15:12 Reserved, must be kept at reset value.
Bits 11:8 EXCH[3:0]: Extremes detector channel selection
These bits select the input channels to be taken by the Extremes detector.
EXCH[y] = 0: Extremes detector does not accept data from channel y
EXCH[y] = 1: Extremes detector accepts data from channel y
Bit 7 Reserved, must be kept at reset value.
Bit 6 CKABIE: Clock absence interrupt enable
0: Detection of channel input clock absence interrupt is disabled
1: Detection of channel input clock absence interrupt is enabled
Please see the explanation of CKABF[3:0] in DFSDM_FLTxISR.
Note: CKABIE is present only in DFSDM_FLT0CR2 register (filter x=0)
Bit 5 SCDIE: Short-circuit detector interrupt enable
0: short-circuit detector interrupt is disabled
1: short-circuit detector interrupt is enabled
Please see the explanation of SCDF[3:0] in DFSDM_FLTxISR.
Note: SCDIE is present only in DFSDM_FLT0CR2 register (filter x=0)
Bit 4 AWDIE: Analog watchdog interrupt enable
0: Analog watchdog interrupt is disabled
1: Analog watchdog interrupt is enabled
Please see the explanation of AWDF in DFSDM_FLTxISR.
Bit 3 ROVRIE: Regular data overrun interrupt enable
0: Regular data overrun interrupt is disabled
1: Regular data overrun interrupt is enabled
Please see the explanation of ROVRF in DFSDM_FLTxISR.
Bit 2 JOVRIE: Injected data overrun interrupt enable
0: Injected data overrun interrupt is disabled
1: Injected data overrun interrupt is enabled
Please see the explanation of JOVRF in DFSDM_FLTxISR.
Bit 1 REOCIE: Regular end of conversion interrupt enable
0: Regular end of conversion interrupt is disabled
1: Regular end of conversion interrupt is enabled
Please see the explanation of REOCF in DFSDM_FLTxISR.
Bit 0 JEOCIE: Injected end of conversion interrupt enable
0: Injected end of conversion interrupt is disabled
1: Injected end of conversion interrupt is enabled
Please see the explanation of JEOCF in DFSDM_FLTxISR.

RM0402 Rev 6 385/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

14.8.3 DFSDM filter x interrupt and status register (DFSDM_FLTxISR)

Address offset: 0x108 + 0x80 * x, (x = 0 to 1)
Reset value: 0x00FF 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. SCDF[3:0] Res. Res. Res. Res. CKABF[3:0]

r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. RCIP JCIP Res. Res. Res. Res. Res. Res. Res. Res. AWDF ROVRF JOVRF REOCF JEOCF

r r r r r r r

Bits 31:28 Reserved, must be kept at reset value.

Bits 27:24 SCDF[3:0]: short-circuit detector flag
SDCF[y]=0: No short-circuit detector event occurred on channel y
SDCF[y]=1: The short-circuit detector counter reaches, on channel y, the value programmed in the
DFSDM_CHyAWSCDR registers
This bit is set by hardware. It can be cleared by software using the corresponding CLRSCDF[y] bit in
the DFSDM_FLTxICR register. SCDF[y] is cleared also by hardware when CHEN[y] = 0 (given
channel is disabled).
Note: SCDF[3:0] is present only in DFSDM_FLT0ISR register (filter x=0)
Bits 23:20 Reserved, must be kept at reset value.
Bits 19:16 CKABF[3:0]: Clock absence flag
CKABF[y]=0: Clock signal on channel y is present.
CKABF[y]=1: Clock signal on channel y is not present.
Given y bit is set by hardware when clock absence is detected on channel y. It is held at
CKABF[y]=1 state by hardware when CHEN=0 (see DFSDM_CHyCFGR1 register). It is held at
CKABF[y]=1 state by hardware when the transceiver is not yet synchronized.It can be cleared by
software using the corresponding CLRCKABF[y] bit in the DFSDM_FLTxICR register.
Note: CKABF[3:0] is present only in DFSDM_FLT0ISR register (filter x=0)
Bit 15 Reserved, must be kept at reset value.
Bit 14 RCIP: Regular conversion in progress status
0: No request to convert the regular channel has been issued
1: The conversion of the regular channel is in progress or a request for a regular conversion is
pending
A request to start a regular conversion is ignored when RCIP=1.
Bit 13 JCIP: Injected conversion in progress status
0: No request to convert the injected channel group (neither by software nor by trigger) has been
issued
1: The conversion of the injected channel group is in progress or a request for a injected conversion
is pending, due either to ‘1’ being written to JSWSTART or to a trigger detection
A request to start an injected conversion is ignored when JCIP=1.
Bits 12:5 Reserved, must be kept at reset value.

386/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bit 4 AWDF: Analog watchdog

0: No Analog watchdog event occurred
1: The analog watchdog block detected voltage which crosses the value programmed in the
DFSDM_FLTxAWLTR or DFSDM_FLTxAWHTR registers.
This bit is set by hardware. It is cleared by software by clearing all source flag bits AWHTF[3:0] and
AWLTF[3:0] in DFSDM_FLTxAWSR register (by writing ‘1’ into the clear bits in
DFSDM_FLTxAWCFR register).
Bit 3 ROVRF: Regular conversion overrun flag
0: No regular conversion overrun has occurred
1: A regular conversion overrun has occurred, which means that a regular conversion finished while
REOCF was already ‘1’. RDATAR is not affected by overruns
This bit is set by hardware. It can be cleared by software using the CLRROVRF bit in the
DFSDM_FLTxICR register.
Bit 2 JOVRF: Injected conversion overrun flag
0: No injected conversion overrun has occurred
1: An injected conversion overrun has occurred, which means that an injected conversion finished
while JEOCF was already ‘1’. JDATAR is not affected by overruns
This bit is set by hardware. It can be cleared by software using the CLRJOVRF bit in the
DFSDM_FLTxICR register.
Bit 1 REOCF: End of regular conversion flag
0: No regular conversion has completed
1: A regular conversion has completed and its data may be read
This bit is set by hardware. It is cleared when the software or DMA reads DFSDM_FLTxRDATAR.
Bit 0 JEOCF: End of injected conversion flag
0: No injected conversion has completed
1: An injected conversion has completed and its data may be read
This bit is set by hardware. It is cleared when the software or DMA reads DFSDM_FLTxJDATAR.

Note: For each of the flag bits, an interrupt can be enabled by setting the corresponding bit in
DFSDM_FLTxCR2. If an interrupt is called, the flag must be cleared before exiting the
interrupt service routine.
All the bits of DFSDM_FLTxISR are automatically reset when DFEN=0.

14.8.4 DFSDM filter x interrupt flag clear register (DFSDM_FLTxICR)

Address offset: 0x10C + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. CLRSCDF[3:0] Res. Res. Res. Res. CLRCKABF[3:0]

rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CLRR CLRJ
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
OVRF OVRF
rc_w1 rc_w1

RM0402 Rev 6 387/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bits 31:28 Reserved, must be kept at reset value.

Bits 27:24 CLRSCDF[3:0]: Clear the short-circuit detector flag
CLRSCDF[y]=0: Writing ‘0’ has no effect
CLRSCDF[y]=1: Writing ‘1’ to position y clears the corresponding SCDF[y] bit in the
DFSDM_FLTxISR register
Note: CLRSCDF[3:0] is present only in DFSDM_FLT0ICR register (filter x=0)
Bits 23:20 Reserved, must be kept at reset value.
Bits 19:16 CLRCKABF[3:0]: Clear the clock absence flag
CLRCKABF[y]=0: Writing ‘0’ has no effect
CLRCKABF[y]=1: Writing ‘1’ to position y clears the corresponding CKABF[y] bit in the
DFSDM_FLTxISR register. When the transceiver is not yet synchronized, the clock absence flag is
set and cannot be cleared by CLRCKABF[y].
Note: CLRCKABF[3:0] is present only in DFSDM_FLT0ICR register (filter x=0)
Bits 15:4 Reserved, must be kept at reset value.
Bit 3 CLRROVRF: Clear the regular conversion overrun flag
0: Writing ‘0’ has no effect
1: Writing ‘1’ clears the ROVRF bit in the DFSDM_FLTxISR register
Bit 2 CLRJOVRF: Clear the injected conversion overrun flag
0: Writing ‘0’ has no effect
1: Writing ‘1’ clears the JOVRF bit in the DFSDM_FLTxISR register
Bits 1:0 Reserved, must be kept at reset value.

Note: The bits of DFSDM_FLTxICR are always read as ‘0’.

14.8.5 DFSDM filter x injected channel group selection register

(DFSDM_FLTxJCHGR)
Address offset: 0x110 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0001

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. JCHG[3:0]

rw rw rw rw

388/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bits 31:4 Reserved, must be kept at reset value.

Bits 3:0 JCHG[3:0]: Injected channel group selection
JCHG[y]=0: channel y is not part of the injected group
JCHG[y]=1: channel y is part of the injected group
If JSCAN=1, each of the selected channels is converted, one after another. The lowest channel
(channel 0, if selected) is converted first and the sequence ends at the highest selected channel.
If JSCAN=0, then only one channel is converted from the selected channels, and the channel
selection is moved to the next channel. Writing JCHG, if JSCAN=0, resets the channel selection to
the lowest selected channel.
At least one channel must always be selected for the injected group. Writes causing all JCHG bits to
be zero are ignored.

14.8.6 DFSDM filter x control register (DFSDM_FLTxFCR)

Address offset: 0x114 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

FORD[2:0] Res. Res. Res. FOSR[9:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. IOSR[7:0]

rw rw rw rw rw rw rw rw

Bits 31:29 FORD[2:0]: Sinc filter order

0: FastSinc filter type
1: Sinc1 filter type
2: Sinc2 filter type
3: Sinc3 filter type
4: Sinc4 filter type
5: Sinc5 filter type
6-7: Reserved x
 1 – z –FOSR
Sincx filter type transfer function: H ( z ) =  ----------------------------
-
 1 – z –1 

2
 1 – z –FOSR
FastSinc filter type transfer function: - ⋅ ( 1 + z –( 2 ⋅ FOSR ) )
H ( z ) =  ----------------------------
 1 – z –1 
This bit can only be modified when DFEN=0 (DFSDM_FLTxCR1).
Bits 28:26 Reserved, must be kept at reset value.

RM0402 Rev 6 389/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bits 25:16 FOSR[9:0]: Sinc filter oversampling ratio (decimation rate)

0 - 1023: Defines the length of the Sinc type filter in the range 1 - 1024 (FOSR = FOSR[9:0] +1). This
number is also the decimation ratio of the output data rate from filter.
This bit can only be modified when DFEN=0 (DFSDM_FLTxCR1)
Note: If FOSR = 0, then the filter has no effect (filter bypass).
Bits 15:8 Reserved, must be kept at reset value.
Bits 7:0 IOSR[7:0]: Integrator oversampling ratio (averaging length)
0- 255: The length of the Integrator in the range 1 - 256 (IOSR + 1). Defines how many samples
from Sinc filter will be summed into one output data sample from the integrator. The output data rate
from the integrator will be decreased by this number (additional data decimation ratio).
This bit can only be modified when DFEN=0 (DFSDM_FLTxCR1)
Note: If IOSR = 0, then the Integrator has no effect (Integrator bypass).

14.8.7 DFSDM filter x data register for injected group

(DFSDM_FLTxJDATAR)
Address offset: 0x118 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

JDATA[23:8]

r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

JDATA[7:0] Res. Res. Res. Res. Res. Res. JDATACH[1:0]

r r r r r r r r r r

Bits 31:8 JDATA[23:0]: Injected group conversion data

When each conversion of a channel in the injected group finishes, its resulting data is stored in this
field. The data is valid when JEOCF=1. Reading this register clears the corresponding JEOCF.
Bits 7:2 Reserved, must be kept at reset value.
Bits 1:0 JDATACH[1:0]: Injected channel most recently converted
When each conversion of a channel in the injected group finishes, JDATACH[1:0] is updated to
indicate which channel was converted. Thus, JDATA[23:0] holds the data that corresponds to the
channel indicated by JDATACH[1:0].

Note: DMA may be used to read the data from this register. Half-word accesses may be used to
read only the MSBs of conversion data.
Reading this register also clears JEOCF in DFSDM_FLTxISR. Thus, the firmware must not
read this register if DMA is activated to read data from this register.

390/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.8.8 DFSDM filter x data register for the regular channel

(DFSDM_FLTxRDATAR)
Address offset: 0x11C + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

RDATA[23:8]

r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

RDATA[7:0] Res. Res. Res. RPEND Res. Res. RDATACH[1:0]

r r r r r r r r r r r

Bits 31:8 RDATA[23:0]: Regular channel conversion data

When each regular conversion finishes, its data is stored in this register. The data is valid when
REOCF=1. Reading this register clears the corresponding REOCF.
Bits 7:5 Reserved, must be kept at reset value.
Bit 4 RPEND: Regular channel pending data
Regular data in RDATA[23:0] was delayed due to an injected channel trigger during the conversion
Bits 3:2 Reserved, must be kept at reset value.
Bits 1:0 RDATACH[1:0]: Regular channel most recently converted
When each regular conversion finishes, RDATACH[1:0] is updated to indicate which channel was
converted (because regular channel selection RCH[1:0] in DFSDM_FLTxCR1 register can be
updated during regular conversion). Thus RDATA[23:0] holds the data that corresponds to the
channel indicated by RDATACH[1:0].

Note: Half-word accesses may be used to read only the MSBs of conversion data.
Reading this register also clears REOCF in DFSDM_FLTxISR.

14.8.9 DFSDM filter x analog watchdog high threshold register

(DFSDM_FLTxAWHTR)
Address offset: 0x120 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

AWHT[23:8]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

AWHT[7:0] Res. Res. Res. Res. BKAWH[3:0]

rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 391/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bits 31:8 AWHT[23:0]: Analog watchdog high threshold

These bits are written by software to define the high threshold for the analog watchdog.
Note: In case channel transceivers monitor (AWFSEL=1), the higher 16 bits (AWHT[23:8]) define the
16-bit threshold as compared with the analog watchdog filter output (because data coming from
the analog watchdog filter are up to a 16-bit resolution). Bits AWHT[7:0] are not taken into
comparison in this case.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 BKAWH[3:0]: Break signal assignment to analog watchdog high threshold event
BKAWH[i] = 0: Break i signal is not assigned to an analog watchdog high threshold event
BKAWH[i] = 1: Break i signal is assigned to an analog watchdog high threshold event

14.8.10 DFSDM filter x analog watchdog low threshold register

(DFSDM_FLTxAWLTR)
Address offset: 0x124 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

AWLT[23:8]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

AWLT[7:0] Res. Res. Res. Res. BKAWL[3:0]

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:8 AWLT[23:0]: Analog watchdog low threshold

These bits are written by software to define the low threshold for the analog watchdog.
Note: In case channel transceivers monitor (AWFSEL=1), only the higher 16 bits (AWLT[23:8]) define
the 16-bit threshold as compared with the analog watchdog filter output (because data coming
from the analog watchdog filter are up to a 16-bit resolution). Bits AWLT[7:0] are not taken into
comparison in this case.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 BKAWL[3:0]: Break signal assignment to analog watchdog low threshold event
BKAWL[i] = 0: Break i signal is not assigned to an analog watchdog low threshold event
BKAWL[i] = 1: Break i signal is assigned to an analog watchdog low threshold event

392/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

14.8.11 DFSDM filter x analog watchdog status register

(DFSDM_FLTxAWSR)
Address offset: 0x128 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. AWHTF[3:0] Res. Res. Res. Res. AWLTF[3:0]

r r r r r r r r

Bits 31:12 Reserved, must be kept at reset value.

Bits 11:8 AWHTF[3:0]: Analog watchdog high threshold flag
AWHTF[y]=1 indicates a high threshold error on channel y. It is set by hardware. It can be cleared by
software using the corresponding CLRAWHTF[y] bit in the DFSDM_FLTxAWCFR register.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 AWLTF[3:0]: Analog watchdog low threshold flag
AWLTF[y]=1 indicates a low threshold error on channel y. It is set by hardware. It can be cleared by
software using the corresponding CLRAWLTF[y] bit in the DFSDM_FLTxAWCFR register.

Note: All the bits of DFSDM_FLTxAWSR are automatically reset when DFEN=0.

14.8.12 DFSDM filter x analog watchdog clear flag register

(DFSDM_FLTxAWCFR)
Address offset: 0x12C + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. CLRAWHTF[3:0] Res. Res. Res. Res. CLRAWLTF[3:0]

rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bits 31:12 Reserved, must be kept at reset value.

RM0402 Rev 6 393/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

Bits 11:8 CLRAWHTF[3:0]: Clear the analog watchdog high threshold flag
CLRAWHTF[y]=0: Writing ‘0’ has no effect
CLRAWHTF[y]=1: Writing ‘1’ to position y clears the corresponding AWHTF[y] bit in the
DFSDM_FLTxAWSR register
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 CLRAWLTF[3:0]: Clear the analog watchdog low threshold flag
CLRAWLTF[y]=0: Writing ‘0’ has no effect
CLRAWLTF[y]=1: Writing ‘1’ to position y clears the corresponding AWLTF[y] bit in the
DFSDM_FLTxAWSR register

14.8.13 DFSDM filter x extremes detector maximum register

(DFSDM_FLTxEXMAX)
Address offset: 0x130 + 0x80 * x, (x = 0 to 1)
Reset value: 0x8000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

EXMAX[23:8]

rs_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

EXMAX[7:0] Res. Res. Res. Res. Res. Res. EXMAXCH[1:0]

rc_r rc_r rc_r rc_r rc_r rc_r rc_r rc_r r r

Bits 31:8 EXMAX[23:0]: Extremes detector maximum value

These bits are set by hardware and indicate the highest value converted by DFSDM_FLTx.
EXMAX[23:0] bits are reset to value (0x800000) by reading of this register.
Bits 7:2 Reserved, must be kept at reset value.
Bits 1:0 EXMAXCH[1:0]: Extremes detector maximum data channel.
These bits contains information about the channel on which the data is stored into EXMAX[23:0].
Bits are cleared by reading of this register.

14.8.14 DFSDM filter x extremes detector minimum register

(DFSDM_FLTxEXMIN)
Address offset: 0x134 + 0x80 * x, (x = 0 to 1)
Reset value: 0x7FFF FF00

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

EXMIN[23:8]

rc_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

EXMIN[7:0] Res. Res. Res. Res. Res. Res. EXMINCH[1:0]

rs_r rs_r rs_r rs_r rs_r rs_r rs_r rs_r r r

394/1163 RM0402 Rev 6

RM0402 Digital filter for sigma delta modulators (DFSDM)

Bits 31:8 EXMIN[23:0]: Extremes detector minimum value

These bits are set by hardware and indicate the lowest value converted by DFSDM_FLTx.
EXMIN[23:0] bits are reset to value (0x7FFFFF) by reading of this register.
Bits 7:2 Reserved, must be kept at reset value.
Bits 1:0 EXMINCH[1:0]: Extremes detector minimum data channel
These bits contain information about the channel on which the data is stored into EXMIN[23:0]. Bits
are cleared by reading of this register.

14.8.15 DFSDM filter x conversion timer register (DFSDM_FLTxCNVTIMR)

Address offset: 0x138 + 0x80 * x, (x = 0 to 1)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

CNVCNT[27:12]

r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

CNVCNT[11:0] Res. Res. Res. Res.

r r r r r r r r r r r r

Bits 31:4 CNVCNT[27:0]: 28-bit timer counting conversion time t = CNVCNT[27:0] / fDFSDMCLK
The timer has an input clock from DFSDM clock (system clock fDFSDMCLK). Conversion time
measurement is started on each conversion start and stopped when conversion finishes (interval
between first and last serial sample). Only in case of filter bypass (FOSR[9:0] = 0) is the conversion
time measurement stopped and CNVCNT[27:0] = 0. The counted time is:
if FAST=0 (or first conversion in continuous mode if FAST=1):
t = [FOSR * (IOSR-1 + FORD) + FORD] / fCKIN..... for Sincx filters
t = [FOSR * (IOSR-1 + 4) + 2] / fCKIN..... for FastSinc filter
if FAST=1 in continuous mode (except first conversion):
t = [FOSR * IOSR] / fCKIN
in case if FOSR = FOSR[9:0]+1 = 1 (filter bypassed, active only integrator):
CNVCNT = 0 (counting is stopped, conversion time: t = IOSR / fCKIN)
where: fCKIN is the channel input clock frequency (on given channel CKINy pin) or input data rate in
case of parallel data input (from CPU/DMA write)
Note: When conversion is interrupted (e.g. by disable/enable selected channel) the timer counts also
this interruption time.
Bits 3:0 Reserved, must be kept at reset value.

RM0402 Rev 6 395/1163

400
Digital filter for sigma delta modulators (DFSDM) RM0402

14.8.16 DFSDM register map

The following table summarizes the DFSDM registers.

Table 91. DFSDM register map and reset values

Register
Offset 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name

DATPACK[1:0]

DATMPX[1:0]
CKOUTSRC
DFSDMEN

SPICKSEL
CHINSEL

SITP[1:0]
CKABEN
SCDEN
CHEN
DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.

Res.

[1:0]
CKOUTDIV[7:0]
0x00 CH0CFGR1

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_

Res.
Res.
Res.
OFFSET[23:0] DTRBS[4:0]
0x04 CH0CFGR2
reset value 0 0
AWFORD

DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
Res.
Res.
Res.
[1:0]

AWFOSR[4:0] BKSCD[3:0] SCDT[7:0]

0x08 CH0AWSCDR

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res. WDATA[15:0]
0x0C CH0WDATR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
INDAT1[15:0] INDAT0[15:0]
0x10 CH0DATINR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x14 -
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Reserved
0x1C
DATPACK[1:0]

DATMPX[1:0]

SPICKSEL
CHINSEL

SITP[1:0]
CKABEN
SCDEN
DFSDM_ CHEN
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.

Res.

[1:0]
0x20 CH1CFGR1

reset value 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
OFFSET[23:0] DTRBS[4:0]
0x24 CH1CFGR2
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
AWFORD[1:0]

DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
Res.
Res.
Res.

AWFOSR[4:0] BKSCD[3:0] SCDT[7:0]

0x28 CH1AWSCDR

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

WDATA[15:0]
0x2C CH1WDATR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
INDAT1[15:0] INDAT0[15:0]
0x30 CH1DATINR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x34 -
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Reserved
0x3C

396/1163 RM0402 Rev 6

 0x70
0x68
0x64
0x60
0x50
0x48
0x44
0x40

0x6C
0x5C
0x4C

0xFC

0x100
0x74 -
0x54 -
Offset
RM0402

name

DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_

Reserved
Reserved

FLT0CR1

reset value
reset value
reset value
reset value
reset value
reset value
reset value
reset value
reset value
reset value
reset value
Register

CH3CFGR2
CH3CFGR1
CH2CFGR2
CH2CFGR1

CH3WDATR
CH2WDATR

CH3DATINR
CH2DATINR

CH3AWSCDR
CH2AWSCDR

0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 31

0
0
0
0
0
AWFSEL Res. Res. Res. Res. Res. Res. Res. Res. 30

0
0
0
0
0
FAST Res. Res. Res. Res. Res. Res. Res. Res. 29

0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 28

0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 27

0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 26

0
0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. 25
RCH[1:0]

0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 24

0
0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. 23

AWFORD[1:0] AWFORD[1:0]

0
0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res.

INDAT1[15:0]
INDAT1[15:0]
22

0
0
0
0
0

RDMAEN Res. Res. Res. Res. Res. Res. Res. Res. 21

0
0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. 20

0
0
0
0
0
0
0

RSYNC Res. Res. Res. Res. Res. Res. 19

0
0
0
0
0
0
0

RCONT Res. Res. Res. Res. Res. Res.

OFFSET[23:0]
OFFSET[23:0]

18

0
0
0
0
0
0
0

RSW START Res. Res. Res. Res. Res. Res.

RM0402 Rev 6
17

AWFOSR[4:0]
AWFOSR[4:0]

0
0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
0

0
0
0
0
0
0

Res. Res. Res. 15

DATPACK[1:0] DATPACK[1:0]

0
0
0
0
0

0
0
0
0
0
0

Res. Res. 14
JEXTEN[1:0]

0
0
0
0

0
0
0
0
0
0
0

Res. Res. 13
DATMPX[1:0] DATMPX[1:0]

BKSCD[3:0]
BKSCD[3:0]

0
0
0
0
0

0
0
0
0
0

Res. Res. Res. 12

0
0
0
0

0
0

Res. Res. Res. Res. Res. Res. Res. 11

0
0
0
0
0

0
0

Res. Res. Res. Res. Res. Res. 10

0
0
0
0
0

0
0

JEXTSEL[2:0] Res. Res. Res. Res. Res. Res. 9

0
0
0
0
0

0
0
0
0

Res. Res. CHINSEL Res. Res. CHINSEL 8

Table 91. DFSDM register map and reset values (continued)

0
0
0
0

0
0
0
0
0
0

Res. Res. CHEN Res. CHEN 7

0
0
0
0

0
0
0
0
0
0

Res. Res. CKABEN Res. CKABEN

WDATA[15:0]
WDATA[15:0]

INDAT0[15:0]
INDAT0[15:0]

0
0
0
0
0

0
0
0
0
0
0

JDMAEN Res. SCDEN Res. SCDEN 5

0
0
0
0
0

0
0
0
0

JSCAN Res. Res. Res. Res. 4

DTRBS[4:0]
DTRBS[4:0]

0
0
0
0
0

0
0
0
0
0
0

JSYNC Res. Res. 3

SPICKSEL[1:0] SPICKSEL[1:0]

SCDT[7:0]
SCDT[7:0]

0
0

0
0
0
0
0
0

Res. Res. Res. Res. Res. 2

0
0
0

0
0
0
0
0
0

JSW START Res. Res. Res. Res. 1

SITP[1:0] SITP[1:0]

0
0
0

0
0
0
0
0
0

DFEN Res. Res. Res. Res. 0

Digital filter for sigma delta modulators (DFSDM)

397/1163
400
Digital filter for sigma delta modulators (DFSDM) RM0402

Table 91. DFSDM register map and reset values (continued)

Register
Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name

ROVRIE

REOCIE
CKABIE

JOVRIE

JEOCIE
AWDIE
SCDIE
DFSDM_

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

Res.
AWDCH[3:0] EXCH[3:0]
0x104 FLT0CR2

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ROVRF

REOCF
JOVRF

JEOCF
AWDF
DFSDM_

RCIP
JCIP
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
SCDF[3:0] CKABF[3:0]
0x108 FLT0ISR

reset value 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0

CLR ROVRF
CLR JOVRF
DFSDM_ CLRSCDF CLRCKABF
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
0x10C FLT0ICR [3:0] [3:0]

reset value 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
JCHG[3:0]
0x110 FLT0JCHGR
reset value 0 0 0 1
FORD[2:0]

DFSDM_
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
FOSR[9:0] IOSR[7:0]
0x114 FLT0FCR

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

JDATACH [1:0]
DFSDM_

Res.
Res.
Res.
Res.
Res.
Res.
JDATA[23:0]
0x118 FLT0JDATAR

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RDATACH[1:0]
RPEND
DFSDM_
Res.
Res.
Res.

Res.
Res.
RDATA[23:0]
0x11C FLT0RDATAR

reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.

AWHT[23:0] BKAWH[3:0]
0x120 FLT0AWHTR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.

AWLT[23:0] BKAWL[3:0]
0x124 FLT0AWLTR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

AWHTF[3:0] AWLTF[3:0]
0x128 FLT0AWSR
reset value 0 0 0 0 0 0 0 0
DFSDM_ CLRAWHTF CLRAWLTF
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.

0x12C FLT0AWCFR [3:0] [3:0]

reset value 0 0 0 0 0 0 0 0
EXMAXCH[1:0]

DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.

EXMAX[23:0]
0x130 FLT0EXMAX

reset value 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

398/1163 RM0402 Rev 6

 0x198
0x194
0x190
0x188
0x184
0x180
0x138
0x134

0x1A0
0x19C
0x18C
0x17C
0x13C -
Offset
RM0402

name

FLT1ISR

FLT1ICR

DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_
DFSDM_

Reserved

FLT1CR2
FLT1CR1

FLT1FCR
reset value
reset value

reset value
reset value
reset value
reset value
reset value
reset value
reset value
reset value
reset value
Register

FLT0EXMIN

FLT1JCHGR

FLT1AWHTR
FLT1JDATAR

FLT1RDATAR
FLT0CNVTIMR

0
0
0
0
0
0
Res. Res. Res. Res. Res. Res. 31

0
0
0
0
0
0
1
FORD[2:0] Res. Res. Res. Res. AWFSEL Res. 30

0
0
0
0
0
0
1
Res. Res. Res. Res. FAST Res. 29

0
0
0
0
1
Res. Res. Res. Res. Res. Res. Res. 28

0
0
0
0
1

Res. Res. Res. Res. Res. Res. Res. 27

0
0
0
0
1

Res. Res. Res. Res. Res. Res. Res. 26

0
0
0
0
0
0
1

Res. Res. Res. Res. Res. 25

RCH[1:0]

0
0
0
0
0
0
1

Res. Res. Res. Res. Res. 24

0
0
0
0
0
1

Res. Res. Res. Res. Res. Res. 23

0
0
0
0
0
1

Res. Res. Res. Res. Res. Res. 22

0
0
0
0
0
0
1

Res. Res. Res. Res. RDMAEN Res. 21

0
0
0
0
0
1

Res. Res. Res. Res. Res. Res. 20

FOSR[9:0]

0
0
0
0
0
0
0
1

Res. Res. Res. RSYNC Res. 19

AWHT[23:0]
JDATA[23:0]

0
0
0
0
0
0
0
1
EXMIN[23:0]

RDATA[23:0]
Res. Res. Res. RCONT Res. 18

0
0
0
0
0
0
0
1

Res. Res. Res. RSW START Res.

RM0402 Rev 6
17

0
0
0
0
0
0
1

AWDCH[3:0]
Res. Res. Res. Res. Res. 16
CNVCNT[27:0]

0
0
0
0
1

Res. Res. Res. Res. Res. Res. Res. 15

0
0
0
0
0
1

0
Res. Res. Res. RCIP Res. Res. 14
JEXTEN[1:0]

0
0
0
0

0
0
1

Res. Res. Res. JCIP Res. Res. 13

0
0
0
0
1

Res. Res. Res. Res. Res. Res. Res. 12

0
0
0
0
1

Res. Res. Res. Res. Res. Res. 11

0
0
0
0
0
1

Res. Res. Res. Res. Res. 10

0
0
0
0
0
1

Res. Res. Res. Res. JEXTSEL[2:0] Res. 9

EXCH[3:0]

0
0
0
0
0
1

Res. Res. Res. Res. Res. 8

Table 91. DFSDM register map and reset values (continued)

0
0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 7

0
0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 6

0
0
0

Res. Res. Res. Res. Res. Res. Res. JDMAEN Res. Res. 5

0
0
0
0
0

Res. RPEND Res. Res. Res. AWDF AWDIE JSCAN Res. Res. 4

0
0
0
0

0
0
0

Res. Res. CLR ROVRF ROVRF ROVRIE JSYNC Res. Res. Res. 3

IOSR[7:0]

0
0
0
0

0
0

Res. Res. CLR JOVRF JOVRF JOVRIE Res. Res. Res. Res. 2

0
0
0
0
0
0
0
0

Res. REOCF REOCIE JSW START Res. Res. 1

RDATACH[1:0] JDATACH[1:0] EXMINCH[1:0]
JCHG[3:0]

0
0
0
0
0
0
0
0

BKAWH[3:0]
Res. JEOCF JEOCIE DFEN Res. Res. 0
Digital filter for sigma delta modulators (DFSDM)

399/1163
400
Digital filter for sigma delta modulators (DFSDM) RM0402

Table 91. DFSDM register map and reset values (continued)

Register
Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name
DFSDM_

Res.
Res.
Res.
Res.
AWLT[23:0] BKAWL[3:0]
0x1A4 FLT1AWLTR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DFSDM_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
AWHTF[3:0] AWLTF[3:0]
0x1A8 FLT1AWSR
reset value 0 0 0 0 0 0 0 0
DFSDM_ CLRAWHTF CLRAWLTF
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
0x1AC FLT1AWCFR [3:0] [3:0]
reset value 0 0 0 0 0 0 0 0

EXMAXCH[1:0]
DFSDM_

Res.
Res.
Res.
Res.
Res.
Res.
EXMAX[23:0]
0x1B0 FLT1EXMAX

reset value 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

EXMINCH[1:0]
DFSDM_

Res.
Res.
Res.
Res.
Res.
Res.
FLT1EXMIN EXMIN[23:0]
0x1B4

reset value 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0
DFSDM_

Res.
Res.
Res.
Res.
CNVCNT[27:0]
0x1B8 FLT1CNVTIMR
reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0x1BC -
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Reserved
0x2FC

Refer to Section 2.2 on page 49 for the register boundary addresses.

400/1163 RM0402 Rev 6

RM0402 True random number generator (RNG)

15 True random number generator (RNG)

15.1 Introduction
The RNG is a true random number generator that continuously provides 32-bit entropy
samples, based on an analog noise source. It can be used by the application as a live
entropy source to build a NIST compliant Deterministic Random Bit Generator (DRBG).
The RNG true random number generator has been tested using NIST statistical test suite
SP800-22 rev1a (April 2010).

15.2 RNG main features

• The RNG delivers 32-bit true random numbers, produced by an analog entropy source
post-processed with linear-feedback shift registers (LFSR).
• It produces one 32-bit random samples every 42 RNG clock cycles(dedicated clock).
• It allows embedded continuous basic health tests with associated error management
– Includes too low sampling clock detection and repetition count tests.
• It can be disabled to reduce power consumption.
• It has an AMBA AHB slave peripheral, accessible through 32-bit word single accesses
only (else an AHB bus error is generated). Warning! any write not equal to 32 bits might
corrupt the register content.

RM0402 Rev 6 401/1163

412
True random number generator (RNG) RM0402

15.3 RNG functional description

15.3.1 RNG block diagram

Figure 75 shows the RNG block diagram.

Figure 75. RNG block diagram

32-bit AHB Bus

True RNG
Banked Registers

control RNG_CR
AHB
interface data RNG_DR

Alarms
rng_it status RNG_SR 16-bit

rng_hclk AHB clock domain

Fault detection Data shift reg
Clock checker
Post-processing logic 16-bit

8-bit LFSR (x2)

rng_clk
2-bit
Sampling &
Normalization (x 2)
RNG clock domain

Analog Analog
noise noise
en_osc source 1 source 2
Analog noise source

MSv42096V2

15.3.2 RNG internal signals

Table 92 describes a list of useful-to-know internal signals available at the RNG level, not at
the STM32 product level (on pads).

Table 92. RNG internal input/output signals

Signal name Signal type Description

rng_it Digital output RNG global interrupt request

rng_hclk Digital input AHB clock
rng_clk Digital input RNG dedicated clock, asynchronous to rng_hclk

402/1163 RM0402 Rev 6

RM0402 True random number generator (RNG)

15.3.3 Random number generation

The true random number generator (RNG) delivers truly random data through its AHB
interface at deterministic intervals. Within its boundary the RNG implements the entropy
source model pictured on Figure 76, and provides three main functions to the application:
• Collects the bitstring output of the entropy source box
• Obtains samples of the noise source for validation purpose
• Collects error messages from continuous health tests

Figure 76. Entropy source model

Error Output
message

Conditioning
Heath (optional)
tests
Raw data

Post-processing
(optional)

Digitization Output
(raw data or
digitized noise
Noise Source source)

Entropy source

MSv42095V1

The main components of the RNG are:

• A source of physical randomness (analog noise source)
• A digitization stage for this analog noise source
• A stage delivering post-processed noise source (raw data)
• An output buffer for the raw data. If further cryptographic conditioning is required by the
application it will need to be performed by software.
• An optional output for the digitized noise source (unbuffered, on digital pads)
• Basic health tests on the digitized noise source
The components pictured above are detailed hereafter:

RM0402 Rev 6 403/1163

412
True random number generator (RNG) RM0402

Noise source
The noise source is the component that contains the non-deterministic, entropy-providing
activity that is ultimately responsible for the uncertainty associated with the bitstring output
by the entropy source. It is composed of:
• Two analog noise sources, each based on three XORed free-running ring oscillator
outputs. It is possible to disable those analog oscillators to save power, as described in
Section 15.3.8: RNG low-power usage.
• A sampling stage of these outputs clocked by a dedicated clock input (rng_clk),
delivering a 2-bit raw data output.
This noise source sampling is independent to the AHB interface clock frequency (rng_hclk).
Note: In Section 15.6: RNG entropy source validation recommended RNG clock frequencies are
given.

Post processing
The sample values obtained from a true random noise source consist of 2-bit bitstrings.
Because this noise source output is biased, the RNG implements a post-processing
component that reduces that bias to a tolerable level.
The RNG post-processing consists of two stages, applied to each noise source bits:
• The RNG takes half of the bits from the sampled noise source, and half of the bits from
inverted sampled noise source. Thus, if the source generates more ‘1’ than ‘0’ (or the
opposite), it is filtered
• A linear feedback shift register (LFSR) performs a whitening process, producing 8-bit
strings.
This component is clocked by the RNG clock.
The times required between two random number generations, and between the RNG
initialization and availability of first sample are described in Section 15.5: RNG processing
time.

Output buffer
The RNG_DR data output register can store up to two 16-bit words which have been output
from the post-processing component (LFSR). In order to read back 32-bit random samples it
is required to wait 42 RNG clock cycles.
Whenever a random number is available through the RNG_DR register the DRDY flag
transitions from “0” to “1”. This flag remains high until output buffer becomes empty after
reading one word from the RNG_DR register.
Note: When interrupts are enabled an interrupt is generated when this data ready flag transitions
from “0” to “1”. Interrupt is then cleared automatically by the RNG as explained above.

404/1163 RM0402 Rev 6

RM0402 True random number generator (RNG)

Health checks
This component ensures that the entire entropy source (with its noise source) starts then
operates as expected, obtaining assurance that failures are caught quickly and with a high
probability and reliability.
The RNG implements the following health check features.
1. Continuous health tests, running indefinitely on the output of the noise source
– Repetition count test, flagging an error when:
a) One of the noise source has provided more than 64 consecutive bits at a constant
value (“0” or “1”)
b) One of the noise sources has delivered more than 32 consecutive occurrence of
two bits patterns (“01” or “10”)
2. Vendor specific continuous test
– Real-time “too slow” sampling clock detector, flagging an error when one RNG
clock cycle is smaller than AHB clock cycle divided by 16.
The CECS and SECS status bits in the RNG_SR register indicate when an error condition is
detected, as detailed in Section 15.3.7: Error management.
Note: An interrupt can be generated when an error is detected.

15.3.4 RNG initialization

When a hardware reset occurs the following chain of events occurs:
1. The analog noise source is enabled, and logic starts sampling the analog output after
four RNG clock cycles, filling LFSR shift register and associated 16-bit post-processing
shift register.
2. The output buffer is refilled automatically according to the RNG usage.
The associated initialization time can be found in Section 15.5: RNG processing time.

15.3.5 RNG operation

Normal operations
To run the RNG using interrupts the following steps are recommended:
1. Enable the interrupts by setting the IE bit in the RNG_CR register. At the same time
enable the RNG by setting the bit RNGEN=1.
2. An interrupt is now generated when a random number is ready or when an error
occurs. Therefore at each interrupt, check that:
– No error occurred. The SEIS and CEIS bits should be set to 0 in the RNG_SR
register.
– A random number is ready. The DRDY bit must be set to 1 in the RNG_SR
register.
– If above two conditions are true the content of the RNG_DR register can be read.

RM0402 Rev 6 405/1163

412
True random number generator (RNG) RM0402

To run the RNG in polling mode following steps are recommended:

1. Enable the random number generation by setting the RNGEN bit to “1” in the RNG_CR
register.
2. Read the RNG_SR register and check that:
– No error occurred (the SEIS and CEIS bits should be set to 0)
– A random number is ready (the DRDY bit should be set to 1)
3. If above conditions are true read the content of the RNG_DR register.
Note: When data is not ready (DRDY=”0”) RNG_DR returns zero.
It is recommended to always verify that RNG_DR is different from zero. Because when it is
the case a seed error occurred between RNG_SR polling and RND_DR output reading (rare
event).

Low-power operations
If the power consumption is a concern to the application, low-power strategies can be used,
as described in Section 15.3.8: RNG low-power usage.

Software post-processing
If a NIST approved DRBG with 128 bits of security strength is required an approved random
generator software must be built around the RNG true random number generator.

15.3.6 RNG clocking

The RNG runs on two different clocks: the AHB bus clock and a dedicated RNG clock.
The AHB clock is used to clock the AHB banked registers and the post-processing
component. The RNG clock is used for noise source sampling. Recommended clock
configurations are detailed in Section 15.6: RNG entropy source validation.
Note: When the CED bit in the RNG_CR register is set to “0”, the RNG clock frequency should be
higher than AHB clock frequency divided by 16, otherwise the clock checker always flags a
clock error (CECS=1 in the RNG_SR register).
See Section 15.3.1: RNG block diagram for details (AHB and RNG clock domains).

15.3.7 Error management

In parallel to random number generation an health check block verifies the correct noise
source behavior and the frequency of the RNG source clock as detailed in this section.
Associated error state is also described.

Clock error detection

When the clock error detection is enabled (CED = 0) and if the RNG clock frequency is too
low, the RNG sets to “1” both the CEIS and CECS bits to indicate that a clock error
occurred. In this case, the application should check that the RNG clock is configured
correctly (see Section 15.3.6: RNG clocking) and then it must clear the CEIS bit interrupt
flag. The CECS bit is automatically cleared when clocking condition is normal.
Note: The clock error has no impact on generated random numbers, i.e. application can still read
RNG_DR register.
CEIS is set only when CECS is set to “1” by RNG.

406/1163 RM0402 Rev 6

RM0402 True random number generator (RNG)

Noise source error detection

When a noise source (or seed) error occurs, the RNG stops generating random numbers
and sets to “1” both SEIS and SECS bits to indicate that a seed error occurred. If a value is
available in the RNG_DR register, it must not be used as it may not have enough entropy.
In order to fully recover from a seed error application must clear the SEIS bit by writing it to
“0”, then clear and set the RNGEN bit to reinitialize and restart the RNG.

15.3.8 RNG low-power usage

If power consumption is a concern, the RNG can be disabled as soon as the DRDY bit is set
to “1” by setting the RNGEN bit to “0” in the RNG_CR register. The 32-bit random value
stored in the RNG_DR register will be still be available. If a new random is needed the
application will need to re-enable the RNG and wait for 42+4 RNG clock cycles.
When disabling the RNG the user deactivates all the analog seed generators, whose power
consumption is given in the datasheet electrical characteristics section.

15.4 RNG interrupts

In the RNG an interrupt can be produced on the following events:
• Data ready flag
• Seed error, see Section 15.3.7: Error management
• Clock error, see Section 15.3.7: Error management
Dedicated interrupt enable control bits are available as shown in Table 93.

Table 93. RNG interrupt requests

Interrupt acronym Interrupt event Event flag Enable control bit Interrupt clear method

Data ready flag DRDY IE None (automatic)

RNG Seed error flag SEIS IE Write 0 to SEIS
Clock error flag CEIS IE Write 0 to CEIS

The user can enable or disable the above interrupt sources individually by changing the
mask bits or the general interrupt control bit IE in the RNG_CR register. The status of the
individual interrupt sources can be read from the RNG_SR register.
Note: Interrupts are generated only when RNG is enabled.

15.5 RNG processing time

The RNG can produce one 32-bit random numbers every 42 RNG clock cycles.
After enabling or re-enabling the RNG using the RNGEN bit it takes 46 RNG clock cycles
before random data are available.

RM0402 Rev 6 407/1163

412
True random number generator (RNG) RM0402

15.6 RNG entropy source validation

15.6.1 Introduction
In order to assess the amount of entropy available from the RNG, STMicroelectronics has
tested the peripheral using NIST SP800-22 rev1a statistical tests. The results can be
provided on demand or the customer can reproduce the tests.
For more information on running this NIST statistical test suite, refer to STM32
microcontrollers random number generation validation using NIST statistical test suite
application note (AN4230), available on STMicroelectronics website.

15.6.2 Validation conditions

STMicroelectronics has tested the RNG true random number generator in the following
conditions:
• RNG clock rng_clk= 48 MHz (CED bit = ’0’ in RNG_CR register) and rng_clk = 400 kHz
(CED bit = ‘1’ in RNG_CR register).

15.6.3 Data collection

In order to run statistical tests it is required to collect samples from the entropy source at raw
data level as well as at the output of the entropy source.Contact STMicroelectronics if above
samples need to be retrieved for your product.

408/1163 RM0402 Rev 6

RM0402 True random number generator (RNG)

15.7 RNG registers

The RNG is associated with a control register, a data register and a status register.

15.7.1 RNG control register (RNG_CR)

Address offset: 0x000
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. CED Res. IE RNGEN Res. Res.
rw rw rw

Bits 31:6 Reserved, must be kept at reset value.

Bit 5 CED: Clock error detection
0: Clock error detection is enable
1: Clock error detection is disable
The clock error detection cannot be enabled nor disabled on-the-fly when the RNG is
enabled, i.e. to enable or disable CED the RNG must be disabled.

Bit 4 Reserved, must be kept at reset value.

Bit 3 IE: Interrupt Enable
0: RNG Interrupt is disabled
1: RNG Interrupt is enabled. An interrupt is pending as soon as DRDY=’1’, SEIS=’1’ or
CEIS=1 in the RNG_SR register.
Bit 2 RNGEN: True random number generator enable
0: True random number generator is disabled. Analog noise sources are powered off and
logic clocked by the RNG clock is gated.
1: True random number generator is enabled.
Bits 1:0 Reserved, must be kept at reset value.

RM0402 Rev 6 409/1163

412
True random number generator (RNG) RM0402

15.7.2 RNG status register (RNG_SR)

Address offset: 0x004
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. SEIS CEIS Res. Res. SECS CECS DRDY
rc_w0 rc_w0 r r r

Bits 31:7 Reserved, must be kept at reset value.

Bit 6 SEIS: Seed error interrupt status
This bit is set at the same time as SECS. It is cleared by writing 0. Writing 1 has no effect.
0: No faulty sequence detected
1: At least one faulty sequence has been detected. See SECS bit description for details.
An interrupt is pending if IE = 1 in the RNG_CR register.
Bit 5 CEIS: Clock error interrupt status
This bit is set at the same time as CECS. It is cleared by writing 0. Writing 1 has no effect.
0: The RNG clock is correct (fRNGCLK > fHCLK/16)
1: The RNG has been detected too slow (fRNGCLK < fHCLK/16)
An interrupt is pending if IE = 1 in the RNG_CR register.
Bits 4:3 Reserved, must be kept at reset value.
Bit 2 SECS: Seed error current status
0: No faulty sequence has currently been detected. If the SEIS bit is set, this means that a
faulty sequence was detected and the situation has been recovered.
1: One of the noise source has provided more than 64 consecutive bits at a constant value
(“0” or “1”), or more than 32 consecutive occurrence of two bits patterns (“01” or “10”)
Bit 1 CECS: Clock error current status
0: The RNG clock is correct (fRNGCLK> fHCLK/16). If the CEIS bit is set, this means that a
slow clock was detected and the situation has been recovered.
1: The RNG clock is too slow (fRNGCLK< fHCLK/16).
Note: CECS bit is valid only if the CED bit in the RNG_CR register is set to 0.
Bit 0 DRDY: Data Ready
0: The RNG_DR register is not yet valid, no random data is available.
1: The RNG_DR register contains valid random data.
Once the RNG_DR register has been read, this bit returns to 0 until a new random value is
generated.
If IE=1 in the RNG_CR register, an interrupt is generated when DRDY=1.

410/1163 RM0402 Rev 6

RM0402 True random number generator (RNG)

15.7.3 RNG data register (RNG_DR)

Address offset: 0x008
Reset value: 0x0000 0000
The RNG_DR register is a read-only register that delivers a 32-bit random value when read.
After being read this register delivers a new random value after 42 periods of RNG clock if
the output FIFO is empty.
The content of this register is valid when DRDY=1 and value is not 0x0, even if RNGEN=0.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RNDATA[31:16]
r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RNDATA[15:0]
r r r r r r r r r r r r r r r r

Bits 31:0 RNDATA[31:0]: Random data

32-bit random data which are valid when DRDY=1. When DRDY=0 RNDATA value is zero.
It is recommended to always verify that RNG_DR is different from zero. Because when it is
the case a seed error occurred between RNG_SR polling and RND_DR output reading
(rare event).

RM0402 Rev 6 411/1163

412
True random number generator (RNG) RM0402

15.7.4 RNG register map

Table 94 gives the RNG register map and reset values.

Table 94. RNG register map and reset map

Offset Register name

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
RNGEN
CED
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
Res.
RNG_CR

IE
0x000
Reset value 0 0 0

DRDY
CECS
SECS
CEIS
SEIS
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
RNG_SR
0x004
Reset value 0 0 0 0 0
RNG_DR RNDATA[31:0]
0x008
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 for the register boundary addresses.

412/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16 Advanced-control timers (TIM1&TIM8)

16.1 TIM1&TIM8 introduction

The advanced-control timers (TIM1&TIM8) consist of a 16-bit auto-reload counter driven by
a programmable prescaler.
It may be used for a variety of purposes, including measuring the pulse length of input
signals (input capture) or generating output waveforms (output compare, PWM,
complementary PWM with dead-time insertion).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The advanced-control (TIM1&TIM8) and general-purpose (TIMx) timers are completely
independent, and do not share any resources. They can be synchronized together as
described in Section 16.3.20.

16.2 TIM1&TIM8 main features

TIM1&TIM8 timer features include:
• 16-bit up, down, up/down auto-reload counter.
• 16-bit programmable prescaler allowing dividing (also “on the fly”) the counter clock
frequency either by any factor between 1 and 65536.
• Up to 4 independent channels for:
– Input Capture
– Output Compare
– PWM generation (Edge and Center-aligned Mode)
– One-pulse mode output
• Complementary outputs with programmable dead-time
• Synchronization circuit to control the timer with external signals and to interconnect
several timers together.
• Repetition counter to update the timer registers only after a given number of cycles of
the counter.
• Break input to put the timer’s output signals in reset state or in a known state.
• Interrupt/DMA generation on the following events:
– Update: counter overflow/underflow, counter initialization (by software or
internal/external trigger)
– Trigger event (counter start, stop, initialization or count by internal/external trigger)
– Input capture
– Output compare
– Break input
• Supports incremental (quadrature) encoder and Hall-sensor circuitry for positioning
purposes
• Trigger input for external clock or cycle-by-cycle current management

RM0402 Rev 6 413/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 77. Advanced-control timer block diagram

Internal clock (CK_INT)

CK_TIM18 from RCC
TIMx_ETR

ETR Trigger
Polarity selection, ETRF controller
Edge detector and Prescaler ETRP
TRGO To other timers
Input filter To DAC and ADC
ITR0
TGI
ITR1
ITR2 TRGI Slave mode
TRC controller
ITR3
TIF_ED

Reset,
UI
Enable,
Up/Down,
TI1FP1 Count REP Register
Encoder
interface U

TIMx_CH1
TI2FP2

OC1

TIMx_CH1N
AutoReload
U
Register

OC1N
Repetition counter
CK_PSC PSC CK_CNT CNT

TIMx_CH2
(prescaler) (counter) DTG[7:0] registers
TIMx_CH1

CC4I CC4I
TI1 TI1FP1

OC2
Input filter & IC1 IC1PS Capture/Compare OC1REF Output
Prescaler DTG

TIMx_CH2N
Edge detector 1 Register control
TI1FP2
U
TRC CC3I CC3I
TIMx_CH2

TI2FP1
TI2 Input filter & IC2 IC2PS Capture/Compare OC2REF
Prescaler DTG Output
Edge detector 2 Register control
TI2FP2 OC2N
U

TIMx_CH3N TIMx_CH3
TRC CC2I CC2I
TIMx_CH3

TI3FP3 OC3
TI3 Input filter & IC3 IC3PS Capture/Compare OC3REF
Prescaler DTG Output
Edge detector 3 Register control
TI3FP4 U

OC3N
TRC CC1I CC1I
TIMx_CH4

TI4FP3
TI4 Input filter & IC4 IC4PS Capture/Compare OC4REF Output
Prescaler control
Edge detector 4 Register
TI4FP4
U
TRC
OC4

TIMx_CH4
TIMx_BKIN

BRK BI
Polarity selection

Clock failure event from clock controller

CSS (Clock Security System)

Interrupt & DMA output

Event
MS39906V2

414/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3 TIM1&TIM8 functional description

16.3.1 Time-base unit

The main block of the programmable advanced-control timer is a 16-bit counter with its
related auto-reload register. The counter can count up, down or both up and down. The
counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
• Counter register (TIMx_CNT)
• Prescaler register (TIMx_PSC)
• Auto-reload register (TIMx_ARR)
• Repetition counter register (TIMx_RCR)
The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow (or underflow when downcounting) and if the UDIS bit equals 0 in the
TIMx_CR1 register. It can also be generated by software. The generation of the update
event is described in detailed for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the counter starts counting 1 clock cycle after setting the CEN bit in the TIMx_CR1
register.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as this control register is buffered. The new prescaler ratio is
taken into account at the next update event.
Figure 78 and Figure 79 give some examples of the counter behavior when the prescaler
ratio is changed on the fly:

RM0402 Rev 6 415/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 78. Counter timing diagram with prescaler division change from 1 to 2

CK_PSC

CEN

Timerclock = CK_CNT

Counter register F7 F8 F9 FA FB FC 00 01 02 03

Update event (UEV)

Prescaler control register 0 1

Write a new value in TIMx_PSC

Prescaler buffer 0 1

Prescaler counter 0 0 1 0 1 0 1 0 1
MS31076V2

Figure 79. Counter timing diagram with prescaler division change from 1 to 4

CK_PSC

CEN

Timerclock = CK_CNT

F7 F8 F9 FA FB FC 00 01
Counter register

Update event (UEV)

0 3
Prescaler control register

Write a new value in TIMx_PSC

0 3
Prescaler buffer

0 0 1 2 3 0 1 2 3
Prescaler counter

MS31077V2

416/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3.2 Counter modes

Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
If the repetition counter is used, the update event (UEV) is generated after upcounting is
repeated for the number of times programmed in the repetition counter register plus one
(TIMx_RCR+1). Else the update event is generated at each counter overflow.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event.
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter restarts from 0, as well as the counter of the prescaler (but the
prescale rate does not change). In addition, if the URS bit (update request selection) in
TIMx_CR1 register is set, setting the UG bit generates an update event UEV but without
setting the UIF flag (thus no interrupt or DMA request is sent). This is to avoid generating
both update and capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The repetition counter is reloaded with the content of TIMx_RCR register,
• The auto-reload shadow register is updated with the preload value (TIMx_ARR),
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.

Figure 80. Counter timing diagram, internal clock divided by 1

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31078V3

RM0402 Rev 6 417/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 81. Counter timing diagram, internal clock divided by 2

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0034 0035 0036 0000 0001 0002 0003

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31079V3

Figure 82. Counter timing diagram, internal clock divided by 4

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0035 0036 0000 0001

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31080V3

Figure 83. Counter timing diagram, internal clock divided by N

CK_PSC

Timerclock = CK_CNT

Counter register 1F 20 00

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31081V3

418/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 84. Counter timing diagram, update event when ARPE=0

(TIMx_ARR not preloaded)

CK_PSC

CEN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FF 36

Write a new value in TIMx_ARR

MS31082V3

Figure 85. Counter timing diagram, update event when ARPE=1

(TIMx_ARR preloaded)

CK_PSC

CEN

Timerclock = CK_CNT

Counter register F0 F1 F2 F3 F4 F5 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)
Auto-reload preload
F5 36
register

Auto-reload shadow
register F5 36

Write a new value in TIMx_ARR MS31083V2

RM0402 Rev 6 419/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Downcounting mode
In downcounting mode, the counter counts from the auto-reload value (content of the
TIMx_ARR register) down to 0, then restarts from the auto-reload value and generates a
counter underflow event.
If the repetition counter is used, the update event (UEV) is generated after downcounting is
repeated for the number of times programmed in the repetition counter register plus one
(TIMx_RCR+1). Else the update event is generated at each counter underflow.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event.
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter restarts from the current auto-reload value, whereas the counter of the
prescaler restarts from 0 (but the prescale rate doesn’t change).
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The repetition counter is reloaded with the content of TIMx_RCR register
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register)
• The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that the auto-reload is updated before the counter is
reloaded, so that the next period is the expected one
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.

420/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 86. Counter timing diagram, internal clock divided by 1

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 05 04 03 02 01 00 36 35 34 33 32 31 30 2F

Counter underflow
(cnt_udf)

Update event (UEV)

Update interrupt flag

(UIF)

MS31184V1

Figure 87. Counter timing diagram, internal clock divided by 2

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0002 0001 0000 0036 0035 0034 0033

Counter underflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31185V1

RM0402 Rev 6 421/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 88. Counter timing diagram, internal clock divided by 4

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0001 0000 0000 0001

Counter underflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31186V1

Figure 89. Counter timing diagram, internal clock divided by N

CK_PSC

Timerclock = CK_CNT

Counter register 20 1F 00 36

Counter underflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31187V1

422/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 90. Counter timing diagram, update event when repetition counter is not used

CK_PSC

CEN

Timerclock = CK_CNT

Counter register 05 04 03 02 01 00 36 35 34 33 32 31 30 2F

Counter underflow

Update event (UEV)

Update interrupt flag

(UIF)

Auto-reload preload
register FF 36

Write a new value in TIMx_ARR

MS31188V1

Center-aligned mode (up/down counting)

In center-aligned mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register) – 1, generates a counter overflow event, then counts from the auto-
reload value down to 1 and generates a counter underflow event. Then it restarts counting
from 0.
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are not equal to
'00'. The Output compare interrupt flag of channels configured in output is set when: the
counter counts down (Center aligned mode 1, CMS = "01"), the counter counts up (Center
aligned mode 2, CMS = "10") the counter counts up and down (Center aligned mode 3,
CMS = "11").
In this mode, the DIR direction bit in the TIMx_CR1 register cannot be written. It is updated
by hardware and gives the current direction of the counter.
The update event can be generated at each counter overflow and at each counter underflow
or by setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event. In this case, the counter restarts counting from
0, as well as the counter of the prescaler.
The UEV update event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter continues counting up and down, based on the current auto-reload
value.
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an UEV update event but without setting the UIF flag (thus no interrupt or

RM0402 Rev 6 423/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The repetition counter is reloaded with the content of TIMx_RCR register
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register)
• The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that if the update source is a counter overflow, the auto-
reload is updated before the counter is reloaded, so that the next period is the expected
one (the counter is loaded with the new value).
The following figures show some examples of the counter behavior for different clock
frequencies.

Figure 91. Counter timing diagram, internal clock divided by 1, TIMx_ARR = 0x6

CK_PSC

CEN

Timerclock = CK_CNT

Counter register 04 03 02 01 00 01 02 03 04 05 06 05 04 03

Counter underflow

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31189V2

1. Here, center-aligned mode 1 is used (for more details refer to Section 16.4: TIM1&TIM8 registers).

Figure 92. Counter timing diagram, internal clock divided by 2

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0003 0002 0001 0000 0001 0002 0003

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

MS31190V2

424/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 93. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0034 0035 0036 0035

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31191V2

1. Center-aligned mode 2 or 3 is used with an UIF on overflow.

Figure 94. Counter timing diagram, internal clock divided by N

CK_PSC

Timerclock = CK_CNT

Counter register 20 1F 01 00

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

MS31192V2

RM0402 Rev 6 425/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 95. Counter timing diagram, update event with ARPE=1 (counter underflow)

CK_PSC

CEN

Timer clock = CK_CNT

Counter register 06 05 04 03 02 01 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FD 36

Write a new value in TIMx_ARR

Auto-reload active register FD 36

MS31193V2

Figure 96. Counter timing diagram, update event with ARPE=1 (counter overflow)

CK_PSC

CEN
Timer clock = CK_CNT

Counter register F7 F8 F9 FA FB FC 36 35 34 33 32 31 30 2F

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FD 36

Write a new value in TIMx_ARR

Auto-reload active register FD 36

MS31194V2

16.3.3 Repetition counter

Section 16.3.1: Time-base unit describes how the update event (UEV) is generated with
respect to the counter overflows/underflows. It is actually generated only when the repetition
counter has reached zero. This can be useful when generating PWM signals.
This means that data are transferred from the preload registers to the shadow registers
(TIMx_ARR auto-reload register, TIMx_PSC prescaler register, but also TIMx_CCRx
capture/compare registers in compare mode) every N+1 counter overflows or underflows,
where N is the value in the TIMx_RCR repetition counter register.

426/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

The repetition counter is decremented:

• At each counter overflow in upcounting mode,
• At each counter underflow in downcounting mode,
• At each counter overflow and at each counter underflow in center-aligned mode.
Although this limits the maximum number of repetition to 128 PWM cycles, it makes it
possible to update the duty cycle twice per PWM period. When refreshing compare
registers only once per PWM period in center-aligned mode, maximum resolution is
2xTck, due to the symmetry of the pattern.
The repetition counter is an auto-reload type; the repetition rate is maintained as defined by
the TIMx_RCR register value (refer to Figure 97). When the update event is generated by
software (by setting the UG bit in TIMx_EGR register) or by hardware through the slave
mode controller, it occurs immediately whatever the value of the repetition counter is and the
repetition counter is reloaded with the content of the TIMx_RCR register.
In center-aligned mode, for odd values of RCR, the update event occurs either on the
overflow or on the underflow depending on when the RCR register was written and when
the counter was started. If the RCR was written before starting the counter, the UEV occurs
on the overflow. If the RCR was written after starting the counter, the UEV occurs on the
underflow. For example for RCR = 3, the UEV is generated on each 4th overflow or
underflow event depending on when RCR was written.

Figure 97. Update rate examples depending on mode and TIMx_RCR register settings

Counter-aligned mode Edge-aligned mode

Counter Upcounting Downcounting
TIMx_CNT

TIMx_RCR = 0
UEV

TIMx_RCR = 1
UEV

UEV
TIMx_RCR = 2

TIMx_RCR = 3 UEV

TIMx_RCR = 3
and
UEV
re-synchronization
(by SW) (by SW) (by SW)

UEV Update event: Preload registers transferred to active registers and update interrupt generated
Update Event if the repetition counter underflow occurs when the counter is equal to the auto-reload value.

MSv31195V1

RM0402 Rev 6 427/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

16.3.4 Clock selection

The counter clock can be provided by the following clock sources:
• Internal clock (CK_INT)
• External clock mode1: external input pin
• External clock mode2: external trigger input ETR
• Internal trigger inputs (ITRx): using one timer as prescaler for another timer, for
example, Timer 1 can be configured to act as a prescaler for Timer 2. Refer to Using
one timer as prescaler for another timer for more details.

Internal clock source (CK_INT)

If the slave mode controller is disabled (SMS=000), then the CEN, DIR (in the TIMx_CR1
register) and UG bits (in the TIMx_EGR register) are actual control bits and can be changed
only by software (except UG which remains cleared automatically). As soon as the CEN bit
is written to 1, the prescaler is clocked by the internal clock CK_INT.
Figure 98 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.

Figure 98. Control circuit in normal mode, internal clock divided by 1

Internal clock

CEN=CNT_EN

UG

CNT_INIT

Counter clock = CK_CNT = CK_PSC

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

MS31085V2

External clock source mode 1

This mode is selected when SMS=111 in the TIMx_SMCR register. The counter can count at
each rising or falling edge on a selected input.

428/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 99. TI2 external clock connection example

TIMx_SMCR
TS[2:0]

or TI2F or
TI1F or Encoder
ITRx mode
0xx
TI1_ED
100 TRGI External clock
TI1FP1 mode 1 CK_PSC
TI2F_Rising 101
TI2 Edge 0 TI2FP2 ETRF External clock
Filter 110
detector 1 ETRF mode 2
TI2F_Falling 111
CK_INT Internal clock
mode
ICF[3:0] CC2P (internal clock)

TIMx_CCMR1 TIMx_CCER
ECE SMS[2:0]
TIMx_SMCR

MS31196V1

For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:
1. Configure channel 2 to detect rising edges on the TI2 input by writing CC2S = ‘01’ in
the TIMx_CCMR1 register.
2. Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=0000).
3. Select rising edge polarity by writing CC2P=0 and CC2NP=0 in the TIMx_CCER
register.
4. Configure the timer in external clock mode 1 by writing SMS=111 in the TIMx_SMCR
register.
5. Select TI2 as the trigger input source by writing TS=110 in the TIMx_SMCR register.
6. Enable the counter by writing CEN=1 in the TIMx_CR1 register.
Note: The capture prescaler is not used for triggering, so it does not need to be configured.
When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.

RM0402 Rev 6 429/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 100. Control circuit in external clock mode 1

TI2

CNT_EN

Counter clock = CK_CNT = CK_PSC

Counter register 34 35 36

TIF

Write TIF=0

MS31087V2

External clock source mode 2

This mode is selected by writing ECE=1 in the TIMx_SMCR register.
The counter can count at each rising or falling edge on the external trigger input ETR.
Figure 101 gives an overview of the external trigger input block.

Figure 101. External trigger input block

or TI2F or
TI1F or Encoder
mode

TRGI External clock

ETR mode 1 CK_PSC
0
Divider ETRP
ETR pin Filter ETRF External clock
/1, /2, /4, /8
1 fDTS downcounter mode 2
CK_INT Internal clock
ETP ETPS[1:0] ETF[3:0] mode
(internal clock)
TIMx_SMCR TIMx_SMCR TIMx_SMCR

ECE SMS[2:0]
TIMx_SMCR

MS33116V1

For example, to configure the upcounter to count each 2 rising edges on ETR, use the
following procedure:

430/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

1. As no filter is needed in this example, write ETF[3:0]=0000 in the TIMx_SMCR register.

2. Set the prescaler by writing ETPS[1:0]=01 in the TIMx_SMCR register
3. Select rising edge detection on the ETR pin by writing ETP=0 in the TIMx_SMCR
register
4. Enable external clock mode 2 by writing ECE=1 in the TIMx_SMCR register.
5. Enable the counter by writing CEN=1 in the TIMx_CR1 register.
The counter counts once each 2 ETR rising edges.
The delay between the rising edge on ETR and the actual clock of the counter is due to the
resynchronization circuit on the ETRP signal.

Figure 102. Control circuit in external clock mode 2

f CK_INT

CNT_EN

ETR

ETRP

ETRF

Counter clock =
CK_INT =CK_PSC

Counter register 34 35 36

MS33111V2

16.3.5 Capture/compare channels

Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
Figure 103 to Figure 106 give an overview of one Capture/Compare channel.
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).

RM0402 Rev 6 431/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 103. Capture/compare channel (example: channel 1 input stage)

TI1F_ED
To the slave mode controller

TI1 TI1F_Rising
Filter TI1F Edge 0 TI1FP1
fDTS downcounter TI1F_Falling 01
detector 1
TI2FP1 IC1 Divider IC1PS
10
/1, /2, /4, /8
ICF[3:0] CC1P/CC1NP TRC
11
TIMx_CCMR1 TIMx_CCER (from slave mode
TI2F_Rising controller)
0
(from channel 2)
CC1S[1:0] ICPS[1:0] CC1E
TI2F_Falling
1
(from channel 2) TIMx_CCMR1 TIMx_CCER

MS33115V1

The output stage generates an intermediate waveform which is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.

Figure 104. Capture/compare channel 1 main circuit

APB Bus

MCU-peripheral interface
(if 16-bit)

8 8
high

S write CCR1H
low

Read CCR1H S write_in_progress

read_in_progress write CCR1L
Read CCR1L Capture/compare preload register R
R Output CC1S[1]
capture_transfer compare_transfer mode
CC1S[0]
CC1S[1] Input
mode OC1PE
CC1S[0] Capture/compare shadow register OC1PE
UEV
TIM1_CCMR1
Comparator (from time
IC1PS Capture
base unit)
CC1E CNT>CCR1
Counter CNT=CCR1
CC1G
TIM1_EGR

MS31089V2

432/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 105. Output stage of capture/compare channel (channels 1 to 3)

To the master mode

controller
ETRF 0 Output OC1
enable
‘0’ 1
x0 circuit
OC1REF OC1REFC
01
OC1_DT CC1P
CNT>CCR1 Output 11
Output Dead-time TIM1_CCER
mode
CNT=CCR1 selector generator
controller OC1N_DT
11
10 0
(1) Output OC1N
OCxREF ‘0’ 0x enable
OC5REF 1 circuit

CC1NE CC1E TIM1_CCER

OC1CE OC1M[3:0] DTG[7:0] CC1NE CC1E CC1NP MOE OSSI OSSR

TIM1_CCMR1 TIM1_BDTR TIM1_CCER TIM1_CCER TIM1_BDTR

MS35909V1

Figure 106. Output stage of capture/compare channel (channel 4)

To the master
mode controller
ETR 0
Output
OC4
enable
1 circuit
CNT>CCR4 Output
mode CC4P
CNT>CCR4 OC4REF
controller TIM1_CCER

CC4E TIM1_CCER

MOE OSSI TIM1_BDTR

OC1M[3:0]
TIM1_CCMR2 OIS4 TIM1_CR2

MS37370V1

The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.

RM0402 Rev 6 433/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

16.3.6 Input capture mode

In Input capture mode, the Capture/Compare Registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a
capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be
cleared by software by writing it to ‘0’ or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when it is written with 0.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
• Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to 01 in the TIMx_CCMR1 register. As soon as CC1S becomes different from 00,
the channel is configured in input and the TIMx_CCR1 register becomes read-only.
• Program the appropriate input filter duration in relation with the signal connected to the
timer (by programming ICxF bits in the TIMx_CCMRx register if the input is a TIx input).
Let’s imagine that, when toggling, the input signal is not stable during at must 5 internal
clock cycles. We must program a filter duration longer than these 5 clock cycles. We
can validate a transition on TI1 when 8 consecutive samples with the new level have
been detected (sampled at fDTS frequency). Then write IC1F bits to 0011 in the
TIMx_CCMR1 register.
• Select the edge of the active transition on the TI1 channel by writing CC1P and CC1NP
bits to 0 in the TIMx_CCER register (rising edge in this case).
• Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to ‘00’ in the
TIMx_CCMR1 register).
• Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.
• If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register, and/or the DMA request by setting the CC1DE bit in the
TIMx_DIER register.
When an input capture occurs:
• The TIMx_CCR1 register gets the value of the counter on the active transition.
• CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.
• An interrupt is generated depending on the CC1IE bit.
• A DMA request is generated depending on the CC1DE bit.
In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note: IC interrupt and/or DMA requests can be generated by software by setting the
corresponding CCxG bit in the TIMx_EGR register.

434/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3.7 PWM input mode

This mode is a particular case of input capture mode. The procedure is the same except:
• Two ICx signals are mapped on the same TIx input.
• These 2 ICx signals are active on edges with opposite polarity.
• One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.
For example, one can measure the period (in TIMx_CCR1 register) and the duty cycle (in
TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure (depending
on CK_INT frequency and prescaler value):
• Select the active input for TIMx_CCR1: write the CC1S bits to 01 in the TIMx_CCMR1
register (TI1 selected).
• Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): write the CC1P and CC1NP bits to ‘0’ (active on rising edge).
• Select the active input for TIMx_CCR2: write the CC2S bits to 10 in the TIMx_CCMR1
register (TI1 selected).
• Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): write the CC2P
and CC2NP bits to ‘1’ (active on falling edge).
• Select the valid trigger input: write the TS bits to 101 in the TIMx_SMCR register
(TI1FP1 selected).
• Configure the slave mode controller in reset mode: write the SMS bits to 100 in the
TIMx_SMCR register.
• Enable the captures: write the CC1E and CC2E bits to ‘1’ in the TIMx_CCER register.

Figure 107. PWM input mode timing

TI1

TIMx_CNT 0004 0000 0001 0002 0003 0004 0000

TIMx_CCR1 0004

TIMx_CCR2 0002

IC1 capture IC2 capture IC1 capture

pulse width period
IC2 capture
measurement measurement
reset counter
ai15413

16.3.8 Forced output mode

In output mode (CCxS bits = 00 in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx/OCxN) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.

RM0402 Rev 6 435/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

To force an output compare signal (OCXREF/OCx) to its active level, one just needs to write
101 in the OCxM bits in the corresponding TIMx_CCMRx register. Thus OCXREF is forced
high (OCxREF is always active high) and OCx get opposite value to CCxP polarity bit.
For example: CCxP=0 (OCx active high) => OCx is forced to high level.
The OCxREF signal can be forced low by writing the OCxM bits to 100 in the TIMx_CCMRx
register.
Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt and DMA requests can be sent
accordingly. This is described in the output compare mode section below.

16.3.9 Output compare mode

This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
• Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=000), be set
active (OCxM=001), be set inactive (OCxM=010) or can toggle (OCxM=011) on match.
• Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).
• Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).
• Sends a DMA request if the corresponding enable bit is set (CCxDE bit in the
TIMx_DIER register, CCDS bit in the TIMx_CR2 register for the DMA request
selection).
The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on OCxREF and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One Pulse mode).
Procedure:
1. Select the counter clock (internal, external, prescaler).
2. Write the desired data in the TIMx_ARR and TIMx_CCRx registers.
3. Set the CCxIE bit if an interrupt request is to be generated.
4. Select the output mode. For example:
– Write OCxM = 011 to toggle OCx output pin when CNT matches CCRx
– Write OCxPE = 0 to disable preload register
– Write CCxP = 0 to select active high polarity
– Write CCxE = 1 to enable the output
5. Enable the counter by setting the CEN bit in the TIMx_CR1 register.
The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=’0’, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 108.

436/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 108. Output compare mode, toggle on OC1.

Write B201h in the CC1R register

TIM1_CNT 0039 003A 003B B200 B201

TIM1_CCR1 003A B201

OC1REF= OC1

Match detected on CCR1

Interrupt generated if enabled
MS31092V2

16.3.10 PWM mode

Pulse Width Modulation mode allows to generate a signal with a frequency determined by
the value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing ‘110’ (PWM mode 1) or ‘111’ (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The corresponding preload register must be enabled by setting the
OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register (in
upcounting or center-aligned modes) by setting the ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, all registers must be initialized by setting the UG bit in
the TIMx_EGR register.
OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register. It
can be programmed as active high or active low. OCx output is enabled by a combination of
the CCxE, CCxNE, MOE, OSSI and OSSR bits (TIMx_CCER and TIMx_BDTR registers).
Refer to the TIMx_CCER register description for more details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CCRx ≤ TIMx_CNT or TIMx_CNT ≤ TIMx_CCRx (depending on the direction
of the counter).
The timer is able to generate PWM in edge-aligned mode or center-aligned mode
depending on the CMS bits in the TIMx_CR1 register.

RM0402 Rev 6 437/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

PWM edge-aligned mode

• Upcounting configuration
Upcounting is active when the DIR bit in the TIMx_CR1 register is low. Refer to
Upcounting mode.
In the following example, we consider PWM mode 1. The reference PWM signal
OCxREF is high as long as TIMx_CNT < TIMx_CCRx else it becomes low. If the
compare value in TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR)
then OCxREF is held at ‘1’. If the compare value is 0 then OCxRef is held at ‘0’.
Figure 109 shows some edge-aligned PWM waveforms in an example where
TIMx_ARR=8.

Figure 109. Edge-aligned PWM waveforms (ARR=8)

Counter register 0 1 2 3 4 5 6 7 8 0 1

OCXREF
CCRx=4
CCxIF

OCXREF
CCRx=8
CCxIF

OCXREF ‘1’
CCRx>8
CCxIF

OCXREF ‘0’
CCRx=0
CCxIF

MS31093V1

• Downcounting configuration
Downcounting is active when DIR bit in TIMx_CR1 register is high. Refer to
Downcounting mode.
In PWM mode 1, the reference signal OCxRef is low as long as
TIMx_CNT > TIMx_CCRx else it becomes high. If the compare value in TIMx_CCRx is
greater than the auto-reload value in TIMx_ARR, then OCxREF is held at ‘1’. 0% PWM
is not possible in this mode.

PWM center-aligned mode

Center-aligned mode is active when the CMS bits in TIMx_CR1 register are different from
‘00’ (all the remaining configurations having the same effect on the OCxRef/OCx signals).
The compare flag is set when the counter counts up, when it counts down or both when it
counts up and down depending on the CMS bits configuration. The direction bit (DIR) in the
TIMx_CR1 register is updated by hardware and must not be changed by software. Refer to
Center-aligned mode (up/down counting).

438/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 110 shows some center-aligned PWM waveforms in an example where:

• TIMx_ARR=8,
• PWM mode is the PWM mode 1,
• The flag is set when the counter counts down corresponding to the center-aligned
mode 1 selected for CMS=01 in TIMx_CR1 register.

Figure 110. Center-aligned PWM waveforms (ARR=8)

Counter register 0 1 2 3 4 5 6 7 8 7 6 5 4 3 2 1 0 1

OCxREF
CCRx = 4

CCxIF CMS=01
CMS=10
CMS=11

OCxREF
CCRx=7

CMS=10 or 11
CCxIF
‘1’
OCxREF
CCRx=8

CCxIF CMS=01
CMS=10
CMS=11
‘1’
OCxREF
CCRx>8

CCxIF CMS=01
CMS=10
CMS=11
‘0’
OCxREF
CCRx=0

CCxIF CMS=01
CMS=10
CMS=11

AI14681b

RM0402 Rev 6 439/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Hints on using center-aligned mode:

• When starting in center-aligned mode, the current up-down configuration is used. It
means that the counter counts up or down depending on the value written in the DIR bit
in the TIMx_CR1 register. Moreover, the DIR and CMS bits must not be changed at the
same time by the software.
• Writing to the counter while running in center-aligned mode is not recommended as it
can lead to unexpected results. In particular:
– The direction is not updated if a value greater than the auto-reload value is written
in the counter (TIMx_CNT>TIMx_ARR). For example, if the counter was counting
up, it continues to count up.
– The direction is updated if 0 or the TIMx_ARR value is written in the counter but no
Update Event UEV is generated.
• The safest way to use center-aligned mode is to generate an update by software
(setting the UG bit in the TIMx_EGR register) just before starting the counter and not to
write the counter while it is running.

16.3.11 Complementary outputs and dead-time insertion

The advanced-control timers (TIM1&TIM8) can output two complementary signals and
manage the switching-off and the switching-on instants of the outputs.
This time is generally known as dead-time and it has to be adjusted depending on the
devices that are connected to the outputs and their characteristics (intrinsic delays of level-
shifters, delays due to power switches...)
The polarity of the outputs (main output OCx or complementary OCxN) can be selected
independently for each output. This is done by writing to the CCxP and CCxNP bits in the
TIMx_CCER register.
The complementary signals OCx and OCxN are activated by a combination of several
control bits: the CCxE and CCxNE bits in the TIMx_CCER register and the MOE, OISx,
OISxN, OSSI and OSSR bits in the TIMx_BDTR and TIMx_CR2 registers. Refer to Table 97
for more details. In particular, the dead-time is activated when switching to the IDLE state
(MOE falling down to 0).
Dead-time insertion is enabled by setting both CCxE and CCxNE bits, and the MOE bit if the
break circuit is present. DTG[7:0] bits of the TIMx_BDTR register are used to control the
dead-time generation for all channels. From a reference waveform OCxREF, it generates 2
outputs OCx and OCxN. If OCx and OCxN are active high:
• The OCx output signal is the same as the reference signal except for the rising edge,
which is delayed relative to the reference rising edge.
• The OCxN output signal is the opposite of the reference signal except for the rising
edge, which is delayed relative to the reference falling edge.
If the delay is greater than the width of the active output (OCx or OCxN) then the
corresponding pulse is not generated.
The following figures show the relationships between the output signals of the dead-time
generator and the reference signal OCxREF. (we suppose CCxP=0, CCxNP=0, MOE=1,
CCxE=1 and CCxNE=1 in these examples).

440/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Figure 111. Complementary output with dead-time insertion.

OCxREF

OCx
delay
OCxN
delay

MS31095V1

Figure 112. Dead-time waveforms with delay greater than the negative pulse.

OCxREF

OCx

delay
OCxN

MS31096V1

Figure 113. Dead-time waveforms with delay greater than the positive pulse.

OCxREF

OCx

OCxN

delay

MS31097V1

The dead-time delay is the same for each of the channels and is programmable with the
DTG bits in the TIMx_BDTR register. Refer to Section 16.4.18: TIM1&TIM8 break and dead-
time register (TIMx_BDTR) for delay calculation.

Re-directing OCxREF to OCx or OCxN

In output mode (forced, output compare or PWM), OCxREF can be re-directed to the OCx
output or to OCxN output by configuring the CCxE and CCxNE bits in the TIMx_CCER
register.
This allows to send a specific waveform (such as PWM or static active level) on one output
while the complementary remains at its inactive level. Other alternative possibilities are to

RM0402 Rev 6 441/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

have both outputs at inactive level or both outputs active and complementary with dead-
time.
Note: When only OCxN is enabled (CCxE=0, CCxNE=1), it is not complemented and becomes
active as soon as OCxREF is high. For example, if CCxNP=0 then OCxN=OCxRef. On the
other hand, when both OCx and OCxN are enabled (CCxE=CCxNE=1) OCx becomes
active when OCxREF is high whereas OCxN is complemented and becomes active when
OCxREF is low.

16.3.12 Using the break function

When using the break function, the output enable signals and inactive levels are modified
according to additional control bits (MOE, OSSI and OSSR bits in the TIMx_BDTR register,
OISx and OISxN bits in the TIMx_CR2 register). In any case, the OCx and OCxN outputs
cannot be set both to active level at a given time. Refer to Table 97 for more details.
The break source can be either the break input pin or a clock failure event, generated by the
Clock Security System (CSS), from the Reset Clock Controller. For further information on
the Clock Security System, refer to Section 6.2.7: Clock security system (CSS).
When exiting from reset, the break circuit is disabled and the MOE bit is low. The break
function can be enabled by setting the BKE bit in the TIMx_BDTR register. The break input
polarity can be selected by configuring the BKP bit in the same register. BKE and BKP can
be modified at the same time. When the BKE and BKP bits are written, a delay of 1 APB
clock cycle is applied before the writing is effective. Consequently, it is necessary to wait 1
APB clock period to correctly read back the bit after the write operation.
Because MOE falling edge can be asynchronous, a resynchronization circuit has been
inserted between the actual signal (acting on the outputs) and the synchronous control bit
(accessed in the TIMx_BDTR register). It results in some delays between the asynchronous
and the synchronous signals. In particular, when writing MOE to 1 whereas it was low, user
must insert a delay (dummy instruction) before reading it correctly. This is because user
writes the asynchronous signal and reads the synchronous signal.
When a break occurs (selected level on the break input):
• The MOE bit is cleared asynchronously, putting the outputs in inactive state, idle state
or in reset state (selected by the OSSI bit). This feature functions even if the MCU
oscillator is off.
• Each output channel is driven with the level programmed in the OISx bit in the
TIMx_CR2 register as soon as MOE=0. If OSSI=0 then the timer releases the enable
output else the enable output remains high.
• When complementary outputs are used:
– The outputs are first put in reset state inactive state (depending on the polarity).
This is done asynchronously so that it works even if no clock is provided to the
timer.
– If the timer clock is still present, then the dead-time generator is reactivated in
order to drive the outputs with the level programmed in the OISx and OISxN bits
after a dead-time. Even in this case, OCx and OCxN cannot be driven to their

442/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

active level together. Note that because of the resynchronization on MOE, the
dead-time duration is a bit longer than usual (around 2 ck_tim clock cycles).
– If OSSI=0 then the timer releases the enable outputs else the enable outputs
remain or become high as soon as one of the CCxE or CCxNE bits is high.
• The break status flag (BIF bit in the TIMx_SR register) is set. An interrupt can be
generated if the BIE bit in the TIMx_DIER register is set. A DMA request can be sent if
the BDE bit in the TIMx_DIER register is set.
• If the AOE bit in the TIMx_BDTR register is set, the MOE bit is automatically set again
at the next update event UEV. This can be used to perform a regulation, for instance.
Else, MOE remains low until it is written with 1 again. In this case, it can be used for
security and the break input can be connected to an alarm from power drivers, thermal
sensors or any security components.
Note: The break inputs is acting on level. Thus, the MOE cannot be set while the break input is
active (neither automatically nor by software). In the meantime, the status flag BIF cannot
be cleared.
The break can be generated by the BRK input which has a programmable polarity and an
enable bit BKE in the TIMx_BDTR Register.
There are two solutions to generate a break:
• By using the BRK input which has a programmable polarity and an enable bit BKE in
the TIMx_BDTR register
• By software through the BG bit of the TIMx_EGR register.
In addition to the break input and the output management, a write protection has been
implemented inside the break circuit to safeguard the application. It allows to freeze the
configuration of several parameters (dead-time duration, OCx/OCxN polarities and state
when disabled, OCxM configurations, break enable and polarity). The protection can be
selected among 3 levels with the LOCK bits in the TIMx_BDTR register. Refer to
Section 16.4.18: TIM1&TIM8 break and dead-time register (TIMx_BDTR). The LOCK bits
can be written only once after an MCU reset.
Figure 114 shows an example of behavior of the outputs in response to a break.

RM0402 Rev 6 443/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 114. Output behavior in response to a break.

BREAK (MOE )

OCxREF

OCx
(OCxN not implemented, CCxP=0, OISx=1)

OCx
(OCxN not implemented, CCxP=0, OISx=0)

OCx
(OCxN not implemented, CCxP=1, OISx=1)

OCx
(OCxN not implemented, CCxP=1, OISx=0)

OCx

OCxN delay delay delay

(CCxE=1, CCxP=0, OISx=0, CCxNE=1, CCxNP=0, OISxN=1)

OCx

OCxN delay delay delay

(CCxE=1, CCxP=0, OISx=1, CCxNE=1, CCxNP=1, OISxN=1)

OCx

OCxN delay
(CCxE=1, CCxP=0, OISx=0, CCxNE=0, CCxNP=0, OISxN=1)

OCx

OCxN delay
(CCxE=1, CCxP=0, OISx=1, CCxNE=0, CCxNP=0, OISxN=0)

OCx

OCxN
(CCxE=1, CCxP=0, CCxNE=0, CCxNP=0, OISx=OISxN=0 or OISx=OISxN=1)

MS31098V1

444/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3.13 Clearing the OCxREF signal on an external event

The OCxREF signal for a given channel can be driven Low by applying a High level to the
ETRF input (OCxCE enable bit of the corresponding TIMx_CCMRx register set to ‘1’). The
OCxREF signal remains Low until the next update event, UEV, occurs.
This function can only be used in output compare and PWM modes, and does not work in
forced mode.
For example, the ETR signal can be connected to the output of a comparator to be used for
current handling. In this case, the ETR must be configured as follow:
1. The External Trigger Prescaler should be kept off: bits ETPS[1:0] of the TIMx_SMCR
register set to ‘00’.
2. The external clock mode 2 must be disabled: bit ECE of the TIMx_SMCR register set to
‘0’.
3. The External Trigger Polarity (ETP) and the External Trigger Filter (ETF) can be
configured according to the user needs.
Figure 115 shows the behavior of the OCxREF signal when the ETRF Input becomes High,
for both values of the enable bit OCxCE. In this example, the timer TIMx is programmed in
PWM mode.

Figure 115. Clearing TIMx OCxREF

(CCRx)
Counter (CNT)

ETRF

OCxREF (OCxCE = ‘0’)

OCxREF (OCxCE = ‘1’)

ETRF becomes high ETRF still high

MSv35889V1

Note: In case of a PWM with a 100% duty cycle (if CCRx>ARR), then OCxREF is enabled again at
the next counter overflow.

RM0402 Rev 6 445/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

16.3.14 6-step PWM generation

When complementary outputs are used on a channel, preload bits are available on the
OCxM, CCxE and CCxNE bits. The preload bits are transferred to the shadow bits at the
COM commutation event. Thus one can program in advance the configuration for the next
step and change the configuration of all the channels at the same time. COM can be
generated by software by setting the COM bit in the TIMx_EGR register or by hardware (on
TRGI rising edge).
A flag is set when the COM event occurs (COMIF bit in the TIMx_SR register), which can
generate an interrupt (if the COMIE bit is set in the TIMx_DIER register) or a DMA request
(if the COMDE bit is set in the TIMx_DIER register).
Figure 116 describes the behavior of the OCx and OCxN outputs when a COM event
occurs, in 3 different examples of programmed configurations.

Figure 116. 6-step generation, COM example (OSSR=1)

counter (CNT) (CCRx)

OCxREF
Write COM to 1

COM event
CCxE=1 write OCxM to 100 CCxE=1
CCxNE=0 CCxNE=0
OCxM=100 (forced inactive) OCxM=100
OCx
Example 1
OCxN

Write CCxNE to 1
and OCxM to 101 CCxE=0
CCxE=1 CCxNE=1
CCxNE=0 OCxM=101
OCx OCxM=100 (forced inactive)
Example 2
OCxN
write CCxNE to 0
CCxE=1
and OCxM to 100 CCxE=1
CCxNE=0 CCxNE=0
OCxM=100 (forced inactive) OCxM=100
OCx
Example 3
OCxN

ai14910

446/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3.15 One-pulse mode

One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. One-pulse mode is selected
by setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically
at the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be:
• In upcounting: CNT < CCRx ≤ ARR (in particular, 0 < CCRx)
• In downcounting: CNT > CCRx

Figure 117. Example of one pulse mode.

TI2

OC1REF
OC1

TIM1_ARR
Counter

TIM1_CCR1

0 t
tDELAY tPULSE
MS31099V2

For example one may want to generate a positive pulse on OC1 with a length of tPULSE and
after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.
Let’s use TI2FP2 as trigger 1:
• Map TI2FP2 to TI2 by writing CC2S=’01’ in the TIMx_CCMR1 register.
• TI2FP2 must detect a rising edge, write CC2P=’0’ and CC2NP=’0’ in the TIMx_CCER
register.
• Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=’110’ in
the TIMx_SMCR register.
• TI2FP2 is used to start the counter by writing SMS to ‘110’ in the TIMx_SMCR register
(trigger mode).

RM0402 Rev 6 447/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
• The tDELAY is defined by the value written in the TIMx_CCR1 register.
• The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR1).
• Let’s say one want to build a waveform with a transition from ‘0’ to ‘1’ when a compare
match occurs and a transition from ‘1’ to ‘0’ when the counter reaches the auto-reload
value. To do this PWM mode 2 must be enabled by writing OC1M=111 in the
TIMx_CCMR1 register. Optionally the preload registers can be enabled by writing
OC1PE=’1’ in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case one has to write the compare value in the TIMx_CCR1 register, the auto-reload
value in the TIMx_ARR register, generate an update by setting the UG bit and wait for
external trigger event on TI2. CC1P is written to ‘0’ in this example.
In our example, the DIR and CMS bits in the TIMx_CR1 register should be low.
Since only 1 pulse (Single mode) is needed, a 1 must be written in the OPM bit in the
TIMx_CR1 register to stop the counter at the next update event (when the counter rolls over
from the auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0',
so the Repetitive Mode is selected.
Particular case: OCx fast enable:
In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
If one wants to output a waveform with the minimum delay, the OCxFE bit can be set in the
TIMx_CCMRx register. Then OCxRef (and OCx) are forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.

16.3.16 Encoder interface mode

To select Encoder Interface mode write SMS=‘001’ in the TIMx_SMCR register if the
counter is counting on TI2 edges only, SMS=’010’ if it is counting on TI1 edges only and
SMS=’011’ if it is counting on both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the CC1P and CC2P bits in the TIMx_CCER
register. When needed, the input filter can be programmed as well. CC1NP and CC2NP
must be kept low.
The two inputs TI1 and TI2 are used to interface to an incremental encoder. Refer to
Table 95. The counter is clocked by each valid transition on TI1FP1 or TI2FP2 (TI1 and TI2
after input filter and polarity selection, TI1FP1=TI1 if not filtered and not inverted,
TI2FP2=TI2 if not filtered and not inverted) assuming that it is enabled (CEN bit in
TIMx_CR1 register written to ‘1’). The sequence of transitions of the two inputs is evaluated
and generates count pulses as well as the direction signal. Depending on the sequence the
counter counts up or down, the DIR bit in the TIMx_CR1 register is modified by hardware
accordingly. The DIR bit is calculated at each transition on any input (TI1 or TI2), whatever
the counter is counting on TI1 only, TI2 only or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This
means that the counter just counts continuously between 0 and the auto-reload value in the
TIMx_ARR register (0 to ARR or ARR down to 0 depending on the direction). So the

448/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

TIMx_ARR must be configured before starting. In the same way, the capture, compare,
prescaler, repetition counter, trigger output features continue to work as normal. Encoder
mode and External clock mode 2 are not compatible and must not be selected together.
In this mode, the counter is modified automatically following the speed and the direction of
the incremental encoder and its content, therefore, always represents the encoder’s
position. The count direction correspond to the rotation direction of the connected sensor.
Table 95 summarizes the possible combinations, assuming TI1 and TI2 do not switch at the
same time.

Table 95. Counting direction versus encoder signals

Level on opposite signal TI1FP1 signal TI2FP2 signal
Active edge (TI1FP1 for TI2,
TI2FP2 for TI1) Rising Falling Rising Falling

Counting on High Down Up No Count No Count

TI1 only Low Up Down No Count No Count

Counting on High No Count No Count Up Down

TI2 only Low No Count No Count Down Up

Counting on High Down Up Up Down

TI1 and TI2 Low Up Down Down Up

An external incremental encoder can be connected directly to the MCU without external
interface logic. However, comparators are normally be used to convert the encoder’s
differential outputs to digital signals. This greatly increases noise immunity. The third
encoder output which indicate the mechanical zero position, may be connected to an
external interrupt input and trigger a counter reset.
Figure 118 gives an example of counter operation, showing count signal generation and
direction control. It also shows how input jitter is compensated where both edges are
selected. This might occur if the sensor is positioned near to one of the switching points. For
this example we assume that the configuration is the following:
• CC1S=’01’ (TIMx_CCMR1 register, TI1FP1 mapped on TI1).
• CC2S=’01’ (TIMx_CCMR2 register, TI1FP2 mapped on TI2).
• CC1P=’0’, CC1NP=’0’, and IC1F = ‘0000’ (TIMx_CCER register, TI1FP1 non-inverted,
TI1FP1=TI1).
• CC2P=’0’, CC2NP=’0’, and IC2F = ‘0000’ (TIMx_CCER register, TI1FP2 non-inverted,
TI1FP2= TI2).
• SMS=’011’ (TIMx_SMCR register, both inputs are active on both rising and falling
edges).
• CEN=’1’ (TIMx_CR1 register, Counter enabled).

RM0402 Rev 6 449/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 118. Example of counter operation in encoder interface mode.

forward jitter backward jitter forward

TI1

TI2

Counter

up down up

MS33107V1

Figure 119 gives an example of counter behavior when TI1FP1 polarity is inverted (same
configuration as above except CC1P=’1’).

Figure 119. Example of encoder interface mode with TI1FP1 polarity inverted.
forward jitter backward jitter forward

TI1

TI2

Counter

down up down

MS33108V1

The timer, when configured in Encoder Interface mode provides information on the sensor’s
current position. Dynamic information can be obtained (speed, acceleration, deceleration)
by measuring the period between two encoder events using a second timer configured in
capture mode. The output of the encoder which indicates the mechanical zero can be used
for this purpose. Depending on the time between two events, the counter can also be read
at regular times. This can be done by latching the counter value into a third input capture
register if available (then the capture signal must be periodic and can be generated by
another timer). when available, it is also possible to read its value through a DMA request
generated by a real-time clock.

450/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3.17 Timer input XOR function

The TI1S bit in the TIMx_CR2 register allows the input filter of channel 1 to be connected to
the output of a XOR gate, combining the three input pins TIMx_CH1, TIMx_CH2 and
TIMx_CH3.
The XOR output can be used with all the timer input functions such as trigger or input
capture. An example of this feature used to interface Hall sensors is given in
Section 16.3.18 below.

16.3.18 Interfacing with Hall sensors

This is done using the advanced-control timers (TIM1 or TIM8) to generate PWM signals to
drive the motor and another timer TIMx (TIM2, TIM3, TIM4 or TIM5) referred to as
“interfacing timer” in Figure 120. The “interfacing timer” captures the 3 timer input pins
(TIMx_CH1, TIMx_CH2, and TIMx_CH3) connected through a XOR to the TI1 input channel
(selected by setting the TI1S bit in the TIMx_CR2 register).
The slave mode controller is configured in reset mode; the slave input is TI1F_ED. Thus,
each time one of the 3 inputs toggles, the counter restarts counting from 0. This creates a
time base triggered by any change on the Hall inputs.
On the “interfacing timer”, capture/compare channel 1 is configured in capture mode,
capture signal is TRC (see Figure 103). The captured value, which corresponds to the time
elapsed between 2 changes on the inputs, gives information about motor speed.
The “interfacing timer” can be used in output mode to generate a pulse which changes the
configuration of the channels of the advanced-control timer (TIM1 or TIM8) (by triggering a
COM event). The TIM1 timer is used to generate PWM signals to drive the motor. To do this,
the interfacing timer channel must be programmed so that a positive pulse is generated
after a programmed delay (in output compare or PWM mode). This pulse is sent to the
advanced-control timer (TIM1 or TIM8) through the TRGO output.
Example: one wants to change the PWM configuration of the advanced-control timer TIM1
after a programmed delay each time a change occurs on the Hall inputs connected to one of
the TIMx timers.
• Configure 3 timer inputs ORed to the TI1 input channel by writing the TI1S bit in the
TIMx_CR2 register to ‘1’,
• Program the time base: write the TIMx_ARR to the max value (the counter must be
cleared by the TI1 change. Set the prescaler to get a maximum counter period longer
than the time between 2 changes on the sensors,
• Program channel 1 in capture mode (TRC selected): write the CC1S bits in the
TIMx_CCMR1 register to ‘11’. The digital filter can also be programmed if needed,
• Program channel 2 in PWM 2 mode with the desired delay: write the OC2M bits to ‘111’
and the CC2S bits to ‘00’ in the TIMx_CCMR1 register,
• Select OC2REF as trigger output on TRGO: write the MMS bits in the TIMx_CR2
register to ‘101’,
In the advanced-control timer TIM1, the right ITR input must be selected as trigger input, the
timer is programmed to generate PWM signals, the capture/compare control signals are
preloaded (CCPC=1 in the TIMx_CR2 register) and the COM event is controlled by the
trigger input (CCUS=1 in the TIMx_CR2 register). The PWM control bits (CCxE, OCxM) are
written after a COM event for the next step (this can be done in an interrupt subroutine
generated by the rising edge of OC2REF).

RM0402 Rev 6 451/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Figure 120 describes this example.

Figure 120. Example of Hall sensor interface

TIH1

TIH2
Interfacing timer

TIH3

Counter (CNT)
(CCR2)

CCR1 C7A3 C7A8 C794 C7A5 C7AB C796

TRGO=OC2REF

COM

OC1

OC1N
Advanced-control timers

OC2

OC2N

OC3

OC3N

Write CCxE, CCxNE

and OCxM for next step
ai17335b

452/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.3.19 TIMx and external trigger synchronization

The TIMx timer can be synchronized with an external trigger in several modes: Reset mode,
Gated mode and Trigger mode.

Slave mode: Reset mode

The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
• Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC1S bits
select the input capture source only, CC1S = 01 in the TIMx_CCMR1 register. Write
CC1P=0 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and detect
rising edges only).
• Configure the timer in reset mode by writing SMS=100 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
• Start the counter by writing CEN=1 in the TIMx_CR1 register.
The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request, or a DMA
request can be sent if enabled (depending on the TIE and TDE bits in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
resynchronization circuit on TI1 input.

Figure 121. Control circuit in reset mode

TI1

UG

Counter clock = ck_cnt = ck_psc

Counter register 30 31 32 33 34 35 36 00 01 02 03 00 01 02 03

TIF

MS31401V1

RM0402 Rev 6 453/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Slave mode: Gated mode

The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
• Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC1S bits
select the input capture source only, CC1S=01 in TIMx_CCMR1 register. Write
CC1P=1 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and detect
low level only).
• Configure the timer in gated mode by writing SMS=101 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
• Enable the counter by writing CEN=1 in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=0, whatever is the trigger input level).
The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
resynchronization circuit on TI1 input.

Figure 122. Control circuit in gated mode

TI1

CNT_EN

Counter clock = ck_cnt = ck_psc

Counter register 30 31 32 33 34 35 36 37 38

TIF

Write TIF=0
MS31402V3

454/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Slave mode: Trigger mode

The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
• Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC2F=0000). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC2S bits
are configured to select the input capture source only, CC2S=01 in TIMx_CCMR1
register. Write CC2P=1 and CC2NP=0 in TIMx_CCER register to validate the polarity
(and detect low level only).
• Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI2 as the input source by writing TS=110 in TIMx_SMCR register.
When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.
The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on TI2 input.

Figure 123. Control circuit in trigger mode

TI2

CNT_EN

Counter clock = ck_cnt = ck_psc

Counter register 34 35 36 37 38

TIF
MS31403V2

Slave mode: external clock mode 2 + trigger mode

The external clock mode 2 can be used in addition to another slave mode (except external
clock mode 1 and encoder mode). In this case, the ETR signal is used as external clock
input, and another input can be selected as trigger input (in reset mode, gated mode or
trigger mode). It is recommended not to select ETR as TRGI through the TS bits of
TIMx_SMCR register.
In the following example, the upcounter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
1. Configure the external trigger input circuit by programming the TIMx_SMCR register as
follows:
– ETF = 0000: no filter
– ETPS = 00: prescaler disabled
– ETP = 0: detection of rising edges on ETR and ECE=1 to enable the external clock
mode 2.

RM0402 Rev 6 455/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

2. Configure the channel 1 as follows, to detect rising edges on TI:

– IC1F=0000: no filter.
– The capture prescaler is not used for triggering and does not need to be
configured.
– CC1S=01 in TIMx_CCMR1 register to select only the input capture source
– CC1P=0 and CC1NP=’0’ in TIMx_CCER register to validate the polarity (and
detect rising edge only).
3. Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
A rising edge on TI1 enables the counter and sets the TIF flag. The counter then counts on
ETR rising edges.
The delay between the rising edge of the ETR signal and the actual reset of the counter is
due to the resynchronization circuit on ETRP input.

Figure 124. Control circuit in external clock mode 2 + trigger mode

TI1

CEN/CNT_EN

ETR

Counter clock = CK_CNT = CK_PSC

Counter register 34 35 36

TIF

MS33110V1

16.3.20 Timer synchronization

The TIM timers are linked together internally for timer synchronization or chaining. Refer to
Section 17.3.15: Timer synchronization on page 516 for details.

16.3.21 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the TIMx
counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBG module. For more details, refer to Section 30.16.2: Debug support
for timers, watchdog, bxCAN and I2C.

456/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.4 TIM1&TIM8 registers

Refer to Section 1.2: List of abbreviations for registers for a list of abbreviations used in
register descriptions.
The peripheral registers must be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-word (16 bits) or words (32 bits).

16.4.1 TIM1&TIM8 control register 1 (TIMx_CR1)

Address offset: 0x00
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. CKD[1:0] ARPE CMS[1:0] DIR OPM URS UDIS CEN
rw rw rw rw rw rw rw rw rw rw

Bits 15:10 Reserved, must be kept at reset value.

Bits 9:8 CKD[1:0]: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and the
dead-time and sampling clock (tDTS)used by the dead-time generators and the digital filters
(ETR, TIx),
00: tDTS=tCK_INT
01: tDTS=2*tCK_INT
10: tDTS=4*tCK_INT
11: Reserved, do not program this value
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:5 CMS[1:0]: Center-aligned mode selection
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(DIR).
01: Center-aligned mode 1. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting down.
10: Center-aligned mode 2. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting up.
11: Center-aligned mode 3. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
both when the counter is counting up or down.
Note: It is not allowed to switch from edge-aligned mode to center-aligned mode as long as
the counter is enabled (CEN=1)
Bit 4 DIR: Direction
0: Counter used as upcounter
1: Counter used as downcounter
Note: This bit is read only when the timer is configured in Center-aligned mode or Encoder
mode.

RM0402 Rev 6 457/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bit 3 OPM: One pulse mode

0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the bit CEN)
Bit 2 URS: Update request source
This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generate an update interrupt or DMA request if enabled.
These events can be:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC, CCRx). However the counter and the prescaler are reinitialized if the UG bit is
set or if a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: External clock, gated mode and encoder mode can work only if the CEN bit has been
previously set by software. However trigger mode can set the CEN bit automatically by
hardware.

16.4.2 TIM1&TIM8 control register 2 (TIMx_CR2)

Address offset: 0x04
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. OIS4 OIS3N OIS3 OIS2N OIS2 OIS1N OIS1 TI1S MMS[2:0] CCDS CCUS Res. CCPC
rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 Reserved, must be kept at reset value.

Bit 14 OIS4: Output Idle state 4 (OC4 output)
refer to OIS1 bit
Bit 13 OIS3N: Output Idle state 3 (OC3N output)
refer to OIS1N bit
Bit 12 OIS3: Output Idle state 3 (OC3 output)
refer to OIS1 bit

458/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bit 11 OIS2N: Output Idle state 2 (OC2N output)

refer to OIS1N bit
Bit 10 OIS2: Output Idle state 2 (OC2 output)
refer to OIS1 bit
Bit 9 OIS1N: Output Idle state 1 (OC1N output)
0: OC1N=0 after a dead-time when MOE=0
1: OC1N=1 after a dead-time when MOE=0
Note: This bit can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).
Bit 8 OIS1: Output Idle state 1 (OC1 output)
0: OC1=0 (after a dead-time if OC1N is implemented) when MOE=0
1: OC1=1 (after a dead-time if OC1N is implemented) when MOE=0
Note: This bit can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).
Bit 7 TI1S: TI1 selection
0: The TIMx_CH1 pin is connected to TI1 input
1: The TIMx_CH1, CH2 and CH3 pins are connected to the TI1 input (XOR combination)
Bits 6:4 MMS[2:0]: Master mode selection
These bits allow to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as trigger output (TRGO). If the
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter Enable signal CNT_EN is used as trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer is
enable. The Counter Enable signal is generated by a logic OR between CEN control bit and
the trigger input when configured in gated mode. When the Counter Enable signal is
controlled by the trigger input, there is a delay on TRGO, except if the master/slave mode is
selected (see the MSM bit description in TIMx_SMCR register).
010: Update - The update event is selected as trigger output (TRGO). For instance a master
timer can then be used as a prescaler for a slave timer.
011: Compare Pulse - The trigger output send a positive pulse when the CC1IF flag is to be
set (even if it was already high), as soon as a capture or a compare match occurred.
(TRGO).
100: Compare - OC1REF signal is used as trigger output (TRGO)
101: Compare - OC2REF signal is used as trigger output (TRGO)
110: Compare - OC3REF signal is used as trigger output (TRGO)
111: Compare - OC4REF signal is used as trigger output (TRGO)
Bit 3 CCDS: Capture/compare DMA selection
0: CCx DMA request sent when CCx event occurs
1: CCx DMA requests sent when update event occurs
Bit 2 CCUS: Capture/compare control update selection
0: When capture/compare control bits are preloaded (CCPC=1), they are updated by setting
the COMG bit only
1: When capture/compare control bits are preloaded (CCPC=1), they are updated by setting
the COMG bit or when an rising edge occurs on TRGI
Note: This bit acts only on channels that have a complementary output.

RM0402 Rev 6 459/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bit 1 Reserved, must be kept at reset value.

Bit 0 CCPC: Capture/compare preloaded control
0: CCxE, CCxNE and OCxM bits are not preloaded
1: CCxE, CCxNE and OCxM bits are preloaded, after having been written, they are updated
only when a commutation event (COM) occurs (COMG bit set or rising edge detected on
TRGI, depending on the CCUS bit).
Note: This bit acts only on channels that have a complementary output.

16.4.3 TIM1&TIM8 slave mode control register (TIMx_SMCR)

Address offset: 0x08
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ETP ECE ETPS[1:0] ETF[3:0] MSM TS[2:0] Res. SMS[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 ETP: External trigger polarity

This bit selects whether ETR or ETR is used for trigger operations
0: ETR is non-inverted, active at high level or rising edge.
1: ETR is inverted, active at low level or falling edge.
Bit 14 ECE: External clock enable
This bit enables External clock mode 2.
0: External clock mode 2 disabled
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
Note: 1: Setting the ECE bit has the same effect as selecting external clock mode 1 with
TRGI connected to ETRF (SMS=111 and TS=111).
2: It is possible to simultaneously use external clock mode 2 with the following slave
modes: reset mode, gated mode and trigger mode. Nevertheless, TRGI must not be
connected to ETRF in this case (TS bits must not be 111).
3: If external clock mode 1 and external clock mode 2 are enabled at the same time,
the external clock input is ETRF.
Bits 13:12 ETPS[1:0]: External trigger prescaler
External trigger signal ETRP frequency must be at most 1/4 of TIMxCLK frequency. A
prescaler can be enabled to reduce ETRP frequency. It is useful when inputting fast external
clocks.
00: Prescaler OFF
01: ETRP frequency divided by 2
10: ETRP frequency divided by 4
11: ETRP frequency divided by 8

460/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bits 11:8 ETF[3:0]: External trigger filter

This bit-field then defines the frequency used to sample ETRP signal and the length of the
digital filter applied to ETRP. The digital filter is made of an event counter in which N
consecutive events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8

Bit 7 MSM: Master/slave mode

0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful if we
want to synchronize several timers on a single external event.
Bits 6:4 TS[2:0]: Trigger selection
This bit-field selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
001: Internal Trigger 1 (ITR1)
010: Internal Trigger 2 (ITR2)
011: Internal Trigger 3 (ITR3)
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
110: Filtered Timer Input 2 (TI2FP2)
111: External Trigger input (ETRF)
See Table 62: TIMx internal trigger connection for more details on ITRx meaning for each
Timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=000) to
avoid wrong edge detections at the transition.
Bit 3 Reserved, must be kept at reset value.

RM0402 Rev 6 461/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bits 2:0 SMS: Slave mode selection

When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input Control register and Control Register
description.
000: Slave mode disabled - if CEN = ‘1’ then the prescaler is clocked directly by the internal
clock.
001: Encoder mode 1 - Counter counts up/down on TI2FP2 edge depending on TI1FP1
level.
010: Encoder mode 2 - Counter counts up/down on TI1FP1 edge depending on TI2FP2
level.
011: Encoder mode 3 - Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset Mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers.
101: Gated Mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both start and stop of
the counter are controlled.
110: Trigger Mode - The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled.
111: External Clock Mode 1 - Rising edges of the selected trigger (TRGI) clock the counter.
Note: The gated mode must not be used if TI1F_ED is selected as the trigger input
(TS=’100’). Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the
gated mode checks the level of the trigger signal.

Table 96. TIMx Internal trigger connection

Slave TIM ITR0 (TS = 000) ITR1 (TS = 001) ITR2 (TS = 010) ITR3 (TS = 011)

TIM1 TIM5 TIM2 TIM3 TIM4

TIM8 TIM1 TIM2 TIM4 TIM5

16.4.4 TIM1&TIM8 DMA/interrupt enable register (TIMx_DIER)

Address offset: 0x0C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. TDE COMDE CC4DE CC3DE CC2DE CC1DE UDE BIE TIE COMIE CC4IE CC3IE CC2IE CC1IE UIE
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 Reserved, must be kept at reset value.

Bit 14 TDE: Trigger DMA request enable
0: Trigger DMA request disabled
1: Trigger DMA request enabled
Bit 13 COMDE: COM DMA request enable
0: COM DMA request disabled
1: COM DMA request enabled
Bit 12 CC4DE: Capture/Compare 4 DMA request enable
0: CC4 DMA request disabled
1: CC4 DMA request enabled

462/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bit 11 CC3DE: Capture/Compare 3 DMA request enable

0: CC3 DMA request disabled
1: CC3 DMA request enabled
Bit 10 CC2DE: Capture/Compare 2 DMA request enable
0: CC2 DMA request disabled
1: CC2 DMA request enabled
Bit 9 CC1DE: Capture/Compare 1 DMA request enable
0: CC1 DMA request disabled
1: CC1 DMA request enabled
Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled
1: Update DMA request enabled
Bit 7 BIE: Break interrupt enable
0: Break interrupt disabled
1: Break interrupt enabled
Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled
1: Trigger interrupt enabled
Bit 5 COMIE: COM interrupt enable
0: COM interrupt disabled
1: COM interrupt enabled
Bit 4 CC4IE: Capture/Compare 4 interrupt enable
0: CC4 interrupt disabled
1: CC4 interrupt enabled
Bit 3 CC3IE: Capture/Compare 3 interrupt enable
0: CC3 interrupt disabled
1: CC3 interrupt enabled
Bit 2 CC2IE: Capture/Compare 2 interrupt enable
0: CC2 interrupt disabled
1: CC2 interrupt enabled
Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled
1: CC1 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled

RM0402 Rev 6 463/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

16.4.5 TIM1&TIM8 status register (TIMx_SR)

Address offset: 0x10
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. CC4OF CC3OF CC2OF CC1OF Res. BIF TIF COMIF CC4IF CC3IF CC2IF CC1IF UIF
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0

Bits 15:13 Reserved, must be kept at reset value.

Bit 12 CC4OF: Capture/Compare 4 overcapture flag
refer to CC1OF description
Bit 11 CC3OF: Capture/Compare 3 overcapture flag
refer to CC1OF description
Bit 10 CC2OF: Capture/Compare 2 overcapture flag
refer to CC1OF description
Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bit 8 Reserved, must be kept at reset value.
Bit 7 BIF: Break interrupt flag
This flag is set by hardware as soon as the break input goes active. It can be cleared by
software if the break input is not active.
0: No break event occurred.
1: An active level has been detected on the break input.
Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred.
1: Trigger interrupt pending.
Bit 5 COMIF: COM interrupt flag
This flag is set by hardware on COM event (when Capture/compare Control bits - CCxE,
CCxNE, OCxM - have been updated). It is cleared by software.
0: No COM event occurred.
1: COM interrupt pending.
Bit 4 CC4IF: Capture/Compare 4 interrupt flag
refer to CC1IF description
Bit 3 CC3IF: Capture/Compare 3 interrupt flag
refer to CC1IF description

464/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bit 2 CC2IF: Capture/Compare 2 interrupt flag

refer to CC1IF description
Bit 1 CC1IF: Capture/Compare 1 interrupt flag
If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value, with some
exception in center-aligned mode (refer to the CMS bits in the TIMx_CR1 register
description). It is cleared by software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF
bit goes high on the counter overflow (in upcounting and up/down-counting modes) or
underflow (in downcounting mode)
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred
1: The counter value has been captured in TIMx_CCR1 register (An edge has been
detected on IC1 which matches the selected polarity)
Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
–At overflow or underflow regarding the repetition counter value (update if repetition
counter = 0) and if the UDIS=0 in the TIMx_CR1 register.
–When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=0
and UDIS=0 in the TIMx_CR1 register.
–When CNT is reinitialized by a trigger event (refer to Section 16.4.3: TIM1&TIM8 slave
mode control register (TIMx_SMCR)), if URS=0 and UDIS=0 in the TIMx_CR1 register.

16.4.6 TIM1&TIM8 event generation register (TIMx_EGR)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. BG TG COMG CC4G CC3G CC2G CC1G UG
w w w w w w w w

Bits 15:8 Reserved, must be kept at reset value.

Bit 7 BG: Break generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A break event is generated. MOE bit is cleared and BIF flag is set. Related interrupt or
DMA transfer can occur if enabled.
Bit 6 TG: Trigger generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: The TIF flag is set in TIMx_SR register. Related interrupt or DMA transfer can occur if
enabled.

RM0402 Rev 6 465/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bit 5 COMG: Capture/Compare control update generation

This bit can be set by software, it is automatically cleared by hardware
0: No action
1: When CCPC bit is set, it allows to update CCxE, CCxNE and OCxM bits
Note: This bit acts only on channels having a complementary output.
Bit 4 CC4G: Capture/Compare 4 generation
refer to CC1G description
Bit 3 CC3G: Capture/Compare 3 generation
refer to CC1G description
Bit 2 CC2G: Capture/Compare 2 generation
refer to CC1G description
Bit 1 CC1G: Capture/Compare 1 generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
CC1IF flag is set, Corresponding interrupt or DMA request is sent if enabled.
If channel CC1 is configured as input:
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt or DMA request is sent if enabled. The CC1OF flag is set if the
CC1IF flag was already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Reinitialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared if
the center-aligned mode is selected or if DIR=0 (upcounting), else it takes the auto-reload
value (TIMx_ARR) if DIR=1 (downcounting).

466/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.4.7 TIM1&TIM8 capture/compare mode register 1 (TIMx_CCMR1)

Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. So one must take care that the same bit
can have a different meaning for the input stage and for the output stage.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC2 OC2 OC2 OC1 OC1 OC1
OC2M[2:0] OC1M[2:0]
CE PE FE CC2S[1:0] CE PE FE CC1S[1:0]
IC2F[3:0] IC2PSC[1:0] IC1F[3:0] IC1PSC[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Output compare mode:

Bit 15 OC2CE: Output Compare 2 clear enable

Bits 14:12 OC2M[2:0]: Output Compare 2 mode
Bit 11 OC2PE: Output Compare 2 preload enable
Bit 10 OC2FE: Output Compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = ‘0’ in TIMx_CCER).
Bit 7 OC1CE: Output Compare 1 clear enable
OC1CE: Output Compare 1 Clear Enable
0: OC1Ref is not affected by the ETRF Input
1: OC1Ref is cleared as soon as a High level is detected on ETRF input

RM0402 Rev 6 467/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bits 6:4 OC1M: Output Compare 1 mode

These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas OC1 and OC1N active level depends on
CC1P and CC1NP bits.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNT<TIMx_CCR1
else inactive. In downcounting, channel 1 is inactive (OC1REF=‘0’) as long as
TIMx_CNT>TIMx_CCR1 else active (OC1REF=’1’).
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNT<TIMx_CCR1
else active. In downcounting, channel 1 is active as long as TIMx_CNT>TIMx_CCR1 else
inactive.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=’00’ (the channel is configured in
output).
2: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode to
“PWM” mode.
3: On channels having a complementary output, this bit field is preloaded. If the CCPC
bit is set in the TIMx_CR2 register then the OC1M active bits take the new value from
the preloaded bits only when a COM event is generated.
Bit 3 OC1PE: Output Compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the new
value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=’00’ (the channel is configured in
output).
2: The PWM mode can be used without validating the preload register only in one pulse
mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output Compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is 5
clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. Then, OC is
set to the compare level independently from the result of the comparison. Delay to sample
the trigger input and to activate CC1 output is reduced to 3 clock cycles. OCFE acts only if
the channel is configured in PWM1 or PWM2 mode.

468/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bits 1:0 CC1S: Capture/Compare 1 selection

This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = ‘0’ in TIMx_CCER).

Input capture mode

Bits 15:12 IC2F: Input capture 2 filter

Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
Bits 9:8 CC2S: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if an
internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = ‘0’ in TIMx_CCER).
Bits 7:4 IC1F[3:0]: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter applied
to TI1. The digital filter is made of an event counter in which N consecutive events are needed to
validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8

RM0402 Rev 6 469/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bits 3:2 IC1PSC: Input capture 1 prescaler

This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 Selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if an
internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = ‘0’ in TIMx_CCER).

16.4.8 TIM1&TIM8 capture/compare mode register 2 (TIMx_CCMR2)

Address offset: 0x1C
Reset value: 0x0000
Refer to the above CCMR1 register description.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC4 OC4 OC4 OC3 OC3 OC3
OC4M[2:0] OC3M[2:0]
CE PE FE CC4S[1:0] CE. PE FE CC3S[1:0]
IC4F[3:0] IC4PSC[1:0] IC3F[3:0] IC3PSC[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Output compare mode

Bit 15 OC4CE: Output compare 4 clear enable

Bits 14:12 OC4M: Output compare 4 mode
Bit 11 OC4PE: Output compare 4 preload enable
Bit 10 OC4FE: Output compare 4 fast enable
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = ‘0’ in TIMx_CCER).
Bit 7 OC3CE: Output compare 3 clear enable
Bits 6:4 OC3M: Output compare 3 mode

470/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bit 3 OC3PE: Output compare 3 preload enable

Bit 2 OC3FE: Output compare 3 fast enable
Bits 1:0 CC3S: Capture/Compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = ‘0’ in TIMx_CCER).

Input capture mode

Bits 15:12 IC4F: Input capture 4 filter

Bits 11:10 IC4PSC: Input capture 4 prescaler
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = ‘0’ in TIMx_CCER).
Bits 7:4 IC3F: Input capture 3 filter
Bits 3:2 IC3PSC: Input capture 3 prescaler
Bits 1:0 CC3S: Capture/compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = ‘0’ in TIMx_CCER).

16.4.9 TIM1&TIM8 capture/compare enable register (TIMx_CCER)

Address offset: 0x20
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. CC4P CC4E CC3NP CC3NE CC3P CC3E CC2NP CC2NE CC2P CC2E CC1NP CC1NE CC1P CC1E
rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:14 Reserved, must be kept at reset value.

Bit 13 CC4P: Capture/Compare 4 output polarity
refer to CC1P description

RM0402 Rev 6 471/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bit 12 CC4E: Capture/Compare 4 output enable

refer to CC1E description
Bit 11 CC3NP: Capture/Compare 3 complementary output polarity
refer to CC1NP description
Bit 10 CC3NE: Capture/Compare 3 complementary output enable
refer to CC1NE description
Bit 9 CC3P: Capture/Compare 3 output polarity
refer to CC1P description
Bit 8 CC3E: Capture/Compare 3 output enable
refer to CC1E description
Bit 7 CC2NP: Capture/Compare 2 complementary output polarity
refer to CC1NP description
Bit 6 CC2NE: Capture/Compare 2 complementary output enable
refer to CC1NE description
Bit 5 CC2P: Capture/Compare 2 output polarity
refer to CC1P description
Bit 4 CC2E: Capture/Compare 2 output enable
refer to CC1E description
Bit 3 CC1NP: Capture/Compare 1 complementary output polarity
CC1 channel configured as output:
0: OC1N active high.
1: OC1N active low.
CC1 channel configured as input:
This bit is used in conjunction with CC1P to define the polarity of TI1FP1 and TI2FP1. Refer
to CC1P description.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1NP active bit takes the new value from the
preloaded bit only when a Commutation event is generated.
Note: This bit is not writable as soon as LOCK level 2 or 3 has been programmed (LOCK bits
in TIMx_BDTR register) and CC1S=”00” (the channel is configured in output).
Bit 2 CC1NE: Capture/Compare 1 complementary output enable
0: Off - OC1N is not active. OC1N level is then function of MOE, OSSI, OSSR, OIS1, OIS1N
and CC1E bits.
1: On - OC1N signal is output on the corresponding output pin depending on MOE, OSSI,
OSSR, OIS1, OIS1N and CC1E bits.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1NE active bit takes the new value from the
preloaded bit only when a Commutation event is generated.

472/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bit 1 CC1P: Capture/Compare 1 output polarity

CC1 channel configured as output:
0: OC1 active high
1: OC1 active low
CC1 channel configured as input:
CC1NP/CC1P bits select the active polarity of TI1FP1 and TI2FP1 for trigger or capture
operations.
00: non-inverted/rising edge
The circuit is sensitive to TIxFP1 rising edge (capture or trigger operations in reset, external
clock or trigger mode), TIxFP1 is not inverted (trigger operation in gated mode or encoder
mode).
01: inverted/falling edge
The circuit is sensitive to TIxFP1 falling edge (capture or trigger operations in reset, external
clock or trigger mode), TIxFP1 is inverted (trigger operation in gated mode or encoder
mode).
10: reserved, do not use this configuration.
11: non-inverted/both edges
The circuit is sensitive to both TIxFP1 rising and falling edges (capture or trigger operations
in reset, external clock or trigger mode), TIxFP1 is not inverted (trigger operation in gated
mode). This configuration must not be used in encoder mode.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1P active bit takes the new value from the
preloaded bit only when a Commutation event is generated.
Note: This bit is not writable as soon as LOCK level 2 or 3 has been programmed (LOCK bits
in TIMx_BDTR register).
Bit 0 CC1E: Capture/Compare 1 output enable
CC1 channel configured as output:
0: Off - OC1 is not active. OC1 level is then function of MOE, OSSI, OSSR, OIS1, OIS1N
and CC1NE bits.
1: On - OC1 signal is output on the corresponding output pin depending on MOE, OSSI,
OSSR, OIS1, OIS1N and CC1NE bits.
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled.
1: Capture enabled.
Note: On channels having a complementary output, this bit is preloaded. If the CCPC bit is
set in the TIMx_CR2 register then the CC1E active bit takes the new value from the
preloaded bit only when a Commutation event is generated.

RM0402 Rev 6 473/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Table 97. Output control bits for complementary OCx and OCxN channels
with break feature
Control bits Output states(1)
MOE OSSI OSSR CCxE CCxNE
OCx output state OCxN output state
bit bit bit bit bit
Output Disabled (not driven by
Output Disabled (not driven by the timer)
0 the timer)
OCxN=0, OCxN_EN=0
OCx=0, OCx_EN=0
0
Output Disabled (not driven by
OCxREF + Polarity OCxN=OCxREF xor
1 the timer)
CCxNP, OCxN_EN=1
OCx=0, OCx_EN=0
0 OCxREF + Polarity
Output Disabled (not driven by the timer)
0 OCx=OCxREF xor CCxP,
OCxN=0, OCxN_EN=0
OCx_EN=1
1 Complementary to OCREF (not
OCREF + Polarity + dead-time OCREF)
1
OCx_EN=1 + Polarity + dead-time
1 X OCxN_EN=1
Output Disabled (not driven by
Output Disabled (not driven by the timer)
0 the timer)
OCxN=CCxNP, OCxN_EN=0
OCx=CCxP, OCx_EN=0
0
Off-State (output enabled with OCxREF + Polarity
1 inactive state) OCxN=OCxREF xor CCxNP,
OCx=CCxP, OCx_EN=1 OCxN_EN=1
1
OCxREF + Polarity Off-State (output enabled with inactive
0 OCx=OCxREF xor CCxP, state)
OCx_EN=1 OCxN=CCxNP, OCxN_EN=1
1
Complementary to OCREF (not
OCREF + Polarity + dead-time
1 OCREF) + Polarity + dead-time
OCx_EN=1
OCxN_EN=1
Output Disabled (not driven by
Output Disabled (not driven by the timer)
0 the timer)
0 OCxN=CCxNP, OCxN_EN=0
OCx=CCxP, OCx_EN=0

0 1 Output Disabled (not driven by the timer)

0 Asynchronously: OCx=CCxP, OCx_EN=0, OCxN=CCxNP, OCxN_EN=0
Then if the clock is present: OCx=OISx and OCxN=OISxN after a dead-
1
1 time, assuming that OISx and OISxN do not correspond to OCX and
OCxN both in active state.
0 X
Output Disabled (not driven by
Output Disabled (not driven by the timer)
0 the timer)
0 OCxN=CCxNP, OCxN_EN=0
OCx=CCxP, OCx_EN=0

1 1 Off-State (output enabled with inactive state)

0 Asynchronously: OCx=CCxP, OCx_EN=1, OCxN=CCxNP, OCxN_EN=1
Then if the clock is present: OCx=OISx and OCxN=OISxN after a dead-
1
1 time, assuming that OISx and OISxN do not correspond to OCX and
OCxN both in active state
1. When both outputs of a channel are not used (CCxE = CCxNE = 0), the OISx, OISxN, CCxP and CCxNP bits must be kept
cleared.

474/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Note: The state of the external I/O pins connected to the complementary OCx and OCxN channels
depends on the OCx and OCxN channel state and the GPIO registers.

16.4.10 TIM1&TIM8 counter (TIMx_CNT)

Address offset: 0x24
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CNT[15:0]: Counter value

16.4.11 TIM1&TIM8 prescaler (TIMx_PSC)

Address offset: 0x28
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 PSC[15:0]: Prescaler value

The counter clock frequency (CK_CNT) is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

16.4.12 TIM1&TIM8 auto-reload register (TIMx_ARR)

Address offset: 0x2C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 ARR[15:0]: Auto-reload value

ARR is the value to be loaded in the actual auto-reload register.
Refer to Section 16.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.

RM0402 Rev 6 475/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

16.4.13 TIM1&TIM8 repetition counter register (TIMx_RCR)

Address offset: 0x30
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. REP[7:0]
rw rw rw rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bits 7:0 REP[7:0]: Repetition counter value
These bits allow the user to set-up the update rate of the compare registers (i.e. periodic
transfers from preload to active registers) when preload registers are enable, as well as the
update interrupt generation rate, if this interrupt is enable.
Each time the REP_CNT related downcounter reaches zero, an update event is generated
and it restarts counting from REP value. As REP_CNT is reloaded with REP value only at
the repetition update event U_RC, any write to the TIMx_RCR register is not taken in
account until the next repetition update event.
It means in PWM mode (REP+1) corresponds to:
– the number of PWM periods in edge-aligned mode
– the number of half PWM period in center-aligned mode.

16.4.14 TIM1&TIM8 capture/compare register 1 (TIMx_CCR1)

Address offset: 0x34
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR1[15:0]: Capture/Compare 1 value

If channel CC1 is configured as output:
CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(bit OC1PE). Else the preload value is copied in the active capture/compare 1 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.

If channel CC1 is configured as input:

CCR1 is the counter value transferred by the last input capture 1 event (IC1).

476/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.4.15 TIM1&TIM8 capture/compare register 2 (TIMx_CCR2)

Address offset: 0x38
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR2[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR2[15:0]: Capture/Compare 2 value

If channel CC2 is configured as output:
CCR2 is the value to be loaded in the actual capture/compare 2 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR2 register
(bit OC2PE). Else the preload value is copied in the active capture/compare 2 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC2 output.
If channel CC2 is configured as input:
CCR2 is the counter value transferred by the last input capture 2 event (IC2).

16.4.16 TIM1&TIM8 capture/compare register 3 (TIMx_CCR3)

Address offset: 0x3C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR3[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR3[15:0]: Capture/Compare value

If channel CC3 is configured as output:
CCR3 is the value to be loaded in the actual capture/compare 3 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR3 register
(bit OC3PE). Else the preload value is copied in the active capture/compare 3 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC3 output.

If channel CC3 is configured as input:

CCR3 is the counter value transferred by the last input capture 3 event (IC3).

RM0402 Rev 6 477/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

16.4.17 TIM1&TIM8 capture/compare register 4 (TIMx_CCR4)

Address offset: 0x40
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR4[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR4[15:0]: Capture/Compare value

If channel CC4 is configured as output:
CCR4 is the value to be loaded in the actual capture/compare 4 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR4 register
(bit OC4PE). Else the preload value is copied in the active capture/compare 4 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC4 output.
If channel CC4 is configured as input:
CCR4 is the counter value transferred by the last input capture 4 event (IC4).

16.4.18 TIM1&TIM8 break and dead-time register (TIMx_BDTR)

Address offset: 0x44
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MOE AOE BKP BKE OSSR OSSI LOCK[1:0] DTG[7:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Note: As the bits AOE, BKP, BKE, OSSI, OSSR and DTG[7:0] can be write-locked depending on
the LOCK configuration, it can be necessary to configure all of them during the first write
access to the TIMx_BDTR register.

Bit 15 MOE: Main output enable

This bit is cleared asynchronously by hardware as soon as the break input is active. It is set
by software or automatically depending on the AOE bit. It is acting only on the channels
which are configured in output.
0: OC and OCN outputs are disabled or forced to idle state.
1: OC and OCN outputs are enabled if their respective enable bits are set (CCxE, CCxNE in
TIMx_CCER register).
See OC/OCN enable description for more details (Section 16.4.9: TIM1&TIM8
capture/compare enable register (TIMx_CCER) on page 471).
Bit 14 AOE: Automatic output enable
0: MOE can be set only by software
1: MOE can be set by software or automatically at the next update event (if the break input is
not be active)
Note: This bit can not be modified as long as LOCK level 1 has been programmed (LOCK bits
in TIMx_BDTR register).

478/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

Bit 13 BKP: Break polarity

0: Break input BRK is active low
1: Break input BRK is active high
Note: This bit can not be modified as long as LOCK level 1 has been programmed (LOCK bits
in TIMx_BDTR register).
Note: Any write operation to this bit takes a delay of 1 APB clock cycle to become effective.
Bit 12 BKE: Break enable
0: Break inputs (BRK and CSS clock failure event) disabled
1; Break inputs (BRK and CSS clock failure event) enabled
Note: This bit cannot be modified when LOCK level 1 has been programmed (LOCK bits in
TIMx_BDTR register).
Note: Any write operation to this bit takes a delay of 1 APB clock cycle to become effective.
Bit 11 OSSR: Off-state selection for Run mode
This bit is used when MOE=1 on channels having a complementary output which are
configured as outputs. OSSR is not implemented if no complementary output is implemented
in the timer.
See OC/OCN enable description for more details (Section 16.4.9: TIM1&TIM8
capture/compare enable register (TIMx_CCER) on page 471).
0: When inactive, OC/OCN outputs are disabled (OC/OCN enable output signal=0).
1: When inactive, OC/OCN outputs are enabled with their inactive level as soon as CCxE=1
or CCxNE=1. Then, OC/OCN enable output signal=1
Note: This bit can not be modified as soon as the LOCK level 2 has been programmed (LOCK
bits in TIMx_BDTR register).
Bit 10 OSSI: Off-state selection for Idle mode
This bit is used when MOE=0 on channels configured as outputs.
See OC/OCN enable description for more details (Section 16.4.9: TIM1&TIM8
capture/compare enable register (TIMx_CCER) on page 471).
0: When inactive, OC/OCN outputs are disabled (OC/OCN enable output signal=0).
1: When inactive, OC/OCN outputs are forced first with their idle level as soon as CCxE=1 or
CCxNE=1. OC/OCN enable output signal=1)
Note: This bit can not be modified as soon as the LOCK level 2 has been programmed (LOCK
bits in TIMx_BDTR register).
Bits 9:8 LOCK[1:0]: Lock configuration
These bits offer a write protection against software errors.
00: LOCK OFF - No bit is write protected.
01: LOCK Level 1 = DTG bits in TIMx_BDTR register, OISx and OISxN bits in TIMx_CR2
register and BKE/BKP/AOE bits in TIMx_BDTR register can no longer be written.
10: LOCK Level 2 = LOCK Level 1 + CC Polarity bits (CCxP/CCxNP bits in TIMx_CCER
register, as long as the related channel is configured in output through the CCxS bits) as well
as OSSR and OSSI bits can no longer be written.
11: LOCK Level 3 = LOCK Level 2 + CC Control bits (OCxM and OCxPE bits in
TIMx_CCMRx registers, as long as the related channel is configured in output through the
CCxS bits) can no longer be written.
Note: The LOCK bits can be written only once after the reset. Once the TIMx_BDTR register
has been written, their content is frozen until the next reset.

RM0402 Rev 6 479/1163

483
Advanced-control timers (TIM1&TIM8) RM0402

Bits 7:0 DTG[7:0]: Dead-time generator setup

This bit-field defines the duration of the dead-time inserted between the complementary
outputs. DT correspond to this duration.
DTG[7:5]=0xx => DT=DTG[7:0]x tdtg with tdtg=tDTS.
DTG[7:5]=10x => DT=(64+DTG[5:0])xtdtg with Tdtg=2xtDTS.
DTG[7:5]=110 => DT=(32+DTG[4:0])xtdtg with Tdtg=8xtDTS.
DTG[7:5]=111 => DT=(32+DTG[4:0])xtdtg with Tdtg=16xtDTS.
Example if TDTS=125ns (8MHz), dead-time possible values are:
0 to 15875 ns by 125 ns steps,
16 us to 31750 ns by 250 ns steps,
32 us to 63us by 1 us steps,
64 us to 126 us by 2 us steps
Note: This bit-field can not be modified as long as LOCK level 1, 2 or 3 has been programmed
(LOCK bits in TIMx_BDTR register).

16.4.19 TIM1&TIM8 DMA control register (TIMx_DCR)

Address offset: 0x48
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. DBL[4:0] Res. Res. Res. DBA[4:0]
rw rw rw rw rw rw rw rw rw rw

Bits 15:13 Reserved, must be kept at reset value.

Bits 12:8 DBL[4:0]: DMA burst length
This 5-bit vector defines the number of DMA transfers (the timer detects a burst transfer
when a read or a write access to the TIMx_DMAR register address is performed).
the TIMx_DMAR address)
00000: 1 transfer
00001: 2 transfers
00010: 3 transfers
...
10001: 18 transfers
Bits 7:5 Reserved, must be kept at reset value.
Bits 4:0 DBA[4:0]: DMA base address
This 5-bits vector defines the base-address for DMA transfers (when read/write access are
done through the TIMx_DMAR address). DBA is defined as an offset starting from the
address of the TIMx_CR1 register.
Example:
00000: TIMx_CR1,
00001: TIMx_CR2,
00010: TIMx_SMCR,
...
Example: Let us consider the following transfer: DBL = 7 transfers and DBA = TIMx_CR1. In
this case the transfer is done to/from 7 registers starting from the TIMx_CR1 address.

480/1163 RM0402 Rev 6

RM0402 Advanced-control timers (TIM1&TIM8)

16.4.20 TIM1&TIM8 DMA address for full transfer (TIMx_DMAR)

Address offset: 0x4C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMAB[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 DMAB[15:0]: DMA register for burst accesses

A read or write operation to the DMAR register accesses the register located at the address
(TIMx_CR1 address) + (DBA + DMA index) x 4
where TIMx_CR1 address is the address of the control register 1, DBA is the DMA base
address configured in TIMx_DCR register, DMA index is automatically controlled by the
DMA transfer, and ranges from 0 to DBL (DBL configured in TIMx_DCR).

Example of how to use the DMA burst feature

In this example the timer DMA burst feature is used to update the contents of the CCRx
registers (x = 2, 3, 4) with the DMA transferring half words into the CCRx registers.
This is done in the following steps:
1. Configure the corresponding DMA channel as follows:
– DMA channel peripheral address is the DMAR register address
– DMA channel memory address is the address of the buffer in the RAM containing
the data to be transferred by DMA into CCRx registers.
– Number of data to transfer = 3 (See note below).
– Circular mode disabled.
2. Configure the DCR register by configuring the DBA and DBL bit fields as follows:
DBL = 3 transfers, DBA = 0xE.
3. Enable the TIMx update DMA request (set the UDE bit in the DIER register).
4. Enable TIMx
5. Enable the DMA channel
Note: This example is for the case where every CCRx register to be updated once. If every CCRx
register is to be updated twice for example, the number of data to transfer should be 6. Let's
take the example of a buffer in the RAM containing data1, data2, data3, data4, data5 and
data6. The data is transferred to the CCRx registers as follows: on the first update DMA
request, data1 is transferred to CCR2, data2 is transferred to CCR3, data3 is transferred to
CCR4 and on the second update DMA request, data4 is transferred to CCR2, data5 is
transferred to CCR3 and data6 is transferred to CCR4.

RM0402 Rev 6 481/1163

483
 0x30
0x28
0x24
0x20
0x18
0x14
0x10
0x08
0x04
0x00

0x2C
0x1C
0x0C
Offset

482/1163
16.4.21

mode
mode
mode
mode
TIMx_SR
TIMx_CR2
TIMx_CR1

TIMx_PSC
TIMx_CNT

TIMx_RCR
TIMx_ARR
TIMx_EGR
Register

TIMx_DIER

Reset value
Reset value
Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

TIMx_CCER
TIMx_SMCR

Input Capture
Input Capture

TIMx_CCMR2
TIMx_CCMR2
TIMx_CCMR1
TIMx_CCMR1

Output Compare
Output Compare
below:

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 28
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 27
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 25
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 24
Advanced-control timers (TIM1&TIM8)

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 23
TIM1&TIM8 register map

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 21
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 18
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

RM0402 Rev 6
17
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
0
0
0
0
0

Res. Res. O24CE OC2CE Res. Res. Res. ETP Res. Res. 15

0
0
0
0
0
0
0
0
0
0

Res. Res. Res. Res. TDE ECE OIS4 Res. 14

0
0
0
0
0
0
0
0
0
0
0

Res. CC4P Res. Res. COMDE OIS3N Res. 13

[2:0]
[2:0]

IC4F[3:0]
IC2F[3:0]

OC4M
OC2M
[1:0]

0
0
0
0
0
0
0
0
0
0
0
0

Res. CC4E Res. CC4OF CC4DE OIS3 Res.

ETPS

12

0
0
0
0
0
0
0
0
0
0
0
0

Res. CC3NP OC4PE OC2PE Res. CC3OF CC3DE OIS2N Res. 11

IC4
IC2

[1:0]
[1:0]

PSC
PSC

0
0
0
0
0
0
0
0
0
0
0
0

Res. CC3NE OC4FE OC2FE Res. CC2OF CC2DE OIS2 Res. 10

Table 98. TIM1&TIM8 register map and reset values

0
0
0
0
0
0
0
0
0
0
0
0
0

Res. CC3P Res. CC1OF CC1DE OIS1N 9

ETF[3:0]

[1:0]
[1:0]
[1:0]
[1:0]
[1:0]
CKD

0
0
0
0
0
0
0
0
0
0
0
0

Res. CC3E Res. Res. UDE OIS1

CC4S
CC4S
CC2S
CC2S

0
0
0
0
0
0
0
0

0
0
0
0
0
0
0

CC2NP OC3CE OC1CE BG BIF BIE MSM TI1S ARPE 7

PSC[15:0]
CNT[15:0]

ARR[15:0]

0
0
0
0
0
0
0

0
0
0
0
0
0
0
0

CC2NE TG TIF TIE 6

[1:0]
CMS

0
0
0
0
0
0
0
0
0
0
0
0

0
0
0

CC2P COMG COMIF COMIE 5

[2:0]
[2:0]

IC3F[3:0]
IC1F[3:0]

OC3M
OC1M
TS[2:0]

0
0
0
0
0
0
0
0
0
0
0

0
0
0
0

CC2E CC4G CC4IF CC4IE DIR

MMS[2:0]

0
0
0
0
0
0
0
0

0
0
0
0

0
0

CC1NP OC3PE OC1PE CC3G CC3IF CC3IE Res. CCDS OPM 3

REP[7:0]
IC3
IC1

[1:0]
[1:0]

PSC
PSC

0
0
0
0
0
0
0
0
0
0
0

0
0
0
0

CC1NE OC3FE OC1FE CC2G CC2IF CC2IE CCUS URS 2

0
0
0
0
0
0
0
0
0

0
0
0
0
0

CC1P CC1G CC1IF CC1IE Res. UDIS 1

[1:0]
[1:0]
[1:0]
[1:0]

0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
TIM1&TIM8 registers are mapped as 16-bit addressable registers as described in the table

CC1E UG UIF UIE CCPC CEN

SMS[2:0]

CC3S
CC3S
CC1S
CC1S

0
RM0402
RM0402 Advanced-control timers (TIM1&TIM8)

Table 98. TIM1&TIM8 register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
TIMx_CCR1 CCR1[15:0]
0x34
Reset value Res. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIMx_CCR2 CCR2[15:0]
0x38
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIMx_CCR3 CCR3[15:0]
0x3C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIMx_CCR4 CCR4[15:0]
0x40
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

OSSR
LOCK

OSSI
MOE
AOE
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

BKP
BKE
TIMx_BDTR DT[7:0]
0x44 [1:0]
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
TIMx_DCR DBL[4:0] DBA[4:0]
0x48
Reset value 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

TIMx_DMAR DMAB[15:0]
0x4C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

RM0402 Rev 6 483/1163

483
General-purpose timers (TIM2 to TIM5) RM0402

17 General-purpose timers (TIM2 to TIM5)

17.1 TIM2 to TIM5 introduction

The general-purpose timers consist of a 16-bit or 32-bit auto-reload counter driven by a
programmable prescaler.
They may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare and PWM).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The timers are completely independent, and do not share any resources. They can be
synchronized together as described in Section 17.3.15.

17.2 TIM2 to TIM5 main features

General-purpose TIMx timer features include:
• 16-bit (TIM3 and TIM4) or 32-bit (TIM2 and TIM5) up, down, up/down auto-reload
counter.
• 16-bit programmable prescaler used to divide (also “on the fly”) the counter clock
frequency by any factor between 1 and 65536.
• Up to 4 independent channels for:
– Input capture
– Output compare
– PWM generation (Edge- and Center-aligned modes)
– One-pulse mode output
• Synchronization circuit to control the timer with external signals and to interconnect
several timers.
• Interrupt/DMA generation on the following events:
– Update: counter overflow/underflow, counter initialization (by software or
internal/external trigger)
– Trigger event (counter start, stop, initialization or count by internal/external trigger)
– Input capture
– Output compare
• Supports incremental (quadrature) encoder and hall-sensor circuitry for positioning
purposes
• Trigger input for external clock or cycle-by-cycle current management

484/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 125. General-purpose timer block diagram

Internal clock (CK_INT)

TIMxCLK from RCC Trigger
ETRF controller
ETR Polarity selection & edge ETRP TRGO
TIMx_ETR Input filter
detector & prescaler to other timers
to DAC/ADC
ITR0 TGI
ITR1 ITR
Slave
ITR2 TRC TRGI controller Reset, enable, up, count
ITR3
mode
TI1F_ED

TI1FP1 Encoder
TI2FP2 interface

U
Auto-reload register UI
Stop, clear or up/down
U
CK_PSC PSC CK_CNT +/- CNT counter
prescaler
CC1I U CC1I
XOR TI1FP1 OC1REF
TI1 Input filter & IC1 IC1PS Output OC1 TIMx_CH1
TI1FP2 Prescaler Capture/Compare 1 register
edge detector control
TIMx_CH1 TRC
CC2I
U CC2I
TI2FP1
TI2 Input filter & IC2 Output OC2
TIMx_CH2 TI2FP2 Prescaler
IC2PS Capture/Compare 2 register OC2REF TIMx_CH2
edge detector control
TRC
CC3I CC3I
U
TI3FP3 OC3REF
TI3 Input filter & IC3 IC3PS Output OC3 TIMx_CH3
TI3FP4 Prescaler Capture/Compare 3 register
TIMx_CH3 edge detector control
TRC CC4I
U CC4I
TI4FP3
TI4 Input filter & IC4 Output OC4
TIMx_CH4 TI4FP4 Prescaler
IC4PS Capture/Compare 4 register OC4REF TIMx_CH4
edge detector control
TRC

ETRF

Notes:
Preload registers transferred
Reg
to active registers on U event
according to control bit

Event

Interrupt & DMA output

MS19673V1

17.3 TIM2 to TIM5 functional description

17.3.1 Time-base unit

The main block of the programmable timer is a 16-bit/32-bit counter with its related auto-
reload register. The counter can count up. The counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.

RM0402 Rev 6 485/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

The time-base unit includes:

• Counter Register (TIMx_CNT)
• Prescaler Register (TIMx_PSC):
• Auto-Reload Register (TIMx_ARR)
The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow (or underflow when downcounting) and if the UDIS bit equals 0 in the
TIMx_CR1 register. It can also be generated by software. The generation of the update
event is described in detail for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the actual counter enable signal CNT_EN is set 1 clock cycle after CEN.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit/32-bit register (in the TIMx_PSC
register). It can be changed on the fly as this control register is buffered. The new prescaler
ratio is taken into account at the next update event.
Figure 126 and Figure 127 give some examples of the counter behavior when the prescaler
ratio is changed on the fly:

Figure 126. Counter timing diagram with prescaler division change from 1 to 2

CK_PSC

CNT_EN

Timerclock = CK_CNT
Counter register F7 F8 F9 FA FB FC 00 01 02 03

Update event (UEV)

Prescaler control register 0 1

Write a new value in TIMx_PSC

Prescaler buffer 0 1

Prescaler counter 0 0 1 0 1 0 1 0 1

MS35833V1

486/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 127. Counter timing diagram with prescaler division change from 1 to 4

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register F7 F8 F9 FA FB FC 00 01

Update event (UEV)

Prescaler control register 0 3

Write a new value in TIMx_PSC

Prescaler buffer 0 3

Prescaler counter 0 0 1 2 3 0 1 2 3

MS35834V1

17.3.2 Counter modes

Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
An Update event can be generated at each counter overflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller).
The UEV event can be disabled by software by setting the UDIS bit in TIMx_CR1 register.
This is to avoid updating the shadow registers while writing new values in the preload
registers. Then no update event occurs until the UDIS bit has been written to 0. However,
the counter restarts from 0, as well as the counter of the prescaler (but the prescale rate
does not change). In addition, if the URS bit (update request selection) in TIMx_CR1
register is set, setting the UG bit generates an update event UEV but without setting the UIF
flag (thus no interrupt or DMA request is sent). This is to avoid generating both update and
capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register)
• The auto-reload shadow register is updated with the preload value (TIMx_ARR)
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.

RM0402 Rev 6 487/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 128. Counter timing diagram, internal clock divided by 1

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS35836V1

Figure 129. Counter timing diagram, internal clock divided by 2

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 0034 0035 0036 0000 0001 0002 0003

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS35835V1

Figure 130. Counter timing diagram, internal clock divided by 4

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 0035 0036 0000 0001

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MSv37301V1

488/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 131. Counter timing diagram, internal clock divided by N

CK_INT

Timerclock = CK_CNT

Counter register 1F 20 00

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MSv37302V1

Figure 132. Counter timing diagram, Update event when ARPE=0 (TIMx_ARR not
preloaded)

CK_INT

CNT_EN

Timerclock = CK_CNT
Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload register FF 36

Write a new value in TIMx_ARR

MSv37303V1

RM0402 Rev 6 489/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 133. Counter timing diagram, Update event when ARPE=1 (TIMx_ARR
preloaded)

CK_PSC

CNT_EN
Timerclock = CK_CNT

Counter register F0 F1 F2 F3 F4 F5 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register F5 36

Auto-reload shadow register F5 36

Write a new value in TIMx_ARR

MSv37304V1

Downcounting mode
In downcounting mode, the counter counts from the auto-reload value (content of the
TIMx_ARR register) down to 0, then restarts from the auto-reload value and generates a
counter underflow event.
An Update event can be generate at each counter underflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller)
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until UDIS bit has been written to 0.
However, the counter restarts from the current auto-reload value, whereas the counter of the
prescaler restarts from 0 (but the prescale rate doesn’t change).
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupts when
clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).
• The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that the auto-reload is updated before the counter is
reloaded, so that the next period is the expected one.
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.

490/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 134. Counter timing diagram, internal clock divided by 1

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 05 04 03 02 01 00 36 35 34 33 32 31 30 2F

Counter underflow (cnt_udf)

Update event (UEV)

Update interrupt flag (UIF)

MSv37305V1

Figure 135. Counter timing diagram, internal clock divided by 2

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 0002 0001 0000 0036 0035 0034 0033

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

MSv37306V1

Figure 136. Counter timing diagram, internal clock divided by 4

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 0001 0000 0000 0001

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

MSv37307V1

RM0402 Rev 6 491/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 137. Counter timing diagram, internal clock divided by N

CK_INT

Timerclock = CK_CNT

Counter register 20 1F 00 36

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS37340V1

Figure 138. Counter timing diagram, Update event

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 05 04 03 02 01 00 36 35 34 33 32 31 30 2F

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FF 36

Write a new value in TIMx_ARR

MS37341V1

Center-aligned mode (up/down counting)

In center-aligned mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register) – 1, generates a counter overflow event, then counts from the auto-
reload value down to 1 and generates a counter underflow event. Then it restarts counting
from 0.
Center-aligned mode is active when the CMS bits in TIMx_CR1 register are not equal to
'00'. The Output compare interrupt flag of channels configured in output is set when: the
counter counts down (Center aligned mode 1, CMS = "01"), the counter counts up (Center
aligned mode 2, CMS = "10") the counter counts up and down (Center aligned mode 3,
CMS = "11").

492/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

In this mode, the direction bit (DIR from TIMx_CR1 register) cannot be written. It is updated
by hardware and gives the current direction of the counter.
The update event can be generated at each counter overflow and at each counter underflow
or by setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller) also generates an update event. In this case, the counter restarts counting from
0, as well as the counter of the prescaler.
The UEV update event can be disabled by software by setting the UDIS bit in TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter continues counting up and down, based on the current auto-reload
value.
In addition, if the URS bit (update request selection) in TIMx_CR1 register is set, setting the
UG bit generates an update event UEV but without setting the UIF flag (thus no interrupt or
DMA request is sent). This is to avoid generating both update and capture interrupt when
clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).
• The auto-reload active register is updated with the preload value (content of the
TIMx_ARR register). Note that if the update source is a counter overflow, the auto-
reload is updated before the counter is reloaded, so that the next period is the expected
one (the counter is loaded with the new value).
The following figures show some examples of the counter behavior for different clock
frequencies.

Figure 139. Counter timing diagram, internal clock divided by 1, TIMx_ARR=0x6

CK_INT

CNT_EN

Timerclock = CK_CNT
Counter register 04 03 02 01 00 01 02 03 04 05 06 05 04 03

Counter underflow

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS37342V1

1. Here, center-aligned mode 1 is used (for more details refer to Section 17.4.1: TIMx control register 1 (TIMx_CR1)
on page 522).

RM0402 Rev 6 493/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 140. Counter timing diagram, internal clock divided by 2

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 0003 0002 0001 0000 0001 0002 0003

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

MS37343V1

Figure 141. Counter timing diagram, internal clock divided by 4, TIMx_ARR=0x36

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 0034 0035 0036 0035

Counter overflow (cnt_ovf)

Update event (UEV)

Update interrupt flag (UIF)

MS37344V1

1. Center-aligned mode 2 or 3 is used with an UIF on overflow.

Figure 142. Counter timing diagram, internal clock divided by N

CK_INT

Timerclock = CK_CNT

Counter register 20 1F 01 00

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

MS37345V1

494/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 143. Counter timing diagram, Update event with ARPE=1 (counter underflow)

CK_INT

CNT_EN

Timerclock = CK_CNT

Counter register 06 05 04 03 02 01 00 01 02 03 04 05 06 07

Counter underflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FD 36

Write a new value in TIMx_ARR

Auto-reload active register FD 36

MS37360V1

Figure 144. Counter timing diagram, Update event with ARPE=1 (counter overflow)

CK_INT

CNT_EN

Timer clock = CK_CNT

Counter register F7 F8 F9 FA FB FC 36 35 34 33 32 31 30 2F

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FD 36

Write a new value in TIMx_ARR

Auto-reload active register FD 36

MS37361V1

RM0402 Rev 6 495/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.3.3 Clock selection

The counter clock can be provided by the following clock sources:
• Internal clock (CK_INT)
• External clock mode1: external input pin (TIx)
• External clock mode2: external trigger input (ETR) available on TIM2, TIM3 and TIM4
only.
• Internal trigger inputs (ITRx): using one timer as prescaler for another timer, for
example, Timer can be configured to act as a prescaler for Timer 2. Refer to Using one
timer as prescaler for another for more details.

Internal clock source (CK_INT)

If the slave mode controller is disabled (SMS=000 in the TIMx_SMCR register), then the
CEN, DIR (in the TIMx_CR1 register) and UG bits (in the TIMx_EGR register) are actual
control bits and can be changed only by software (except UG which remains cleared
automatically). As soon as the CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
Figure 145 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.

Figure 145. Control circuit in normal mode, internal clock divided by 1

Internal clock

CEN=CNT_EN

UG

CNT_INIT

Counter clock = CK_CNT = CK_PSC

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

MS31085V2

External clock source mode 1

This mode is selected when SMS=111 in the TIMx_SMCR register. The counter can count at
each rising or falling edge on a selected input.

496/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 146. TI2 external clock connection example

TIMx_SMCR
TS[2:0]

or TI2F or
TI1F or Encoder
ITRx mode
0xx
TI1_ED
100 TRGI External clock
TI1FP1 mode 1 CK_PSC
TI2F_Rising 101
TI2 Edge 0 TI2FP2 ETRF External clock
Filter 110
detector 1 ETRF mode 2
TI2F_Falling 111
CK_INT Internal clock
mode
ICF[3:0] CC2P (internal clock)

TIMx_CCMR1 TIMx_CCER
ECE SMS[2:0]
TIMx_SMCR

MS31196V1

For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:
1. Configure channel 2 to detect rising edges on the TI2 input by writing CC2S= ‘01 in the
TIMx_CCMR1 register.
2. Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=0000).
Note: The capture prescaler is not used for triggering, so it does not need to be configured.
3. Select rising edge polarity by writing CC2P=0 and CC2NP=0 in the TIMx_CCER
register.
4. Configure the timer in external clock mode 1 by writing SMS=111 in the TIMx_SMCR
register.
5. Select TI2 as the input source by writing TS=110 in the TIMx_SMCR register.
6. Enable the counter by writing CEN=1 in the TIMx_CR1 register.
When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.

RM0402 Rev 6 497/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 147. Control circuit in external clock mode 1

TI2

CNT_EN

Counter clock = CK_CNT = CK_PSC

Counter register 34 35 36

TIF

Write TIF=0

MS31087V2

External clock source mode 2

This mode is selected by writing ECE=1 in the TIMx_SMCR register.
The counter can count at each rising or falling edge on the external trigger input ETR.
Figure 148 gives an overview of the external trigger input block.

Figure 148. External trigger input block

or TI2F or
TI1F or Encoder
mode
TRGI External clock
ETR CK_PSC
mode 1
0
ETR pin Divider ETRP Filter ETRF External clock
1 /1, /2, /4, /8 downcounter mode 2
CK_INT CK_INT Internal clock
ETP ETPS[1:0] ETF[3:0] (internal clock) mode
TIMx_SMCR TIMx_SMCR TIMx_SMCR
ECE SMS[2:0]
TIMx_SMCR
MS37365V1

For example, to configure the upcounter to count each 2 rising edges on ETR, use the
following procedure:
1. As no filter is needed in this example, write ETF[3:0]=0000 in the TIMx_SMCR register.
2. Set the prescaler by writing ETPS[1:0]=01 in the TIMx_SMCR register
3. Select rising edge detection on the ETR pin by writing ETP=0 in the TIMx_SMCR
register
4. Enable external clock mode 2 by writing ECE=1 in the TIMx_SMCR register.
5. Enable the counter by writing CEN=1 in the TIMx_CR1 register.
The counter counts once each 2 ETR rising edges.

498/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

The delay between the rising edge on ETR and the actual clock of the counter is due to the
resynchronization circuit on the ETRP signal.

Figure 149. Control circuit in external clock mode 2

CK_INT

CNT_EN

ETR

ETRP

ETRF

Counter clock = CK_INT =CK_PSC

Counter register 34 35 36

MS37362V1

17.3.4 Capture/compare channels

Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
The following figure gives an overview of one Capture/Compare channel.
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).

RM0402 Rev 6 499/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 150. Capture/compare channel (example: channel 1 input stage)

TI1F_ED
To the slave mode controller

TI1 TI1F_Rising
Filter TI1F Edge 0 TI1FP1
fDTS downcounter TI1F_Falling 01
detector 1
TI2FP1 IC1 Divider IC1PS
10
/1, /2, /4, /8
ICF[3:0] CC1P/CC1NP TRC
11
TIMx_CCMR1 TIMx_CCER (from slave mode
TI2F_Rising controller)
0
(from channel 2)
CC1S[1:0] ICPS[1:0] CC1E
TI2F_Falling
1
(from channel 2) TIMx_CCMR1 TIMx_CCER

MS33115V1

The output stage generates an intermediate waveform which is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.

Figure 151. Capture/compare channel 1 main circuit

APB Bus

MCU-peripheral interface
(if 16-bit)

8 8
high

S write CCR1H
low

Read CCR1H S write_in_progress

read_in_progress write CCR1L
Read CCR1L Capture/compare preload register R
R Output CC1S[1]
capture_transfer compare_transfer mode
CC1S[0]
CC1S[1] Input
mode OC1PE
CC1S[0] Capture/compare shadow register OC1PE
UEV
TIMx_CCMR1
Comparator (from time
IC1PS Capture
base unit)
CC1E CNT>CCR1
Counter CNT=CCR1
CC1G
TIMx_EGR

MS33144V1

500/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 152. Output stage of capture/compare channel (channel 1)

OCREF_CLR
0
OCREF_CLR_INT
ETRF 1
To the master mode 0
controller Output OC1
OCCS Enable
1 Circuit
TIMx_SMCR
CC1P
CNT > CCR1
Output mode oc1ref TIMx_CCER
CNT = CCR1 controller

CC1E

OC1M[2:0] TIMx_CCER

TIMx_CCMR1 ai17187

The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.

17.3.5 Input capture mode

In Input capture mode, the Capture/Compare Registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a
capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be
cleared by software by writing it to 0 or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when it is written with 0.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
• Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to 01 in the TIMx_CCMR1 register. As soon as CC1S becomes different from 00,
the channel is configured in input and the TIMx_CCR1 register becomes read-only.
• Program the appropriate input filter duration in relation with the signal connected to the
timer (by programming the ICxF bits in the TIMx_CCMRx register if the input is one of
the TIx inputs). Let’s imagine that, when toggling, the input signal is not stable during at
must 5 internal clock cycles. We must program a filter duration longer than these 5
clock cycles. We can validate a transition on TI1 when 8 consecutive samples with the

RM0402 Rev 6 501/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

new level have been detected (sampled at fDTS frequency). Then write IC1F bits to
0011 in the TIMx_CCMR1 register.
• Select the edge of the active transition on the TI1 channel by writing the CC1P and
CC1NP bits to 00 in the TIMx_CCER register (rising edge in this case).
• Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to 00 in the
TIMx_CCMR1 register).
• Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.
• If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register, and/or the DMA request by setting the CC1DE bit in the
TIMx_DIER register.
When an input capture occurs:
• The TIMx_CCR1 register gets the value of the counter on the active transition.
• CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.
• An interrupt is generated depending on the CC1IE bit.
• A DMA request is generated depending on the CC1DE bit.
In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note: IC interrupt and/or DMA requests can be generated by software by setting the
corresponding CCxG bit in the TIMx_EGR register.

17.3.6 PWM input mode

This mode is a particular case of input capture mode. The procedure is the same except:
• Two ICx signals are mapped on the same TIx input.
• These 2 ICx signals are active on edges with opposite polarity.
• One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.

502/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

For example, one can measure the period (in TIMx_CCR1 register) and the duty cycle (in
TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure (depending
on CK_INT frequency and prescaler value):
• Select the active input for TIMx_CCR1: write the CC1S bits to 01 in the TIMx_CCMR1
register (TI1 selected).
• Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): write the CC1P to ‘0’ and the CC1NP bit to ‘0’ (active on rising edge).
• Select the active input for TIMx_CCR2: write the CC2S bits to 10 in the TIMx_CCMR1
register (TI1 selected).
• Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): write the CC2P
bit to ‘1’ and the CC2NP bit to ’0’(active on falling edge).
• Select the valid trigger input: write the TS bits to 101 in the TIMx_SMCR register
(TI1FP1 selected).
• Configure the slave mode controller in reset mode: write the SMS bits to 100 in the
TIMx_SMCR register.
• Enable the captures: write the CC1E and CC2E bits to ‘1 in the TIMx_CCER register.

Figure 153. PWM input mode timing

TI1

TIMx_CNT 0004 0000 0001 0002 0003 0004 0000

TIMx_CCR1 0004

TIMx_CCR2 0002

IC1 capture IC2 capture IC1 capture

pulse width period
IC2 capture
measurement measurement
reset counter
ai15413

17.3.7 Forced output mode

In output mode (CCxS bits = 00 in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.
To force an output compare signal (ocxref/OCx) to its active level, one just needs to write
101 in the OCxM bits in the corresponding TIMx_CCMRx register. Thus ocxref is forced high
(OCxREF is always active high) and OCx get opposite value to CCxP polarity bit.
e.g.: CCxP=0 (OCx active high) => OCx is forced to high level.
ocxref signal can be forced low by writing the OCxM bits to 100 in the TIMx_CCMRx
register.

RM0402 Rev 6 503/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt and DMA requests can be sent
accordingly. This is described in the Output Compare Mode section.

17.3.8 Output compare mode

This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
• Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=000), be set
active (OCxM=001), be set inactive (OCxM=010) or can toggle (OCxM=011) on match.
• Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).
• Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).
• Sends a DMA request if the corresponding enable bit is set (CCxDE bit in the
TIMx_DIER register, CCDS bit in the TIMx_CR2 register for the DMA request
selection).
The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on ocxref and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One-pulse mode).
Procedure:
1. Select the counter clock (internal, external, prescaler).
2. Write the desired data in the TIMx_ARR and TIMx_CCRx registers.
3. Set the CCxIE and/or CCxDE bits if an interrupt and/or a DMA request is to be
generated.
4. Select the output mode. For example, one must write OCxM=011, OCxPE=0, CCxP=0
and CCxE=1 to toggle OCx output pin when CNT matches CCRx, CCRx preload is not
used, OCx is enabled and active high.
5. Enable the counter by setting the CEN bit in the TIMx_CR1 register.
The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=0, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 154.

504/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Figure 154. Output compare mode, toggle on OC1

Write B201h in the CC1R register

TIMx_CNT 0039 003A 003B B200 B201

TIMx_CCR1 003A B201

OC1REF = OC1

Match detected on CCR1

Interrupt generated if enabled
MS37363V1

17.3.9 PWM mode

Pulse width modulation mode allows to generate a signal with a frequency determined by
the value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing 110 (PWM mode 1) or ‘111 (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The corresponding preload register must be enabled by setting the
OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register by
setting the ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, all registers must be initialized by setting the UG bit in
the TIMx_EGR register.
OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register. It
can be programmed as active high or active low. OCx output is enabled by the CCxE bit in
the TIMx_CCER register. Refer to the TIMx_CCERx register description for more details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CCRx ≤ TIMx_CNT or TIMx_CNT ≤ TIMx_CCRx (depending on the direction
of the counter). However, to comply with the ETRF (OCREF can be cleared by an external
event through the ETR signal until the next PWM period), the OCREF signal is asserted
only:
• When the result of the comparison changes, or
• When the output compare mode (OCxM bits in TIMx_CCMRx register) switches from
the “frozen” configuration (no comparison, OCxM=‘000) to one of the PWM modes
(OCxM=‘110 or ‘111).
This forces the PWM by software while the timer is running.
The timer is able to generate PWM in edge-aligned mode or center-aligned mode
depending on the CMS bits in the TIMx_CR1 register.

RM0402 Rev 6 505/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

PWM edge-aligned mode

Upcounting configuration
Upcounting is active when the DIR bit in the TIMx_CR1 register is low. Refer to Upcounting
mode on page 487.
In the following example, we consider PWM mode 1. The reference PWM signal OCxREF is
high as long as TIMx_CNT <TIMx_CCRx else it becomes low. If the compare value in
TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR) then OCxREF is held at ‘1.
If the compare value is 0 then OCxREF is held at ‘0. Figure 155 shows some edge-aligned
PWM waveforms in an example where TIMx_ARR=8.

Figure 155. Edge-aligned PWM waveforms (ARR=8)

Counter register 0 1 2 3 4 5 6 7 8 0 1

OCXREF
CCRx=4
CCxIF

OCXREF
CCRx=8
CCxIF

OCXREF ‘1’
CCRx>8
CCxIF

OCXREF ‘0’
CCRx=0
CCxIF

MS31093V1

Downcounting configuration
Downcounting is active when DIR bit in TIMx_CR1 register is high. Refer to Downcounting
mode on page 490.
In PWM mode 1, the reference signal ocxref is low as long as TIMx_CNT>TIMx_CCRx else
it becomes high. If the compare value in TIMx_CCRx is greater than the auto-reload value in
TIMx_ARR, then ocxref is held at ‘1. 0% PWM is not possible in this mode.

PWM center-aligned mode

Center-aligned mode is active when the CMS bits in TIMx_CR1 register are different from
‘00 (all the remaining configurations having the same effect on the ocxref/OCx signals). The
compare flag is set when the counter counts up, when it counts down or both when it counts
up and down depending on the CMS bits configuration. The direction bit (DIR) in the

506/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

TIMx_CR1 register is updated by hardware and must not be changed by software. Refer to
Center-aligned mode (up/down counting) on page 492.
Figure 156 shows some center-aligned PWM waveforms in an example where:
• TIMx_ARR=8,
• PWM mode is the PWM mode 1,
• The flag is set when the counter counts down corresponding to the center-aligned
mode 1 selected for CMS=01 in TIMx_CR1 register.

Figure 156. Center-aligned PWM waveforms (ARR=8)

Counter register 0 1 2 3 4 5 6 7 8 7 6 5 4 3 2 1 0 1

OCxREF
CCRx = 4

CCxIF CMS=01
CMS=10
CMS=11

OCxREF
CCRx=7

CMS=10 or 11
CCxIF
‘1’
OCxREF
CCRx=8

CCxIF CMS=01
CMS=10
CMS=11
‘1’
OCxREF
CCRx>8

CCxIF CMS=01
CMS=10
CMS=11
‘0’
OCxREF
CCRx=0

CCxIF CMS=01
CMS=10
CMS=11

AI14681b

Hints on using center-aligned mode:

• When starting in center-aligned mode, the current up-down configuration is used. It
means that the counter counts up or down depending on the value written in the DIR bit

RM0402 Rev 6 507/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

in the TIMx_CR1 register. Moreover, the DIR and CMS bits must not be changed at the
same time by the software.
• Writing to the counter while running in center-aligned mode is not recommended as it
can lead to unexpected results. In particular:
– The direction is not updated if a value greater than the auto-reload value is written
in the counter (TIMx_CNT>TIMx_ARR). For example, if the counter was counting
up, it continues to count up.
– The direction is updated if 0 or the TIMx_ARR value is written in the counter but no
Update Event UEV is generated.
• The safest way to use center-aligned mode is to generate an update by software
(setting the UG bit in the TIMx_EGR register) just before starting the counter and not to
write the counter while it is running.

17.3.10 One-pulse mode

One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. One-pulse mode is selected
by setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically
at the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be:
• In upcounting: CNT<CCRx ≤ ARR (in particular, 0<CCRx),
• In downcounting: CNT>CCRx.

Figure 157. Example of one-pulse mode

TI2

OC1REF
OC1

TIM1_ARR
Counter

TIM1_CCR1

0
tDELAY tPULSE t

MS31099V1

For example one may want to generate a positive pulse on OC1 with a length of tPULSE and
after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.

508/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Let’s use TI2FP2 as trigger 1:

• Map TI2FP2 on TI2 by writing CC2S=01 in the TIMx_CCMR1 register.
• TI2FP2 must detect a rising edge, write CC2P=0 and CC2NP=’0’ in the TIMx_CCER
register.
• Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=110 in
the TIMx_SMCR register.
• TI2FP2 is used to start the counter by writing SMS to ‘110 in the TIMx_SMCR register
(trigger mode).
The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
• The tDELAY is defined by the value written in the TIMx_CCR1 register.
• The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR + 1).
• Let’s say one want to build a waveform with a transition from ‘0 to ‘1 when a compare
match occurs and a transition from ‘1 to ‘0 when the counter reaches the auto-reload
value. To do this PWM mode 2 must be enabled by writing OC1M=111 in the
TIMx_CCMR1 register. Optionally the preload registers can be enabled by writing
OC1PE=1 in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case one has to write the compare value in the TIMx_CCR1 register, the auto-reload
value in the TIMx_ARR register, generate an update by setting the UG bit and wait for
external trigger event on TI2. CC1P is written to ‘0 in this example.
In our example, the DIR and CMS bits in the TIMx_CR1 register should be low.
Since only 1 pulse (Single mode) is needed, a 1 must be written in the OPM bit in the
TIMx_CR1 register to stop the counter at the next update event (when the counter rolls over
from the auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0',
so the Repetitive Mode is selected.

Particular case: OCx fast enable:

In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
If one wants to output a waveform with the minimum delay, the OCxFE bit can be set in the
TIMx_CCMRx register. Then OCxRef (and OCx) is forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.

17.3.11 Clearing the OCxREF signal on an external event

The OCxREF signal for a given channel can be driven Low by applying a High level to the
ETRF input (OCxCE enable bit of the corresponding TIMx_CCMRx register set to '1'). The
OCxREF signal remains Low until the next update event, UEV, occurs.
This function can only be used in output compare and PWM modes, and does not work in
forced mode.
For example, the ETR signal can be connected to the output of a comparator to be used for
current handling. In this case, ETR must be configured as follows:

RM0402 Rev 6 509/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

1. The external trigger prescaler should be kept off: bits ETPS[1:0] in the TIMx_SMCR
register are cleared to 00.
2. The external clock mode 2 must be disabled: bit ECE in the TIM1_SMCR register is
cleared to 0.
3. The external trigger polarity (ETP) and the external trigger filter (ETF) can be
configured according to the application’s needs.
Figure 158 shows the behavior of the OCxREF signal when the ETRF input becomes high,
for both values of the OCxCE enable bit. In this example, the timer TIMx is programmed in
PWM mode.

Figure 158. Clearing TIMx OCxREF

(CCRx)

Counter (CNT)

ETRF

OCxREF (OCxCE = ‘0’)

OCxREF (OCxCE = ‘1’)

ETRF becomes high ETRF still high
MS37368V1

1. In case of a PWM with a 100% duty cycle (if CCRx>ARR), OCxREF is enabled again at the next counter
overflow.

17.3.12 Encoder interface mode

To select Encoder Interface mode write SMS=‘001 in the TIMx_SMCR register if the counter
is counting on TI2 edges only, SMS=010 if it is counting on TI1 edges only and SMS=011 if
it is counting on both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the CC1P and CC2P bits in the TIMx_CCER
register. When needed, the input filter can be programmed as well.
The two inputs TI1 and TI2 are used to interface to an incremental encoder. Refer to
Table 99. The counter is clocked by each valid transition on TI1FP1 or TI2FP2 (TI1 and TI2
after input filter and polarity selection, TI1FP1=TI1 if not filtered and not inverted,
TI2FP2=TI2 if not filtered and not inverted) assuming that it is enabled (CEN bit in
TIMx_CR1 register written to ‘1). The sequence of transitions of the two inputs is evaluated
and generates count pulses as well as the direction signal. Depending on the sequence the
counter counts up or down, the DIR bit in the TIMx_CR1 register is modified by hardware
accordingly. The DIR bit is calculated at each transition on any input (TI1 or TI2), whatever
the counter is counting on TI1 only, TI2 only or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This
means that the counter just counts continuously between 0 and the auto-reload value in the
TIMx_ARR register (0 to ARR or ARR down to 0 depending on the direction). So the
TIMx_ARR must be configured before starting. In the same way, the capture, compare,
prescaler, trigger output features continue to work as normal.

510/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

In this mode, the counter is modified automatically following the speed and the direction of
the incremental encoder and its content, therefore, always represents the encoder’s
position. The count direction correspond to the rotation direction of the connected sensor.
The table summarizes the possible combinations, assuming TI1 and TI2 do not switch at the
same time.

Table 99. Counting direction versus encoder signals

Level on opposite TI1FP1 signal TI2FP2 signal
Active edge signal (TI1FP1 for
TI2, TI2FP2 for TI1) Rising Falling Rising Falling

Counting on High Down Up No Count No Count

TI1 only Low Up Down No Count No Count

Counting on High No Count No Count Up Down

TI2 only Low No Count No Count Down Up

Counting on High Down Up Up Down

TI1 and TI2 Low Up Down Down Up

An external incremental encoder can be connected directly to the MCU without external
interface logic. However, comparators are normally be used to convert the encoder’s
differential outputs to digital signals. This greatly increases noise immunity. The third
encoder output which indicate the mechanical zero position, may be connected to an
external interrupt input and trigger a counter reset.
Figure 159 gives an example of counter operation, showing count signal generation and
direction control. It also shows how input jitter is compensated where both edges are
selected. This might occur if the sensor is positioned near to one of the switching points. For
this example we assume that the configuration is the following:
• CC1S= ‘01’ (TIMx_CCMR1 register, TI1FP1 mapped on TI1)
• CC2S= ‘01’ (TIMx_CCMR2 register, TI2FP2 mapped on TI2)
• CC1P= ‘0’, CC1NP = ‘0’, IC1F =’0000’ (TIMx_CCER register, TI1FP1 noninverted,
TI1FP1=TI1)
• CC2P= ‘0’, CC2NP = ‘0’, IC2F =’0000’ (TIMx_CCER register, TI2FP2 noninverted,
TI2FP2=TI2)
• SMS= ‘011’ (TIMx_SMCR register, both inputs are active on both rising and falling
edges)
• CEN = 1 (TIMx_CR1 register, Counter is enabled)

RM0402 Rev 6 511/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 159. Example of counter operation in encoder interface mode

forward jitter backward jitter forward

TI1

TI2

Counter

up down up

MS33107V1

Figure 160 gives an example of counter behavior when TI1FP1 polarity is inverted (same
configuration as above except CC1P=1).

Figure 160. Example of encoder interface mode with TI1FP1 polarity inverted
forward jitter backward jitter forward

TI1

TI2

Counter

down up down

MS33108V1

The timer, when configured in Encoder Interface mode provides information on the sensor’s
current position. Dynamic information can be obtained (speed, acceleration, deceleration)
by measuring the period between two encoder events using a second timer configured in
capture mode. The output of the encoder which indicates the mechanical zero can be used
for this purpose. Depending on the time between two events, the counter can also be read
at regular times. This can be done by latching the counter value into a third input capture
register if available (then the capture signal must be periodic and can be generated by
another timer). when available, it is also possible to read its value through a DMA request
generated by a Real-Time clock.

512/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

17.3.13 Timer input XOR function

The TI1S bit in the TIM_CR2 register, allows the input filter of channel 1 to be connected to
the output of a XOR gate, combining the three input pins TIMx_CH1 to TIMx_CH3.
The XOR output can be used with all the timer input functions such as trigger or input
capture.

17.3.14 Timers and external trigger synchronization

The TIMx Timers can be synchronized with an external trigger in several modes: Reset
mode, Gated mode and Trigger mode.

Slave mode: Reset mode

The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
• Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC1S bits
select the input capture source only, CC1S = 01 in the TIMx_CCMR1 register. Write
CC1P=0 and CC1NP=0 in TIMx_CCER register to validate the polarity (and detect
rising edges only).
• Configure the timer in reset mode by writing SMS=100 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
• Start the counter by writing CEN=1 in the TIMx_CR1 register.
The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request, or a DMA
request can be sent if enabled (depending on the TIE and TDE bits in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
resynchronization circuit on TI1 input.

Figure 161. Control circuit in reset mode

TI1

UG

Counter clock = CK_CNT = CK_PSC

Counter register 30 31 32 33 34 35 36 00 01 02 03 00 01 02 03

TIF
MS37384V1

RM0402 Rev 6 513/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Slave mode: Gated mode

The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
• Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC1F=0000). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC1S bits
select the input capture source only, CC1S=01 in TIMx_CCMR1 register. Write
CC1P=1 in TIMx_CCER register to validate the polarity (and detect low level only).
• Configure the timer in gated mode by writing SMS=101 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
• Enable the counter by writing CEN=1 in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=0, whatever is the trigger input level).
The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
resynchronization circuit on TI1 input.

Figure 162. Control circuit in gated mode

TI1
CNT_EN

Counter clock = CK_CNT = CK_PSC

Counter register 30 31 32 33 34 35 36 37 38

TIF

Write TIF=0
MS37385V1

1. The configuration “CCxP=CCxNP=1” (detection of both rising and falling edges) does not have any effect
in gated mode because gated mode acts on a level and not on an edge.

Slave mode: Trigger mode

The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
• Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC2F=0000). The capture
prescaler is not used for triggering, so it does not need to be configured. CC2S bits are
selecting the input capture source only, CC2S=01 in TIMx_CCMR1 register. Write
CC2P=1 in TIMx_CCER register to validate the polarity (and detect low level only).
• Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI2 as the input source by writing TS=110 in TIMx_SMCR register.
When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.

514/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on TI2 input.

Figure 163. Control circuit in trigger mode

TI2

CNT_EN

Counter clock = CK_CNT = CK_PSC

Counter register 34 35 36 37 38

TIF
MS37386V1

Slave mode: External Clock mode 2 + trigger mode

The external clock mode 2 can be used in addition to another slave mode (except external
clock mode 1 and encoder mode). In this case, the ETR signal is used as external clock
input, and another input can be selected as trigger input when operating in reset mode,
gated mode or trigger mode. It is recommended not to select ETR as TRGI through the TS
bits of TIMx_SMCR register.
In the following example, the upcounter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
1. Configure the external trigger input circuit by programming the TIMx_SMCR register as
follows:
– ETF = 0000: no filter
– ETPS = 00: prescaler disabled
– ETP = 0: detection of rising edges on ETR and ECE=1 to enable the external clock
mode 2.
2. Configure the channel 1 as follows, to detect rising edges on TI:
– IC1F = 0000: no filter.
– The capture prescaler is not used for triggering and does not need to be
configured.
– CC1S = 01 in TIMx_CCMR1 register to select only the input capture source
– CC1P = 0 in TIMx_CCER register to validate the polarity (and detect rising edge
only).
3. Configure the timer in trigger mode by writing SMS=110 in TIMx_SMCR register. Select
TI1 as the input source by writing TS=101 in TIMx_SMCR register.
A rising edge on TI1 enables the counter and sets the TIF flag. The counter then counts on
ETR rising edges.
The delay between the rising edge of the ETR signal and the actual reset of the counter is
due to the resynchronization circuit on ETRP input.

RM0402 Rev 6 515/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 164. Control circuit in external clock mode 2 + trigger mode

TI1

CEN/CNT_EN

ETR

Counter clock = CK_CNT = CK_PSC

Counter register 34 35 36

TIF

MS33110V1

17.3.15 Timer synchronization

The TIMx timers are linked together internally for timer synchronization or chaining. When
one Timer is configured in Master Mode, it can reset, start, stop or clock the counter of
another Timer configured in Slave Mode.
Figure 165 presents an overview of the trigger selection and the master mode selection
blocks.

Using one timer as prescaler for another

Figure 165. Master/Slave timer example

TIM1 TIM2
Clock
MMS TS SMS

UEV
Master Slave
TRGO1 ITR0 CK_PSC
mode mode
Prescaler Counter control control Prescaler Counter

Input trigger
selection

MS37387V1

516/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

For example, Timer 1 can be configured to act as a prescaler for Timer 2. Refer to
Figure 165. To do this:
• Configure Timer 1 in master mode so that it outputs a periodic trigger signal on each
update event UEV. If MMS=010 is written in the TIM1_CR2 register, a rising edge is
output on TRGO1 each time an update event is generated.
• To connect the TRGO1 output of Timer 1 to Timer 2, Timer 2 must be configured in
slave mode using ITR0 as internal trigger. This is selected through the TS bits in the
TIM2_SMCR register (writing TS=000).
• Then the slave mode controller must be put in external clock mode 1 (write SMS=111 in
the TIM2_SMCR register). This causes Timer 2 to be clocked by the rising edge of the
periodic Timer 1 trigger signal (which correspond to the timer 1 counter overflow).
• Finally both timers must be enabled by setting their respective CEN bits (TIMx_CR1
register).
Note: If OCx is selected on Timer 1 as trigger output (MMS=1xx), its rising edge is used to clock
the counter of timer 2.

Using one timer to enable another timer

In this example, we control the enable of Timer 2 with the output compare 1 of Timer 1.
Refer to Figure 165 for connections. Timer 2 counts on the divided internal clock only when
OC1REF of Timer 1 is high. Both counter clock frequencies are divided by 3 by the
prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).
• Configure Timer 1 master mode to send its Output Compare 1 Reference (OC1REF)
signal as trigger output (MMS=100 in the TIM1_CR2 register).
• Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).
• Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).
• Configure Timer 2 in gated mode (SMS=101 in TIM2_SMCR register).
• Enable Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).
• Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).
Note: The counter 2 clock is not synchronized with counter 1, this mode only affects the Timer 2
counter enable signal.

Figure 166. Gating timer 2 with OC1REF of timer 1

CK_INT

TIMER1-OC1REF
TIMER1-CNT FC FD FE FF 00 01

TIMER2-CNT 3045 3046 3047 3048

TIMER2-TIF

Write TIF = 0
MS37388V1

In the example in Figure 166, the Timer 2 counter and prescaler are not initialized before
being started. So they start counting from their current value. It is possible to start from a
given value by resetting both timers before starting Timer 1. Then any value can be written

RM0402 Rev 6 517/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

in the timer counters. The timers can easily be reset by software using the UG bit in the
TIMx_EGR registers.
In the next example, we synchronize Timer 1 and Timer 2. Timer 1 is the master and starts
from 0. Timer 2 is the slave and starts from 0xE7. The prescaler ratio is the same for both
timers. Timer 2 stops when Timer 1 is disabled by writing ‘0 to the CEN bit in the TIM1_CR1
register:
• Configure Timer 1 master mode to send its Output Compare 1 Reference (OC1REF)
signal as trigger output (MMS=100 in the TIM1_CR2 register).
• Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).
• Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).
• Configure Timer 2 in gated mode (SMS=101 in TIM2_SMCR register).
• Reset Timer 1 by writing ‘1 in UG bit (TIM1_EGR register).
• Reset Timer 2 by writing ‘1 in UG bit (TIM2_EGR register).
• Initialize Timer 2 to 0xE7 by writing ‘0xE7’ in the timer 2 counter (TIM2_CNTL).
• Enable Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).
• Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).
• Stop Timer 1 by writing ‘0 in the CEN bit (TIM1_CR1 register).

Figure 167. Gating timer 2 with Enable of timer 1

CK_INT

TIMER1-CEN=CNT_EN

TIMER1-CNT_INIT

TIMER1-CNT 75 00 01 02

TIMER2-CNT AB 00 E7 E8 E9

TIMER2-CNT_INIT

TIMER2-write CNT

TIMER2-TIF

Write TIF = 0
MS37389V1

Using one timer to start another timer

In this example, we set the enable of Timer 2 with the update event of Timer 1. Refer to
Figure 165 for connections. Timer 2 starts counting from its current value (which can be
nonzero) on the divided internal clock as soon as the update event is generated by Timer 1.
When Timer 2 receives the trigger signal its CEN bit is automatically set and the counter

518/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

counts until we write ‘0 to the CEN bit in the TIM2_CR1 register. Both counter clock
frequencies are divided by 3 by the prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).
• Configure Timer 1 master mode to send its Update Event (UEV) as trigger output
(MMS=010 in the TIM1_CR2 register).
• Configure the Timer 1 period (TIM1_ARR registers).
• Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).
• Configure Timer 2 in trigger mode (SMS=110 in TIM2_SMCR register).
• Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).

Figure 168. Triggering timer 2 with update of timer 1

CK_INT

TIMER1-UEV

TIMER1-CNT FD FE FF 00 01 02

TIMER2-CNT 45 46 47 48

TIMER2-CEN=CNT_EN

TIMER2-TIF

Write TIF = 0
MS37390V1

As in the previous example, both counters can be initialized before starting counting.
Figure 169 shows the behavior with the same configuration as in Figure 166, but in trigger
mode instead of gated mode (SMS=110 in the TIM2_SMCR register).

RM0402 Rev 6 519/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Figure 169. Triggering timer 2 with Enable of timer 1

CK_INT

TIMER1-CEN=CNT_EN

TIMER1-CNT_INIT

TIMER1-CNT 75 00 01 02

TIMER2-CNT CD 00 E7 E8 E9 EA

TIMER2-CNT_INIT

TIMER2-write CNT

TIMER2-TIF

Write TIF = 0
MS37391V1

Using one timer as prescaler for another timer

For example, Timer 1 can be configured to act as a prescaler for Timer 2. Refer to
Figure 165 for connections. To do this:
• Configure Timer 1 master mode to send its Update Event (UEV) as trigger output
(MMS=010 in the TIM1_CR2 register). then it outputs a periodic signal on each counter
overflow.
• Configure the Timer 1 period (TIM1_ARR registers).
• Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).
• Configure Timer 2 in external clock mode 1 (SMS=111 in TIM2_SMCR register).
• Start Timer 2 by writing ‘1 in the CEN bit (TIM2_CR1 register).
• Start Timer 1 by writing ‘1 in the CEN bit (TIM1_CR1 register).

Starting 2 timers synchronously in response to an external trigger

In this example, we set the enable of timer 1 when its TI1 input rises, and the enable of
Timer 2 with the enable of Timer 1. Refer to Figure 165 for connections. To ensure the

520/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

counters are aligned, Timer 1 must be configured in Master/Slave mode (slave with respect
to TI1, master with respect to Timer 2):
• Configure Timer 1 master mode to send its Enable as trigger output (MMS=001 in the
TIM1_CR2 register).
• Configure Timer 1 slave mode to get the input trigger from TI1 (TS=100 in the
TIM1_SMCR register).
• Configure Timer 1 in trigger mode (SMS=110 in the TIM1_SMCR register).
• Configure the Timer 1 in Master/Slave mode by writing MSM=1 (TIM1_SMCR register).
• Configure Timer 2 to get the input trigger from Timer 1 (TS=000 in the TIM2_SMCR
register).
• Configure Timer 2 in trigger mode (SMS=110 in the TIM2_SMCR register).
When a rising edge occurs on TI1 (Timer 1), both counters starts counting synchronously on
the internal clock and both TIF flags are set.
Note: In this example both timers are initialized before starting (by setting their respective UG
bits). Both counters starts from 0, but an offset can easily be inserted between them by
writing any of the counter registers (TIMx_CNT). One can see that the master/slave mode
insert a delay between CNT_EN and CK_PSC on timer 1.

Figure 170. Triggering timer 1 and 2 with timer 1 TI1 input

CK_INT

TIMER1-TI1

TIMER1-CEN=CNT_EN

TIMER1-CK_PSC

TIMER1-CNT 00 01 02 03 04 05 06 07 08 09

TIMER1-TIF

TIMER2-CEN=CNT_EN

TIMER2-CK_PSC

TIMER2-CNT 00 01 02 03 04 05 06 07 08 09

TIMER2-TIF
MS37392V1

17.3.16 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core - halted), the
TIMx counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBGMCU module. For more details, refer to Section 33.16.2: Debug
support for timers, watchdog, bxCAN and I2C.

RM0402 Rev 6 521/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4 TIM2 to TIM5 registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The 32-bit peripheral registers have to be written by words (32 bits). All other peripheral
registers have to be written by half-words (16 bits) or words (32 bits). Read accesses can be
done by bytes (8 bits), half-words (16 bits) or words (32 bits).

17.4.1 TIMx control register 1 (TIMx_CR1)

Address offset: 0x00
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. CKD[1:0] ARPE CMS DIR OPM URS UDIS CEN
rw rw rw rw rw rw rw rw rw rw

Bits 15:10 Reserved, must be kept at reset value.

Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (ETR, TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
10: tDTS = 4 × tCK_INT
11: Reserved
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:5 CMS: Center-aligned mode selection
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(DIR).
01: Center-aligned mode 1. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting down.
10: Center-aligned mode 2. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
only when the counter is counting up.
11: Center-aligned mode 3. The counter counts up and down alternatively. Output compare
interrupt flags of channels configured in output (CCxS=00 in TIMx_CCMRx register) are set
both when the counter is counting up or down.
Note: It is not allowed to switch from edge-aligned mode to center-aligned mode as long as
the counter is enabled (CEN=1)
Bit 4 DIR: Direction
0: Counter used as upcounter
1: Counter used as downcounter
Note: This bit is read only when the timer is configured in Center-aligned mode or Encoder
mode.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the bit CEN)

522/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Bit 2 URS: Update request source

This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generate an update interrupt or DMA request if enabled.
These events can be:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC, CCRx). However the counter and the prescaler are reinitialized if the UG bit is
set or if a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: External clock, gated mode and encoder mode can work only if the CEN bit has been
previously set by software. However trigger mode can set the CEN bit automatically by
hardware.
CEN is cleared automatically in one-pulse mode, when an update event occurs.

RM0402 Rev 6 523/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4.2 TIMx control register 2 (TIMx_CR2)

Address offset: 0x04
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. TI1S MMS[2:0] CCDS Res. Res. Res.
rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bit 7 TI1S: TI1 selection
0: The TIMx_CH1 pin is connected to TI1 input
1: The TIMx_CH1, CH2 and CH3 pins are connected to the TI1 input (XOR combination)
Bits 6:4 MMS[2:0]: Master mode selection
These bits allow to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as trigger output (TRGO). If the
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter enable signal, CNT_EN, is used as trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer is
enabled. The Counter Enable signal is generated by a logic OR between CEN control bit
and the trigger input when configured in gated mode.
When the Counter Enable signal is controlled by the trigger input, there is a delay on TRGO,
except if the master/slave mode is selected (see the MSM bit description in TIMx_SMCR
register).
010: Update - The update event is selected as trigger output (TRGO). For instance a master
timer can then be used as a prescaler for a slave timer.
011: Compare Pulse - The trigger output send a positive pulse when the CC1IF flag is to be
set (even if it was already high), as soon as a capture or a compare match occurred.
(TRGO)
100: Compare - OC1REF signal is used as trigger output (TRGO)
101: Compare - OC2REF signal is used as trigger output (TRGO)
110: Compare - OC3REF signal is used as trigger output (TRGO)
111: Compare - OC4REF signal is used as trigger output (TRGO)
Bit 3 CCDS: Capture/compare DMA selection
0: CCx DMA request sent when CCx event occurs
1: CCx DMA requests sent when update event occurs
Bits 2:0 Reserved, must be kept at reset value.

524/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

17.4.3 TIMx slave mode control register (TIMx_SMCR)

Address offset: 0x08
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ETP ECE ETPS[1:0] ETF[3:0] MSM TS[2:0] Res. SMS[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 ETP: External trigger polarity

This bit selects whether ETR or ETR is used for trigger operations
0: ETR is noninverted, active at high level or rising edge
1: ETR is inverted, active at low level or falling edge
Bit 14 ECE: External clock enable
This bit enables External clock mode 2.
0: External clock mode 2 disabled
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
1: Setting the ECE bit has the same effect as selecting external clock mode 1 with TRGI
connected to ETRF (SMS=111 and TS=111).
2: It is possible to simultaneously use external clock mode 2 with the following slave modes:
reset mode, gated mode and trigger mode. Nevertheless, TRGI must not be connected to
ETRF in this case (TS bits must not be 111).
3: If external clock mode 1 and external clock mode 2 are enabled at the same time, the
external clock input is ETRF.
Bits 13:12 ETPS: External trigger prescaler
External trigger signal ETRP frequency must be at most 1/4 of CK_INT frequency. A
prescaler can be enabled to reduce ETRP frequency. It is useful when inputting fast external
clocks.
00: Prescaler OFF
01: ETRP frequency divided by 2
10: ETRP frequency divided by 4
11: ETRP frequency divided by 8
Bits 11:8 ETF[3:0]: External trigger filter
This bit-field then defines the frequency used to sample ETRP signal and the length of the
digital filter applied to ETRP. The digital filter is made of an event counter in which N
consecutive events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8

RM0402 Rev 6 525/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Bit 7 MSM: Master/Slave mode

0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful if we
want to synchronize several timers on a single external event.
Bits 6:4 TS: Trigger selection
This bit-field selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
001: Internal Trigger 1 (ITR1).
010: Internal Trigger 2 (ITR2).
011: Internal Trigger 3 (ITR3).
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
110: Filtered Timer Input 2 (TI2FP2)
111: External Trigger input (ETRF)
See Table 100 for more details on ITRx meaning for each Timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=000) to
avoid wrong edge detections at the transition.
Bit 3 Reserved, must be kept at reset value.
Bits 2:0 SMS: Slave mode selection
When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input Control register and Control Register
description.
000: Slave mode disabled - if CEN = ‘1 then the prescaler is clocked directly by the internal
clock.
001: Encoder mode 1 - Counter counts up/down on TI2FP2 edge depending on TI1FP1
level.
010: Encoder mode 2 - Counter counts up/down on TI1FP1 edge depending on TI2FP2
level.
011: Encoder mode 3 - Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset Mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers.
101: Gated Mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both start and stop of
the counter are controlled.
110: Trigger Mode - The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled.
111: External Clock Mode 1 - Rising edges of the selected trigger (TRGI) clock the counter.
Note: The gated mode must not be used if TI1F_ED is selected as the trigger input (TS=100).
Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the gated mode
checks the level of the trigger signal.

Table 100. TIMx internal trigger connections

Slave TIM ITR0 (TS = 000) ITR1 (TS = 001) ITR2 (TS = 010) ITR3 (TS = 011)

TIM2 TIM1 TIM8 TIM3 TIM4

TIM3 TIM1 TIM2 TIM5 TIM4
TIM4 TIM1 TIM2 TIM3 TIM8
TIM5 TIM2 TIM3 TIM4 TIM8

526/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

17.4.4 TIMx DMA/Interrupt enable register (TIMx_DIER)

Address offset: 0x0C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. TDE Res. CC4DE CC3DE CC2DE CC1DE UDE Res. TIE Res. CC4IE CC3IE CC2IE CC1IE UIE
rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 Reserved, must be kept at reset value.

Bit 14 TDE: Trigger DMA request enable
0: Trigger DMA request disabled.
1: Trigger DMA request enabled.
Bit 13 Reserved, always read as 0
Bit 12 CC4DE: Capture/Compare 4 DMA request enable
0: CC4 DMA request disabled.
1: CC4 DMA request enabled.
Bit 11 CC3DE: Capture/Compare 3 DMA request enable
0: CC3 DMA request disabled.
1: CC3 DMA request enabled.
Bit 10 CC2DE: Capture/Compare 2 DMA request enable
0: CC2 DMA request disabled.
1: CC2 DMA request enabled.
Bit 9 CC1DE: Capture/Compare 1 DMA request enable
0: CC1 DMA request disabled.
1: CC1 DMA request enabled.
Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled.
1: Update DMA request enabled.
Bit 7 Reserved, must be kept at reset value.
Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled.
1: Trigger interrupt enabled.
Bit 5 Reserved, must be kept at reset value.
Bit 4 CC4IE: Capture/Compare 4 interrupt enable
0: CC4 interrupt disabled.
1: CC4 interrupt enabled.
Bit 3 CC3IE: Capture/Compare 3 interrupt enable
0: CC3 interrupt disabled
1: CC3 interrupt enabled

RM0402 Rev 6 527/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Bit 2 CC2IE: Capture/Compare 2 interrupt enable

0: CC2 interrupt disabled
1: CC2 interrupt enabled
Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled
1: CC1 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled

17.4.5 TIMx status register (TIMx_SR)

Address offset: 0x10
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. CC4OF CC3OF CC2OF CC1OF Res. Res. TIF Res. CC4IF CC3IF CC2IF CC1IF UIF
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0

Bits 15:13 Reserved, must be kept at reset value.

Bit 12 CC4OF: Capture/Compare 4 overcapture flag
refer to CC1OF description
Bit 11 CC3OF: Capture/Compare 3 overcapture flag
refer to CC1OF description
Bit 10 CC2OF: Capture/compare 2 overcapture flag
refer to CC1OF description
Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0.
0: No overcapture has been detected
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bits 8:7 Reserved, must be kept at reset value.
Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred
1: Trigger interrupt pending
Bit 5 Reserved, must be kept at reset value.
Bit 4 CC4IF: Capture/Compare 4 interrupt flag
refer to CC1IF description
Bit 3 CC3IF: Capture/Compare 3 interrupt flag
refer to CC1IF description

528/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Bit 2 CC2IF: Capture/Compare 2 interrupt flag

refer to CC1IF description
Bit 1 CC1IF: Capture/compare 1 interrupt flag
If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value, with some
exception in center-aligned mode (refer to the CMS bits in the TIMx_CR1 register
description). It is cleared by software.
0: No match
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF bit
goes high on the counter overflow (in upcounting and up/down-counting modes) or underflow
(in downcounting mode)
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred
1: The counter value has been captured in TIMx_CCR1 register (An edge has been detected
on IC1 which matches the selected polarity)
Bit 0 UIF: Update interrupt flag
″ This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
″ At overflow or underflow (for TIM2 to TIM5) and if UDIS=0 in the TIMx_CR1 register.
″ When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=0
and UDIS=0 in the TIMx_CR1 register.
When CNT is reinitialized by a trigger event (refer to the synchro control register description),
if URS=0 and UDIS=0 in the TIMx_CR1 register.

RM0402 Rev 6 529/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4.6 TIMx event generation register (TIMx_EGR)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. TG Res. CC4G CC3G CC2G CC1G UG
w w w w w w

Bits 15:7 Reserved, must be kept at reset value.

Bit 6 TG: Trigger generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: The TIF flag is set in TIMx_SR register. Related interrupt or DMA transfer can occur if
enabled.
Bit 5 Reserved, must be kept at reset value.
Bit 4 CC4G: Capture/compare 4 generation
refer to CC1G description
Bit 3 CC3G: Capture/compare 3 generation
refer to CC1G description
Bit 2 CC2G: Capture/compare 2 generation
refer to CC1G description
Bit 1 CC1G: Capture/compare 1 generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
CC1IF flag is set, Corresponding interrupt or DMA request is sent if enabled.
If channel CC1 is configured as input:
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt or DMA request is sent if enabled. The CC1OF flag is set if the
CC1IF flag was already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Re-initialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared if
the center-aligned mode is selected or if DIR=0 (upcounting), else it takes the auto-reload
value (TIMx_ARR) if DIR=1 (downcounting).

530/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

17.4.7 TIMx capture/compare mode register 1 (TIMx_CCMR1)

Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. So one must take care that the same bit
can have a different meaning for the input stage and for the output stage.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC2CE OC2M[2:0] OC2PE OC2FE OC1CE OC1M[2:0] OC1PE OC1FE
CC2S[1:0] CC1S[1:0]
IC2F[3:0] IC2PSC[1:0] IC1F[3:0] IC1PSC[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Output compare mode

Bit 15 OC2CE: Output compare 2 clear enable

Bits 14:12 OC2M[2:0]: Output compare 2 mode
Bit 11 OC2PE: Output compare 2 preload enable
Bit 10 OC2FE: Output compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bit 7 OC1CE: Output compare 1 clear enable
OC1CE: Output Compare 1 Clear Enable
0: OC1Ref is not affected by the ETRF input
1: OC1Ref is cleared as soon as a High level is detected on ETRF input

RM0402 Rev 6 531/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Bits 6:4 OC1M: Output compare 1 mode

These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas OC1 and OC1N active level depends
on CC1P and CC1NP bits.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the
counter TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNT<TIMx_CCR1
else inactive. In downcounting, channel 1 is inactive (OC1REF=‘0) as long as
TIMx_CNT>TIMx_CCR1 else active (OC1REF=1).
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNT<TIMx_CCR1
else active. In downcounting, channel 1 is active as long as TIMx_CNT>TIMx_CCR1 else
inactive.
Note: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode
to “PWM” mode.
Bit 3 OC1PE: Output compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
Note: 1: These bits can not be modified as long as LOCK level 3 has been programmed
(LOCK bits in TIMx_BDTR register) and CC1S=00 (the channel is configured in
output).
2: The PWM mode can be used without validating the preload register only in one-
pulse mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. Then, OC
is set to the compare level independently from the result of the comparison. Delay to sample
the trigger input and to activate CC1 output is reduced to 3 clock cycles. OCFE acts only if
the channel is configured in PWM1 or PWM2 mode.
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output.
01: CC1 channel is configured as input, IC1 is mapped on TI1.
10: CC1 channel is configured as input, IC1 is mapped on TI2.
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

532/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Input capture mode

Bits 15:12 IC2F: Input capture 2 filter

Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
Bits 9:8 CC2S: Capture/compare 2 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output.
01: CC2 channel is configured as input, IC2 is mapped on TI2.
10: CC2 channel is configured as input, IC2 is mapped on TI1.
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bits 7:4 IC1F: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter
applied to TI1. The digital filter is made of an event counter in which N consecutive events
are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS
0001: fSAMPLING=fCK_INT, N=2
0010: fSAMPLING=fCK_INT, N=4
0011: fSAMPLING=fCK_INT, N=8
0100: fSAMPLING=fDTS/2, N=6
0101: fSAMPLING=fDTS/2, N=8
0110: fSAMPLING=fDTS/4, N=6
0111: fSAMPLING=fDTS/4, N=8
1000: fSAMPLING=fDTS/8, N=6
1001: fSAMPLING=fDTS/8, N=8
1010: fSAMPLING=fDTS/16, N=5
1011: fSAMPLING=fDTS/16, N=6
1100: fSAMPLING=fDTS/16, N=8
1101: fSAMPLING=fDTS/32, N=5
1110: fSAMPLING=fDTS/32, N=6
1111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=0 (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

RM0402 Rev 6 533/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4.8 TIMx capture/compare mode register 2 (TIMx_CCMR2)

Address offset: 0x1C
Reset value: 0x0000
Refer to the above CCMR1 register description.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OC4CE OC4M[2:0] OC4PE OC4FE OC3CE OC3M[2:0] OC3PE OC3FE
CC4S[1:0] CC3S[1:0]
IC4F[3:0] IC4PSC[1:0] IC3F[3:0] IC3PSC[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Output compare mode

Bit 15 OC4CE: Output compare 4 clear enable

Bits 14:12 OC4M: Output compare 4 mode
Bit 11 OC4PE: Output compare 4 preload enable
Bit 10 OC4FE: Output compare 4 fast enable
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = 0 in TIMx_CCER).
Bit 7 OC3CE: Output compare 3 clear enable
Bits 6:4 OC3M: Output compare 3 mode
Bit 3 OC3PE: Output compare 3 preload enable
Bit 2 OC3FE: Output compare 3 fast enable
Bits 1:0 CC3S: Capture/Compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = 0 in TIMx_CCER).

534/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

Input capture mode

Bits 15:12 IC4F: Input capture 4 filter

Bits 11:10 IC4PSC: Input capture 4 prescaler
Bits 9:8 CC4S: Capture/Compare 4 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC4 channel is configured as output
01: CC4 channel is configured as input, IC4 is mapped on TI4
10: CC4 channel is configured as input, IC4 is mapped on TI3
11: CC4 channel is configured as input, IC4 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC4S bits are writable only when the channel is OFF (CC4E = 0 in TIMx_CCER).
Bits 7:4 IC3F: Input capture 3 filter
Bits 3:2 IC3PSC: Input capture 3 prescaler
Bits 1:0 CC3S: Capture/Compare 3 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC3 channel is configured as output
01: CC3 channel is configured as input, IC3 is mapped on TI3
10: CC3 channel is configured as input, IC3 is mapped on TI4
11: CC3 channel is configured as input, IC3 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: CC3S bits are writable only when the channel is OFF (CC3E = 0 in TIMx_CCER).

17.4.9 TIMx capture/compare enable register (TIMx_CCER)

Address offset: 0x20
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CC4NP Res. CC4P CC4E CC3NP Res. CC3P CC3E CC2NP Res. CC2P CC2E CC1NP Res. CC1P CC1E
rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 CC4NP: Capture/Compare 4 output Polarity.

Refer to CC1NP description
Bit 14 Reserved, must be kept at reset value.
Bit 13 CC4P: Capture/Compare 4 output Polarity.
refer to CC1P description
Bit 12 CC4E: Capture/Compare 4 output enable.
refer to CC1E description
Bit 11 CC3NP: Capture/Compare 3 output Polarity.
refer to CC1NP description
Bit 10 Reserved, must be kept at reset value.
Bit 9 CC3P: Capture/Compare 3 output Polarity.
refer to CC1P description
Bit 8 CC3E: Capture/Compare 3 output enable.
refer to CC1E description

RM0402 Rev 6 535/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

Bit 7 CC2NP: Capture/Compare 2 output Polarity.

refer to CC1NP description
Bit 6 Reserved, must be kept at reset value.
Bit 5 CC2P: Capture/Compare 2 output Polarity.
refer to CC1P description
Bit 4 CC2E: Capture/Compare 2 output enable.
refer to CC1E description
Bit 3 CC1NP: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
CC1NP must be kept cleared in this case.
CC1 channel configured as input:
This bit is used in conjunction with CC1P to define TI1FP1/TI2FP1 polarity. refer to CC1P
description.
Bit 2 Reserved, must be kept at reset value.
Bit 1 CC1P: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
0: OC1 active high
1: OC1 active low
CC1 channel configured as input:
CC1NP/CC1P bits select TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TIxFP1 rising edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is not inverted (trigger in gated mode, encoder mode).
01: inverted/falling edge
Circuit is sensitive to TIxFP1 falling edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is inverted (trigger in gated mode, encoder mode).
10: reserved, do not use this configuration.
11: noninverted/both edges
Circuit is sensitive to both TIxFP1 rising and falling edges (capture, trigger in reset, external
clock or trigger mode), TIxFP1 is not inverted (trigger in gated mode). This configuration
must not be used for encoder mode.
Bit 0 CC1E: Capture/Compare 1 output enable.
CC1 channel configured as output:
0: Off - OC1 is not active
1: On - OC1 signal is output on the corresponding output pin
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled
1: Capture enabled

Table 101. Output control bit for standard OCx channels

CCxE bit OCx output state
0 Output Disabled (OCx=0, OCx_EN=0)
1 OCx=OCxREF + Polarity, OCx_EN=1

Note: The state of the external IO pins connected to the standard OCx channels depends on the
OCx channel state and the GPIO registers.

536/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

17.4.10 TIMx counter (TIMx_CNT)

Address offset: 0x24
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CNT[15:0]: Counter value

17.4.11 TIMx prescaler (TIMx_PSC)

Address offset: 0x28
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 PSC[15:0]: Prescaler value

The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event.

17.4.12 TIMx auto-reload register (TIMx_ARR)

Address offset: 0x2C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 ARR[15:0]: Auto-reload value

ARR is the value to be loaded in the actual auto-reload register.
Refer to the Section 17.3.1: Time-base unit on page 485 for more details about ARR update
and behavior.
The counter is blocked while the auto-reload value is null.

RM0402 Rev 6 537/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4.13 TIMx capture/compare register 1 (TIMx_CCR1)

Address offset: 0x34
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CCR1[31:16] (depending on timers)
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 CCR1[31:16]: High Capture/Compare 1 value (on TIM2 and TIM5).
Bits 15:0 CCR1[15:0]: Low Capture/Compare 1 value
If channel CC1 is configured as output:
CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(bit OC1PE). Else the preload value is copied in the active capture/compare 1 register when
an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.

If channel CC1is configured as input:

CCR1 is the counter value transferred by the last input capture 1 event (IC1).

17.4.14 TIMx capture/compare register 2 (TIMx_CCR2)

Address offset: 0x38
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CCR2[31:16] (depending on timers)
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR2[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 CCR2[31:16]: High Capture/Compare 2 value (on TIM2 and TIM5).
Bits 15:0 CCR2[15:0]: Low Capture/Compare 2 value
If channel CC2 is configured as output:
CCR2 is the value to be loaded in the actual capture/compare 2 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register (bit
OC2PE). Else the preload value is copied in the active capture/compare 2 register when an
update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC2 output.
If channel CC2 is configured as input:
CCR2 is the counter value transferred by the last input capture 2 event (IC2).

538/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

17.4.15 TIMx capture/compare register 3 (TIMx_CCR3)

Address offset: 0x3C
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CCR3[31:16] (depending on timers)
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR3[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 CCR3[31:16]: High Capture/Compare 3 value (on TIM2 and TIM5).
Bits 15:0 CCR3[15:0]: Low Capture/Compare value
If channel CC3 is configured as output:
CCR3 is the value to be loaded in the actual capture/compare 3 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register (bit
OC3PE). Else the preload value is copied in the active capture/compare 3 register when an
update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC3 output.
If channel CC3 is configured as input:
CCR3 is the counter value transferred by the last input capture 3 event (IC3).

17.4.16 TIMx capture/compare register 4 (TIMx_CCR4)

Address offset: 0x40
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CCR4[31:16] (depending on timers)
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR4[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 CCR4[31:16]: High Capture/Compare 4 value (onTIM2 and TIM5).

Bits 15:0 CCR4[15:0]: Low Capture/Compare value
1. if CC4 channel is configured as output (CC4S bits):
CCR4 is the value to be loaded in the actual capture/compare 4 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR register
(bit OC4PE). Else the preload value is copied in the active capture/compare 4 register
when an update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signalled on OC4 output.
2. if CC4 channel is configured as input (CC4S bits in TIMx_CCMR4 register):
CCR4 is the counter value transferred by the last input capture 4 event (IC4).

RM0402 Rev 6 539/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4.17 TIMx DMA control register (TIMx_DCR)

Address offset: 0x48
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. DBL[4:0] Res. Res. Res. DBA[4:0]
rw rw rw rw rw rw rw rw rw rw

Bits 15:13 Reserved, must be kept at reset value.

Bits 12:8 DBL[4:0]: DMA burst length
This 5-bit vector defines the number of DMA transfers (the timer recognizes a burst transfer
when a read or a write access is done to the TIMx_DMAR address).
00000: 1 transfer,
00001: 2 transfers,
00010: 3 transfers,
...
10001: 18 transfers.
Bits 7:5 Reserved, must be kept at reset value.
Bits 4:0 DBA[4:0]: DMA base address
This 5-bit vector defines the base-address for DMA transfers (when read/write access are
done through the TIMx_DMAR address). DBA is defined as an offset starting from the
address of the TIMx_CR1 register.
Example:
00000: TIMx_CR1,
00001: TIMx_CR2,
00010: TIMx_SMCR,
...
Example: Let us consider the following transfer: DBL = 7 transfers & DBA = TIMx_CR1. In this
case the transfer is done to/from 7 registers starting from the TIMx_CR1 address.

17.4.18 TIMx DMA address for full transfer (TIMx_DMAR)

Address offset: 0x4C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMAB[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 DMAB[15:0]: DMA register for burst accesses

A read or write operation to the DMAR register accesses the register located at the address
(TIMx_CR1 address) + (DBA + DMA index) x 4
where TIMx_CR1 address is the address of the control register 1, DBA is the DMA base
address configured in TIMx_DCR register, DMA index is automatically controlled by the
DMA transfer, and ranges from 0 to DBL (DBL configured in TIMx_DCR).

Example of how to use the DMA burst feature

In this example the timer DMA burst feature is used to update the contents of the CCRx
registers (x = 2, 3, 4) with the DMA transferring half words into the CCRx registers.
This is done in the following steps:

540/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM2 to TIM5)

1. Configure the corresponding DMA channel as follows:

– DMA channel peripheral address is the DMAR register address
– DMA channel memory address is the address of the buffer in the RAM containing
the data to be transferred by DMA into CCRx registers.
– Number of data to transfer = 3 (See note below).
– Circular mode disabled.
2. Configure the DCR register by configuring the DBA and DBL bit fields as follows:
DBL = 3 transfers, DBA = 0xE.
3. Enable the TIMx update DMA request (set the UDE bit in the DIER register).
4. Enable TIMx
5. Enable the DMA channel
Note: This example is for the case where every CCRx register to be updated once. If every CCRx
register is to be updated twice for example, the number of data to transfer should be 6. Let's
take the example of a buffer in the RAM containing data1, data2, data3, data4, data5 and
data6. The data is transferred to the CCRx registers as follows: on the first update DMA
request, data1 is transferred to CCR2, data2 is transferred to CCR3, data3 is transferred to
CCR4 and on the second update DMA request, data4 is transferred to CCR2, data5 is
transferred to CCR3 and data6 is transferred to CCR4.

17.4.19 TIM2 option register (TIM2_OR)

Address offset: 0x50
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. ITR1_RMP Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
rw rw

Bits 15:12 Reserved, must be kept at reset value.

Bits 11:10 ITR1_RMP: Internal trigger 1 remap
Set and cleared by software.
00: TIM8_TRGOUT
01: Reserved
10: OTG FS SOF is connected to the TIM2_ITR1 input
11: OTG HS SOF is connected to the TIM2_ITR1 input
Bits 9:0 Reserved, must be kept at reset value.

RM0402 Rev 6 541/1163

544
General-purpose timers (TIM2 to TIM5) RM0402

17.4.20 TIM5 option register (TIM5_OR)

Address offset: 0x50
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. TI4_RMP Res. Res. Res. Res. Res. Res.
rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bits 7:6 TI4_RMP: Timer Input 4 remap
Set and cleared by software.
00: TIM5 Channel4 is connected to the GPIO: Refer to the Alternate function mapping table
in the datasheets.
01: the LSI internal clock is connected to the TIM5_CH4 input for calibration purposes
10: the LSE internal clock is connected to the TIM5_CH4 input for calibration purposes
11: the RTC wakeup interrupt is connected to TIM5_CH4 input for calibration purposes.
Wakeup interrupt should be enabled.
Bits 5:0 Reserved, must be kept at reset value.

542/1163 RM0402 Rev 6

 0x24
0x20
0x18
0x14
0x10
0x08
0x04
0x00

0x1C
0x0C
Offset
RM0402

17.4.21

mode
mode
mode
mode
TIMx_SR
TIMx_CR2
TIMx_CR1

TIMx_CNT
TIMx_EGR
Register

TIMx_DIER

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

TIMx_CCER
TIMx_SMCR

Input Capture
Input Capture

TIMx_CCMR2
TIMx_CCMR2
TIMx_CCMR1
TIMx_CCMR1

Output Compare
Output Compare

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 31

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 30

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 29

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 28

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 27

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
TIMx register map

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 25

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 24

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 23

CNT[31:16]

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 22

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 21

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 20

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 19

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 18

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 17

(TIM2 and TIM5 only, reserved on the other timers)

RM0402 Rev 6
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
0
0
0
0

CC4NP O24CE OC2CE Res. Res. Res. ETP Res. Res. 15

0
0
0
0
0
0
0

Res. Res. Res. TDE ECE Res. Res. 14

0
0
0
0
0
0
0
0

CC4P Res. Res. COMDE Res. Res. 13

[2:0]
[2:0]

IC4F[3:0]
IC2F[3:0]

OC4M
OC2M
[1:0]

0
0
0
0
0
0
0
0
0
ETPS

CC4E Res. CC4OF CC4DE Res. Res. 12

TIMx registers are mapped as described in the table below:

0
0
0
0
0
0
0
0
0

CC3NP OC4PE OC2PE Res. CC3OF CC3DE Res. Res. 11

IC4
IC2

[1:0]
[1:0]

PSC
PSC

0
0
0
0
0
0
0
0

Res. OC4FE OC2FE Res. CC2OF CC2DE Res. Res. 10

0
0
0
0
0
0
0
0
0
0

CC3P Res. CC1OF CC1DE Res. 9

ETF[3:0]
Table 102. TIM2 to TIM5 register map and reset values

[1:0]
[1:0]
[1:0]
[1:0]
[1:0]
CKD

0
0
0
0
0
0
0
0
0

CC4S
CC4S
CC2S
CC2S
CC3E Res. Res. UDE Res. 8

0
0
0
0
0
0
0
0
0

CC2NP OC3CE OC1CE Res. Res. Res. MSM TI1S ARPE 7

CNT[15:0]

0
0
0
0
0
0

0
0
0
0
0

Res. TG TIF TIE 6

[1:0]
CMS

0
0
0
0
0
0
0
0
0

CC2P Res. Res. Res. 5

[2:0]
[2:0]

IC3F[3:0]
IC1F[3:0]

OC3M
OC1M
TS[2:0]

0
0
0
0
0
0
0
0
0
0
0
0

MMS[2:0]

CC2E CC4G CC4IF CC4IE DIR 4

0
0
0
0
0
0
0
0
0
0
0

CC1NP OC3PE OC1PE CC3G CC3IF CC3IE Res. CCDS OPM 3

IC3
IC1

[1:0]
[1:0]

PSC
PSC

0
0
0
0
0
0

0
0
0
0

Res. OC3FE OC1FE CC2G CC2IF CC2IE Res. URS 2

0
0
0
0
0
0
0
0

0
0
0

CC1P CC1G CC1IF CC1IE Res. UDIS 1

[1:0]
[1:0]
[1:0]
[1:0]

0
0
0
0
0
0
0
0

0
0
0

SMS[2:0]

CC3S
CC3S
CC1S
CC1S

CC1E UG UIF UIE Res. CEN 0

543/1163
General-purpose timers (TIM2 to TIM5)

544
General-purpose timers (TIM2 to TIM5) RM0402

Table 102. TIM2 to TIM5 register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIMx_PSC PSC[15:0]
0x28
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ARR[31:16]
TIMx_ARR ARR[15:0]
0x2C (TIM2 and TIM5 only, reserved on the other timers)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
0x30 Reserved

CCR1[31:16]
TIMx_CCR1 CCR1[15:0]
0x34 (TIM2 and TIM5 only, reserved on the other timers)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CCR2[31:16]
TIMx_CCR2 CCR2[15:0]
0x38 (TIM2 and TIM5 only, reserved on the other timers)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CCR3[31:16]
TIMx_CCR3 CCR3[15:0]
0x3C (TIM2 and TIM5 only, reserved on the other timers)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CCR4[31:16]
TIMx_CCR4 CCR4[15:0]
0x40 (TIM2 and TIM5 only, reserved on the other timers)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
0x44 Reserved
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
TIMx_DCR DBL[4:0] DBA[4:0]
0x48
Reset value 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

TIMx_DMAR DMAB[15:0]
0x4C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ITR1_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIM2_OR
RMP
0x50
Reset value 0 0

IT4_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.

TIM5_OR
0x50 RMP

Reset value 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

544/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18 General-purpose timers (TIM9 to TIM14)

18.1 TIM9 to TIM14 introduction

The TIM9 to TIM14 general-purpose timers consist of a 16-bit auto-reload counter driven by
a programmable prescaler.
They may be used for a variety of purposes, including measuring the pulse lengths of input
signals (input capture) or generating output waveforms (output compare, PWM).
Pulse lengths and waveform periods can be modulated from a few microseconds to several
milliseconds using the timer prescaler and the RCC clock controller prescalers.
The TIM9 to TIM14 timers are completely independent, and do not share any resources.
They can be synchronized together as described in Section 18.3.12.

18.2 TIM9 to TIM14 main features

18.2.1 TIM9/TIM12 main features

The features of the TIM9 to TIM14 general-purpose timers include:
• 16-bit auto-reload upcounter
• 16-bit programmable prescaler used to divide the counter clock frequency by any factor
between 1 and 65536 (can be changed “on the fly”)
• Up to 2 independent channels for:
– Input capture
– Output compare
– PWM generation (edge-aligned mode)
– One-pulse mode output
• Synchronization circuit to control the timer with external signals and to interconnect
several timers together
• Interrupt generation on the following events:
– Update: counter overflow, counter initialization (by software or internal trigger)
– Trigger event (counter start, stop, initialization or count by internal trigger)
– Input capture
– Output compare

RM0402 Rev 6 545/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 171. General-purpose timer block diagram (TIM9 and TIM12)

Internal clock (CK_INT)

ITR0
Trigger
ITR1 TGI controller
ITR
ITR2 TRC TRGI
ITR3 Slave
Reset, Enable, Count
mode
TI1F_ED controller

TI1FP1
TI2FP2

U Auto-reload register UI

Stop, Clear
U
CK_PSC PSC CK_CNT CNT
+/-
Prescaler COUNTER
CC1I CC1I
TI1 TI1FP1
Input filter &
IC1 IC1PS U OC1REF output OC1
TIMx_CH1 TI1FP2 Prescaler Capture/Compare 1 register TIMx_CH1
Edge detector control
TRC
CC2I
CC2I
TI2FP1 IC2 IC2PS U
TI2 Input filter & OC2REF output OC2
TIMx_CH2
TIMx_CH2 TI2FP2 Prescaler Capture/Compare 2 register
Edge detector control
TRC

Notes:
Reg Preload registers transferred
to active registers on U event
according to control bit

event

interrupt

ai17190

18.2.2 TIM10/TIM11 and TIM13/TIM14 main features

The features of general-purpose timers TIM10/TIM11 and TIM13/TIM14 include:
• 16-bit auto-reload upcounter
• 16-bit programmable prescaler used to divide the counter clock frequency by any factor
between 1 and 65536 (can be changed “on the fly”)
• independent channel for:
– Input capture
– Output compare
– PWM generation (edge-aligned mode)
– One-pulse mode output
• Interrupt generation on the following events:
– Update: counter overflow, counter initialization (by software)
– Input capture
– Output compare

546/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Figure 172. General-purpose timer block diagram (TIM10/11/13/14)

Internal clock (CK_INT)

Enable
Trigger counter
Controller

U Autoreload register UI

Stop, Clear
U
CK_PSC PSC CK_CNT CNT
+/-
prescaler counter
CC1I CC1I
TI1 TI1FP1 IC1 U
Input filter & IC1PS OC1REF output OC1
TIMx_CH1 Prescaler Capture/Compare 1 register TIMx_CH1
edge detector control

Notes:

Reg Preload registers transferred

to active registers on U event
according to control bit

event

interrupt & DMA output

ai17725c

RM0402 Rev 6 547/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.3 TIM9 to TIM14 functional description

18.3.1 Time-base unit

The main block of the timer is a 16-bit counter with its related auto-reload register. The
counters counts up.
The counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
• Counter register (TIMx_CNT)
• Prescaler register (TIMx_PSC)
• Auto-reload register (TIMx_ARR)
The auto-reload register is preloaded. Writing to or reading from the auto-reload register
accesses the preload register. The content of the preload register are transferred into the
shadow register permanently or at each update event (UEV), depending on the auto-reload
preload enable bit (ARPE) in TIMx_CR1 register. The update event is sent when the counter
reaches the overflow and if the UDIS bit equals 0 in the TIMx_CR1 register. It can also be
generated by software. The generation of the update event is described in details for each
configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in TIMx_CR1 register is set (refer also to the slave mode controller
description to get more details on counter enabling).
Note that the counter starts counting 1 clock cycle after setting the CEN bit in the TIMx_CR1
register.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as this control register is buffered. The new prescaler ratio is
taken into account at the next update event.
Figure 173 and Figure 174 give some examples of the counter behavior when the prescaler
ratio is changed on the fly.

548/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Figure 173. Counter timing diagram with prescaler division change from 1 to 2

CK_PSC

CEN

Timerclock = CK_CNT

Counter register F7 F8 F9 FA FB FC 00 01 02 03

Update event (UEV)

Prescaler control register 0 1

Write a new value in TIMx_PSC

Prescaler buffer 0 1

Prescaler counter 0 0 1 0 1 0 1 0 1
MS31076V2

Figure 174. Counter timing diagram with prescaler division change from 1 to 4

CK_PSC

CEN

Timerclock = CK_CNT

F7 F8 F9 FA FB FC 00 01
Counter register

Update event (UEV)

0 3
Prescaler control register

Write a new value in TIMx_PSC

0 3
Prescaler buffer

0 0 1 2 3 0 1 2 3
Prescaler counter

MS31077V2

RM0402 Rev 6 549/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.3.2 Counter modes

Upcounting mode
In upcounting mode, the counter counts from 0 to the auto-reload value (content of the
TIMx_ARR register), then restarts from 0 and generates a counter overflow event.
Setting the UG bit in the TIMx_EGR register (by software or by using the slave mode
controller on TIM9 and TIM12) also generates an update event.
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This is to avoid updating the shadow registers while writing new values in the
preload registers. Then no update event occurs until the UDIS bit has been written to 0.
However, the counter restarts from 0, as well as the counter of the prescaler (but the
prescale rate does not change). In addition, if the URS bit (update request selection) in
TIMx_CR1 register is set, setting the UG bit generates an update event UEV but without
setting the UIF flag (thus no interrupt is sent). This is to avoid generating both update and
capture interrupts when clearing the counter on the capture event.
When an update event occurs, all the registers are updated and the update flag (UIF bit in
TIMx_SR register) is set (depending on the URS bit):
• The auto-reload shadow register is updated with the preload value (TIMx_ARR),
• The buffer of the prescaler is reloaded with the preload value (content of the TIMx_PSC
register).
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR=0x36.

Figure 175. Counter timing diagram, internal clock divided by 1

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31078V3

550/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Figure 176. Counter timing diagram, internal clock divided by 2

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0034 0035 0036 0000 0001 0002 0003

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31079V3

Figure 177. Counter timing diagram, internal clock divided by 4

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0035 0036 0000 0001

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31080V3

Figure 178. Counter timing diagram, internal clock divided by N

CK_PSC

Timerclock = CK_CNT

Counter register 1F 20 00

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

MS31081V3

RM0402 Rev 6 551/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 179. Counter timing diagram, update event when ARPE=0

(TIMx_ARR not preloaded)

CK_PSC

CEN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag (UIF)

Auto-reload preload register FF 36

Write a new value in TIMx_ARR

MS31082V3

Figure 180. Counter timing diagram, update event when ARPE=1

(TIMx_ARR preloaded)

CK_PSC

CEN

Timerclock = CK_CNT

Counter register F0 F1 F2 F3 F4 F5 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)
Auto-reload preload
F5 36
register

Auto-reload shadow
register F5 36

Write a new value in TIMx_ARR MS31083V2

552/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18.3.3 Clock selection

The counter clock can be provided by the following clock sources:
• Internal clock (CK_INT)
• External clock mode1 (for TIM9 and TIM12): external input pin (TIx)
• Internal trigger inputs (ITRx) (for TIM9 and TIM12): connecting the trigger output from
another timer. Refer to Using one timer as prescaler for another for more details.

Internal clock source (CK_INT)

The internal clock source is the default clock source for TIM10/TIM11 and TIM13/TIM14.
For TIM9 and TIM12, the internal clock source is selected when the slave mode controller is
disabled (SMS=’000’). The CEN bit in the TIMx_CR1 register and the UG bit in the
TIMx_EGR register are then used as control bits and can be changed only by software
(except for UG which remains cleared). As soon as the CEN bit is programmed to 1, the
prescaler is clocked by the internal clock CK_INT.
Figure 181 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.

Figure 181. Control circuit in normal mode, internal clock divided by 1

Internal clock

CEN=CNT_EN

UG

CNT_INIT

Counter clock = CK_CNT = CK_PSC

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

MS31085V2

External clock source mode 1(TIM9 and TIM12)

This mode is selected when SMS=’111’ in the TIMx_SMCR register. The counter can count
at each rising or falling edge on a selected input.

RM0402 Rev 6 553/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 182. TI2 external clock connection example

TIMx_SMCR
TS[2:0]

or TI2F or
TI1F or
ITRx
0xx
TI1_ED
100 TRGI External clock CK_PSC
TI1FP1 mode 1
TI2F_Rising 101
TI2 Edge 0 TI2FP2
Filter 110
detector 1
TI2F_Falling
CK_INT Internal clock
mode
ICF[3:0] CC2P (internal clock)

TIMx_CCMR1 TIMx_CCER
SMS[2:0]
TIMx_SMCR
MS37337V1

For example, to configure the upcounter to count in response to a rising edge on the TI2
input, use the following procedure:
1. Configure channel 2 to detect rising edges on the TI2 input by writing CC2S = ‘01’ in
the TIMx_CCMR1 register.
2. Configure the input filter duration by writing the IC2F[3:0] bits in the TIMx_CCMR1
register (if no filter is needed, keep IC2F=’0000’).
3. Select the rising edge polarity by writing CC2P=’0’ and CC2NP=’0’ in the TIMx_CCER
register.
4. Configure the timer in external clock mode 1 by writing SMS=’111’ in the TIMx_SMCR
register.
5. Select TI2 as the trigger input source by writing TS=’110’ in the TIMx_SMCR register.
6. Enable the counter by writing CEN=’1’ in the TIMx_CR1 register.
Note: The capture prescaler is not used for triggering, so it does not need to be configured.
When a rising edge occurs on TI2, the counter counts once and the TIF flag is set.
The delay between the rising edge on TI2 and the actual clock of the counter is due to the
resynchronization circuit on TI2 input.

Figure 183. Control circuit in external clock mode 1

TI2

CNT_EN

Counter clock = CK_CNT = CK_PSC

Counter register 34 35 36

TIF

Write TIF = 0
MS31087V3

554/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18.3.4 Capture/compare channels

Each Capture/Compare channel is built around a capture/compare register (including a
shadow register), a input stage for capture (with digital filter, multiplexing and prescaler) and
an output stage (with comparator and output control).
Figure 184 to Figure 186 give an overview of one capture/compare channel.
The input stage samples the corresponding TIx input to generate a filtered signal TIxF.
Then, an edge detector with polarity selection generates a signal (TIxFPx) which can be
used as trigger input by the slave mode controller or as the capture command. It is
prescaled before the capture register (ICxPS).

Figure 184. Capture/compare channel (example: channel 1 input stage)

TI1F_ED
To the slave mode controller

TI1 TI1F_Rising
Filter TI1F Edge 0 TI1FP1
fDTS downcounter TI1F_Falling 01
detector 1
TI2FP1 IC1 Divider IC1PS
10
/1, /2, /4, /8
ICF[3:0] CC1P/CC1NP TRC
11
TIMx_CCMR1 TIMx_CCER (from slave mode
TI2F_Rising controller)
0
(from channel 2)
CC1S[1:0] ICPS[1:0] CC1E
TI2F_Falling
1
(from channel 2) TIMx_CCMR1 TIMx_CCER

MS33115V1

The output stage generates an intermediate waveform which is then used for reference:
OCxRef (active high). The polarity acts at the end of the chain.

RM0402 Rev 6 555/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 185. Capture/compare channel 1 main circuit

APB Bus

MCU-peripheral interface

(if 16-bit)
8 8

high
S write CCR1H

low
Read CCR1H S write_in_progress
read_in_progress write CCR1L
Read CCR1L Capture/compare preload register R
R Output CC1S[1]
capture_transfer compare_transfer mode
CC1S[0]
CC1S[1] Input
mode OC1PE
CC1S[0] Capture/compare shadow register OC1PE
UEV
TIM1_CCMR1
Comparator (from time
IC1PS Capture
base unit)
CC1E CNT>CCR1
Counter CNT=CCR1
CC1G
TIM1_EGR

MS31089V2

Figure 186. Output stage of capture/compare channel (channel 1)

To the master
mode controller

CNT > CCR2

Output 0 Output
OC1_REF OC1
CNT = CCR2 mode enable
controller 1 circuit

CC1P

OC2M[2:0] TIMx_CCER CC1E TIMx_CCER

TIMx_CCMR1
ai17720

The capture/compare block is made of one preload register and one shadow register. Write
and read always access the preload register.
In capture mode, captures are actually done in the shadow register, which is copied into the
preload register.
In compare mode, the content of the preload register is copied into the shadow register
which is compared to the counter.

18.3.5 Input capture mode

In Input capture mode, the Capture/Compare Registers (TIMx_CCRx) are used to latch the
value of the counter after a transition detected by the corresponding ICx signal. When a
capture occurs, the corresponding CCXIF flag (TIMx_SR register) is set and an interrupt or
a DMA request can be sent if they are enabled. If a capture occurs while the CCxIF flag was
already high, then the over-capture flag CCxOF (TIMx_SR register) is set. CCxIF can be

556/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

cleared by software by writing it to ‘0’ or by reading the captured data stored in the
TIMx_CCRx register. CCxOF is cleared when it is written with 0.
The following example shows how to capture the counter value in TIMx_CCR1 when TI1
input rises. To do this, use the following procedure:
1. Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the CC1S
bits to ‘01’ in the TIMx_CCMR1 register. As soon as CC1S becomes different from ‘00’,
the channel is configured in input mode and the TIMx_CCR1 register becomes read-
only.
2. Program the appropriate input filter duration in relation with the signal connected to the
timer (by programming the ICxF bits in the TIMx_CCMRx register if the input is one of
the TIx inputs). Let’s imagine that, when toggling, the input signal is not stable during at
must 5 internal clock cycles. We must program a filter duration longer than these 5
clock cycles. We can validate a transition on TI1 when 8 consecutive samples with the
new level have been detected (sampled at fDTS frequency). Then write IC1F bits to
‘0011’ in the TIMx_CCMR1 register.
3. Select the edge of the active transition on the TI1 channel by programming CC1P and
CC1NP bits to ‘00’ in the TIMx_CCER register (rising edge in this case).
4. Program the input prescaler. In our example, we wish the capture to be performed at
each valid transition, so the prescaler is disabled (write IC1PS bits to ‘00’ in the
TIMx_CCMR1 register).
5. Enable capture from the counter into the capture register by setting the CC1E bit in the
TIMx_CCER register.
6. If needed, enable the related interrupt request by setting the CC1IE bit in the
TIMx_DIER register.
When an input capture occurs:
• The TIMx_CCR1 register gets the value of the counter on the active transition.
• CC1IF flag is set (interrupt flag). CC1OF is also set if at least two consecutive captures
occurred whereas the flag was not cleared.
• An interrupt is generated depending on the CC1IE bit.
In order to handle the overcapture, it is recommended to read the data before the
overcapture flag. This is to avoid missing an overcapture which could happen after reading
the flag and before reading the data.
Note: IC interrupt requests can be generated by software by setting the corresponding CCxG bit in
the TIMx_EGR register.

18.3.6 PWM input mode (only for TIM9/12)

This mode is a particular case of input capture mode. The procedure is the same except:
• Two ICx signals are mapped on the same TIx input.
• These 2 ICx signals are active on edges with opposite polarity.
• One of the two TIxFP signals is selected as trigger input and the slave mode controller
is configured in reset mode.
For example, one can measure the period (in TIMx_CCR1 register) and the duty cycle (in
TIMx_CCR2 register) of the PWM applied on TI1 using the following procedure (depending
on CK_INT frequency and prescaler value):

RM0402 Rev 6 557/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

1. Select the active input for TIMx_CCR1: write the CC1S bits to ‘01’ in the TIMx_CCMR1
register (TI1 selected).
2. Select the active polarity for TI1FP1 (used both for capture in TIMx_CCR1 and counter
clear): program the CC1P and CC1NP bits to ‘00’ (active on rising edge).
3. Select the active input for TIMx_CCR2: write the CC2S bits to ‘10’ in the TIMx_CCMR1
register (TI1 selected).
4. Select the active polarity for TI1FP2 (used for capture in TIMx_CCR2): program the
CC2P and CC2NP bits to ‘11’ (active on falling edge).
5. Select the valid trigger input: write the TS bits to ‘101’ in the TIMx_SMCR register
(TI1FP1 selected).
6. Configure the slave mode controller in reset mode: write the SMS bits to ‘100’ in the
TIMx_SMCR register.
7. Enable the captures: write the CC1E and CC2E bits to ‘1’ in the TIMx_CCER register.

Figure 187. PWM input mode timing

TI1

TIMx_CNT 0004 0000 0001 0002 0003 0004 0000

TIMx_CCR1 0004

TIMx_CCR2 0002

IC1 capture IC2 capture IC1 capture

pulse width period
IC2 capture
measurement measurement
reset counter
ai15413

1. The PWM input mode can be used only with the TIMx_CH1/TIMx_CH2 signals due to the fact that only
TI1FP1 and TI2FP2 are connected to the slave mode controller.

18.3.7 Forced output mode

In output mode (CCxS bits = ‘00’ in the TIMx_CCMRx register), each output compare signal
(OCxREF and then OCx) can be forced to active or inactive level directly by software,
independently of any comparison between the output compare register and the counter.
To force an output compare signal (OCXREF/OCx) to its active level, one just needs to write
‘101’ in the OCxM bits in the corresponding TIMx_CCMRx register. Thus OCXREF is forced
high (OCxREF is always active high) and OCx get opposite value to CCxP polarity bit.
For example: CCxP=’0’ (OCx active high) => OCx is forced to high level.
The OCxREF signal can be forced low by writing the OCxM bits to ‘100’ in the
TIMx_CCMRx register.
Anyway, the comparison between the TIMx_CCRx shadow register and the counter is still
performed and allows the flag to be set. Interrupt requests can be sent accordingly. This is
described in the output compare mode section below.

558/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18.3.8 Output compare mode

This function is used to control an output waveform or indicating when a period of time has
elapsed.
When a match is found between the capture/compare register and the counter, the output
compare function:
1. Assigns the corresponding output pin to a programmable value defined by the output
compare mode (OCxM bits in the TIMx_CCMRx register) and the output polarity (CCxP
bit in the TIMx_CCER register). The output pin can keep its level (OCXM=’000’), be set
active (OCxM=’001’), be set inactive (OCxM=’010’) or can toggle (OCxM=’011’) on
match.
2. Sets a flag in the interrupt status register (CCxIF bit in the TIMx_SR register).
3. Generates an interrupt if the corresponding interrupt mask is set (CCXIE bit in the
TIMx_DIER register).
The TIMx_CCRx registers can be programmed with or without preload registers using the
OCxPE bit in the TIMx_CCMRx register.
In output compare mode, the update event UEV has no effect on OCxREF and OCx output.
The timing resolution is one count of the counter. Output compare mode can also be used to
output a single pulse (in One-pulse mode).
Procedure:
1. Select the counter clock (internal, external, prescaler).
2. Write the desired data in the TIMx_ARR and TIMx_CCRx registers.
3. Set the CCxIE bit if an interrupt request is to be generated.
4. Select the output mode. For example:
– Write OCxM = ‘011’ to toggle OCx output pin when CNT matches CCRx
– Write OCxPE = ‘0’ to disable preload register
– Write CCxP = ‘0’ to select active high polarity
– Write CCxE = ‘1’ to enable the output
5. Enable the counter by setting the CEN bit in the TIMx_CR1 register.
The TIMx_CCRx register can be updated at any time by software to control the output
waveform, provided that the preload register is not enabled (OCxPE=’0’, else TIMx_CCRx
shadow register is updated only at the next update event UEV). An example is given in
Figure 188.

RM0402 Rev 6 559/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 188. Output compare mode, toggle on OC1.

Write B201h in the CC1R register

TIM1_CNT 0039 003A 003B B200 B201

TIM1_CCR1 003A B201

OC1REF= OC1

Match detected on CCR1

Interrupt generated if enabled
MS31092V2

18.3.9 PWM mode

Pulse Width Modulation mode allows to generate a signal with a frequency determined by
the value of the TIMx_ARR register and a duty cycle determined by the value of the
TIMx_CCRx register.
The PWM mode can be selected independently on each channel (one PWM per OCx
output) by writing ‘110’ (PWM mode 1) or ‘111’ (PWM mode 2) in the OCxM bits in the
TIMx_CCMRx register. The corresponding preload register must be enabled by setting the
OCxPE bit in the TIMx_CCMRx register, and eventually the auto-reload preload register (in
upcounting or center-aligned modes) by setting the ARPE bit in the TIMx_CR1 register.
As the preload registers are transferred to the shadow registers only when an update event
occurs, before starting the counter, all registers must be initialized by setting the UG bit in
the TIMx_EGR register.
The OCx polarity is software programmable using the CCxP bit in the TIMx_CCER register.
It can be programmed as active high or active low. The OCx output is enabled by the CCxE
bit in the TIMx_CCER register. Refer to the TIMx_CCERx register description for more
details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRx are always compared to determine
whether TIMx_CNT ≤ TIMx_CCRx.
The timer is able to generate PWM in edge-aligned mode only since the counter is
upcounting.

PWM edge-aligned mode

In the following example, we consider PWM mode 1. The reference PWM signal OCxREF is
high as long as TIMx_CNT < TIMx_CCRx else it becomes low. If the compare value in
TIMx_CCRx is greater than the auto-reload value (in TIMx_ARR) then OCxREF is held at
‘1’. If the compare value is 0 then OCxRef is held at ‘0’. Figure 189 shows some edge-
aligned PWM waveforms in an example where TIMx_ARR=8.

560/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Figure 189. Edge-aligned PWM waveforms (ARR=8)

Counter register 0 1 2 3 4 5 6 7 8 0 1

OCXREF
CCRx=4
CCxIF

OCXREF
CCRx=8
CCxIF

OCXREF ‘1’
CCRx>8
CCxIF

OCXREF ‘0’
CCRx=0
CCxIF

MS31093V1

18.3.10 One-pulse mode

One-pulse mode (OPM) is a particular case of the previous modes. It allows the counter to
be started in response to a stimulus and to generate a pulse with a programmable length
after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the
waveform can be done in output compare mode or PWM mode. One-pulse mode is selected
by setting the OPM bit in the TIMx_CR1 register. This makes the counter stop automatically
at the next update event UEV.
A pulse can be correctly generated only if the compare value is different from the counter
initial value. Before starting (when the timer is waiting for the trigger), the configuration must
be as follows:
CNT < CCRx ≤ ARR (in particular, 0 < CCRx)

RM0402 Rev 6 561/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 190. Example of one pulse mode.

TI2

OC1REF
OC1

TIM1_ARR
Counter

TIM1_CCR1

0
tDELAY tPULSE t

MS31099V1

For example one may want to generate a positive pulse on OC1 with a length of tPULSE and
after a delay of tDELAY as soon as a positive edge is detected on the TI2 input pin.
Use TI2FP2 as trigger 1:
1. Map TI2FP2 to TI2 by writing CC2S=’01’ in the TIMx_CCMR1 register.
2. TI2FP2 must detect a rising edge, write CC2P=’0’ and CC2NP = ‘0’ in the TIMx_CCER
register.
3. Configure TI2FP2 as trigger for the slave mode controller (TRGI) by writing TS=’110’ in
the TIMx_SMCR register.
4. TI2FP2 is used to start the counter by writing SMS to ‘110’ in the TIMx_SMCR register
(trigger mode).
The OPM waveform is defined by writing the compare registers (taking into account the
clock frequency and the counter prescaler).
• The tDELAY is defined by the value written in the TIMx_CCR1 register.
• The tPULSE is defined by the difference between the auto-reload value and the compare
value (TIMx_ARR - TIMx_CCR1).
• Let’s say one want to build a waveform with a transition from ‘0’ to ‘1’ when a compare
match occurs and a transition from ‘1’ to ‘0’ when the counter reaches the auto-reload
value. To do this PWM mode 2 must be enabled by writing OC1M=’111’ in the
TIMx_CCMR1 register. Optionally the preload registers can be enabled by writing
OC1PE=’1’ in the TIMx_CCMR1 register and ARPE in the TIMx_CR1 register. In this
case one has to write the compare value in the TIMx_CCR1 register, the auto-reload
value in the TIMx_ARR register, generate an update by setting the UG bit and wait for
external trigger event on TI2. CC1P is written to ‘0’ in this example.
Since only 1 pulse (Single mode) is needed, a 1 must be written in the OPM bit in the
TIMx_CR1 register to stop the counter at the next update event (when the counter rolls over
from the auto-reload value back to 0). When OPM bit in the TIMx_CR1 register is set to '0',
so the Repetitive Mode is selected.

562/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Particular case: OCx fast enable

In One-pulse mode, the edge detection on TIx input set the CEN bit which enables the
counter. Then the comparison between the counter and the compare value makes the
output toggle. But several clock cycles are needed for these operations and it limits the
minimum delay tDELAY min we can get.
If one wants to output a waveform with the minimum delay, the OCxFE bit can be set in the
TIMx_CCMRx register. Then OCxRef (and OCx) are forced in response to the stimulus,
without taking in account the comparison. Its new level is the same as if a compare match
had occurred. OCxFE acts only if the channel is configured in PWM1 or PWM2 mode.

18.3.11 TIM9/12 external trigger synchronization

The TIM9/12 timers can be synchronized with an external trigger in several modes: Reset
mode, Gated mode and Trigger mode.

Slave mode: Reset mode

The counter and its prescaler can be reinitialized in response to an event on a trigger input.
Moreover, if the URS bit from the TIMx_CR1 register is low, an update event UEV is
generated. Then all the preloaded registers (TIMx_ARR, TIMx_CCRx) are updated.
In the following example, the upcounter is cleared in response to a rising edge on TI1 input:
1. Configure the channel 1 to detect rising edges on TI1. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC1F=’0000’). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC1S bits
select the input capture source only, CC1S = ‘01’ in the TIMx_CCMR1 register.
Program CC1P and CC1NP to ‘00’ in TIMx_CCER register to validate the polarity (and
detect rising edges only).
2. Configure the timer in reset mode by writing SMS=’100’ in TIMx_SMCR register. Select
TI1 as the input source by writing TS=’101’ in TIMx_SMCR register.
3. Start the counter by writing CEN=’1’ in the TIMx_CR1 register.
The counter starts counting on the internal clock, then behaves normally until TI1 rising
edge. When TI1 rises, the counter is cleared and restarts from 0. In the meantime, the
trigger flag is set (TIF bit in the TIMx_SR register) and an interrupt request can be sent if
enabled (depending on the TIE bit in TIMx_DIER register).
The following figure shows this behavior when the auto-reload register TIMx_ARR=0x36.
The delay between the rising edge on TI1 and the actual reset of the counter is due to the
resynchronization circuit on TI1 input.

RM0402 Rev 6 563/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Figure 191. Control circuit in reset mode

TI1

UG

Counter clock = ck_cnt = ck_psc

Counter register 30 31 32 33 34 35 36 00 01 02 03 00 01 02 03

TIF

MS31401V2

Slave mode: Gated mode

The counter can be enabled depending on the level of a selected input.
In the following example, the upcounter counts only when TI1 input is low:
1. Configure the channel 1 to detect low levels on TI1. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC1F=’0000’). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC1S bits
select the input capture source only, CC1S=’01’ in TIMx_CCMR1 register. Program
CC1P=’1’ and CC1NP= ‘0’ in TIMx_CCER register to validate the polarity (and detect
low level only).
2. Configure the timer in gated mode by writing SMS=’101’ in TIMx_SMCR register.
Select TI1 as the input source by writing TS=’101’ in TIMx_SMCR register.
3. Enable the counter by writing CEN=’1’ in the TIMx_CR1 register (in gated mode, the
counter doesn’t start if CEN=’0’, whatever is the trigger input level).
The counter starts counting on the internal clock as long as TI1 is low and stops as soon as
TI1 becomes high. The TIF flag in the TIMx_SR register is set both when the counter starts
or stops.
The delay between the rising edge on TI1 and the actual stop of the counter is due to the
resynchronization circuit on TI1 input.

564/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Figure 192. Control circuit in gated mode

TI1

cnt_en

Counter clock = ck_cnt = ck_psc

Counter register 30 31 32 33 34 35 36 37 38

TIF

Write TIF=0

MS31402V1

Slave mode: Trigger mode

The counter can start in response to an event on a selected input.
In the following example, the upcounter starts in response to a rising edge on TI2 input:
1. Configure the channel 2 to detect rising edges on TI2. Configure the input filter duration
(in this example, we do not need any filter, so we keep IC2F=’0000’). The capture
prescaler is not used for triggering, so it does not need to be configured. The CC2S bits
are configured to select the input capture source only, CC2S=’01’ in TIMx_CCMR1
register. Program CC2P=’1’ and CC2NP=’0’ in TIMx_CCER register to validate the
polarity (and detect low level only).
2. Configure the timer in trigger mode by writing SMS=’110’ in TIMx_SMCR register.
Select TI2 as the input source by writing TS=’110’ in TIMx_SMCR register.
When a rising edge occurs on TI2, the counter starts counting on the internal clock and the
TIF flag is set.
The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on TI2 input.

Figure 193. Control circuit in trigger mode

TI2

cnt_en

Counter clock = ck_cnt = ck_psc

Counter register 34 35 36 37 38

TIF

MS31403V1

RM0402 Rev 6 565/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.3.12 Timer synchronization (TIM9/12)

The TIM timers are linked together internally for timer synchronization or chaining. Refer to
Section 17.3.15: Timer synchronization for details.

18.3.13 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the TIMx
counter either continues to work normally or stops, depending on DBG_TIMx_STOP
configuration bit in DBG module. For more details, refer to Section 30.16.2: Debug support
for timers, watchdog, bxCAN and I2C.

18.4 TIM9 and TIM12 registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).

18.4.1 TIM9/12 control register 1 (TIMx_CR1)

Address offset: 0x00
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. CKD[1:0] ARPE Res. Res. Res. OPM URS UDIS CEN
rw rw rw rw rw rw rw

Bits 15:10 Reserved, must be kept at reset value.

Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
10: tDTS = 4 × tCK_INT
11: Reserved
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped on the update event
1: Counter stops counting on the next update event (clearing the CEN bit).

566/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Bit 2 URS: Update request source

This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generates an update interrupt if enabled:
– Counter overflow
– Setting the UG bit
1: Only counter overflow generates an update interrupt if enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable update event (UEV) generation.
0: UEV enabled. An UEV is generated by one of the following events:
– Counter overflow
– Setting the UG bit
Buffered registers are then loaded with their preload values.
1: UEV disabled. No UEV is generated, shadow registers keep their value (ARR, PSC,
CCRx). The counter and the prescaler are reinitialized if the UG bit is set.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
CEN is cleared automatically in one-pulse mode, when an update event occurs.

RM0402 Rev 6 567/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.4.2 TIM9/12 slave mode control register (TIMx_SMCR)

Address offset: 0x08
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. MSM TS[2:0] Res. SMS[2:0]
rw rw rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bit 7 MSM: Master/Slave mode
0: No action
1: The effect of an event on the trigger input (TRGI) is delayed to allow a perfect
synchronization between the current timer and its slaves (through TRGO). It is useful in
order to synchronize several timers on a single external event.
Bits 6:4 TS: Trigger selection
This bitfield selects the trigger input to be used to synchronize the counter.
000: Internal Trigger 0 (ITR0)
001: Internal Trigger 1 (ITR1)
010: Internal Trigger 2 (ITR2)
011: Internal Trigger 3 (ITR3)
100: TI1 Edge Detector (TI1F_ED)
101: Filtered Timer Input 1 (TI1FP1)
110: Filtered Timer Input 2 (TI2FP2)
111: Reserved.
See Table 103 for more details on the meaning of ITRx for each timer.
Note: These bits must be changed only when they are not used (e.g. when SMS=’000’) to
avoid wrong edge detections at the transition.
Bit 3 Reserved, must be kept at reset value.
Bits 2:0 SMS: Slave mode selection
When external signals are selected, the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input (see Input control register and Control register
descriptions.
000: Slave mode disabled - if CEN = 1 then the prescaler is clocked directly by the internal
clock
001: Reserved
010: Reserved
011: Reserved
100: Reset mode - Rising edge of the selected trigger input (TRGI) reinitializes the counter
and generates an update of the registers
101: Gated mode - The counter clock is enabled when the trigger input (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Counter starts and stops
are both controlled
110: Trigger mode - The counter starts on a rising edge of the trigger TRGI (but it is not
reset). Only the start of the counter is controlled
111: External clock mode 1 - Rising edges of the selected trigger (TRGI) clock the counter
Note: The Gated mode must not be used if TI1F_ED is selected as the trigger input
(TS=’100’). Indeed, TI1F_ED outputs 1 pulse for each transition on TI1F, whereas the
Gated mode checks the level of the trigger signal.

568/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Table 103. TIMx internal trigger connections

Slave TIM ITR0 (TS = ‘000’) ITR1 (TS = ‘001’) ITR2 (TS = ‘010’) ITR3 (TS = ‘011’)

TIM9 TIM2 TIM3 TIM10_OC TIM11_OC

TIM12 TIM4 TIM5 TIM13_OC TIM14_OC

18.4.3 TIM9/12 Interrupt enable register (TIMx_DIER)

Address offset: 0x0C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. TIE Res. Res. Res. CC2IE CC1IE UIE
rw rw rw rw

Bits 15:7 Reserved, must be kept at reset value.

Bit 6 TIE: Trigger interrupt enable
0: Trigger interrupt disabled.
1: Trigger interrupt enabled.
Bits 5:3 Reserved, must be kept at reset value.
Bit 2 CC2IE: Capture/Compare 2 interrupt enable
0: CC2 interrupt disabled.
1: CC2 interrupt enabled.
Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled.
1: CC1 interrupt enabled.
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled.
1: Update interrupt enabled.

RM0402 Rev 6 569/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.4.4 TIM9/12 status register (TIMx_SR)

Address offset: 0x10
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. CC2OF CC1OF Res. Res. TIF Res. Res. Res. CC2IF CC1IF UIF
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0

Bits 15:11 Reserved, must be kept at reset value.

Bit 10 CC2OF: Capture/compare 2 overcapture flag
refer to CC1OF description
Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bits 8:7 Reserved, must be kept at reset value.
Bit 6 TIF: Trigger interrupt flag
This flag is set by hardware on trigger event (active edge detected on TRGI input when the
slave mode controller is enabled in all modes but gated mode. It is set when the counter
starts or stops when gated mode is selected. It is cleared by software.
0: No trigger event occurred.
1: Trigger interrupt pending.
Bits 5:3 Reserved, must be kept at reset value.
Bit 2 CC2IF: Capture/Compare 2 interrupt flag
refer to CC1IF description

570/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Bit 1 CC1IF: Capture/compare 1 interrupt flag

If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value. It is cleared by
software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF
bit goes high on the counter overflow.
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred.
1: The counter value has been captured in TIMx_CCR1 register (an edge has been detected
on IC1 which matches the selected polarity).
Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow and if UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if URS=’0’ and
UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by a trigger event (refer to the synchro control register
description), if URS=’0’ and UDIS=’0’ in the TIMx_CR1 register.

RM0402 Rev 6 571/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.4.5 TIM9/12 event generation register (TIMx_EGR)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. TG Res. Res. Res. CC2G CC1G UG
w w w w

Bits 15:7 Reserved, must be kept at reset value.

Bit 6 TG: Trigger generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: The TIF flag is set in the TIMx_SR register. Related interrupt can occur if enabled
Bits 5:3 Reserved, must be kept at reset value.
Bit 2 CC2G: Capture/compare 2 generation
refer to CC1G description
Bit 1 CC1G: Capture/compare 1 generation
This bit is set by software to generate an event, it is automatically cleared by hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
the CC1IF flag is set, the corresponding interrupt is sent if enabled.
If channel CC1 is configured as input:
The current counter value is captured in the TIMx_CCR1 register. The CC1IF flag is set, the
corresponding interrupt is sent if enabled. The CC1OF flag is set if the CC1IF flag was
already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Re-initializes the counter and generates an update of the registers. The prescaler counter
is also cleared and the prescaler ratio is not affected. The counter is cleared.

18.4.6 TIM9/12 capture/compare mode register 1 (TIMx_CCMR1)

Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits in this register have different functions in input and output modes. For a given bit, OCxx
describes its function when the channel is configured in output mode, ICxx describes its
function when the channel is configured in input mode. So one must take care that the same
bit can have different meanings for the input stage and the output stage.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. OC2M[2:0] OC2PE OC2FE Res. OC1M[2:0] OC1PE OC1FE
CC2S[1:0] CC1S[1:0]
IC2F[3:0] IC2PSC[1:0] IC1F[3:0] IC1PSC[1:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

572/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Output compare mode

Bit 15 Reserved, must be kept at reset value.

Bits 14:12 OC2M[2:0]: Output compare 2 mode
Bit 11 OC2PE: Output compare 2 preload enable
Bit 10 OC2FE: Output compare 2 fast enable
Bits 9:8 CC2S[1:0]: Capture/Compare 2 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register
Note: The CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bit 7 Reserved, must be kept at reset value.
Bits 6:4 OC1M: Output compare 1 mode
These bits define the behavior of the output reference signal OC1REF from which OC1 and
OC1N are derived. OC1REF is active high whereas the active levels of OC1 and OC1N
depend on the CC1P and CC1NP bits, respectively.
000: Frozen - The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.(this mode is used to generate a timing
base).
001: Set channel 1 to active level on match. The OC1REF signal is forced high when the
TIMx_CNT counter matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. The OC1REF signal is forced low when the
TIMx_CNT counter matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT=TIMx_CCR1
100: Force inactive level - OC1REF is forced low
101: Force active level - OC1REF is forced high
110: PWM mode 1 - In upcounting, channel 1 is active as long as TIMx_CNT<TIMx_CCR1
else it is inactive. In downcounting, channel 1 is inactive (OC1REF=‘0) as long as
TIMx_CNT>TIMx_CCR1, else it is active (OC1REF=’1’)
111: PWM mode 2 - In upcounting, channel 1 is inactive as long as TIMx_CNT<TIMx_CCR1
else it is active. In downcounting, channel 1 is active as long as TIMx_CNT>TIMx_CCR1
else it is inactive.
Note: In PWM mode 1 or 2, the OCREF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode
to “PWM” mode.

RM0402 Rev 6 573/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Bit 3 OC1PE: Output compare 1 preload enable

0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken into account immediately
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded into the active register at each update event
Note: The PWM mode can be used without validating the preload register only in one-pulse
mode (OPM bit set in the TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on the counter and CCR1 values even when the
trigger is ON. The minimum delay to activate the CC1 output when an edge occurs on the
trigger input is 5 clock cycles
1: An active edge on the trigger input acts like a compare match on the CC1 output. Then,
OC is set to the compare level independently of the result of the comparison. Delay to
sample the trigger input and to activate CC1 output is reduced to 3 clock cycles. OC1FE
acts only if the channel is configured in PWM1 or PWM2 mode.
Bits 1:0 CC1S: Capture/Compare 1 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: The CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

574/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Input capture mode

Bits 15:12 IC2F: Input capture 2 filter

Bits 11:10 IC2PSC[1:0]: Input capture 2 prescaler
Bits 9:8 CC2S: Capture/compare 2 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC2 channel is configured as output
01: CC2 channel is configured as input, IC2 is mapped on TI2
10: CC2 channel is configured as input, IC2 is mapped on TI1
11: CC2 channel is configured as input, IC2 is mapped on TRC. This mode works only if an
internal trigger input is selected through the TS bit (TIMx_SMCR register)
Note: The CC2S bits are writable only when the channel is OFF (CC2E = 0 in TIMx_CCER).
Bits 7:4 IC1F: Input capture 1 filter
This bitfield defines the frequency used to sample the TI1 input and the length of the digital
filter applied to TI1. The digital filter is made of an event counter in which N consecutive
events are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS1000: fSAMPLING=fDTS/8, N=6
0001: fSAMPLING=fCK_INT, N=21001: fSAMPLING=fDTS/8, N=8
0010: fSAMPLING=fCK_INT, N=41010: fSAMPLING=fDTS/16, N=5
0011: fSAMPLING=fCK_INT, N=8 1011: fSAMPLING=fDTS/16, N=6
0100: fSAMPLING=fDTS/2, N=61100: fSAMPLING=fDTS/16, N=8
0101: fSAMPLING=fDTS/2, N=81101: fSAMPLING=fDTS/32, N=5
0110: fSAMPLING=fDTS/4, N=61110: fSAMPLING=fDTS/32, N=6
0111: fSAMPLING=fDTS/4, N=81111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bitfield defines the ratio of the prescaler acting on the CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 selection
This bitfield defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: CC1 channel is configured as input, IC1 is mapped on TI2
11: CC1 channel is configured as input, IC1 is mapped on TRC. This mode is working only if
an internal trigger input is selected through TS bit (TIMx_SMCR register)
Note: The CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

RM0402 Rev 6 575/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.4.7 TIM9/12 capture/compare enable register (TIMx_CCER)

Address offset: 0x20
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. CC2NP Res. CC2P CC2E CC1NP Res. CC1P CC1E
rw rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bit 7 CC2NP: Capture/Compare 2 output Polarity
refer to CC1NP description
Bit 6 Reserved, must be kept at reset value.
Bit 5 CC2P: Capture/Compare 2 output Polarity
refer to CC1P description
Bit 4 CC2E: Capture/Compare 2 output enable
refer to CC1E description
Bit 3 CC1NP: Capture/Compare 1 complementary output Polarity
CC1 channel configured as output: CC1NP must be kept cleared
CC1 channel configured as input: CC1NP is used in conjunction with CC1P to define
TI1FP1/TI2FP1 polarity (refer to CC1P description).
Bit 2 Reserved, must be kept at reset value.
Bit 1 CC1P: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
0: OC1 active high.
1: OC1 active low.
CC1 channel configured as input:
CC1NP/CC1P bits select TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TIxFP1 rising edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is not inverted (trigger in gated mode, encoder mode).
01: inverted/falling edge
Circuit is sensitive to TIxFP1 falling edge (capture, trigger in reset, external clock or trigger
mode), TIxFP1 is inverted (trigger in gated mode, encoder mode).
10: reserved, do not use this configuration.
Note: 11: noninverted/both edges
Circuit is sensitive to both TIxFP1 rising and falling edges (capture, trigger in reset,
external clock or trigger mode), TIxFP1 is not inverted (trigger in gated mode). This
configuration must not be used for encoder mode.
Bit 0 CC1E: Capture/Compare 1 output enable.
CC1 channel configured as output:
0: Off - OC1 is not active.
1: On - OC1 signal is output on the corresponding output pin.
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled.
1: Capture enabled.

576/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Table 104. Output control bit for standard OCx channels

CCxE bit OCx output state

0 Output disabled (OCx=’0’, OCx_EN=’0’)

1 OCx=OCxREF + Polarity, OCx_EN=’1’

Note: The states of the external I/O pins connected to the standard OCx channels depend on the
state of the OCx channel and on the GPIO registers.

18.4.8 TIM9/12 counter (TIMx_CNT)

Address offset: 0x24
Reset value: 0x0000 0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CNT[15:0]: Counter value

18.4.9 TIM9/12 prescaler (TIMx_PSC)

Address offset: 0x28
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 PSC[15:0]: Prescaler value

The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded into the active prescaler register at each update event.

18.4.10 TIM9/12 auto-reload register (TIMx_ARR)

Address offset: 0x2C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 ARR[15:0]: Auto-reload value

ARR is the value to be loaded into the actual auto-reload register.
Refer to Section 18.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.

RM0402 Rev 6 577/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.4.11 TIM9/12 capture/compare register 1 (TIMx_CCR1)

Address offset: 0x34
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR1[15:0]: Capture/Compare 1 value

If channel CC1 is configured as output:
CCR1 is the value to be loaded into the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register
(OC1PE bit). Else the preload value is copied into the active capture/compare 1 register
when an update event occurs.
The active capture/compare register contains the value to be compared to the TIMx_CNT
counter and signaled on the OC1 output.
If channel CC1is configured as input:
CCR1 is the counter value transferred by the last input capture 1 event (IC1).

18.4.12 TIM9/12 capture/compare register 2 (TIMx_CCR2)

Address offset: 0x38
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR2[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR2[15:0]: Capture/Compare 2 value

If channel CC2 is configured as output:
CCR2 is the value to be loaded into the actual capture/compare 2 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR2 register
(OC2PE bit). Else the preload value is copied into the active capture/compare 2 register
when an update event occurs.
The active capture/compare register contains the value to be compared to the TIMx_CNT
counter and signalled on the OC2 output.
If channel CC2 is configured as input:
CCR2 is the counter value transferred by the last input capture 2 event (IC2).

578/1163 RM0402 Rev 6

 0x30
0x28
0x24
0x20
0x18
0x14
0x10
0x08
0x00

0x2C
0x1C
0x0C
Offset
RM0402

18.4.13

mode
mode
TIMx_SR

Reserved
Reserved
TIMx_CR1

TIMx_PSC
TIMx_CNT

TIMx_ARR
TIMx_EGR
Register

TIMx_DIER

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

TIMx_CCER
TIMx_SMCR

Input Capture
TIMx_CCMR1
TIMx_CCMR1
Output Compare
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 28
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 27
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 25
TIM9/12 register map

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 24
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 23
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 21
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 18
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 17

RM0402 Rev 6
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 15

0
0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. 14

0
0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. 13

[2:0]

IC2F[3:0]
OC2M

0
0
0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. 12

0
0
0
0
0
Res. Res. Res. OC2PE Res. Res. Res. Res. Res. 11

IC2

[1:0]
PSC

0
0
0
0
0
0

Res. Res. Res. OC2FE Res. CC2OF Res. Res. Res. 10

Table 105. TIM9/12 register map and reset values

0
0
0
0
0
0
0

Res. Res. Res. Res. CC1OF Res. Res. 9

[1:0]
[1:0]
[1:0]
CKD

0
0
0
0
0
0

CC2S
CC2S

Res. Res. Res. Res. Res. Res. Res. 8

0
0
0
0
0

0
0
0

Res. CC2NP Res. Res. Res. Res. Res. MSM ARPE 7

PSC[15:0]
CNT[15:0]

ARR[15:0]

0
0
0
0
0
0
0
0

Res. Res. Res. TG TIF TIE Res. 6

0
0
0
0
0

0
Res. CC2P Res. Res. Res. Res. Res. 5
[2:0]

IC1F[3:0]
OC1M
TS[2:0]

0
0
0
0
0

0
Res. CC2E Res. Res. Res. Res. Res. 4
TIM9/12 registers are mapped as 16-bit addressable registers as described below:

0
0
0
0
0

0
0

Res. CC1NP Res. OC1PE Res. Res. Res. Res. OPM 3

IC1

[1:0]
PSC

0
0
0
0
0
0
0
0
0
0

Res. Res. Res. OC1FE CC2G CC2IF CC2IE URS 2

0
0
0
0
0

0
0
0
0
0
0

Res. CC1P Res. CC1G CC1IF CC1IE UDIS 1

S
S

[1:0]
[1:0]

CC1
CC1

0
0
0
0
0

0
0
0
0
0
0

SMS[2:0]

Res. CC1E Res. UG UIF UIE CEN 0

579/1163
General-purpose timers (TIM9 to TIM14)

591
General-purpose timers (TIM9 to TIM14) RM0402

Table 105. TIM9/12 register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIMx_CCR1 CCR1[15:0]
0x34
Reset value Res. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
TIMx_CCR2 CCR2[15:0]
0x38
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0x3C to
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Reserved
0x4C

Refer to Section 2.2 on page 49 for the register boundary addresses.

580/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18.5 TIM10/11/13/14 registers

The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).

18.5.1 TIM10/11/13/14 control register 1 (TIMx_CR1)

Address offset: 0x00
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. CKD[1:0] ARPE Res. Res. Res. Res. URS UDIS CEN
rw rw rw rw rw rw

Bits 15:10 Reserved, must be kept at reset value.

Bits 9:8 CKD: Clock division
This bit-field indicates the division ratio between the timer clock (CK_INT) frequency and
sampling clock used by the digital filters (TIx),
00: tDTS = tCK_INT
01: tDTS = 2 × tCK_INT
10: tDTS = 4 × tCK_INT
11: Reserved
Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered
1: TIMx_ARR register is buffered
Bits 6:3 Reserved, must be kept at reset value.
Bit 2 URS: Update request source
This bit is set and cleared by software to select the update interrupt (UEV) sources.
0: Any of the following events generate an UEV if enabled:
– Counter overflow
– Setting the UG bit
1: Only counter overflow generates an UEV if enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable update interrupt (UEV) event
generation.
0: UEV enabled. An UEV is generated by one of the following events:
– Counter overflow
– Setting the UG bit.
Buffered registers are then loaded with their preload values.
1: UEV disabled. No UEV is generated, shadow registers keep their value (ARR, PSC,
CCRx). The counter and the prescaler are reinitialized if the UG bit is set.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled

RM0402 Rev 6 581/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.5.2 TIM10/11/13/14 Interrupt enable register (TIMx_DIER)

Address offset: 0x0C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. CC1IE UIE
rw rw

Bits 15:2 Reserved, must be kept at reset value.

Bit 1 CC1IE: Capture/Compare 1 interrupt enable
0: CC1 interrupt disabled
1: CC1 interrupt enabled
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled
1: Update interrupt enabled

18.5.3 TIM10/11/13/14 status register (TIMx_SR)

Address offset: 0x10
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. CC1OF Res. Res. Res. Res. Res. Res. Res. CC1IF UIF
rc_w0 rc_w0 rc_w0

Bits 15:10 Reserved, must be kept at reset value.

Bit 9 CC1OF: Capture/Compare 1 overcapture flag
This flag is set by hardware only when the corresponding channel is configured in input
capture mode. It is cleared by software by writing it to ‘0’.
0: No overcapture has been detected.
1: The counter value has been captured in TIMx_CCR1 register while CC1IF flag was
already set
Bits 8:2 Reserved, must be kept at reset value.

582/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Bit 1 CC1IF: Capture/compare 1 interrupt flag

If channel CC1 is configured as output:
This flag is set by hardware when the counter matches the compare value. It is cleared by
software.
0: No match.
1: The content of the counter TIMx_CNT matches the content of the TIMx_CCR1 register.
When the contents of TIMx_CCR1 are greater than the contents of TIMx_ARR, the CC1IF bit
goes high on the counter overflow.
If channel CC1 is configured as input:
This bit is set by hardware on a capture. It is cleared by software or by reading the
TIMx_CCR1 register.
0: No input capture occurred.
1: The counter value has been captured in TIMx_CCR1 register (an edge has been detected
on IC1 which matches the selected polarity).
Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow and if UDIS=’0’ in the TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in TIMx_EGR register, if
URS=’0’ and UDIS=’0’ in the TIMx_CR1 register.

18.5.4 TIM10/11/13/14 event generation register (TIMx_EGR)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. CC1G UG
w w

Bits 15:2 Reserved, must be kept at reset value.

Bit 1 CC1G: Capture/compare 1 generation
This bit is set by software in order to generate an event, it is automatically cleared by
hardware.
0: No action
1: A capture/compare event is generated on channel 1:
If channel CC1 is configured as output:
CC1IF flag is set, Corresponding interrupt or is sent if enabled.
If channel CC1 is configured as input:
The current value of the counter is captured in TIMx_CCR1 register. The CC1IF flag is set,
the corresponding interrupt is sent if enabled. The CC1OF flag is set if the CC1IF flag was
already high.
Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action
1: Re-initialize the counter and generates an update of the registers. Note that the prescaler
counter is cleared too (anyway the prescaler ratio is not affected). The counter is cleared.

RM0402 Rev 6 583/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.5.5 TIM10/11/13/14 capture/compare mode register 1

(TIMx_CCMR1)
Address offset: 0x18
Reset value: 0x0000
The channels can be used in input (capture mode) or in output (compare mode). The
direction of a channel is defined by configuring the corresponding CCxS bits. All the other
bits of this register have a different function in input and in output mode. For a given bit,
OCxx describes its function when the channel is configured in output, ICxx describes its
function when the channel is configured in input. So the user must take care that the same
bit can have a different meaning for the input stage and for the output stage.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. OC1M[2:0] OC1PE OC1FE
CC1S[1:0]
Res. Res. Res. Res. Res. Res. Res. Res. IC1F[3:0] IC1PSC[1:0]
rw rw rw rw rw rw rw rw

584/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

Output compare mode

Bits 15:7 Reserved, must be kept at reset value.

Bits 6:4 OC1M: Output compare 1 mode
These bits define the behavior of the output reference signal OC1REF from which OC1 is
derived. OC1REF is active high whereas OC1 active level depends on CC1P bit.
000: Frozen. The comparison between the output compare register TIMx_CCR1 and the
counter TIMx_CNT has no effect on the outputs.
001: Set channel 1 to active level on match. OC1REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
010: Set channel 1 to inactive level on match. OC1REF signal is forced low when the
counter TIMx_CNT matches the capture/compare register 1 (TIMx_CCR1).
011: Toggle - OC1REF toggles when TIMx_CNT = TIMx_CCR1.
100: Force inactive level - OC1REF is forced low.
101: Force active level - OC1REF is forced high.
110: PWM mode 1 - Channel 1 is active as long as TIMx_CNT < TIMx_CCR1 else inactive.
111: PWM mode 2 - Channel 1 is inactive as long as TIMx_CNT < TIMx_CCR1 else active.
Note: In PWM mode 1 or 2, the OCREF level changes when the result of the comparison
changes or when the output compare mode switches from frozen to PWM mode.
Bit 3 OC1PE: Output compare 1 preload enable
0: Preload register on TIMx_CCR1 disabled. TIMx_CCR1 can be written at anytime, the
new value is taken in account immediately.
1: Preload register on TIMx_CCR1 enabled. Read/Write operations access the preload
register. TIMx_CCR1 preload value is loaded in the active register at each update event.
Note: The PWM mode can be used without validating the preload register only in one pulse
mode (OPM bit set in TIMx_CR1 register). Else the behavior is not guaranteed.
Bit 2 OC1FE: Output compare 1 fast enable
This bit is used to accelerate the effect of an event on the trigger in input on the CC output.
0: CC1 behaves normally depending on counter and CCR1 values even when the trigger is
ON. The minimum delay to activate CC1 output when an edge occurs on the trigger input is
5 clock cycles.
1: An active edge on the trigger input acts like a compare match on CC1 output. OC is then
set to the compare level independently of the result of the comparison. Delay to sample the
trigger input and to activate CC1 output is reduced to 3 clock cycles. OC1FE acts only if the
channel is configured in PWM1 or PWM2 mode.
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output.
01: CC1 channel is configured as input, IC1 is mapped on TI1.
10:
11:
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

RM0402 Rev 6 585/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

Input capture mode

Bits 15:8 Reserved, must be kept at reset value.

Bits 7:4 IC1F: Input capture 1 filter
This bit-field defines the frequency used to sample TI1 input and the length of the digital filter
applied to TI1. The digital filter is made of an event counter in which N consecutive events
are needed to validate a transition on the output:
0000: No filter, sampling is done at fDTS1000: fSAMPLING=fDTS/8, N=6
0001: fSAMPLING=fCK_INT, N=21001: fSAMPLING=fDTS/8, N=8
0010: fSAMPLING=fCK_INT, N=41010: fSAMPLING=fDTS/16, N=5
0011: fSAMPLING=fCK_INT, N=81011: fSAMPLING=fDTS/16, N=6
0100: fSAMPLING=fDTS/2, N=61100: fSAMPLING=fDTS/16, N=8
0101: fSAMPLING=fDTS/2, N=81101: fSAMPLING=fDTS/32, N=5
0110: fSAMPLING=fDTS/4, N=61110: fSAMPLING=fDTS/32, N=6
0111: fSAMPLING=fDTS/4, N=81111: fSAMPLING=fDTS/32, N=8
Bits 3:2 IC1PSC: Input capture 1 prescaler
This bit-field defines the ratio of the prescaler acting on CC1 input (IC1).
The prescaler is reset as soon as CC1E=’0’ (TIMx_CCER register).
00: no prescaler, capture is done each time an edge is detected on the capture input
01: capture is done once every 2 events
10: capture is done once every 4 events
11: capture is done once every 8 events
Bits 1:0 CC1S: Capture/Compare 1 selection
This bit-field defines the direction of the channel (input/output) as well as the used input.
00: CC1 channel is configured as output
01: CC1 channel is configured as input, IC1 is mapped on TI1
10: Reserved
11: Reserved
Note: CC1S bits are writable only when the channel is OFF (CC1E = 0 in TIMx_CCER).

586/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18.5.6 TIM10/11/13/14 capture/compare enable register

(TIMx_CCER)
Address offset: 0x20
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. CC1NP Res. CC1P CC1E
rw rw rw

Bits 15:4 Reserved, must be kept at reset value.

Bit 3 CC1NP: Capture/Compare 1 complementary output Polarity.
CC1 channel configured as output: CC1NP must be kept cleared.
CC1 channel configured as input: CC1NP bit is used in conjunction with CC1P to define
TI1FP1 polarity (refer to CC1P description).
Bit 2 Reserved, must be kept at reset value.
Bit 1 CC1P: Capture/Compare 1 output Polarity.
CC1 channel configured as output:
0: OC1 active high
1: OC1 active low
CC1 channel configured as input:
The CC1P bit selects TI1FP1 and TI2FP1 polarity for trigger or capture operations.
00: noninverted/rising edge
Circuit is sensitive to TI1FP1 rising edge (capture mode), TI1FP1 is not inverted.
01: inverted/falling edge
Circuit is sensitive to TI1FP1 falling edge (capture mode), TI1FP1 is inverted.
10: reserved, do not use this configuration.
11: noninverted/both edges
Circuit is sensitive to both TI1FP1 rising and falling edges (capture mode), TI1FP1 is not
inverted.
Bit 0 CC1E: Capture/Compare 1 output enable.
CC1 channel configured as output:
0: Off - OC1 is not active
1: On - OC1 signal is output on the corresponding output pin
CC1 channel configured as input:
This bit determines if a capture of the counter value can actually be done into the input
capture/compare register 1 (TIMx_CCR1) or not.
0: Capture disabled
1: Capture enabled

Table 106. Output control bit for standard OCx channels

CCxE bit OCx output state
0 Output Disabled (OCx=’0’, OCx_EN=’0’)
1 OCx=OCxREF + Polarity, OCx_EN=’1’

Note: The state of the external I/O pins connected to the standard OCx channels depends on the
OCx channel state and the GPIO registers.

RM0402 Rev 6 587/1163

591
General-purpose timers (TIM9 to TIM14) RM0402

18.5.7 TIM10/11/13/14 counter (TIMx_CNT)

Address offset: 0x24
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CNT[15:0]: Counter value

18.5.8 TIM10/11/13/14 prescaler (TIMx_PSC)

Address offset: 0x28
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 PSC[15:0]: Prescaler value

The counter clock frequency CK_CNT is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event.

18.5.9 TIM10/11/13/14 auto-reload register (TIMx_ARR)

Address offset: 0x2C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 ARR[15:0]: Auto-reload value

ARR is the value to be loaded in the actual auto-reload register.
Refer to Section 18.3.1: Time-base unit for more details about ARR update and behavior.
The counter is blocked while the auto-reload value is null.

588/1163 RM0402 Rev 6

RM0402 General-purpose timers (TIM9 to TIM14)

18.5.10 TIM10/11/13/14 capture/compare register 1 (TIMx_CCR1)

Address offset: 0x34
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CCR1[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CCR1[15:0]: Capture/Compare 1 value

If channel CC1 is configured as output:
CCR1 is the value to be loaded in the actual capture/compare 1 register (preload value).
It is loaded permanently if the preload feature is not selected in the TIMx_CCMR1 register (bit
OC1PE). Else the preload value is copied in the active capture/compare 1 register when an
update event occurs.
The active capture/compare register contains the value to be compared to the counter
TIMx_CNT and signaled on OC1 output.

If channel CC1is configured as input:

CCR1 is the counter value transferred by the last input capture 1 event (IC1).

18.5.11 TIM11 option register 1 (TIM11_OR)

Address offset: 0x50
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. TI1_RMP[1:0]
rw

Bits 15:2 Reserved, must be kept at reset value.

Bits 1:0 TI1_RMP[1:0]: TIM11 Input 1 remapping capability
Set and cleared by software.
00,01,11: TIM11 Channel1 is connected to the GPIO (refer to the Alternate function mapping
table in the datasheets).
10: HSE_RTC clock (HSE divided by programmable prescaler) is connected to the
TIM11_CH1 input for measurement purposes.

RM0402 Rev 6 589/1163

591
 0x30
0x28
0x24
0x20
0x18
0x14
0x10
0x08
0x00

0x2C
0x1C
0x0C
Offset

590/1163
18.5.12

mode
mode
TIMx_SR

Reserved
Reserved
TIMx_CR1

TIMx_PSC
TIMx_CNT
TIMx_EGR

TIMx_ARR
Register

TIMx_DIER

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

TIMx_CCER
TIMx_SMCR

Input capture
TIMx_CCMR1
TIMx_CCMR1
Output compare
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 28
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 27
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 25
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 24
General-purpose timers (TIM9 to TIM14)

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 23
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 21
TIM10/11/13/14 register map

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 18
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 17

RM0402 Rev 6
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 15

0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 14

0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 13

0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 12

0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 11

0
0
0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 10

0
0
0
0
0

Res. Res. Res. Res. Res. Res. CC1OF Res. Res. 9

[1:0]
CKD

0
0
0
0
Table 107. TIM10/11/13/14 register map and reset values

Res. Res. Res. Res. Res. Res. Res. Res. Res. 8

0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. ARPE 7

PSC[15:0]
CNT[15:0]

ARR[15:0]

0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 6

0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 5

[2:0]

IC1F[3:0]
OC1M

0
0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 4

0
0
0
0
0
0

Res. CC1NP Res. OC1PE Res. Res. Res. Res. Res. 3

IC1

[1:0]
PSC

0
0
0
0
0
0

Res. Res. Res. OC1FE Res. Res. Res. Res. URS 2

0
0
0
0
0
0
0
0
0
0

Res. CC1P Res. CC1G CC1IF CC1IE Res. UDIS 1

[1:0]
[1:0]

0
0
0
0
0
0
0
0
0
0

CC1S
CC1S
TIMx registers are mapped as 16-bit addressable registers as described in the table below:

Res. CC1E Res. UG UIF UIE Res. CEN 0

RM0402
 0x50
0x34

0x4C
0x38 to
Offset
RM0402

Reserved

TIMx_OR
Register

Reset value

Reset value
TIMx_CCR1

Res. Res. Res. 31

Res. Res. Res. 30
Res. Res. Res. 29
Res. Res. Res. 28
Res. Res. Res. 27
Res. Res. Res. 26
Res. Res. Res. 25
Res. Res. Res. 24
Res. Res. Res. 23
Res. Res. Res. 22
Res. Res. Res. 21
Res. Res. Res. 20
Res. Res. Res. 19
Res. Res. Res. 18
Res. Res. Res. 17

RM0402 Rev 6
Res. Res. Res. 16
0

Res. Res. 15
0

Res. Res. 14
0

Res. Res. 13
0

Res. Res. 12
0

Res. Res. 11
0

Res. Res. 10
0

Res. Res. 9
0

Res. Res. 8
Refer to Section 2.2.2 on page 50 for the register boundary addresses.
0

Res. Res. 7
0
CCR1[15:0]

Res. Res. 6
Table 107. TIM10/11/13/14 register map and reset values (continued)

Res. Res. 5
0

Res. Res. 4
0

Res. Res. 3
0

Res. Res. 2
0

Res. 1
TI1_RMP
0

Res. 0

591/1163
General-purpose timers (TIM9 to TIM14)

591
Basic timers (TIM6/7) RM0402

19 Basic timers (TIM6/7)

19.1 Introduction
The basic timers TIM6, TIM7 consist of a 16-bit auto-reload counter driven by a
programmable prescaler.

19.2 TIM6/7 main features

Basic timer (TIM6/TIM7) features include:
• 16-bit auto-reload upcounter
• 16-bit programmable prescaler used to divide (also “on the fly”) the counter clock
frequency by any factor between 1 and 65536
• Interrupt/DMA generation on the update event: counter overflow

Figure 194. Basic timer block diagram

Trigger TRGO
Internal clock (CK_INT) Controller
TIMxCLK from RCC

Control Reset, enable, Count

Auto-reload register
U UI
Stop, clear or up
U
CK_PSC PSC CK_CNT +
prescaler CNT counter

Notes:
Preload registers transferred
Reg
to active registers on U event
according to control bit

Event

Interrupt & DMA output

MSv34754V1

592/1163 RM0402 Rev 6

RM0402 Basic timers (TIM6/7)

19.3 TIM6/7 functional description

19.3.1 Time-base unit

The main block of the programmable timer is a 16-bit upcounter with its related auto-reload
register. The counter clock can be divided by a prescaler.
The counter, the auto-reload register and the prescaler register can be written or read by
software. This is true even when the counter is running.
The time-base unit includes:
• Counter Register (TIMx_CNT)
• Prescaler Register (TIMx_PSC)
• Auto-Reload Register (TIMx_ARR)
The auto-reload register is preloaded. The preload register is accessed each time an
attempt is made to write or read the auto-reload register. The contents of the preload
register are transferred into the shadow register permanently or at each update event UEV,
depending on the auto-reload preload enable bit (ARPE) in the TIMx_CR1 register. The
update event is sent when the counter reaches the overflow value and if the UDIS bit equals
0 in the TIMx_CR1 register. It can also be generated by software. The generation of the
update event is described in detail for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the
counter enable bit (CEN) in the TIMx_CR1 register is set.
Note that the actual counter enable signal CNT_EN is set 1 clock cycle after CEN.

Prescaler description
The prescaler can divide the counter clock frequency by any factor between 1 and 65536. It
is based on a 16-bit counter controlled through a 16-bit register (in the TIMx_PSC register).
It can be changed on the fly as the TIMx_PSC control register is buffered. The new
prescaler ratio is taken into account at the next update event.
Figure 195 and Figure 196 give some examples of the counter behavior when the prescaler
ratio is changed on the fly.

RM0402 Rev 6 593/1163

604
Basic timers (TIM6/7) RM0402

Figure 195. Counter timing diagram with prescaler division change from 1 to 2

CK_PSC

CEN

Timerclock = CK_CNT

Counter register F7 F8 F9 FA FB FC 00 01 02 03

Update event (UEV)

Prescaler control register 0 1

Write a new value in TIMx_PSC

Prescaler buffer 0 1

Prescaler counter 0 0 1 0 1 0 1 0 1
MS31076V2

Figure 196. Counter timing diagram with prescaler division change from 1 to 4

CK_PSC

CEN

Timerclock = CK_CNT

F7 F8 F9 FA FB FC 00 01
Counter register

Update event (UEV)

0 3
Prescaler control register

Write a new value in TIMx_PSC

0 3
Prescaler buffer

0 0 1 2 3 0 1 2 3
Prescaler counter

MS31077V2

594/1163 RM0402 Rev 6

RM0402 Basic timers (TIM6/7)

19.3.2 Counting mode

The counter counts from 0 to the auto-reload value (contents of the TIMx_ARR register),
then restarts from 0 and generates a counter overflow event.
An update event can be generate at each counter overflow or by setting the UG bit in the
TIMx_EGR register (by software or by using the slave mode controller).
The UEV event can be disabled by software by setting the UDIS bit in the TIMx_CR1
register. This avoids updating the shadow registers while writing new values into the preload
registers. In this way, no update event occurs until the UDIS bit has been written to 0,
however, the counter and the prescaler counter both restart from 0 (but the prescale rate
does not change). In addition, if the URS (update request selection) bit in the TIMx_CR1
register is set, setting the UG bit generates an update event UEV, but the UIF flag is not set
(so no interrupt or DMA request is sent).
When an update event occurs, all the registers are updated and the update flag (UIF bit in
the TIMx_SR register) is set (depending on the URS bit):
• The buffer of the prescaler is reloaded with the preload value (contents of the
TIMx_PSC register)
• The auto-reload shadow register is updated with the preload value (TIMx_ARR)
The following figures show some examples of the counter behavior for different clock
frequencies when TIMx_ARR = 0x36.

Figure 197. Counter timing diagram, internal clock divided by 1

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31078V2

RM0402 Rev 6 595/1163

604
Basic timers (TIM6/7) RM0402

Figure 198. Counter timing diagram, internal clock divided by 2

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0034 0035 0036 0000 0001 0002 0003

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31079V2

Figure 199. Counter timing diagram, internal clock divided by 4

CK_PSC

CNT_EN

Timerclock = CK_CNT

Counter register 0035 0036 0000 0001

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31080V2

596/1163 RM0402 Rev 6

RM0402 Basic timers (TIM6/7)

Figure 200. Counter timing diagram, internal clock divided by N

CK_PSC

Timerclock = CK_CNT

Counter register 1F 20 00

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)

MS31081V2

Figure 201. Counter timing diagram, update event when ARPE = 0 (TIMx_ARR not
preloaded)

CK_PSC

CEN

Timerclock = CK_CNT

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)

Auto-reload preload
register FF 36

Write a new value in TIMx_ARR

MS31082V2

RM0402 Rev 6 597/1163

604
Basic timers (TIM6/7) RM0402

Figure 202. Counter timing diagram, update event when ARPE=1 (TIMx_ARR
preloaded)

CK_PSC

CEN

Timerclock = CK_CNT

Counter register F0 F1 F2 F3 F4 F5 00 01 02 03 04 05 06 07

Counter overflow

Update event (UEV)

Update interrupt flag

(UIF)
Auto-reload preload
F5 36
register

Auto-reload shadow
register F5 36

Write a new value in TIMx_ARR MS31083V2

19.3.3 Clock source

The counter clock is provided by the Internal clock (CK_INT) source.
The CEN (in the TIMx_CR1 register) and UG bits (in the TIMx_EGR register) are actual
control bits and can be changed only by software (except for UG that remains cleared
automatically). As soon as the CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
Figure 203 shows the behavior of the control circuit and the upcounter in normal mode,
without prescaler.

598/1163 RM0402 Rev 6

RM0402 Basic timers (TIM6/7)

Figure 203. Control circuit in normal mode, internal clock divided by 1

Internal clock

CEN=CNT_EN

UG

CNT_INIT

Counter clock = CK_CNT = CK_PSC

Counter register 31 32 33 34 35 36 00 01 02 03 04 05 06 07

MS31085V2

19.3.4 Debug mode

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core - halted), the
TIMx counter either continues to work normally or stops, depending on the
DBG_TIMx_STOP configuration bit in the DBG module. For more details, refer to
Section 30.16.2: Debug support for timers, watchdog, bxCAN and I2C.

RM0402 Rev 6 599/1163

604
Basic timers (TIM6/7) RM0402

19.4 TIM6/7 registers

Refer to Section 1.2: List of abbreviations for registers for a list of abbreviations used in
register descriptions.
The peripheral registers have to be written by half-words (16 bits) or words (32 bits). Read
accesses can be done by bytes (8 bits), half-words (16 bits) or words (32 bits).

19.4.1 TIM6/7 control register 1 (TIMx_CR1)

Address offset: 0x00
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. ARPE Res. Res. Res. OPM URS UDIS CEN
rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bit 7 ARPE: Auto-reload preload enable
0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 OPM: One-pulse mode
0: Counter is not stopped at update event
1: Counter stops counting at the next update event (clearing the CEN bit).
Bit 2 URS: Update request source
This bit is set and cleared by software to select the UEV event sources.
0: Any of the following events generates an update interrupt or DMA request if enabled.
These events can be:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
1: Only counter overflow/underflow generates an update interrupt or DMA request if
enabled.
Bit 1 UDIS: Update disable
This bit is set and cleared by software to enable/disable UEV event generation.
0: UEV enabled. The Update (UEV) event is generated by one of the following events:
– Counter overflow/underflow
– Setting the UG bit
– Update generation through the slave mode controller
Buffered registers are then loaded with their preload values.
1: UEV disabled. The Update event is not generated, shadow registers keep their value
(ARR, PSC). However the counter and the prescaler are reinitialized if the UG bit is set or if
a hardware reset is received from the slave mode controller.
Bit 0 CEN: Counter enable
0: Counter disabled
1: Counter enabled
Note: Gated mode can work only if the CEN bit has been previously set by software.
However trigger mode can set the CEN bit automatically by hardware.
CEN is cleared automatically in one-pulse mode, when an update event occurs.

600/1163 RM0402 Rev 6

RM0402 Basic timers (TIM6/7)

19.4.2 TIM6/7 control register 2 (TIMx_CR2)

Address offset: 0x04
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. MMS[2:0] Res. Res. Res. Res.

rw rw rw

Bits 15:7 Reserved, must be kept at reset value.

Bits 6:4 MMS: Master mode selection
These bits are used to select the information to be sent in master mode to slave timers for
synchronization (TRGO). The combination is as follows:
000: Reset - the UG bit from the TIMx_EGR register is used as a trigger output (TRGO). If
reset is generated by the trigger input (slave mode controller configured in reset mode) then
the signal on TRGO is delayed compared to the actual reset.
001: Enable - the Counter enable signal, CNT_EN, is used as a trigger output (TRGO). It is
useful to start several timers at the same time or to control a window in which a slave timer
is enabled. The Counter Enable signal is generated by a logic OR between CEN control bit
and the trigger input when configured in gated mode.
When the Counter Enable signal is controlled by the trigger input, there is a delay on TRGO,
except if the master/slave mode is selected (see the MSM bit description in the TIMx_SMCR
register).
010: Update - The update event is selected as a trigger output (TRGO). For instance a
master timer can then be used as a prescaler for a slave timer.
Bits 3:0 Reserved, must be kept at reset value.

19.4.3 TIM6/7 DMA/Interrupt enable register (TIMx_DIER)

Address offset: 0x0C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. UDE Res. Res. Res. Res. Res. Res. Res. UIE

rw rw

Bits 15:9 Reserved, must be kept at reset value.

Bit 8 UDE: Update DMA request enable
0: Update DMA request disabled.
1: Update DMA request enabled.
Bits 7:1 Reserved, must be kept at reset value.
Bit 0 UIE: Update interrupt enable
0: Update interrupt disabled.
1: Update interrupt enabled.

RM0402 Rev 6 601/1163

604
Basic timers (TIM6/7) RM0402

19.4.4 TIM6/7 status register (TIMx_SR)

Address offset: 0x10
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. UIF

rc_w0

Bits 15:1 Reserved, must be kept at reset value.

Bit 0 UIF: Update interrupt flag
This bit is set by hardware on an update event. It is cleared by software.
0: No update occurred.
1: Update interrupt pending. This bit is set by hardware when the registers are updated:
– At overflow or underflow regarding the repetition counter value and if UDIS = 0 in the
TIMx_CR1 register.
– When CNT is reinitialized by software using the UG bit in the TIMx_EGR register, if URS = 0
and UDIS = 0 in the TIMx_CR1 register.

19.4.5 TIM6/7 event generation register (TIMx_EGR)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. UG

Bits 15:1 Reserved, must be kept at reset value.

Bit 0 UG: Update generation
This bit can be set by software, it is automatically cleared by hardware.
0: No action.
1: Re-initializes the timer counter and generates an update of the registers. Note that the
prescaler counter is cleared too (but the prescaler ratio is not affected).

19.4.6 TIM6/7 counter (TIMx_CNT)

Address offset: 0x24
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CNT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CNT[15:0]: Counter value

602/1163 RM0402 Rev 6

RM0402 Basic timers (TIM6/7)

19.4.7 TIM6/7 prescaler (TIMx_PSC)

Address offset: 0x28
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PSC[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 PSC[15:0]: Prescaler value

The counter clock frequency (CK_CNT) is equal to fCK_PSC / (PSC[15:0] + 1).
PSC contains the value to be loaded in the active prescaler register at each update event
(including when the counter is cleared through UG bit of TIMx_EGR register or through
trigger controller when configured in “reset mode”).

19.4.8 TIM6/7 auto-reload register (TIMx_ARR)

Address offset: 0x2C
Reset value: 0xFFFF
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ARR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 ARR[15:0]: Auto-reload value

ARR is the value to be loaded into the actual auto-reload register.
Refer to Section 19.3.1: Time-base unit on page 593 for more details about ARR update and
behavior.
The counter is blocked while the auto-reload value is null.

RM0402 Rev 6 603/1163

604
Basic timers (TIM6/7) RM0402

19.4.9 TIM6/7 register map

TIMx registers are mapped as 16-bit addressable registers as described in the table below:

Table 108. TIM6/7 register map and reset values

Offset Register

15

14

13

12

10
11

0
ARPE

UDIS
OPM

CEN
URS
Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.
TIMx_CR1
0x00
Reset value 0 0 0 0 0
Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.
TIMx_CR2 MMS[2:0]
0x04
Reset value 0 0 0

0x08 Res.

UDE
Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

UIE
TIMx_DIER
0x0C
Reset value 0 0
Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

UIF
TIMx_SR
0x10
Reset value 0
Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

Res.

UG
TIMx_EGR
0x14
Reset value 0

0x18 Res.

0x1C Res.

0x20 Res.

TIMx_CNT CNT[15:0]
0x24
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

TIMx_PSC PSC[15:0]
0x28
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

TIMx_ARR ARR[15:0]
0x2C
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

Refer to Section 2.2 on page 49 for the register boundary addresses.

604/1163 RM0402 Rev 6

RM0402 Independent watchdog (IWDG)

20 Independent watchdog (IWDG)

20.1 IWDG introduction

The devices feature two embedded watchdog peripherals that offer a combination of high
safety level, timing accuracy and flexibility of use. Both watchdog peripherals (Independent
and Window) serve to detect and resolve malfunctions due to software failure, and to trigger
system reset or an interrupt (window watchdog only) when the counter reaches a given
timeout value.
The independent watchdog (IWDG) is clocked by its own dedicated low-speed clock (LSI)
and thus stays active even if the main clock fails. The window watchdog (WWDG) clock is
prescaled from the APB1 clock and has a configurable time-window that can be
programmed to detect abnormally late or early application behavior.
The IWDG is best suited for applications that require the watchdog to run as a totally
independent process outside the main application, but have lower timing accuracy
constraints. The WWDG is best suited for applications that require the watchdog to react
within an accurate timing window. For further information on the window watchdog, refer to
Section 21: Window watchdog (WWDG).

20.2 IWDG main features

• Free-running downcounter
• Clocked from an independent RC oscillator (can operate in Standby and Stop modes)
• Reset (if watchdog activated) when the downcounter value of 0x000 is reached

20.3 IWDG functional description

Figure 204 shows the functional blocks of the independent watchdog module.
When the independent watchdog is started by writing the value 0xCCCC in the Key register
(IWDG_KR), the counter starts counting down from the reset value of 0xFFF. When it
reaches the end of count value (0x000) a reset signal is generated (IWDG reset).
Whenever the key value 0xAAAA is written in the IWDG_KR register, the IWDG_RLR value
is reloaded in the counter and the watchdog reset is prevented.

20.3.1 Hardware watchdog

If the “Hardware watchdog” feature is enabled through the device option bits, the watchdog
is automatically enabled at power-on, and will generate a reset unless the Key register is
written by the software before the counter reaches end of count.

20.3.2 Register access protection

Write access to the IWDG_PR and IWDG_RLR registers is protected. To modify them, you
must first write the code 0x5555 in the IWDG_KR register. A write access to this register
with a different value will break the sequence and register access will be protected again.
This implies that it is the case of the reload operation (writing 0xAAAA).

RM0402 Rev 6 605/1163

610
Independent watchdog (IWDG) RM0402

A status register is available to indicate that an update of the prescaler or the down-counter
reload value is on going.

20.3.3 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the IWDG
counter either continues to work normally or stops, depending on DBG_IWDG_STOP
configuration bit in DBG module. For more details, refer to Section 30.16.4: Debug MCU
APB1 freeze register (DBGMCU_APB1_FZ).

Figure 204. Independent watchdog block diagram

CORE
Prescaler register Status register Reload register Key register
IWDG_PR IWDG_SR IWDG_RLR IWDG_KR

12-bit reload value

LSI 8-bit
(40 kHz) prescaler
12-bit downcounter IWDG reset
VDD voltage domain

MS19944V2

Note: The watchdog function is implemented in the VDD voltage domain that is still functional in
Stop and Standby modes.

Table 109. Min/max IWDG timeout period at 32 kHz (LSI)(1)

Min timeout (ms) RL[11:0]= Max timeout (ms) RL[11:0]=
Prescaler divider PR[2:0] bits
0x000 0xFFF

/4 0 0.125 512
/8 1 0.25 1024
/16 2 0.5 2048
/32 3 1 4096
/64 4 2 8192
/128 5 4 16384
/256 6 8 32768
1. These timings are given for a 32 kHz clock but the microcontroller internal RC frequency can vary. Refers
to LSI oscillator characteristics table in device datasheet for from max and min values.

606/1163 RM0402 Rev 6

RM0402 Independent watchdog (IWDG)

20.4 IWDG registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

20.4.1 Key register (IWDG_KR)

Address offset: 0x00
Reset value: 0x0000 0000 (reset by Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
KEY[15:0]
w w w w w w w w w w w w w w w w

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 KEY[15:0]: Key value (write only, read 0000h)
These bits must be written by software at regular intervals with the key value AAAAh,
otherwise the watchdog generates a reset when the counter reaches 0.
Writing the key value 5555h to enable access to the IWDG_PR and IWDG_RLR registers
(see Section 20.3.2)
Writing the key value CCCCh starts the watchdog (except if the hardware watchdog option is
selected)

RM0402 Rev 6 607/1163

610
Independent watchdog (IWDG) RM0402

20.4.2 Prescaler register (IWDG_PR)

Address offset: 0x04
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PR[2:0]
rw rw rw

Bits 31:3 Reserved, must be kept at reset value.

Bits 2:0 PR[2:0]: Prescaler divider
These bits are write access protected seeSection 20.3.2. They are written by software to
select the prescaler divider feeding the counter clock. PVU bit of IWDG_SR must be reset in
order to be able to change the prescaler divider.
000: divider /4
001: divider /8
010: divider /16
011: divider /32
100: divider /64
101: divider /128
110: divider /256
111: divider /256
Note: Reading this register returns the prescaler value from the VDD voltage domain. This
value may not be up to date/valid if a write operation to this register is ongoing. For this
reason the value read from this register is valid only when the PVU bit in the IWDG_SR
register is reset.

608/1163 RM0402 Rev 6

RM0402 Independent watchdog (IWDG)

20.4.3 Reload register (IWDG_RLR)

Address offset: 0x08
Reset value: 0x0000 0FFF (reset by Standby mode)
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. RL[11:0]
rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:12 Reserved, must be kept at reset value.

Bits11:0 RL[11:0]: Watchdog counter reload value
These bits are write access protected see Section 20.3.2. They are written by software to
define the value to be loaded in the watchdog counter each time the value AAAAh is written
in the IWDG_KR register. The watchdog counter counts down from this value. The timeout
period is a function of this value and the clock prescaler. Refer to Table 109.
The RVU bit in the IWDG_SR register must be reset in order to be able to change the reload
value.
Note: Reading this register returns the reload value from the VDD voltage domain. This value
may not be up to date/valid if a write operation to this register is ongoing on this
register. For this reason the value read from this register is valid only when the RVU bit
in the IWDG_SR register is reset.

20.4.4 Status register (IWDG_SR)

Address offset: 0x0C
Reset value: 0x0000 0000 (not reset by Standby mode)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. RVU PVU
r r

Bits 31:2 Reserved, must be kept at reset value.

Bit 1 RVU: Watchdog counter reload value update
This bit is set by hardware to indicate that an update of the reload value is ongoing. It is reset
by hardware when the reload value update operation is completed in the VDD voltage domain
(takes up to 5 RC 40 kHz cycles).
Reload value can be updated only when RVU bit is reset.
Bit 0 PVU: Watchdog prescaler value update
This bit is set by hardware to indicate that an update of the prescaler value is ongoing. It is
reset by hardware when the prescaler update operation is completed in the VDD voltage
domain (takes up to 5 RC 40 kHz cycles).
Prescaler value can be updated only when PVU bit is reset.

RM0402 Rev 6 609/1163

610
Independent watchdog (IWDG) RM0402

Note: If several reload values or prescaler values are used by application, it is mandatory to wait
until RVU bit is reset before changing the reload value and to wait until PVU bit is reset
before changing the prescaler value. However, after updating the prescaler and/or the
reload value it is not necessary to wait until RVU or PVU is reset before continuing code
execution (even in case of low-power mode entry, the write operation is taken into account
and will complete)

20.4.5 IWDG register map

The following table gives the IWDG register map and reset values.

Table 110. IWDG register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
IWDG_KR KEY[15:0]
0x00
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
IWDG_PR PR[2:0]
0x04
Reset value 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
IWDG_RLR RL[11:0]
0x08
Reset value 1 1 1 1 1 1 1 1 1 1 1 1

RVU
PVU
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
IWDG_SR
0x0C
Reset value 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

610/1163 RM0402 Rev 6

RM0402 Window watchdog (WWDG)

21 Window watchdog (WWDG)

21.1 WWDG introduction

The window watchdog is used to detect the occurrence of a software fault, usually
generated by external interference or by unforeseen logical conditions, which causes the
application program to abandon its normal sequence. The watchdog circuit generates an
MCU reset on expiry of a programmed time period, unless the program refreshes the
contents of the downcounter before the T6 bit becomes cleared. An MCU reset is also
generated if the 7-bit downcounter value (in the control register) is refreshed before the
downcounter has reached the window register value. This implies that the counter must be
refreshed in a limited window.

21.2 WWDG main features

• Programmable free-running downcounter
• Conditional reset
– Reset (if watchdog activated) when the downcounter value becomes less than
0x40
– Reset (if watchdog activated) if the downcounter is reloaded outside the window
(see Figure 206)
• Early wakeup interrupt (EWI): triggered (if enabled and the watchdog activated) when
the downcounter is equal to 0x40.

21.3 WWDG functional description

If the watchdog is activated (the WDGA bit is set in the WWDG_CR register) and when the
7-bit downcounter (T[6:0] bits) rolls over from 0x40 to 0x3F (T6 becomes cleared), it initiates
a reset. If the software reloads the counter while the counter is greater than the value stored
in the window register, then a reset is generated.

RM0402 Rev 6 611/1163

617
Window watchdog (WWDG) RM0402

Figure 205. Watchdog block diagram

RESET
Watchdog configuration register (WWDG_CFR)

- W6 W5 W4 W3 W2 W1 W0
comparator
=1 when
T6:0 > W6:0

Write WWDG_CR

Watchdog control register (WWDG_CR)

WDGA T6 T5 T4 T3 T2 T1 T0
7-bit downcounter (CNT)
PCLK1
(from RCC clock controller)

/4096 WDG prescaler

(WDGTB)

MSv37226V2

The application program must write in the WWDG_CR register at regular intervals during
normal operation to prevent an MCU reset. This operation must occur only when the counter
value is lower than the window register value. The value to be stored in the WWDG_CR
register must be between 0xFF and 0xC0.

Enabling the watchdog

The watchdog is always disabled after a reset. It is enabled by setting the WDGA bit in the
WWDG_CR register, then it cannot be disabled again except by a reset.

Controlling the downcounter

This downcounter is free-running, counting down even if the watchdog is disabled. When
the watchdog is enabled, the T6 bit must be set to prevent generating an immediate reset.
The T[5:0] bits contain the number of increments which represents the time delay before the
watchdog produces a reset. The timing varies between a minimum and a maximum value
due to the unknown status of the prescaler when writing to the WWDG_CR register (see
Figure 206). The Configuration register (WWDG_CFR) contains the high limit of the window:
To prevent a reset, the downcounter must be reloaded when its value is lower than the
window register value and greater than 0x3F. Figure 206 describes the window watchdog
process.
Note: The T6 bit can be used to generate a software reset (the WDGA bit is set and the T6 bit is
cleared).

Advanced watchdog interrupt feature

The Early Wakeup Interrupt (EWI) can be used if specific safety operations or data logging
must be performed before the actual reset is generated. The EWI interrupt is enabled by
setting the EWI bit in the WWDG_CFR register. When the downcounter reaches the value
0x40, an EWI interrupt is generated and the corresponding interrupt service routine (ISR)
can be used to trigger specific actions (such as communications or data logging), before
resetting the device.

612/1163 RM0402 Rev 6

RM0402 Window watchdog (WWDG)

In some applications, the EWI interrupt can be used to manage a software system check
and/or system recovery/graceful degradation, without generating a WWDG reset. In this
case, the corresponding interrupt service routine (ISR) should reload the WWDG counter to
avoid the WWDG reset, then trigger the required actions.
The EWI interrupt is cleared by writing '0' to the EWIF bit in the WWDG_SR register.
Note: When the EWI interrupt cannot be served, e.g. due to a system lock in a higher priority task,
the WWDG reset will eventually be generated.

21.4 How to program the watchdog timeout

The formula in Figure 206 must be used to calculate the WWDG timeout.

Warning: When writing to the WWDG_CR register, always write 1 in the

T6 bit to avoid generating an immediate reset.

Figure 206. Window watchdog timing diagram

T[6:0] CNT downcounter

W[6:0]

0x3F

Refresh not allowed Refresh allowed Time

T6 bit

RESET

ai17101c

The formula to calculate the timeout value is given by:

WDGTB[1:0]
t WWDG = t PCLK1 × 4096 × 2 × ( T5:0] + 1 ) ( ms )

where:
tWWDG: WWDG timeout
tPCLK1: APB1 clock period measured in ms
4096: value corresponding to internal divider.

RM0402 Rev 6 613/1163

617
Window watchdog (WWDG) RM0402

As an example, let us assume APB1 frequency is equal to 24 MHz, WDGTB[1:0] is set to 3

and T[5:0] is set to 63:
3
t WWDG = 1 ⁄ 24000 × 4096 × 2 × ( 63 + 1 ) = 21.85 ms

Refer to the datasheets for the minimum and maximum values of the tWWDG.

21.5 Debug mode

When the microcontroller enters debug mode (Cortex®-M4 with FPU core halted), the
WWDG counter either continues to work normally or stops, depending on
DBG_WWDG_STOP configuration bit in DBGMCU module. For more details, refer to
Section 30.16.4: Debug MCU APB1 freeze register (DBGMCU_APB1_FZ).

614/1163 RM0402 Rev 6

RM0402 Window watchdog (WWDG)

21.6 WWDG registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

21.6.1 Control register (WWDG_CR)

Address offset: 0x00
Reset value: 0x0000 007F

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. WDGA T[6:0]
rs rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 WDGA: Activation bit
This bit is set by software and only cleared by hardware after a reset. When WDGA = 1, the
watchdog can generate a reset.
0: Watchdog disabled
1: Watchdog enabled
Bits 6:0 T[6:0]: 7-bit counter (MSB to LSB)
These bits contain the value of the watchdog counter. It is decremented every (4096 x
2WDGTB[1:0]) PCLK1 cycles. A reset is produced when it rolls over from 0x40 to 0x3F (T6
becomes cleared).

RM0402 Rev 6 615/1163

617
Window watchdog (WWDG) RM0402

21.6.2 Configuration register (WWDG_CFR)

Address offset: 0x04
Reset value: 0x0000 007F

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. EWI WDGTB[1:0] W[6:0]
rs rw rw

Bits 31:10 Reserved, must be kept at reset value.

Bit 9 EWI: Early wakeup interrupt
When set, an interrupt occurs whenever the counter reaches the value 0x40. This interrupt is
only cleared by hardware after a reset.
Bits 8:7 WDGTB[1:0]: Timer base
The time base of the prescaler can be modified as follows:
00: CK Counter Clock (PCLK1 div 4096) div 1
01: CK Counter Clock (PCLK1 div 4096) div 2
10: CK Counter Clock (PCLK1 div 4096) div 4
11: CK Counter Clock (PCLK1 div 4096) div 8
Bits 6:0 W[6:0]: 7-bit window value
These bits contain the window value to be compared to the downcounter.

21.6.3 Status register (WWDG_SR)

Address offset: 0x08
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. EWIF
rc_w0

Bits 31:1 Reserved, must be kept at reset value.

Bit 0 EWIF: Early wakeup interrupt flag
This bit is set by hardware when the counter has reached the value 0x40. It must be cleared
by software by writing ‘0’. A write of ‘1’ has no effect. This bit is also set if the interrupt is not
enabled.

616/1163 RM0402 Rev 6

 0x08
0x04
0x00
Offset
21.6.4
RM0402

Register

Reset value
Reset value
Reset value

WWDG_SR
WWDG_CR

WWDG_CFR
Res. Res. Res. 31
Res. Res. Res. 30
Res. Res. Res. 29
Res. Res. Res. 28
Res. Res. Res. 27
Res. Res. Res. 26
Res. Res. Res. 25
WWDG register map

Res. Res. Res. 24

Res. Res. Res. 23
Res. Res. Res. 22
Res. Res. Res. 21
Res. Res. Res. 20
Res. Res. Res. 19
Res. Res. Res. 18

RM0402 Rev 6
Res. Res. Res. 17
Res. Res. Res. 16
Res. Res. Res. 15
Res. Res. Res. 14
Res. Res. Res. 13
Res. Res. Res. 12
Res. Res. Res. 11
Table 111. WWDG register map and reset values

Res. Res. Res. 10

0

Res. EWI Res. 9

The following table gives the WWDG register map and reset values.

Refer to Section 2.2 on page 49 for the register boundary addresses.

0

Res. WDGTB1 Res. 8

0
0

Res. WDGTB0 WDGA 7

1
1

Res. 6
1
1

Res. 5
1
1

Res. 4
1
1

Res. 3
T[6:0]

W[6:0]

1
1

Res. 2
1
1

Res. 1
0
1
1

EWIF 0

617/1163
Window watchdog (WWDG)

617
Real-time clock (RTC) RM0402

22 Real-time clock (RTC)

22.1 Introduction
The real-time clock (RTC) is an independent BCD timer/counter. The RTC provides a
time-of-day clock/calendar, two programmable alarm interrupts, and a periodic
programmable wakeup flag with interrupt capability. The RTC also includes an automatic
wakeup unit to manage low power modes.
Two 32-bit registers contain the seconds, minutes, hours (12- or 24-hour format), day (day
of week), date (day of month), month, and year, expressed in binary coded decimal format
(BCD). The sub-seconds value is also available in binary format.
Compensations for 28-, 29- (leap year), 30-, and 31-day months are performed
automatically. Daylight saving time compensation can also be performed.
Additional 32-bit registers contain the programmable alarm subseconds, seconds, minutes,
hours, day, and date.
A digital calibration feature is available to compensate for any deviation in crystal oscillator
accuracy.
After backup domain reset, all RTC registers are protected against possible parasitic write
accesses.
As long as the supply voltage remains in the operating range, the RTC never stops,
regardless of the device status (Run mode, low power mode or under reset).

22.2 RTC main features

The RTC unit main features are the following (see Figure 207):
• Calendar with subseconds, seconds, minutes, hours (12 or 24 format), day (day of
week), date (day of month), month, and year.
• Daylight saving compensation programmable by software.
• Two programmable alarms with interrupt function. The alarms can be triggered by any
combination of the calendar fields.
• Automatic wakeup unit generating a periodic flag that triggers an automatic wakeup
interrupt.
• Reference clock detection: a more precise second source clock (50 or 60 Hz) can be
used to enhance the calendar precision.
• Accurate synchronization with an external clock using the subsecond shift feature.
• Maskable interrupts/events:
– Alarm A
– Alarm B
– Wakeup interrupt
– Timestamp
– Tamper detection
• Digital calibration circuit (periodic counter correction)
– 5 ppm accuracy

618/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

– 0.95 ppm accuracy, obtained in a calibration window of several seconds

• Timestamp function for event saving (1 event)
• Tamper detection:
– 2 tamper events with configurable filter and internal pull-up.
• 20 backup registers (80 bytes). The backup registers are reset when a tamper
detection event occurs.
• Alternate function output (RTC_OUT) which selects one of the following two outputs:
– RTC_CALIB: 512 Hz or 1 Hz clock output (with an LSE frequency of 32.768 kHz).
This output is enabled by setting the COE bit in the RTC_CR register. It is routed
to the device RTC_AF1 function.
– RTC_ALARM (Alarm A, Alarm B or wakeup).
This output is selected by configuring the OSEL[1:0] bits in the RTC_CR register.
It is routed to the device RTC_AF1 function.
• RTC alternate function inputs:
– RTC_TS: timestamp event detection. It is routed to the device RTC_AF1.
– RTC_TAMP1: TAMPER1 event detection. It is routed to the device RTC_AF1.
– RTC_REFIN: reference clock input (usually the mains, 50 or 60 Hz).

Figure 207. RTC block diagram

RTC_TS Time
512 Hz stamp TSF
registers
1 Hz RTC_CALIB
Ck_apre RTC_OUT
Output
(default 256 Hz)
control RTC_AF1
Alarm A
Coarse (RTC_ALRMAR, ALRAF
Smooth RTC_PRER RTC_PRER RTC_ALRMASSR
Calibration calibration ck_spre =
Asynchronous Synchronous registers)
RTC_CAL RTC_CAL (default
7-bit prescaler 15-bit prescaler
R IBR 1 Hz)
(default = 128) (default = 256) Calendar
Shadow Shadow
RTC_ALARM
LSE Registers Registers
(32.768 Hz) (RTC_SSR) (RTC_TR/DR)
HSE_RTC RTCCLK
(4 MHz max) Alarm B
LSI WUCKSEL[1:0] (RTC_ALRMBR, = ALRBF
RTC_ALRMBSSR
Prescaler registers)
RTC_WUTR
/ 2, 4, 8, 16 WUTF
16-bit wakeup
auto-reload timer
Backup and
RTC tamper
RTC_TAMP1
control
registers
TAMPE

TSE

MSv39307V2

RM0402 Rev 6 619/1163

655
Real-time clock (RTC) RM0402

22.3 RTC functional description

22.3.1 Clock and prescalers

The RTC clock source (RTCCLK) is selected through the clock controller among the LSE
clock, the LSI oscillator clock, and the HSE clock. For more information on the RTC clock
source configuration, refer to Section 6: Reset and clock control (RCC) for STM32F412xx.
A programmable prescaler stage generates a 1 Hz clock which is used to update the
calendar. To minimize power consumption, the prescaler is split into 2 programmable
prescalers (see Figure 207: RTC block diagram):
• A 7-bit asynchronous prescaler configured through the PREDIV_A bits of the
RTC_PRER register.
• A 15-bit synchronous prescaler configured through the PREDIV_S bits of the
RTC_PRER register.
Note: When both prescalers are used, it is recommended to configure the asynchronous prescaler
to a high value to minimize consumption.
The asynchronous prescaler division factor is set to 128, and the synchronous division
factor to 256, to obtain an internal clock frequency of 1 Hz (ck_spre) with an LSE frequency
of 32.768 kHz.
The minimum division factor is 1 and the maximum division factor is 222.
This corresponds to a maximum input frequency of around 4 MHz.
fck_apre is given by the following formula:
f RTCCLK
f CK_APRE = --------------------------------------
-
PREDIV_A + 1

The ck_apre clock is used to clock the binary RTC_SSR subseconds downcounter. When it
reaches 0, RTC_SSR is reloaded with the content of PREDIV_S.
fck_spre is given by the following formula:
f RTCCLK
f CK_SPRE = ----------------------------------------------------------------------------------------------
( PREDIV_S + 1 ) × ( PREDIV_A + 1 )

The ck_spre clock can be used either to update the calendar or as timebase for the 16-bit
wakeup auto-reload timer. To obtain short timeout periods, the 16-bit wakeup auto-reload
timer can also run with the RTCCLK divided by the programmable 4-bit asynchronous
prescaler (see Section 22.3.4 for details).

22.3.2 Real-time clock and calendar

The RTC calendar time and date registers are accessed through shadow registers which
are synchronized with PCLK1 (APB1 clock). They can also be accessed directly in order to
avoid waiting for the synchronization duration.
• RTC_SSR for the subseconds
• RTC_TR for the time
• RTC_DR for the date

620/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Every two RTCCLK periods, the current calendar value is copied into the shadow registers,
and the RSF bit of RTC_ISR register is set (see Section 22.6.4). The copy is not performed
in Stop and Standby mode. When exiting these modes, the shadow registers are updated
after up to two RTCCLK periods.
When the application reads the calendar registers, it accesses the content of the shadow
registers.It is possible to make a direct access to the calendar registers by setting the
BYPSHAD control bit in the RTC_CR register. By default, this bit is cleared, and the user
accesses the shadow registers.
When reading the RTC_SSR, RTC_TR or RTC_DR registers in BYPSHAD=0 mode, the
frequency of the APB clock (fAPB) must be at least 7 times the frequency of the RTC clock
(fRTCCLK).
The shadow registers are reset by system reset.

22.3.3 Programmable alarms

The RTC unit provides two programmable alarms, Alarm A and Alarm B.
The programmable alarm functions are enabled through the ALRAIE and ALRBIE bits in the
RTC_CR register. The ALRAF and ALRBF flags are set to 1 if the calendar subseconds,
seconds, minutes, hours, date or day match the values programmed in the alarm registers
RTC_ALRMASSR/RTC_ALRMAR and RTC_ALRMBSSR/RTC_ALRMBR, respectively.
Each calendar field can be independently selected through the MSKx bits of the
RTC_ALRMAR and RTC_ALRMBR registers, and through the MASKSSx bits of the
RTC_ALRMASSR and RTC_ALRMBSSR registers. The alarm interrupts are enabled
through the ALRAIE and ALRBIE bits in the RTC_CR register.
Alarm A and Alarm B (if enabled by bits OSEL[1:0] in RTC_CR register) can be routed to the
RTC_ALARM output. RTC_ALARM polarity can be configured through bit POL in the
RTC_CR register.
Caution: If the seconds field is selected (MSK0 bit reset in RTC_ALRMAR or RTC_ALRMBR), the
synchronous prescaler division factor set in the RTC_PRER register must be at least 3 to
ensure correct behavior.

22.3.4 Periodic auto-wakeup

The periodic wakeup flag is generated by a 16-bit programmable auto-reload down-counter.
The wakeup timer range can be extended to 17 bits.
The wakeup function is enabled through the WUTE bit in the RTC_CR register.
The wakeup timer clock input can be:
• RTC clock (RTCCLK) divided by 2, 4, 8, or 16.
When RTCCLK is LSE(32.768 kHz), this allows to configure the wakeup interrupt
period from 122 µs to 32 s, with a resolution down to 61µs.
• ck_spre (usually 1 Hz internal clock)
When ck_spre frequency is 1Hz, this allows to achieve a wakeup time from 1 s to
around 36 hours with one-second resolution. This large programmable time range is
divided in 2 parts:
– from 1s to 18 hours when WUCKSEL [2:1] = 10
– and from around 18h to 36h when WUCKSEL[2:1] = 11. In this last case 216 is
added to the 16-bit counter current value.When the initialization sequence is

RM0402 Rev 6 621/1163

655
Real-time clock (RTC) RM0402

complete (see Programming the wakeup timer), the timer starts counting
down.When the wakeup function is enabled, the down-counting remains active in
low power modes. In addition, when it reaches 0, the WUTF flag is set in the
RTC_ISR register, and the wakeup counter is automatically reloaded with its
reload value (RTC_WUTR register value).
The WUTF flag must then be cleared by software.
When the periodic wakeup interrupt is enabled by setting the WUTIE bit in the RTC_CR2
register, it can exit the device from low power modes.
The periodic wakeup flag can be routed to the RTC_ALARM output provided it has been
enabled through bits OSEL[1:0] of RTC_CR register. RTC_ALARM polarity can be
configured through the POL bit in the RTC_CR register.
System reset, as well as low power modes (Sleep, Stop and Standby) have no influence on
the wakeup timer.

22.3.5 RTC initialization and configuration

RTC register access
The RTC registers are 32-bit registers. The APB interface introduces 2 wait-states in RTC
register accesses except on read accesses to calendar shadow registers when
BYPSHAD=0.

RTC register write protection

After system reset, the RTC registers are protected against parasitic write access with the
DBP bit of the PWR power control register (PWR_CR). The DBP bit must be set to enable
RTC registers write access.
After backup domain reset, all the RTC registers are write-protected. Writing to the RTC
registers is enabled by writing a key into the Write Protection register, RTC_WPR.
The following steps are required to unlock the write protection on all the RTC registers
except for RTC_ISR[13:8], RTC_TAFCR, and RTC_BKPxR.
1. Write ‘0xCA’ into the RTC_WPR register.
2. Write ‘0x53’ into the RTC_WPR register.
Writing a wrong key reactivates the write protection.
The protection mechanism is not affected by system reset.

Calendar initialization and configuration

To program the initial time and date calendar values, including the time format and the
prescaler configuration, the following sequence is required:
1. Set INIT bit to 1 in the RTC_ISR register to enter initialization mode. In this mode, the
calendar counter is stopped and its value can be updated.
2. Poll INITF bit of in the RTC_ISR register. The initialization phase mode is entered when
INITF is set to 1. It takes from 1 to 2 RTCCLK clock cycles (due to clock
synchronization).
3. To generate a 1 Hz clock for the calendar counter, program first the synchronous
prescaler factor in RTC_PRER register, and then program the asynchronous prescaler

622/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

factor. Even if only one of the two fields needs to be changed, 2 separate write
accesses must be performed to the RTC_PRER register.
4. Load the initial time and date values in the shadow registers (RTC_TR and RTC_DR),
and configure the time format (12 or 24 hours) through the FMT bit in the RTC_CR
register.
5. Exit the initialization mode by clearing the INIT bit. The actual calendar counter value is
then automatically loaded and the counting restarts after 4 RTCCLK clock cycles.
When the initialization sequence is complete, the calendar starts counting.
Note: After a system reset, the application can read the INITS flag in the RTC_ISR register to
check if the calendar has been initialized or not. If this flag equals 0, the calendar has not
been initialized since the year field is set at its backup domain reset default value (0x00).
To read the calendar after initialization, the software must first check that the RSF flag is set
in the RTC_ISR register.

Daylight saving time

The daylight saving time management is performed through bits SUB1H, ADD1H, and BKP
of the RTC_CR register.
Using SUB1H or ADD1H, the software can subtract or add one hour to the calendar in one
single operation without going through the initialization procedure.
In addition, the software can use the BKP bit to memorize this operation.

Programming the alarm

A similar procedure must be followed to program or update the programmable alarm (Alarm
A or Alarm B):
1. Clear ALRAE or ALRBIE in RTC_CR to disable Alarm A or Alarm B.
2. Poll ALRAWF or ALRBWF in RTC_ISR until it is set to make sure the access to alarm
registers is allowed. This takes 1 to 2 RTCCLK clock cycles (due to clock
synchronization).
3. Program the Alarm A or Alarm B registers (RTC_ALRMASSR/RTC_ALRMAR or
RTC_ALRMBSSR/RTC_ALRMBR).
4. Set ALRAE or ALRBIE in the RTC_CR register to enable Alarm A or Alarm B again.
Note: Each change of the RTC_CR register is taken into account after 1 to 2 RTCCLK clock cycles
due to clock synchronization.

Programming the wakeup timer

The following sequence is required to configure or change the wakeup timer auto-reload
value (WUT[15:0] in RTC_WUTR):
1. Clear WUTE in RTC_CR to disable the wakeup timer.
2. Poll WUTWF until it is set in RTC_ISR to make sure the access to wakeup auto-reload
counter and to WUCKSEL[2:0] bits is allowed. It takes 1 to 2 RTCCLK clock cycles
(due to clock synchronization).
3. Program the wakeup auto-reload value WUT[15:0] and the wakeup clock selection
(WUCKSEL[2:0] bits in RTC_CR).Set WUTE in RTC_CR to enable the timer again.
The wakeup timer restarts down-counting. Due to clock synchronization, the WUTWF
bit is cleared up to 2 RTCCLK clocks cycles after WUTE is cleared.

RM0402 Rev 6 623/1163

655
Real-time clock (RTC) RM0402

22.3.6 Reading the calendar

When BYPSHAD control bit is cleared in the RTC_CR register
To read the RTC calendar registers (RTC_SSR, RTC_TR and RTC_DR) properly, the APB1
clock frequency (fPCLK1) must be equal to or greater than seven times the fRTCCLK RTC
clock frequency. This ensures a secure behavior of the synchronization mechanism.
If the APB1 clock frequency is less than seven times the RTC clock frequency, the software
must read the calendar time and date registers twice. If the second read of the RTC_TR
gives the same result as the first read, this ensures that the data is correct. Otherwise a third
read access must be done. In any case the APB1 clock frequency must never be lower than
the RTC clock frequency.
The RSF bit is set in RTC_ISR register each time the calendar registers are copied into the
RTC_SSR, RTC_TR and RTC_DR shadow registers. The copy is performed every two
RTCCLK cycles. To ensure consistency between the 3 values, reading either RTC_SSR or
RTC_TR locks the values in the higher-order calendar shadow registers until RTC_DR is
read. In case the software makes read accesses to the calendar in a time interval smaller
than 2 RTCCLK periods: RSF must be cleared by software after the first calendar read, and
then the software must wait until RSF is set before reading again the RTC_SSR, RTC_TR
and RTC_DR registers.
After waking up from low power mode (Stop or Standby), RSF must be cleared by software.
The software must then wait until it is set again before reading the RTC_SSR, RTC_TR and
RTC_DR registers.
The RSF bit must be cleared after wakeup and not before entering low power mode.
Note: After a system reset, the software must wait until RSF is set before reading the RTC_SSR,
RTC_TR and RTC_DR registers. Indeed, a system reset resets the shadow registers to
their default values.
After an initialization (refer to Calendar initialization and configuration): the software must
wait until RSF is set before reading the RTC_SSR, RTC_TR and RTC_DR registers.
After synchronization (refer to Section 22.3.8): the software must wait until RSF is set before
reading the RTC_SSR, RTC_TR and RTC_DR registers.

When the BYPSHAD control bit is set in the RTC_CR register (bypass shadow
registers)
Reading the calendar registers gives the values from the calendar counters directly, thus
eliminating the need to wait for the RSF bit to be set. This is especially useful after exiting
from low power modes (STOP or Standby), since the shadow registers are not updated
during these modes.
When the BYPSHAD bit is set to 1, the results of the different registers might not be
coherent with each other if an RTCCLK edge occurs between two read accesses to the
registers. Additionally, the value of one of the registers may be incorrect if an RTCCLK edge
occurs during the read operation. The software must read all the registers twice, and then
compare the results to confirm that the data is coherent and correct. Alternatively, the
software can just compare the two results of the least-significant calendar register.
Note: While BYPSHAD=1, instructions which read the calendar registers require one extra APB
cycle to complete.

624/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.3.7 Resetting the RTC

The calendar shadow registers (RTC_SSR, RTC_TR and RTC_DR) and some bits of the
RTC status register (RTC_ISR) are reset to their default values by all available system reset
sources.
On the contrary, the following registers are resetted to their default values by a backup
domain reset and are not affected by a system reset: the RTC current calendar registers,
the RTC control register (RTC_CR), the prescaler register (RTC_PRER), the RTC
calibration registers (RTC_CALIBR or RTC_CALR), the RTC shift register (RTC_SHIFTR),
the RTC timestamp registers (RTC_TSSSR, RTC_TSTR and RTC_TSDR), the RTC tamper
and alternate function configuration register (RTC_TAFCR), the RTC backup registers
(RTC_BKPxR), the wakeup timer register (RTC_WUTR), the Alarm A and Alarm B registers
(RTC_ALRMASSR/RTC_ALRMAR and RTC_ALRMBSSR/RTC_ALRMBR).
In addition, when clocked by LSE, the RTC keeps on running under system reset if the reset
source is different from the Backup domain reset one (refer to RCC for details about the list
of RTC clock sources not affected by system reset). When a backup domain reset occurs,
the RTC is stopped and all the RTC registers are set to their reset values.

22.3.8 RTC synchronization

The RTC can be synchronized to a remote clock with a high degree of precision. After
reading the sub-second field (RTC_SSR or RTC_TSSSR), a calculation can be made of the
precise offset between the times being maintained by the remote clock and the RTC. The
RTC can then be adjusted to eliminate this offset by “shifting” its clock by a fraction of a
second using RTC_SHIFTR.
RTC_SSR contains the value of the synchronous prescaler’s counter. This allows one to
calculate the exact time being maintained by the RTC down to a resolution of
1 / (PREDIV_S + 1) seconds. As a consequence, the resolution can be improved by
increasing the synchronous prescaler value (PREDIV_S[14:0]. The maximum resolution
allowed (30.52 μs with a 32768 Hz clock) is obtained with PREDIV_S set to 0x7FFF.
However, increasing PREDIV_S means that PREDIV_A must be decreased in order to
maintain the synchronous prescaler’s output at 1 Hz. In this way, the frequency of the
asynchronous prescaler’s output increases, which may increase the RTC dynamic
consumption.
The RTC can be finely adjusted using the RTC shift control register (RTC_SHIFTR). Writing
to RTC_SHIFTR can shift (either delay or advance) the clock by up to a second with a
resolution of 1 / (PREDIV_S + 1) seconds. The shift operation consists of adding the
SUBFS[14:0] value to the synchronous prescaler counter SS[15:0]: this will delay the clock.
If at the same time the ADD1S bit is set, this results in adding one second and at the same
time subtracting a fraction of second, so this will advance the clock.
Caution: Before initiating a shift operation, the user must check that SS[15] = 0 in order to ensure that
no overflow will occur.
As soon as a shift operation is initiated by a write to the RTC_SHIFTR register, the SHPF
flag is set by hardware to indicate that a shift operation is pending. This bit is cleared by
hardware as soon as the shift operation has completed.
Caution: This synchronization feature is not compatible with the reference clock detection feature:
firmware must not write to RTC_SHIFTR when REFCKON=1.

RM0402 Rev 6 625/1163

655
Real-time clock (RTC) RM0402

22.3.9 RTC reference clock detection

The RTC calendar update can be synchronized to a reference clock RTC_REFIN, usually
the mains (50 or 60 Hz). The RTC_REFIN reference clock should have a higher precision
than the 32.768 kHz LSE clock. When the RTC_REFIN detection is enabled (REFCKON bit
of RTC_CR set to 1), the calendar is still clocked by the LSE, and RTC_REFIN is used to
compensate for the imprecision of the calendar update frequency (1 Hz).
Each 1 Hz clock edge is compared to the nearest reference clock edge (if one is found
within a given time window). In most cases, the two clock edges are properly aligned. When
the 1 Hz clock becomes misaligned due to the imprecision of the LSE clock, the RTC shifts
the 1 Hz clock a bit so that future 1 Hz clock edges are aligned. Thanks to this mechanism,
the calendar becomes as precise as the reference clock.
The RTC detects if the reference clock source is present by using the 256 Hz clock
(ck_apre) generated from the 32.768 kHz quartz. The detection is performed during a time
window around each of the calendar updates (every 1 s). The window equals 7 ck_apre
periods when detecting the first reference clock edge. A smaller window of 3 ck_apre
periods is used for subsequent calendar updates.
Each time the reference clock is detected in the window, the asynchronous prescaler which
outputs the ck_apre clock is forced to reload. This has no effect when the reference clock
and the 1 Hz clock are aligned because the prescaler is being reloaded at the same
moment. When the clocks are not aligned, the reload shifts future 1 Hz clock edges a little
for them to be aligned with the reference clock.
If the reference clock halts (no reference clock edge occurred during the 3 ck_apre window),
the calendar is updated continuously based solely on the LSE clock. The RTC then waits for
the reference clock using a large 7 ck_apre period detection window centered on the
ck_spre edge.
When the reference clock detection is enabled, PREDIV_A and PREDIV_S must be set to
their default values:
• PREDIV_A = 0x007F
• PREDIV_S = 0x00FF
Note: The reference clock detection is not available in Standby mode.
Caution: The reference clock detection feature cannot be used in conjunction with the coarse digital
calibration: RTC_CALIBR must be kept at 0x0000 0000 when REFCKON=1.

22.3.10 RTC coarse digital calibration

Two digital calibration methods are available: coarse and smooth calibration. To perform
coarse calibration refer to Section 22.6.7: RTC calibration register (RTC_CALIBR).
The two calibration methods are not intended to be used together, the application must
select one of the two methods. Coarse calibration is provided for compatibly reasons. To
perform smooth calibration refer to Section 22.3.11: RTC smooth digital calibration and to
Section 22.6.16: RTC calibration register (RTC_CALR)
The coarse digital calibration can be used to compensate crystal inaccuracy by adding
(positive calibration) or masking (negative calibration) clock cycles at the output of the
asynchronous prescaler (ck_apre).
Positive and negative calibration are selected by setting the DCS bit in RTC_CALIBR
register to ‘0’ and ‘1’, respectively.

626/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

When positive calibration is enabled (DCS = ‘0’), 2 ck_apre cycles are added every minute
(around 15360 ck_apre cycles) for 2xDC minutes. This causes the calendar to be updated
sooner, thereby adjusting the effective RTC frequency to be a bit higher.
When negative calibration is enabled (DCS = ‘1’), 1 ck_apre cycle is removed every minute
(around 15360 ck_apre cycles) for 2xDC minutes. This causes the calendar to be updated
later, thereby adjusting the effective RTC frequency to be a bit lower.
DC is configured through bits DC[4:0] of RTC_CALIBR register. This number ranges from 0
to 31 corresponding to a time interval (2xDC) ranging from 0 to 62.
The coarse digital calibration can be configured only in initialization mode, and starts when
the INIT bit is cleared. The full calibration cycle lasts 64 minutes. The first 2xDC minutes of
the 64 -minute cycle are modified as just described.
Negative calibration can be performed with a resolution of about 2 ppm while positive
calibration can be performed with a resolution of about 4 ppm. The maximum calibration
ranges from -63 ppm to 126 ppm.
The calibration can be performed either on the LSE or on the HSE clock.
Caution: Digital calibration may not work correctly if PREDIV_A < 6.

Case of RTCCLK=32.768 kHz and PREDIV_A+1=128

The following description assumes that ck_apre frequency is 256 Hz obtained with an LSE
clock nominal frequency of 32.768 kHz, and PREDIV_A set to 127 (default value).
The ck_spre clock frequency is only modified during the first 2xDC minutes of the 64-minute
cycle. For example, when DC equals 1, only the first 2 minutes are modified. This means
that the first 2xDC minutes of each 64-minute cycle have, once per minute, one second
either shortened by 256 or lengthened by 128 RTCCLK cycles, given that each ck_apre
cycle represents 128 RTCCLK cycles (with PREDIV_A+1=128).
Therefore each calibration step has the effect of adding 512 or subtracting 256 oscillator
cycles for every 125829120 RTCCLK cycles (64min x 60 s/min x 32768 cycles/s). This is
equivalent to +4.069 ppm or-2.035 ppm per calibration step. As a result, the calibration
resolution is +10.5 or -5.27 seconds per month, and the total calibration ranges from +5.45
to -2.72 minutes per month.
In order to measure the clock deviation, a 512 Hz clock is output for calibration.Refer to
Section 22.3.14: Calibration clock output.

22.3.11 RTC smooth digital calibration

RTC frequency can be digitally calibrated with a resolution of about 0.954 ppm with a range
from -487.1 ppm to +488.5 ppm. The correction of the frequency is performed using series
of small adjustments (adding and/or subtracting individual RTCCLK pulses). These
adjustments are fairly well distributed so that the RTC is well calibrated even when observed
over short durations of time.
The smooth digital calibration is performed during a cycle of about 220 RTCCLK pulses, or
32 seconds when the input frequency is 32768 Hz. This cycle is maintained by a 20-bit
counter, cal_cnt[19:0], clocked by RTCCLK.

RM0402 Rev 6 627/1163

655
Real-time clock (RTC) RM0402

The smooth calibration register (RTC_CALR) specifies the number of RTCCLK clock cycles
to be masked during the 32-second cycle:
• Setting the bit CALM[0] to 1 causes exactly one pulse to be masked during the 32-
second cycle.
• Setting CALM[1] to 1 causes two additional cycles to be masked
• Setting CALM[2] to 1 causes four additional cycles to be masked
• and so on up to CALM[8] set to 1 which causes 256 clocks to be masked.
Note: CALM[8:0] (RTC_CALRx) specifies the number of RTCCLK pulses to be masked during the
32-second cycle. Setting the bit CALM[0] to ‘1’ causes exactly one pulse to be masked
during the 32-second cycle at the moment when cal_cnt[19:0] is 0x80000; CALM[1]=1
causes two other cycles to be masked (when cal_cnt is 0x40000 and 0xC0000); CALM[2]=1
causes four other cycles to be masked (cal_cnt = 0x20000/0x60000/0xA0000/ 0xE0000);
and so on up to CALM[8]=1 which causes 256 clocks to be masked (cal_cnt = 0xXX800).
While CALM allows the RTC frequency to be reduced by up to 487.1 ppm with fine
resolution, the bit CALP can be used to increase the frequency by 488.5 ppm. Setting CALP
to ‘1’ effectively inserts an extra RTCCLK pulse every 211 RTCCLK cycles, which means
that 512 clocks are added during every 32-second cycle.
Using CALM together with CALP, an offset ranging from -511 to +512 RTCCLK cycles can
be added during the 32-second cycle, which translates to a calibration range of -487.1 ppm
to +488.5 ppm with a resolution of about 0.954 ppm.
The formula to calculate the effective calibrated frequency (FCAL) given the input frequency
(FRTCCLK) is as follows:
FCAL = FRTCCLK x [1 + (CALP x 512 - CALM) / (220 + CALM - CALP x 512)]

Calibration when PREDIV_A<3

The CALP bit can not be set to 1 when the asynchronous prescaler value (PREDIV_A bits in
RTC_PRER register) is less than 3. If CALP was already set to 1 and PREDIV_A bits are
set to a value less than 3, CALP is ignored and the calibration operates as if CALP was
equal to 0.
To perform a calibration with PREDIV_A less than 3, the synchronous prescaler value
(PREDIV_S) should be reduced so that each second is accelerated by 8 RTCCLK clock
cycles, which is equivalent to adding 256 clock cycles every 32 seconds. As a result,
between 255 and 256 clock pulses (corresponding to a calibration range from 243.3 to
244.1 ppm) can effectively be added during each 32-second cycle using only the CALM bits.
With a nominal RTCCLK frequency of 32768 Hz, when PREDIV_A equals 1 (division factor
of 2), PREDIV_S should be set to 16379 rather than 16383 (4 less). The only other
interesting case is when PREDIV_A equals 0, PREDIV_S should be set to 32759 rather
than 32767 (8 less).
If PREDIV_S is reduced in this way, the formula given the effective frequency of the
calibrated input clock is as follows:
FCAL = FRTCCLK x [1 + (256 - CALM) / (220 + CALM - 256)]
In this case, CALM[7:0] equals 0x100 (the midpoint of the CALM range) is the correct
setting if RTCCLK is exactly 32768.00 Hz.

628/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Verifying the RTC calibration

RTC precision is performed by measuring the precise frequency of RTCCLK and calculating
the correct CALM value and CALP values. An optional 1 Hz output is provided to allow
applications to measure and verify the RTC precision.
Measuring the precise frequency of the RTC over a limited interval can result in a
measurement error of up to 2 RTCCLK clock cycles over the measurement period,
depending on how the digital calibration cycle is aligned with the measurement period.
However, this measurement error can be eliminated if the measurement period is the same
length as the calibration cycle period. In this case, the only error observed is the error due to
the resolution of the digital calibration.
• By default, the calibration cycle period is 32 seconds.
Using this mode and measuring the accuracy of the 1 Hz output over exactly 32
seconds guarantees that the measure is within 0.477 ppm (0.5 RTCCLK cycles over 32
seconds, due to the limitation of the calibration resolution).
• CALW16 bit of the RTC_CALR register can be set to 1 to force a 16- second calibration
cycle period.
In this case, the RTC precision can be measured during 16 seconds with a maximum
error of 0.954 ppm (0.5 RTCCLK cycles over 16 seconds). However, since the
calibration resolution is reduced, the long term RTC precision is also reduced to 0.954
ppm: CALM[0] bit is stuck at 0 when CALW16 is set to 1.
• CALW8 bit of the RTC_CALR register can be set to 1 to force a 8- second calibration
cycle period.
In this case, the RTC precision can be measured during 8 seconds with a maximum
error of 1.907 ppm (0.5 RTCCLK cycles over 8s). The long term RTC precision is also
reduced to 1.907 ppm: CALM[1:0] bits are stuck at 00 when CALW8 is set to 1.

Re-calibration on-the-fly
The calibration register (RTC_CALR) can be updated on-the-fly while RTC_ISR/INITF=0, by
using the follow process:
1. Poll the RTC_ISR/RECALPF (re-calibration pending flag).
2. If it is set to 0, write a new value to RTC_CALR, if necessary. RECALPF is then
automatically set to 1
3. Within three ck_apre cycles after the write operation to RTC_CALR, the new calibration
settings take effect.

22.3.12 Timestamp function

Timestamp is enabled by setting the TSE bit of RTC_CR register to 1.
The calendar is saved in the timestamp registers (RTC_TSSSR, RTC_TSTR, RTC_TSDR)
when a timestamp event is detected on the pin to which the TIMESTAMP alternate function
is mapped. When a timestamp event occurs, the timestamp flag bit (TSF) in RTC_ISR
register is set.
By setting the TSIE bit in the RTC_CR register, an interrupt is generated when a timestamp
event occurs.

RM0402 Rev 6 629/1163

655
Real-time clock (RTC) RM0402

If a new timestamp event is detected while the timestamp flag (TSF) is already set, the
timestamp overflow flag (TSOVF) flag is set and the timestamp registers (RTC_TSTR and
RTC_TSDR) maintain the results of the previous event.
Note: TSF is set 2 ck_apre cycles after the timestamp event occurs due to synchronization
process.
There is no delay in the setting of TSOVF. This means that if two timestamp events are
close together, TSOVF can be seen as '1' while TSF is still '0'. As a consequence, it is
recommended to poll TSOVF only after TSF has been set.
Caution: If a timestamp event occurs immediately after the TSF bit is supposed to be cleared, then
both TSF and TSOVF bits are set. To avoid masking a timestamp event occurring at the
same moment, the application must not write ‘0’ into TSF bit unless it has already read it to
‘1’.
Optionally, a tamper event can cause a timestamp to be recorded. See the description of the
TAMPTS control bit in Section 22.6.17: RTC tamper and alternate function configuration
register (RTC_TAFCR). If the timestamp event is on the same pin as a tamper event
configured in filtered mode (TAMPFLT set to a non-zero value), the timestamp on tamper
detection event mode must be selected by setting TAMPTS='1' in RTC_TAFCR register.

TIMESTAMP alternate function

The TIMESTAMP additional function is mapped to RTC_AF1.

22.3.13 Tamper detection

Two tamper detection inputs are available. They can be configured either for edge detection,
or for level detection with filtering.

RTC backup registers

The backup registers (RTC_BKPxR) are twenty 32-bit registers for storing 80 bytes of user
application data. They are implemented in the backup domain that remains powered-on by
VBAT when the VDD power is switched off. They are not reset by system reset or when the
device wakes up from Standby mode. They are reset by a backup domain reset
The backup registers are reset when a tamper detection event occurs (see Section 22.6.20:
RTC backup registers (RTC_BKPxR) and Tamper detection initialization on page 630.

Tamper detection initialization

Each tamper detection input is associated with the TAMP1F/TAMP2F flags in the RTC_ISR2
register. Each input can be enabled by setting the corresponding TAMP1E/TAMP2E bits to 1
in the RTC_TAFCR register.
A tamper detection event resets all backup registers (RTC_BKPxR).
By setting the TAMPIE bit in the RTC_TAFCR register, an interrupt is generated when a
tamper detection event occurs.

Timestamp on tamper event

With TAMPTS set to ‘1 , any tamper event causes a timestamp to occur. In this case, either
the TSF bit or the TSOVF bit are set in RTC_ISR, in the same manner as if a normal
timestamp event occurs. The affected tamper flag register (TAMP1F, TAMP2F) is set at the
same time that TSF or TSOVF is set.

630/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Edge detection on tamper inputs

If the TAMPFLT bits are “00”, the TAMPER pins generate tamper detection events
(RTC_TAMP[2:1]) when either a rising edge is observed or an falling edge is observed
depending on the corresponding TAMPxTRG bit. The internal pull-up resistors on the
TAMPER inputs are deactivated when edge detection is selected.
Caution: To avoid losing tamper detection events, the signal used for edge detection is logically
ANDed with TAMPxE in order to detect a tamper detection event in case it occurs before the
TAMPERx pin is enabled.
• When TAMPxTRG = 0: if the TAMPERx alternate function is already high before
tamper detection is enabled (TAMPxE bit set to 1), a tamper event is detected as soon
as TAMPERx is enabled, even if there was no rising edge on TAMPERx after TAMPxE
was set.
• When TAMPxTRG = 1: if the TAMPERx alternate function is already low before tamper
detection is enabled, a tamper event is detected as soon as TAMPERx is enabled
(even if there was no falling edge on TAMPERx after TAMPxE was set.
After a tamper event has been detected and cleared, the TAMPERx alternate function
should be disabled and then re-enabled (TAMPxE set to 1) before re-programming the
backup registers (RTC_BKPxR). This prevents the application from writing to the backup
registers while the TAMPERx value still indicates a tamper detection. This is equivalent to a
level detection on the TAMPERx alternate function.
Note: Tamper detection is still active when VDD power is switched off. To avoid unwanted resetting
of the backup registers, the pin to which the TAMPER alternate function is mapped should
be externally tied to the correct level.

Level detection with filtering on tamper inputs

Level detection with filtering is performed by setting TAMPFLT to a non-zero value. A tamper
detection event is generated when either 2, 4, or 8 (depending on TAMPFLT) consecutive
samples are observed at the level designated by the TAMPxTRG bits
(TAMP1TRG/TAMP2TRG).
The TAMPER inputs are pre-charged through the I/O internal pull-up resistance before its
state is sampled, unless disabled by setting TAMPPUDIS to 1,The duration of the precharge
is determined by the TAMPPRCH bits, allowing for larger capacitances on the tamper
inputs.
The trade-off between tamper detection latency and power consumption through the pull-up
can be optimized by using TAMPFREQ to determine the frequency of the sampling for level
detection.
Note: Refer to the datasheets for the electrical characteristics of the pull-up resistors.

TAMPER alternate function detection

The TAMPER1 additional function is mapped to RTC_AF1 pin.

22.3.14 Calibration clock output

When the COE bit is set to 1 in the RTC_CR register, a reference clock is provided on the
RTC_CALIB device output. If the COSEL bit in the RTC_CR register is reset and
PREDIV_A = 0x7F, the RTC_CALIB frequency is fRTCCLK/64. This corresponds to a
calibration output at 512 Hz for an RTCCLK frequency at 32.768 kHz.

RM0402 Rev 6 631/1163

655
Real-time clock (RTC) RM0402

The RTC_CALIB output is not impacted by the calibration value programmed in

RTC_CALIBR register. The RTC_CALIB duty cycle is irregular: there is a light jitter on falling
edges. It is therefore recommended to use rising edges.
If COSEL is set and “PREDIV_S+1” is a non-zero multiple of 256 (i.e: PREDIV_S[7:0] =
0xFF), the RTC_CALIB frequency is fRTCCLK/(256 * (PREDIV_A+1)). This corresponds to a
calibration output at 1 Hz for prescaler default values (PREDIV_A = Ox7F, PREDIV_S =
0xFF), with an RTCCLK frequency at 32.768 kHz.

Calibration alternate function output

When the COE bit in the RTC_CR register is set to 1, the calibration alternate function
(RTC_CALIB) is enabled on RTC_AF1.
Note: When RTC_CALIB or RTC_ALARM is selected, RTC_AF1 is automatically configured in
output alternate function.

22.3.15 Alarm output

Three functions can be selected on Alarm output: ALRAF, ALRBF and WUTF. These
functions reflect the contents of the corresponding flags in the RTC_ISR register.
The OSEL[1:0] control bits in the RTC_CR register are used to activate the alarm alternate
function output (RTC_ALARM) in RTC_AF1, and to select the function which is output on
RTC_ALARM.
The polarity of the output is determined by the POL control bit in RTC_CR so that the
opposite of the selected flag bit is output when POL is set to 1.

Alarm alternate function output

RTC_ALARM can be configured in output open drain or output push-pull using the control
bit ALARMOUTTYPE in the RTC_TAFCR register.
Note: Once RTC_ALARM is enabled, it has priority over RTC_CALIB (COE bit is don't care on
RTC_AF1).
When RTC_CALIB or RTC_ALARM is selected, RTC_AF1 is automatically configured in
output alternate function.

22.4 RTC and low power modes

Table 112. Effect of low power modes on RTC
Mode Description

No effect
Sleep
RTC interrupts cause the device to exit the Sleep mode.
The RTC remains active when the RTC clock source is LSE or LSI. RTC alarm, RTC
Stop tamper event, RTC time stamp event, and RTC Wakeup cause the device to exit the Stop
mode.
The RTC remains active when the RTC clock source is LSE or LSI. RTC alarm, RTC
Standby tamper event, RTC time stamp event, and RTC Wakeup cause the device to exit the
Standby mode.

632/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.5 RTC interrupts

All RTC interrupts are connected to the EXTI controller.
To enable the RTC Alarm interrupt, the following sequence is required:
1. Configure and enable the EXTI Line 17 in interrupt mode and select the rising edge
sensitivity.
2. Configure and enable the RTC_Alarm IRQ channel in the NVIC.
3. Configure the RTC to generate RTC alarms (Alarm A or Alarm B).
To enable the RTC Wakeup interrupt, the following sequence is required:
1. Configure and enable the EXTI Line 22 in interrupt mode and select the rising edge
sensitivity.
2. Configure and enable the RTC_WKUP IRQ channel in the NVIC.
3. Configure the RTC to generate the RTC wakeup timer event.
To enable the RTC Tamper interrupt, the following sequence is required:
1. Configure and enable the EXTI Line 21 in interrupt mode and select the rising edge
sensitivity.
2. Configure and Enable the TAMP_STAMP IRQ channel in the NVIC.
3. Configure the RTC to detect the RTC tamper event.
To enable the RTC TimeStamp interrupt, the following sequence is required:
1. Configure and enable the EXTI Line 21 in interrupt mode and select the rising edge
sensitivity.
2. Configure and Enable the TAMP_STAMP IRQ channel in the NVIC.
3. Configure the RTC to detect the RTC timestamp event.

Table 113. Interrupt control bits

Enable Exit the Exit the Exit the
Interrupt event Event flag control Sleep Stop Standby
bit mode mode mode

Alarm A ALRAF ALRAIE yes yes(1) yes(1)

Alarm B ALRBF ALRBIE yes yes(1) yes(1)
Wakeup WUTF WUTIE yes yes(1) yes(1)
TimeStamp TSF TSIE yes yes(1) yes(1)
Tamper1 detection TAMP1F TAMPIE yes yes(1) yes(1)
1. Wakeup from STOP and Standby modes is possible only when the RTC clock source is LSE or LSI.

RM0402 Rev 6 633/1163

655
Real-time clock (RTC) RM0402

22.6 RTC registers

Refer to Section 1.2 on page 45 of this reference manual for a list of abbreviations used in
register descriptions.
The peripheral registers have to be accessed by words (32 bits).

22.6.1 RTC time register (RTC_TR)

The RTC_TR is the calendar time shadow register. This register must be written in
initialization mode only. Refer to Calendar initialization and configuration and Reading the
calendar.
Address offset: 0x00
Backup domain reset value: 0x0000 0000
System reset: 0x0000 0000 when BYPSHAD = 0. Not affected when BYPSHAD = 1.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. PM HT[1:0] HU[3:0]
rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. MNT[2:0] MNU[3:0] Res. ST[2:0] SU[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31-24 Reserved, must be kept at reset value

Bit 23 Reserved, must be kept at reset value.
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format
Bits 19:16 HU[3:0]: Hour units in BCD format
Bit 15 Reserved, must be kept at reset value.
Bits 14:12 MNT[2:0]: Minute tens in BCD format
Bits 11:8 MNU[3:0]: Minute units in BCD format
Bit 7 Reserved, must be kept at reset value.
Bits 6:4 ST[2:0]: Second tens in BCD format
Bits 3:0 SU[3:0]: Second units in BCD format

Note: This register is write protected. The write access procedure is described in RTC register
write protection.

634/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.6.2 RTC date register (RTC_DR)

The RTC_DR is the calendar date shadow register. This register must be written in
initialization mode only. Refer to Calendar initialization and configuration and Reading the
calendar.
Address offset: 0x04
Backup domain reset value: 0x0000_2101
System reset: 0x0000 2101 when BYPSHAD = 0. Not affected when BYPSHAD = 1.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. YT[3:0] YU[3:0]
rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WDU[2:0] MT MU[3:0] Res. Res. DT[1:0] DU[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31-24 Reserved, must be kept at reset value

Bits 23:20 YT[3:0]: Year tens in BCD format
Bits 19:16 YU[3:0]: Year units in BCD format
Bits 15:13 WDU[2:0]: Week day units
000: forbidden
001: Monday
...
111: Sunday
Bit 12 MT: Month tens in BCD format
Bits 11:8 MU: Month units in BCD format
Bits 7:6 Reserved, must be kept at reset value.
Bits 5:4 DT[1:0]: Date tens in BCD format
Bits 3:0 DU[3:0]: Date units in BCD format

Note: This register is write protected. The write access procedure is described in RTC register
write protection.

RM0402 Rev 6 635/1163

655
Real-time clock (RTC) RM0402

22.6.3 RTC control register (RTC_CR)

Address offset: 0x08
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. COE OSEL[1:0] POL COSEL BKP SUB1H ADD1H
rw rw rw rw rw rw w w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TSIE WUTIE ALRBIE ALRAIE TSE WUTE ALRBE ALRAE DCE FMT BYPSHAD REFCKON TSEDGE WUCKSEL[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:24 Reserved, must be kept at reset value.

Bit 23 COE: Calibration output enable
This bit enables the RTC_CALIB output
0: Calibration output disabled
1: Calibration output enabled
Bits 22:21 OSEL[1:0]: Output selection
These bits are used to select the flag to be routed to RTC_ALARM output
00: Output disabled
01: Alarm A output enabled
10:Alarm B output enabled
11: Wakeup output enabled
Bit 20 POL: Output polarity
This bit is used to configure the polarity of RTC_ALARM output
0: The pin is high when ALRAF/ALRBF/WUTF is asserted (depending on OSEL[1:0])
1: The pin is low when ALRAF/ALRBF/WUTF is asserted (depending on OSEL[1:0]).
Bit 19 COSEL: Calibration output selection
When COE=1, this bit selects which signal is output on RTC_CALIB.
0: Calibration output is 512 Hz
1: Calibration output is 1 Hz
These frequencies are valid for RTCCLK at 32.768 kHz and prescalers at their default
values (PREDIV_A=127 and PREDIV_S=255). Refer to Section 22.3.14: Calibration clock
output
Bit 18 BKP: Backup
This bit can be written by the user to memorize whether the daylight saving time change has
been performed or not.
Bit 17 SUB1H: Subtract 1 hour (winter time change)
When this bit is set outside initialization mode, 1 hour is subtracted to the calendar time if the
current hour is not 0. This bit is always read as 0.
Setting this bit has no effect when current hour is 0.
0: No effect
1: Subtracts 1 hour to the current time. This can be used for winter time change.

636/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Bit 16 ADD1H: Add 1 hour (summer time change)

When this bit is set outside initialization mode, 1 hour is added to the calendar time. This bit
is always read as 0.
0: No effect
1: Adds 1 hour to the current time. This can be used for summer time change
Bit 15 TSIE: Timestamp interrupt enable
0: Timestamp Interrupt disable
1: Timestamp Interrupt enable
Bit 14 WUTIE: Wakeup timer interrupt enable
0: Wakeup timer interrupt disabled
1: Wakeup timer interrupt enabled
Bit 13 ALRBIE: Alarm B interrupt enable
0: Alarm B Interrupt disable
1: Alarm B Interrupt enable
Bit 12 ALRAIE: Alarm A interrupt enable
0: Alarm A interrupt disabled
1: Alarm A interrupt enabled
Bit 11 TSE: Time stamp enable
0: Time stamp disable
1: Time stamp enable
Bit 10 WUTE: Wakeup timer enable
0: Wakeup timer disabled
1: Wakeup timer enabled
Bit 9 ALRBE: Alarm B enable
0: Alarm B disabled
1: Alarm B enabled
Bit 8 ALRAE: Alarm A enable
0: Alarm A disabled
1: Alarm A enabled
Bit 7 DCE: Coarse digital calibration enable
0: Digital calibration disabled
1: Digital calibration enabled
PREDIV_A must be 6 or greater
Bit 6 FMT: Hour format
0: 24 hour/day format
1: AM/PM hour format
Bit 5 BYPSHAD: Bypass the shadow registers
0: Calendar values (when reading from RTC_SSR, RTC_TR, and RTC_DR) are taken from
the shadow registers, which are updated once every two RTCCLK cycles.
1: Calendar values (when reading from RTC_SSR, RTC_TR, and RTC_DR) are taken
directly from the calendar counters.
Note: If the frequency of the APB1 clock is less than seven times the frequency of RTCCLK,
BYPSHAD must be set to ‘1’.

RM0402 Rev 6 637/1163

655
Real-time clock (RTC) RM0402

Bit 4 REFCKON: Reference clock detection enable (50 or 60 Hz)

0: Reference clock detection disabled
1: Reference clock detection enabled
Note: PREDIV_S must be 0x00FF.
Bit 3 TSEDGE: Timestamp event active edge
0: TIMESTAMP rising edge generates a timestamp event
1: TIMESTAMP falling edge generates a timestamp event
TSE must be reset when TSEDGE is changed to avoid unwanted TSF setting
Bits 2:0 WUCKSEL[2:0]: Wakeup clock selection
000: RTC/16 clock is selected
001: RTC/8 clock is selected
010: RTC/4 clock is selected
011: RTC/2 clock is selected
10x: ck_spre (usually 1 Hz) clock is selected
11x: ck_spre (usually 1 Hz) clock is selected and 216 is added to the WUT counter value
(see note below)

Note: WUT = Wakeup unit counter value. WUT = (0x0000 to 0xFFFF) + 0x10000 added when
WUCKSEL[2:1 = 11].
Bits 7, 6 and 4 of this register can be written in initialization mode only (RTC_ISR/INITF = 1).
Bits 2 to 0 of this register can be written only when RTC_CR WUTE bit = 0 and RTC_ISR
WUTWF bit = 1.
It is recommended not to change the hour during the calendar hour increment as it could
mask the incrementation of the calendar hour.
ADD1H and SUB1H changes are effective in the next second.
This register is write protected. The write access procedure is described in RTC register
write protection.

22.6.4 RTC initialization and status register (RTC_ISR)

Address offset: 0x0C
Backup domain reset value: 0x0000 0007
System reset value: Not affected except INIT, INITF and RSF which are cleared to 0.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. RECALPF
r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res TAMP1F TSOVF TSF WUTF ALRBF ALRAF INIT INITF RSF INITS SHPF WUT WF ALRB WF ALRA WF
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rw r rc_w0 r r r r r

Bits 31:17 Reserved, must be kept at reset value

Bit 16 RECALPF: Recalibration pending Flag
The RECALPF status flag is automatically set to ‘1’ when software writes to the RTC_CALR
register, indicating that the RTC_CALR register is blocked. When the new calibration
settings are taken into account, this bit returns to ‘0’. Refer to Re-calibration on-the-fly.
Bits 15:14 Reserved, must be kept at reset value.

638/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Bit 13 TAMP1F: Tamper detection flag

This flag is set by hardware when a tamper detection event is detected.
It is cleared by software writing 0.
Bit 12 TSOVF: Timestamp overflow flag
This flag is set by hardware when a timestamp event occurs while TSF is already set.
This flag is cleared by software by writing 0. It is recommended to check and then clear
TSOVF only after clearing the TSF bit. Otherwise, an overflow might not be noticed if a
timestamp event occurs immediately before the TSF bit is cleared.
Bit 11 TSF: Timestamp flag
This flag is set by hardware when a timestamp event occurs.
This flag is cleared by software by writing 0.
Bit 10 WUTF: Wakeup timer flag
This flag is set by hardware when the wakeup auto-reload counter reaches 0.
This flag is cleared by software by writing 0.
This flag must be cleared by software at least 1.5 RTCCLK periods before WUTF is set to 1
again.
Bit 9 ALRBF: Alarm B flag
This flag is set by hardware when the time/date registers (RTC_TR and RTC_DR) match the
Alarm B register (RTC_ALRMBR).
This flag is cleared by software by writing 0.
Bit 8 ALRAF: Alarm A flag
This flag is set by hardware when the time/date registers (RTC_TR and RTC_DR) match the
Alarm A register (RTC_ALRMAR).
This flag is cleared by software by writing 0.
Bit 7 INIT: Initialization mode
0: Free running mode
1: Initialization mode used to program time and date register (RTC_TR and RTC_DR), and
prescaler register (RTC_PRER). Counters are stopped and start counting from the new
value when INIT is reset.
Bit 6 INITF: Initialization flag
When this bit is set to 1, the RTC is in initialization state, and the time, date and prescaler
registers can be updated.
0: Calendar registers update is not allowed
1: Calendar registers update is allowed.
Bit 5 RSF: Registers synchronization flag
This bit is set by hardware each time the calendar registers are copied into the shadow
registers (RTC_SSRx, RTC_TRx and RTC_DRx). This bit is cleared by hardware in
initialization mode, while a shift operation is pending (SHPF=1), or when in bypass shadow
register mode (BYPSHAD=1). This bit can also be cleared by software.
0: Calendar shadow registers not yet synchronized
1: Calendar shadow registers synchronized
Bit 4 INITS: Initialization status flag
This bit is set by hardware when the calendar year field is different from 0 (backup domain
reset value state).
0: Calendar has not been initialized
1: Calendar has been initialized

RM0402 Rev 6 639/1163

655
Real-time clock (RTC) RM0402

Bit 3 SHPF: Shift operation pending

0: No shift operation is pending
1: A shift operation is pending
This flag is set by hardware as soon as a shift operation is initiated by a write to the
RTC_SHIFTR. It is cleared by hardware when the corresponding shift operation has been
executed. Writing to SHPF has no effect.
Bit 2 WUTWF: Wakeup timer write flag
This bit is set by hardware up to 2 RTCCLK cycles after the WUTE bit has been set to 0 in
RTC_CR. It is cleared up to 2 RTCCLK cycles after the WUTE bit has been set to 1. The
wakeup timer values can be changed when WUTE bit is cleared and WUTWF is set.
0: Wakeup timer configuration update not allowed
1: Wakeup timer configuration update allowed
Bit 1 ALRBWF: Alarm B write flag
This bit is set by hardware when Alarm B values can be changed, after the ALRBIE bit has
been set to 0 in RTC_CR.
It is cleared by hardware in initialization mode.
0: Alarm B update not allowed
1: Alarm B update allowed.
Bit 0 ALRAWF: Alarm A write flag
This bit is set by hardware when Alarm A values can be changed, after the ALRAE bit has
been set to 0 in RTC_CR.
It is cleared by hardware in initialization mode.
0: Alarm A update not allowed
1: Alarm A update allowed

Note: The ALRAF, ALRBF, WUTF and TSF bits are cleared 2 APB clock cycles after programming
them to 0.
This register is write protected (except for RTC_ISR[13:8] bits). The write access procedure
is described in RTC register write protection.

22.6.5 RTC prescaler register (RTC_PRER)

Address offset: 0x10
Backup domain reset value: 0x007F 00FF
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. PREDIV_A[6:0]
rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. PREDIV_S[14:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

640/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Bits 31:23 Reserved, must be kept at reset value

Bits 22:16 PREDIV_A[6:0]: Asynchronous prescaler factor
This is the asynchronous division factor:
ck_apre frequency = RTCCLK frequency/(PREDIV_A+1)
Bit 15 Reserved, must be kept at reset value.
Bits 14:0 PREDIV_S[14:0]: Synchronous prescaler factor
This is the synchronous division factor:
ck_spre frequency = ck_apre frequency/(PREDIV_S+1)

Note: This register must be written in initialization mode only. The initialization must be performed
in two separate write accesses. Refer to Calendar initialization and configuration
This register is write protected. The write access procedure is described in RTC register
write protection.

22.6.6 RTC wakeup timer register (RTC_WUTR)

Address offset: 0x14
Backup domain reset value: 0x0000 FFFF
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WUT[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value

Bits 15:0 WUT[15:0]: Wakeup auto-reload value bits
When the wakeup timer is enabled (WUTE set to 1), the WUTF flag is set every (WUT[15:0]
+ 1) ck_wut cycles. The ck_wut period is selected through WUCKSEL[2:0] bits of the
RTC_CR register
When WUCKSEL[2] = 1, the wakeup timer becomes 17-bits and WUCKSEL[1] effectively
becomes WUT[16] the most-significant bit to be reloaded into the timer.
Note: The first assertion of WUTF occurs (WUT+1) ck_wut cycles after WUTE is set. Setting
WUT[15:0] to 0x0000 with WUCKSEL[2:0] =011 (RTCCLK/2) is forbidden.

Note: This register can be written only when WUTWF is set to 1 in RTC_ISR.
This register is write protected. The write access procedure is described in RTC register
write protection.

22.6.7 RTC calibration register (RTC_CALIBR)

Address offset: 0x18
Backup domain reset value: 0x0000 0000
System reset: not affected

RM0402 Rev 6 641/1163

655
Real-time clock (RTC) RM0402

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. DCS Res. Res. DC[4:0]
rw rw rw rw rw rw

Bits 31:8 Reserved, must be kept at reset value

Bit 7 DCS: Digital calibration sign
0: Positive calibration: calendar update frequency is increased
1: Negative calibration: calendar update frequency is decreased
Bits 6:5 Reserved, must be kept at reset value.
Bits 4:0 DC[4:0]: Digital calibration
DCS = 0 (positive calibration)
00000: +0 ppm
00001: +4 ppm (rounded value)
00010: +8 ppm (rounded value)
..
11111: +126 ppm (rounded value)
DCS = 1 (negative calibration)
00000: -0 ppm
00001: -2 ppm (rounded value)
00010: -4 ppm (rounded value)
..
11111: - 63 ppm (rounded value)
Refer to Case of RTCCLK=32.768 kHz and PREDIV_A+1=128 for the exact step value.

Note: This register can be written in initialization mode only (RTC_ISR/INITF = ‘1’).
This register is write protected. The write access procedure is described in RTC register
write protection.

642/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.6.8 RTC alarm A register (RTC_ALRMAR)

Address offset: 0x1C
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MSK4 WDSEL DT[1:0] DU[3:0] MSK3 PM HT[1:0] HU[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MSK2 MNT[2:0] MNU[3:0] MSK1 ST[2:0] SU[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 MSK4: Alarm A date mask

0: Alarm A set if the date/day match
1: Date/day don’t care in Alarm A comparison
Bit 30 WDSEL: Week day selection
0: DU[3:0] represents the date units
1: DU[3:0] represents the week day. DT[1:0] is don’t care.
Bits 29:28 DT[1:0]: Date tens in BCD format.
Bits 27:24 DU[3:0]: Date units or day in BCD format.
Bit 23 MSK3: Alarm A hours mask
0: Alarm A set if the hours match
1: Hours don’t care in Alarm A comparison
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format.
Bits 19:16 HU[3:0]: Hour units in BCD format.
Bit 15 MSK2: Alarm A minutes mask
0: Alarm A set if the minutes match
1: Minutes don’t care in Alarm A comparison
Bits 14:12 MNT[2:0]: Minute tens in BCD format.
Bits 11:8 MNU[3:0]: Minute units in BCD format.
Bit 7 MSK1: Alarm A seconds mask
0: Alarm A set if the seconds match
1: Seconds don’t care in Alarm A comparison
Bits 6:4 ST[2:0]: Second tens in BCD format.
Bits 3:0 SU[3:0]: Second units in BCD format.

Note: This register can be written only when ALRAWF is set to 1 in RTC_ISR, or in initialization
mode.
This register is write protected. The write access procedure is described in RTC register
write protection.

RM0402 Rev 6 643/1163

655
Real-time clock (RTC) RM0402

22.6.9 RTC alarm B register (RTC_ALRMBR)

Address offset: 0x20
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
MSK4 WDSEL DT[1:0] DU[3:0] MSK3 PM HT[1:0] HU[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
MSK2 MNT[2:0] MNU[3:0] MSK1 ST[2:0] SU[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 MSK4: Alarm B date mask

0: Alarm B set if the date and day match
1: Date and day don’t care in Alarm B comparison
Bit 30 WDSEL: Week day selection
0: DU[3:0] represents the date units
1: DU[3:0] represents the week day. DT[1:0] is don’t care.
Bits 29:28 DT[1:0]: Date tens in BCD format
Bits 27:24 DU[3:0]: Date units or day in BCD format
Bit 23 MSK3: Alarm B hours mask
0: Alarm B set if the hours match
1: Hours don’t care in Alarm B comparison
Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format
Bits 19:16 HU[3:0]: Hour units in BCD format
Bit 15 MSK2: Alarm B minutes mask
0: Alarm B set if the minutes match
1: Minutes don’t care in Alarm B comparison
Bits 14:12 MNT[2:0]: Minute tens in BCD format
Bits 11:8 MNU[3:0]: Minute units in BCD format
Bit 7 MSK1: Alarm B seconds mask
0: Alarm B set if the seconds match
1: Seconds don’t care in Alarm B comparison
Bits 6:4 ST[2:0]: Second tens in BCD format
Bits 3:0 SU[3:0]: Second units in BCD format

Note: This register can be written only when ALRBWF is set to 1 in RTC_ISR, or in initialization
mode.
This register is write protected. The write access procedure is described in RTC register
write protection.

644/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.6.10 RTC write protection register (RTC_WPR)

Address offset: 0x24
Backup domain reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. KEY
w w w w w w w w

Bits 31:8 Reserved, must be kept at reset value.

Bits 7:0 KEY: Write protection key
This byte is written by software.
Reading this byte always returns 0x00.
Refer to RTC register write protection for a description of how to unlock RTC register write
protection.

22.6.11 RTC sub second register (RTC_SSR)

Address offset: 0x28
Backup domain reset value: 0x0000 0000
System reset: 0x0000 0000 when BYPSHAD = 0. Not affected when BYPSHAD = 1
.
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SS[15:0]
r r r r r r r r r r r r r r r r

Bits 31:16 Reserved, must be kept at reset value

Bits 15:0 SS: Sub second value
SS[15:0] is the value in the synchronous prescaler’s counter. The fraction of a second is
given by the formula below:
Second fraction = ( PREDIV_S - SS ) / ( PREDIV_S + 1 )
Note: SS can be larger than PREDIV_S only after a shift operation. In that case, the correct
time/date is one second less than as indicated by RTC_TR/RTC_DR.

RM0402 Rev 6 645/1163

655
Real-time clock (RTC) RM0402

22.6.12 RTC shift control register (RTC_SHIFTR)

Address offset: 0x2C
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ADD1S Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
w
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. SUBFS[14:0]
w w w w w w w w w w w w w w w

Bit 31 ADD1S: Add one second

0: No effect
1: Add one second to the clock/calendar
This bit is write only and is always read as zero. Writing to this bit has no effect when a shift
operation is pending (when SHPF=1, in RTC_ISR).
This function is intended to be used with SUBFS (see description below) in order to
effectively add a fraction of a second to the clock in an atomic operation.
Bits 30:15 Reserved, must be kept at reset value
Bits 14:0 SUBFS: Subtract a fraction of a second
These bits are write only and is always read as zero. Writing to this bit has no effect when a
shift operation is pending (when SHPF=1, in RTC_ISR).
The value which is written to SUBFS is added to the synchronous prescaler’s counter. Since
this counter counts down, this operation effectively subtracts from (delays) the clock by:
Delay (seconds) = SUBFS / ( PREDIV_S + 1 )
A fraction of a second can effectively be added to the clock (advancing the clock) when the
ADD1S function is used in conjunction with SUBFS, effectively advancing the clock by:
Advance (seconds) = ( 1 - ( SUBFS / ( PREDIV_S + 1 ) ) ) .
Note: Writing to SUBFS causes RSF to be cleared. Software can then wait until RSF=1 to be
sure that the shadow registers have been updated with the shifted time.
Refer to Section 22.3.8: RTC synchronization.

Note: This register is write protected. The write access procedure is described in RTC register
write protection

646/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.6.13 RTC time stamp time register (RTC_TSTR)

Address offset: 0x30
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. PM HT[1:0] HU[3:0]
r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. MNT[2:0] MNU[3:0] Res. ST[2:0] SU[3:0]
r r r r r r r r r r r r r r

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 PM: AM/PM notation
0: AM or 24-hour format
1: PM
Bits 21:20 HT[1:0]: Hour tens in BCD format.
Bits 19:16 HU[3:0]: Hour units in BCD format.
Bit 15 Reserved, must be kept at reset value.
Bits 14:12 MNT[2:0]: Minute tens in BCD format.
Bits 11:8 MNU[3:0]: Minute units in BCD format.
Bit 7 Reserved, must be kept at reset value.
Bits 6:4 ST[2:0]: Second tens in BCD format.
Bits 3:0 SU[3:0]: Second units in BCD format.

Note: The content of this register is valid only when TSF is set to 1 in RTC_ISR. It is cleared when
TSF bit is reset.

22.6.14 RTC time stamp date register (RTC_TSDR)

Address offset: 0x34
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
WDU[1:0] MT MU[3:0] Res. Res. DT[1:0] DU[3:0]
r r r r r r r r r r r r r r

RM0402 Rev 6 647/1163

655
Real-time clock (RTC) RM0402

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:13 WDU[1:0]: Week day units
Bit 12 MT: Month tens in BCD format
Bits 11:8 MU[3:0]: Month units in BCD format
Bits 7:6 Reserved, must be kept at reset value.
Bits 5:4 DT[1:0]: Date tens in BCD format
Bits 3:0 DU[3:0]: Date units in BCD format

Note: The content of this register is valid only when TSF is set to 1 in RTC_ISR. It is cleared when
TSF bit is reset.

22.6.15 RTC timestamp sub second register (RTC_TSSSR)

Address offset: 0x38
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SS[15:0]
r r r r r r r r r r r r r r r r

Bits 31:16 Reserved

Bits 15:0 SS: Sub second value
SS[15:0] is the value of the synchronous prescaler’s counter when the timestamp event
occurred.

Note: The content of this register is valid only when RTC_ISR/TSF is set. It is cleared when the
RTC_ISR/TSF bit is reset.

22.6.16 RTC calibration register (RTC_CALR)

Address offset: 0x3C
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CALP CALW8 CALW16 Res. Res. Res. Res. CALM[8:0]
rw rw rw r r r r rw rw rw rw rw rw rw rw rw

648/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Bits 31:16 Reserved, must be kept at reset value

Bit 15 CALP: Increase frequency of RTC by 488.5 ppm
0: No RTCCLK pulses are added.
1: One RTCCLK pulse is effectively inserted every 211 pulses (frequency increased by
488.5 ppm).
This feature is intended to be used in conjunction with CALM, which lowers the frequency of
the calendar with a fine resolution. if the input frequency is 32768 Hz, the number of
RTCCLK pulses added during a 32-second window is calculated as follows: (512 * CALP) -
CALM.
Refer to Section 22.3.11: RTC smooth digital calibration.
Bit 14 CALW8: Use an 8-second calibration cycle period
When CALW8 is set to ‘1’, the 8-second calibration cycle period is selected.
CALM[1:0] are stuck at “00” when CALW8=’1’.
Refer to Section 22.3.11: RTC smooth digital calibration.
Bit 13 CALW16: Use a 16-second calibration cycle period
When CALW16 is set to ‘1’, the 16-second calibration cycle period is selected. This bit must
not be set to ‘1’ if CALW8=1.
Note: CALM[0] is stuck at ‘0’ when CALW16=’1’.
Refer to Section 22.3.11: RTC smooth digital calibration.
Bits 12:9 Reserved, must be kept at reset value
Bits 8:0 CALM[8:0]: Calibration minus
The frequency of the calendar is reduced by masking CALM out of 220 RTCCLK pulses (32
seconds if the input frequency is 32768 Hz). This decreases the frequency of the calendar
with a resolution of 0.9537 ppm.
To increase the frequency of the calendar, this feature should be used in conjunction with
CALP.
See Section 22.3.11: RTC smooth digital calibration on page 627.

Note: This register is write protected. The write access procedure is described in RTC register
write protection.

22.6.17 RTC tamper and alternate function configuration register

(RTC_TAFCR)
Address offset: 0x40
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALARMOUT TSIN TAMP1I
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
TYPE SEL NSEL
rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TAMP TAMP TAMP TAMP
TAMPTS Res. Res. Res. Res. TAMPIE TAMP1TRG TAMP1E
PUDIS PRCH[1:0] FLT[1:0] FREQ[2:0]
rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 649/1163

655
Real-time clock (RTC) RM0402

Bits 31:19 Reserved, must be kept at reset value. Always read as 0.

Bit 18 ALARMOUTTYPE: RTC_ALARM output type
0: RTC_ALARM is an open-drain output
1: RTC_ALARM is a push-pull output
Bit 17 TSINSEL: TIMESTAMP mapping
0: RTC_AF1 used as TIMESTAMP
1: Reserved
Bit 16 TAMP1INSEL: TAMPER1 mapping
0: RTC_AF1 used as TAMPER1
1: Reserved
Note: TAMP1E must be reset when TAMP1INSEL is changed to avoid unwanted setting of
TAMP1F.
Bit 15 TAMPPUDIS: TAMPER pull-up disable
This bit determines if each of the tamper pins are pre-charged before each sample.
0: Precharge tamper pins before sampling (enable internal pull-up)
1: Disable precharge of tamper pins
Note:
Bits 14:13 TAMPPRCH[1:0]: Tamper precharge duration
These bit determines the duration of time during which the pull-up/is activated before each
sample. TAMPPRCH is valid for each of the tamper inputs.
0x0: 1 RTCCLK cycle
0x1: 2 RTCCLK cycles
0x2: 4 RTCCLK cycles
0x3: 8 RTCCLK cycles
Bits 12:11 TAMPFLT[1:0]: Tamper filter count
These bits determines the number of consecutive samples at the specified level
(TAMP*TRG) necessary to activate a Tamper event. TAMPFLT is valid for each of the tamper
inputs.
0x0: Tamper is activated on edge of tamper input transitions to the active level (no internal
pull-up on tamper input).
0x1: Tamper is activated after 2 consecutive samples at the active level.
0x2: Tamper is activated after 4 consecutive samples at the active level.
0x3: Tamper is activated after 8 consecutive samples at the active level.
Bits 10:8 TAMPFREQ[2:0]: Tamper sampling frequency
Determines the frequency at which each of the tamper inputs are sampled.
0x0: RTCCLK / 32768 (1 Hz when RTCCLK = 32768 Hz)
0x1: RTCCLK / 16384 (2 Hz when RTCCLK = 32768 Hz)
0x2: RTCCLK / 8192 (4 Hz when RTCCLK = 32768 Hz)
0x3: RTCCLK / 4096 (8 Hz when RTCCLK = 32768 Hz)
0x4: RTCCLK / 2048 (16 Hz when RTCCLK = 32768 Hz)
0x5: RTCCLK / 1024 (32 Hz when RTCCLK = 32768 Hz)
0x6: RTCCLK / 512 (64 Hz when RTCCLK = 32768 Hz)
0x7: RTCCLK / 256 (128 Hz when RTCCLK = 32768 Hz)
Bit 7 TAMPTS: Activate timestamp on tamper detection event
0: Tamper detection event does not cause a timestamp to be saved
1: Save timestamp on tamper detection event
TAMPTS is valid even if TSE=0 in the RTC_CR register.

650/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

Bits 6:3 Reserved. Always read as 0.

Bit 2 TAMPIE: Tamper interrupt enable
0: Tamper interrupt disabled
1: Tamper interrupt enabled
Bit 1 TAMP1TRG: Active level for tamper 1
if TAMPFLT != 00:
0: TAMPER1 staying low triggers a tamper detection event.
1: TAMPER1 staying high triggers a tamper detection event.
if TAMPFLT = 00:
0: TAMPER1 rising edge triggers a tamper detection event.
1: TAMPER1 falling edge triggers a tamper detection event.
Caution: When TAMPFLT = 0, TAMP1E must be reset when TAMP1TRG is changed to avoid
spuriously setting TAMP1F.
Bit 0 TAMP1E: Tamper 1 detection enable
0: Tamper 1 detection disabled
1: Tamper 1 detection enabled

22.6.18 RTC alarm A sub second register (RTC_ALRMASSR)

Address offset: 0x44
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. MASKSS[3:0] Res. Res. Res. Res. Res. Res. Res. Res.
rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. SS[14:0]
rw rw rw rw rw rw rw rw rw rw rw rw w rw rw

Bits 31:28 Reserved, must be kept at reset value

Bits 27:24 MASKSS[3:0]: Mask the most-significant bits starting at this bit
0: No comparison on sub seconds for Alarm A. The alarm is set when the seconds unit is
incremented (assuming that the rest of the fields match).
1: SS[14:1] are don’t care in Alarm A comparison. Only SS[0] is compared.
2: SS[14:2] are don’t care in Alarm A comparison. Only SS[1:0] are compared.
3: SS[14:3] are don’t care in Alarm A comparison. Only SS[2:0] are compared.
...
12: SS[14:12] are don’t care in Alarm A comparison. SS[11:0] are compared.
13: SS[14:13] are don’t care in Alarm A comparison. SS[12:0] are compared.
14: SS[14] is don’t care in Alarm A comparison. SS[13:0] are compared.
15: All 15 SS bits are compared and must match to activate alarm.
The overflow bits of the synchronous counter (bits 15) is never compared. This bit can be
different from 0 only after a shift operation.

RM0402 Rev 6 651/1163

655
Real-time clock (RTC) RM0402

Bits 23:15 Reserved, must be kept at reset value

Bits 14:0 SS[14:0]: Sub seconds value
This value is compared with the contents of the synchronous prescaler’s counter to
determine if Alarm A is to be activated. Only bits 0 up MASKSS-1 are compared.

Note: This register can be written only when ALRAE is reset in RTC_CR register, or in initialization
mode.
This register is write protected. The write access procedure is described in RTC register
write protection on page 622

22.6.19 RTC alarm B sub second register (RTC_ALRMBSSR)

Address offset: 0x48
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. MASKSS[3:0] Res. Res. Res. Res. Res. Res. Res. Res.
r r r r rw rw rw rw r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. SS[14:0]
r rw rw rw rw rw rw rw rw rw rw rw rw w rw rw

Bits 31:28 Reserved, must be kept at reset value

Bits 27:24 MASKSS[3:0]: Mask the most-significant bits starting at this bit
0x0: No comparison on sub seconds for Alarm B. The alarm is set when the seconds unit is
incremented (assuming that the rest of the fields match).
0x1: SS[14:1] are don’t care in Alarm B comparison. Only SS[0] is compared.
0x2: SS[14:2] are don’t care in Alarm B comparison. Only SS[1:0] are compared.
0x3: SS[14:3] are don’t care in Alarm B comparison. Only SS[2:0] are compared.
...
0xC: SS[14:12] are don’t care in Alarm B comparison. SS[11:0] are compared.
0xD: SS[14:13] are don’t care in Alarm B comparison. SS[12:0] are compared.
0xE: SS[14] is don’t care in Alarm B comparison. SS[13:0] are compared.
0xF: All 15 SS bits are compared and must match to activate alarm.
The overflow bits of the synchronous counter (bits 15) is never compared. This bit can be
different from 0 only after a shift operation.
Bits 23:15 Reserved, must be kept at reset value
Bits 14:0 SS[14:0]: Sub seconds value
This value is compared with the contents of the synchronous prescaler’s counter to
determine if Alarm B is to be activated. Only bits 0 up to MASKSS-1 are compared.

Note: This register can be written only when ALRBIE is reset in RTC_CR register, or in
initialization mode.
This register is write protected.The write access procedure is described in RTC register
write protection

652/1163 RM0402 Rev 6

RM0402 Real-time clock (RTC)

22.6.20 RTC backup registers (RTC_BKPxR)

Address offset: 0x50 to 0x9C
Backup domain reset value: 0x0000 0000
System reset: not affected

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
BKP[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BKP[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw w rw rw

Bits 31:0 BKP[31:0]

The application can write or read data to and from these registers.
They are powered-on by VBAT when VDD is switched off, so that they are not reset by
System reset, and their contents remain valid when the device operates in low-power mode.
This register is reset on a tamper detection event, as long as TAMPxF=1.

22.6.21 RTC register map

Table 114. RTC register map and reset values

Offset Register
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
HT
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
PM

RTC_TR HU[3:0] MNT[2:0] MNU[3:0] ST[2:0] SU[3:0]

[1:0]
0x00
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DT
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
MT

RTC_DR YT[3:0] YU[3:0] WDU[2:0] MU[3:0] DU[3:0]

[1:0]
0x04
Reset value 0 0 1 0 0 0 0 1 0 0
REFCKON 0 0 0 1
BYPSHAD

TSEDGE
ALRBIE
ALRAIE
COSEL

ADD1H
SUB1H

ALRBE
ALRAE
WUTIE

WUTE

OSEL WCKSEL
TSIE
COE

DCE
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

FMT
POL

BKP

TSE

RTC_CR
0x08 [1:0] [2:0]

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ALRBWF
ALRAWF
TAMP1F

WUTWF
TSOVF

ALRBF
ALRAF
WUTF

SHPF
INITS
INITF
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

RSF
TSF

INIT

RTC_ISR
0x0C

Reset value 0 0 0 0 0 0 0 0 0 0 0 1 0 1
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

RTC_PRER PREDIV_A[6:0] PREDIV_S[14:0]

0x10
Reset value 1 1 1 1 1 1 1 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

RTC_WUTR WUT[15:0]
0x14
Reset value 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
DCS
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.

RTC_CALIBR DC[4:0]
0x18
Reset value 0 0 0 0 0 0

RM0402 Rev 6 653/1163

655
 0x48
0x44
0x40
0x38
0x30
0x28
0x24
0x20

0x3C
0x2C
0x1C
Offset

654/1163
RTC_
RTC_
RTC_SSR
RTC_WPR
Register

RTC_TSTR

Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

RTC_ CALR

ALRMBSSR
ALRMASSR
RTC_TAFCR
RTC_TSSSR
RTC_SHIFTR
RTC_ALRMBR
RTC_ALRMAR

0
0

0
Res. Res. Res. Res. Res. Res. ADD1S Res. Res. MSK4 MSK4 31
Real-time clock (RTC)

0
0 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. WDSEL WDSEL

0
0

Res. Res. Res. Res. Res. Res. Res. Res. Res. DT

29
DT

[1:0]
[1:0]

0
0

Res. Res. Res. Res. Res. Res. Res. Res. Res. 28

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. 27

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. 26

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. 25

DU[3:0]
DU[3:0]

0
0
0
0

MASKSS[3:0]
MASKSS[3:0]
Res. Res. Res. Res. Res. Res. Res. 24
0
0

Res. Res. Res. Res. Res. Res. Res. Res. Res. MSK3 MSK3 23

0
0
0

Res. Res. Res. Res. Res. PM Res. Res. Res. PM PM 22

0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 21

HT[1:0]
HT
HT

[1:0]
[1:0]

0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 20

0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 19

0
0
0

0
Res. Res. ALARMOUTTYPE Res. Res. Res. Res. Res. 18

RM0402 Rev 6
0
0
0

0
Res. Res. TSINSEL Res. Res. Res. Res. Res. 17

HU[3:0]
HU[3:0]
HU[3:0]

0
0
0

0
Res. Res. TAMP1INSEL Res. Res. Res. Res. Res. 16
0
0

0
0
0
0

Res. Res. TAMPPUDIS CALP Res. Res. Res. MSK2 MSK2 15

0
0
0

0
0
0
0
0
0
0

CALW8 Res. 14
TAMPPRCH[1:0]

0
0
0

0
0
0
0
0
0
0

CALW16 MNT[2:0] Res. 13

0
0
0

0
0
0
0
0
0
MNT[2:0]
MNT[2:0]

Res. Res. 12
TAMPFLT[1:0] 0
0
0

0
0
0
0
0
0

Res. Res. 11
0
0
0

0
0
0
0
0
0

Res. Res. 10
0
0
0

0
0
0
0
0
0

TAMPFREQ[2:0] Res. Res. 9

MNU[3:0]
MNU[3:0]
MNU[3:0]

0
0
0

0
0
0
0
0
0
0
Table 114. RTC register map and reset values (continued)

Res. 8
0
0

0
0
0
0
0
0
0
0

TAMPTS Res. MSK2 MSK1 7

SS[15:0]
SS[15:0]

0
0
0

0
0
0
0
0
0
0

SS[14:0]
SS[14:0]
Res. 6
SUBFS[14:0]

0
0
0

0
0
0
0
0
0
0

Res. 5
ST[2:0]
ST[2:0]
ST[2:0]

0
0
0

0
0
0
0
0
0
0
0

Res. 4
0
0
0

0
0
0
0
0
0
0
0

Res. 3
CALM[8:0]
KEY[7:0]

0
0
0

0
0
0
0
0
0
0
0

TAMPIE 2
0
0
0

0
0
0
0
0
0
0
0

TAMP1ETRG 1
SU[3:0]
SU[3:0]
SU[3:0]

0
0
0

0
0
0
0
0
0
0
0
RM0402

TAMP1E 0
RM0402 Real-time clock (RTC)

Table 114. RTC register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
RTC_BKP0R BKP[31:0]

0x50 Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

to 0x9C
to
BKP[31:0]
RTC_BKP19R

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

Caution: In Table 114, the reset value is the value after a backup domain reset. The majority of the
registers are not affected by a system reset. For more information, refer to Section 22.3.7:
Resetting the RTC.

RM0402 Rev 6 655/1163

655
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23 Fast-mode Plus Inter-integrated circuit (FMPI2C)

interface

23.1 Introduction
The I2C (inter-integrated circuit) bus interface handles communications between the
microcontroller and the serial I2C bus. It provides multimaster capability, and controls all I2C
bus-specific sequencing, protocol, arbitration and timing. It supports Standard-mode (Sm),
Fast-mode (Fm) and Fast-mode Plus (Fm+).
It is also SMBus (system management bus) and PMBus (power management bus)
compatible.
DMA can be used to reduce CPU overload.

23.2 FMPI2C main features

• I2C bus specification rev03 compatibility:
– Slave and master modes
– Multimaster capability
– Standard-mode (up to 100 kHz)
– Fast-mode (up to 400 kHz)
– Fast-mode Plus (up to 1 MHz)
– 7-bit and 10-bit addressing mode
– Multiple 7-bit slave addresses (2 addresses, 1 with configurable mask)
– All 7-bit addresses acknowledge mode
– General call
– Programmable setup and hold times
– Easy to use event management
– Optional clock stretching
– Software reset
• 1-byte buffer with DMA capability
• Programmable analog and digital noise filters

656/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

The following additional features are also available depending on the product
implementation (see Section 23.3: FMPI2C implementation):
• SMBus specification rev 3.0 compatibility:
– Hardware PEC (packet error checking) generation and verification with ACK
control
– Command and data acknowledge control
– Address resolution protocol (ARP) support
– Host and Device support
– SMBus alert
– Timeouts and idle condition detection
• PMBus rev 1.3 standard compatibility
• Independent clock: a choice of independent clock sources allowing the FMPI2C
communication speed to be independent from the PCLK reprogramming

23.3 FMPI2C implementation

This manual describes the full set of features implemented in FMPI2C1.

Table 115. STM32F412xx FMPI2C implementation

I2C features(1) FMPI2C1

7-bit addressing mode X

10-bit addressing mode X
Standard-mode (up to 100 kbit/s) X
Fast-mode (up to 400 kbit/s) X
(2)(up
Fast-mode Plus to 1 Mbit/s) X
Independent clock X
Wakeup from Stop mode -
SMBus/PMBus X
1. X = supported.
2. 20 mA output drive for Fm+ mode is not supported.

23.4 FMPI2C functional description

In addition to receiving and transmitting data, this interface converts it from serial to parallel
format and vice versa. The interrupts are enabled or disabled by software. The interface is
connected to the I2C bus by a data pin (SDA) and by a clock pin (SCL). It can be connected
with a standard (up to 100 kHz), Fast-mode (up to 400 kHz) or Fast-mode Plus (up to
1 MHz) I2C bus.
This interface can also be connected to a SMBus with the data pin (SDA) and clock pin
(SCL).
If SMBus feature is supported: the additional optional SMBus Alert pin (SMBA) is also
available.

RM0402 Rev 6 657/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.4.1 FMPI2C block diagram

The block diagram of the FMPI2C interface is shown in Figure 208.

Figure 208. FMPI2C block diagram

I2CCLK
I2c_ker_ck

Data control
Digital Analog
Shift register noise noise GPIO
filter I2C_SDA
filter logic
SMBUS
PEC
generation/
check
Wakeup
on
address
match Clock control
Master clock
generation Digital Analog
noise noise
Slave clock GPIO I2C_SCL
filter filter
stretching logic

SMBus
Timeout
check

SMBus Alert
control & I2C_SMBA
status

PCLK
I2c_pclk Registers

APB bus

MSv46198V2

The FMPI2C is clocked by an independent clock source which allows the FMPI2C to
operate independently from the PCLK frequency.
For I2C I/Os supporting 20mA output current drive for Fast-mode Plus operation, the driving
capability is enabled through control bits in the system configuration controller (SYSCFG).
Refer to Section 23.3: FMPI2C implementation.

658/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

23.4.2 FMPI2C pins and internal signals

Table 116. FMPI2C input/output pins
Pin name Signal type Description

I2C_SDA Bidirectional I2C data

I2C_SCL Bidirectional I2C clock
I2C_SMBA Bidirectional SMBus alert

Table 117. FMPI2C internal input/output signals

Internal signal name Signal type Description

I2C kernel clock, also named I2CCLK in this

i2c_ker_ck Input
document
i2c_pclk Input I2C APB clock
I2C interrupts, refer to Table 130: FMPI2C
i2c_it Output Interrupt requests for the full list of interrupt
sources
i2c_rx_dma Output I2C receive data DMA request (I2C_RX)
i2c_tx_dma Output I2C transmit data DMA request (I2C_TX)

23.4.3 FMPI2C clock requirements

The FMPI2C kernel is clocked by FMPI2CCLK.
The FMPI2CCLK period tI2CCLK must respect the following conditions:
tI2CCLK < (tLOW - tfilters) / 4 and tI2CCLK < tHIGH
with:
tLOW: SCL low time and tHIGH: SCL high time
tfilters: when enabled, sum of the delays brought by the analog filter and by the digital filter.
Analog filter delay is maximum 260 ns. Digital filter delay is DNF x tI2CCLK.
The PCLK clock period tPCLK must respect the following condition:
tPCLK < 4/3 tSCL
with tSCL: SCL period
Caution: When the FMPI2C kernel is clocked by PCLK, this clock must respect the conditions for
tI2CCLK.

RM0402 Rev 6 659/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.4.4 Mode selection

The interface can operate in one of the four following modes:
• Slave transmitter
• Slave receiver
• Master transmitter
• Master receiver
By default, it operates in slave mode. The interface automatically switches from slave to
master when it generates a START condition, and from master to slave if an arbitration loss
or a STOP generation occurs, allowing multimaster capability.

Communication flow
In Master mode, the FMPI2C interface initiates a data transfer and generates the clock
signal. A serial data transfer always begins with a START condition and ends with a STOP
condition. Both START and STOP conditions are generated in master mode by software.
In Slave mode, the interface is capable of recognizing its own addresses (7 or 10-bit), and
the general call address. The general call address detection can be enabled or disabled by
software. The reserved SMBus addresses can also be enabled by software.
Data and addresses are transferred as 8-bit bytes, MSB first. The first byte(s) following the
START condition contain the address (one in 7-bit mode, two in 10-bit mode). The address
is always transmitted in Master mode.
A 9th clock pulse follows the 8 clock cycles of a byte transfer, during which the receiver must
send an acknowledge bit to the transmitter. Refer to the following figure.

Figure 209. I2C bus protocol

SDA
MSB ACK

SCL
1 2 8 9

Start Stop
condition condition

MS19854V1

Acknowledge can be enabled or disabled by software. The FMPI2C interface addresses can
be selected by software.

23.4.5 FMPI2C initialization

Enabling and disabling the peripheral
The FMPI2C peripheral clock must be configured and enabled in the clock controller.
Then the FMPI2C can be enabled by setting the PE bit in the FMPI2C_CR1 register.

660/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

When the FMPI2C is disabled (PE=0), the I2C performs a software reset. Refer to
Section 23.4.6: Software reset for more details.

Noise filters
Before enabling the FMPI2C peripheral by setting the PE bit in FMPI2C_CR1 register, the
user must configure the noise filters, if needed. By default, an analog noise filter is present
on the SDA and SCL inputs. This analog filter is compliant with the I2C specification which
requires the suppression of spikes with a pulse width up to 50 ns in Fast-mode and Fast-
mode Plus. The user can disable this analog filter by setting the ANFOFF bit, and/or select a
digital filter by configuring the DNF[3:0] bit in the FMPI2C_CR1 register.
When the digital filter is enabled, the level of the SCL or the SDA line is internally changed
only if it remains stable for more than DNF x FMPI2CCLK periods. This allows spikes with a
programmable length of 1 to 15 FMPI2CCLK periods to be suppressed.

Table 118. Comparison of analog vs. digital filters

- Analog filter Digital filter

Pulse width of Programmable length from 1 to 15 I2C peripheral

≥ 50 ns
suppressed spikes clocks

Caution: Changing the filter configuration is not allowed when the FMPI2C is enabled.

RM0402 Rev 6 661/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

FMPI2C timings
The timings must be configured in order to guarantee a correct data hold and setup time,
used in master and slave modes. This is done by programming the PRESC[3:0],
SCLDEL[3:0] and SDADEL[3:0] bits in the FMPI2C_TIMINGR register.
The STM32CubeMX tool calculates and provides the I2C_TIMINGR content in the I2C
configuration window

Figure 210. Setup and hold timings

DATA HOLD TIME
SCL falling edge internal
detection

tSYNC1 SDADEL: SCL stretched low by the I2C

SDA output delay

SCL

SDA

tHD;DAT
Data hold time: in case of transmission, the data is sent on SDA output after
the SDADEL delay, if it is already available in I2C_TXDR.

DATA SETUP TIME

SCLDEL
SCL stretched low by the I2C

SCL

SDA

tSU;DAT
SU;STA
Data setup time: in case of transmission, the SCLDEL counter starts
when the data is sent on SDA output. MSv40108V1
MS49608V1

662/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

• When the SCL falling edge is internally detected, a delay is inserted before sending
SDA output. This delay is tSDADEL = SDADEL x tPRESC + tI2CCLK where tPRESC = (PRESC+1)
x tI2CCLK.
TSDADEL impacts the hold time tHD;DAT.

The total SDA output delay is:

tSYNC1 + {[SDADEL x (PRESC+1) + 1] x tI2CCLK }
tSYNC1 duration depends on these parameters:
– SCL falling slope
– When enabled, input delay brought by the analog filter: tAF(min) < tAF < tAF(max)
– When enabled, input delay brought by the digital filter: tDNF = DNF x tI2CCLK
– Delay due to SCL synchronization to FMPI2CCLK clock (2 to 3 FMPI2CCLK
periods)
In order to bridge the undefined region of the SCL falling edge, the user must program
SDADEL in such a way that:
{tf (max) +tHD;DAT (min) -tAF(min) - [(DNF +3) x tI2CCLK]} / {(PRESC +1) x tI2CCLK } ≤ SDADEL
SDADEL ≤ {tHD;DAT (max) -tAF(max) - [(DNF+4) x tI2CCLK]} / {(PRESC +1) x tI2CCLK }
Note: tAF(min) / tAF(max) are part of the equation only when the analog filter is enabled. Refer to
device datasheet for tAF values.
The maximum tHD;DAT can be 3.45 µs, 0.9 µs and 0.45 µs for Standard-mode, Fast-mode
and Fast-mode Plus, but must be less than the maximum of tVD;DAT by a transition time.
This maximum must only be met if the device does not stretch the LOW period (tLOW) of the
SCL signal. If the clock stretches the SCL, the data must be valid by the set-up time before
it releases the clock.
The SDA rising edge is usually the worst case, so in this case the previous equation
becomes:
SDADEL ≤ {tVD;DAT (max) -tr (max) -260 ns - [(DNF+4) x tI2CCLK]} / {(PRESC +1) x tI2CCLK }.
Note: This condition can be violated when NOSTRETCH=0, because the device stretches SCL
low to guarantee the set-up time, according to the SCLDEL value.
Refer to Table 119: I2C-SMBus specification data setup and hold times for tf, tr, tHD;DAT and
tVD;DAT standard values.
• After tSDADEL delay, or after sending SDA output in case the slave had to stretch the
clock because the data was not yet written in I2C_TXDR register, SCL line is kept at
low level during the setup time. This setup time is tSCLDEL = (SCLDEL+1) x tPRESC where
tPRESC = (PRESC+1) x tI2CCLK.
tSCLDEL impacts the setup time tSU;DAT .

In order to bridge the undefined region of the SDA transition (rising edge usually worst
case), the user must program SCLDEL in such a way that:
{[tr (max) + tSU;DAT (min)] / [(PRESC+1)] x tI2CCLK]} - 1 <= SCLDEL
Refer to Table 119: I2C-SMBus specification data setup and hold times for tr and tSU;DAT
standard values.

RM0402 Rev 6 663/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

The SDA and SCL transition time values to be used are the ones in the application. Using
the maximum values from the standard increases the constraints for the SDADEL and
SCLDEL calculation, but ensures the feature whatever the application.
Note: At every clock pulse, after SCL falling edge detection, the I2C master or slave stretches SCL
low during at least [(SDADEL+SCLDEL+1) x (PRESC+1) + 1] x tI2CCLK, in both transmission
and reception modes. In transmission mode, in case the data is not yet written in I2C_TXDR
when SDADEL counter is finished, the I2C keeps on stretching SCL low until the next data
is written. Then new data MSB is sent on SDA output, and SCLDEL counter starts,
continuing stretching SCL low to guarantee the data setup time.
If NOSTRETCH=1 in slave mode, the SCL is not stretched. Consequently the SDADEL
must be programmed in such a way to guarantee also a sufficient setup time.

Table 119. I2C-SMBus specification data setup and hold times

Standard-mode Fast-mode Fast-mode Plus
SMBus
(Sm) (Fm) (Fm+)
Symbol Parameter Unit
Min. Max Min. Max Min. Max Min. Max

tHD;DAT Data hold time 0 - 0 - 0 - 0.3 -

µs
tVD;DAT Data valid time - 3.45 - 0.9 - 0.45 - -
tSU;DAT Data setup time 250 - 100 - 50 - 250 -
Rise time of both SDA
tr - 1000 - 300 - 120 - 1000
and SCL signals ns
Fall time of both SDA
tf - 300 - 300 - 120 - 300
and SCL signals

Additionally, in master mode, the SCL clock high and low levels must be configured by
programming the PRESC[3:0], SCLH[7:0] and SCLL[7:0] bits in the FMPI2C_TIMINGR
register.
• When the SCL falling edge is internally detected, a delay is inserted before releasing
the SCL output. This delay is tSCLL = (SCLL+1) x tPRESC where tPRESC = (PRESC+1) x
tI2CCLK.
tSCLL impacts the SCL low time tLOW .
• When the SCL rising edge is internally detected, a delay is inserted before forcing the
SCL output to low level. This delay is tSCLH = (SCLH+1) x tPRESC where tPRESC =
(PRESC+1) x tI2CCLK. tSCLH impacts the SCL high time tHIGH .

Refer to FMPI2C master initialization for more details.

Caution: Changing the timing configuration is not allowed when the FMPI2C is enabled.
The FMPI2C slave NOSTRETCH mode must also be configured before enabling the
peripheral. Refer to FMPI2C slave initialization for more details.
Caution: Changing the NOSTRETCH configuration is not allowed when the FMPI2C is enabled.

664/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Figure 211. FMPI2C initialization flowchart

Initial settings

Clear PE bit in FMPI2C_CR1

Configure ANFOFF and DNF[3:0] in

FMPI2C_CR1

Configure PRESC[3:0],

SDADEL[3:0], SCLDEL[3:0], SCLH[7:0],

SCLL[7:0] in FMPI2C_TIMINGR

Configure NOSTRETCH in FMPI2C_CR1

Set PE bit in FMPI2C_CR1

End

MSv35962V1

23.4.6 Software reset

A software reset can be performed by clearing the PE bit in the FMPI2C_CR1 register. In
that case FMPI2C lines SCL and SDA are released. Internal states machines are reset and
communication control bits, as well as status bits come back to their reset value. The
configuration registers are not impacted.
Here is the list of impacted register bits:
1. FMPI2C_CR2 register: START, STOP, NACK
2. FMPI2C_ISR register: BUSY, TXE, TXIS, RXNE, ADDR, NACKF, TCR, TC, STOPF,
BERR, ARLO, OVR
and in addition when the SMBus feature is supported:
1. FMPI2C_CR2 register: PECBYTE
2. FMPI2C_ISR register: PECERR, TIMEOUT, ALERT
PE must be kept low during at least 3 APB clock cycles in order to perform the software
reset. This is ensured by writing the following software sequence: - Write PE=0 - Check
PE=0 - Write PE=1.

RM0402 Rev 6 665/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.4.7 Data transfer

The data transfer is managed through transmit and receive data registers and a shift
register.

Reception
The SDA input fills the shift register. After the 8th SCL pulse (when the complete data byte is
received), the shift register is copied into FMPI2C_RXDR register if it is empty (RXNE=0). If
RXNE=1, meaning that the previous received data byte has not yet been read, the SCL line
is stretched low until FMPI2C_RXDR is read. The stretch is inserted between the 8th and
9th SCL pulse (before the acknowledge pulse).

Figure 212. Data reception

ACK pulse ACK pulse

legend:
SCL SCL
stretch
Shift register xx data1 xx data2 xx

RXNE

rd data0 rd data1

FMPI2C_RXDR data0 data1 data2

MSv35976V1

666/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Transmission
If the FMPI2C_TXDR register is not empty (TXE=0), its content is copied into the shift
register after the 9th SCL pulse (the Acknowledge pulse). Then the shift register content is
shifted out on SDA line. If TXE=1, meaning that no data is written yet in FMPI2C_TXDR,
SCL line is stretched low until FMPI2C_TXDR is written. The stretch is done after the 9th
SCL pulse.

Figure 213. Data transmission

ACK pulse ACK pulse

legend:
SCL
SCL
stretch

data2
data1
Shift register xx xx xx

TXE

wr data1 wr data2

FMPI2C_TXDR data0 data1 data2

MSv35977V1

Hardware transfer management

The FMPI2C has a byte counter embedded in hardware in order to manage byte transfer
and to close the communication in various modes such as:
– NACK, STOP and ReSTART generation in master mode
– ACK control in slave receiver mode
– PEC generation/checking when SMBus feature is supported
The byte counter is always used in master mode. By default it is disabled in slave mode, but
it can be enabled by software by setting the SBC (Slave Byte Control) bit in the
FMPI2C_CR2 register.
The number of bytes to be transferred is programmed in the NBYTES[7:0] bit field in the
FMPI2C_CR2 register. If the number of bytes to be transferred (NBYTES) is greater than
255, or if a receiver wants to control the acknowledge value of a received data byte, the
reload mode must be selected by setting the RELOAD bit in the FMPI2C_CR2 register. In
this mode, the TCR flag is set when the number of bytes programmed in NBYTES is
transferred, and an interrupt is generated if TCIE is set. SCL is stretched as long as TCR
flag is set. TCR is cleared by software when NBYTES is written to a non-zero value.
When the NBYTES counter is reloaded with the last number of bytes, RELOAD bit must be
cleared.

RM0402 Rev 6 667/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

When RELOAD=0 in master mode, the counter can be used in 2 modes:

• Automatic end mode (AUTOEND = ‘1’ in the FMPI2C_CR2 register). In this mode, the
master automatically sends a STOP condition once the number of bytes programmed
in the NBYTES[7:0] bit field is transferred.
• Software end mode (AUTOEND = ‘0’ in the FMPI2C_CR2 register). In this mode,
software action is expected once the number of bytes programmed in the NBYTES[7:0]
bit field is transferred; the TC flag is set and an interrupt is generated if the TCIE bit is
set. The SCL signal is stretched as long as the TC flag is set. The TC flag is cleared by
software when the START or STOP bit is set in the FMPI2C_CR2 register. This mode
must be used when the master wants to send a RESTART condition.
Caution: The AUTOEND bit has no effect when the RELOAD bit is set.

Table 120. FMPI2C configuration

Function SBC bit RELOAD bit AUTOEND bit

Master Tx/Rx NBYTES + STOP x 0 1

Master Tx/Rx + NBYTES + RESTART x 0 0
Slave Tx/Rx
0 x x
all received bytes ACKed
Slave Rx with ACK control 1 1 x

23.4.8 FMPI2C slave mode

FMPI2C slave initialization
In order to work in slave mode, the user must enable at least one slave address. Two
registers FMPI2C_OAR1 and FMPI2C_OAR2 are available in order to program the slave
own addresses OA1 and OA2.
• OA1 can be configured either in 7-bit mode (by default) or in 10-bit addressing mode by
setting the OA1MODE bit in the FMPI2C_OAR1 register.
OA1 is enabled by setting the OA1EN bit in the FMPI2C_OAR1 register.
• If additional slave addresses are required, the 2nd slave address OA2 can be
configured. Up to 7 OA2 LSB can be masked by configuring the OA2MSK[2:0] bits in
the FMPI2C_OAR2 register. Therefore for OA2MSK configured from 1 to 6, only
OA2[7:2], OA2[7:3], OA2[7:4], OA2[7:5], OA2[7:6] or OA2[7] are compared with the
received address. As soon as OA2MSK is not equal to 0, the address comparator for
OA2 excludes the FMPI2C reserved addresses (0000 XXX and 1111 XXX), which are
not acknowledged. If OA2MSK=7, all received 7-bit addresses are acknowledged
(except reserved addresses). OA2 is always a 7-bit address.
These reserved addresses can be acknowledged if they are enabled by the specific
enable bit, if they are programmed in the FMPI2C_OAR1 or FMPI2C_OAR2 register
with OA2MSK=0.
OA2 is enabled by setting the OA2EN bit in the FMPI2C_OAR2 register.
• The general call address is enabled by setting the GCEN bit in the FMPI2C_CR1
register.
When the FMPI2C is selected by one of its enabled addresses, the ADDR interrupt status
flag is set, and an interrupt is generated if the ADDRIE bit is set.

668/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

By default, the slave uses its clock stretching capability, which means that it stretches the
SCL signal at low level when needed, in order to perform software actions. If the master
does not support clock stretching, the FMPI2C must be configured with NOSTRETCH=1 in
the FMPI2C_CR1 register.
After receiving an ADDR interrupt, if several addresses are enabled the user must read the
ADDCODE[6:0] bits in the FMPI2C_ISR register in order to check which address matched.
DIR flag must also be checked in order to know the transfer direction.

Slave clock stretching (NOSTRETCH = 0)

In default mode, the FMPI2C slave stretches the SCL clock in the following situations:
• When the ADDR flag is set: the received address matches with one of the enabled
slave addresses. This stretch is released when the ADDR flag is cleared by software
setting the ADDRCF bit.
• In transmission, if the previous data transmission is completed and no new data is
written in FMPI2C_TXDR register, or if the first data byte is not written when the ADDR
flag is cleared (TXE=1). This stretch is released when the data is written to the
FMPI2C_TXDR register.
• In reception when the FMPI2C_RXDR register is not read yet and a new data reception
is completed. This stretch is released when FMPI2C_RXDR is read.
• When TCR = 1 in Slave Byte Control mode, reload mode (SBC=1 and RELOAD=1),
meaning that the last data byte has been transferred. This stretch is released when
then TCR is cleared by writing a non-zero value in the NBYTES[7:0] field.
• After SCL falling edge detection, the FMPI2C stretches SCL low during
[(SDADEL+SCLDEL+1) x (PRESC+1) + 1] x tI2CCLK.

Slave without clock stretching (NOSTRETCH = 1)

When NOSTRETCH = 1 in the FMPI2C_CR1 register, the FMPI2C slave does not stretch
the SCL signal.
• The SCL clock is not stretched while the ADDR flag is set.
• In transmission, the data must be written in the FMPI2C_TXDR register before the first
SCL pulse corresponding to its transfer occurs. If not, an underrun occurs, the OVR
flag is set in the FMPI2C_ISR register and an interrupt is generated if the ERRIE bit is
set in the FMPI2C_CR1 register. The OVR flag is also set when the first data
transmission starts and the STOPF bit is still set (has not been cleared). Therefore, if
the user clears the STOPF flag of the previous transfer only after writing the first data to
be transmitted in the next transfer, he ensures that the OVR status is provided, even for
the first data to be transmitted.
• In reception, the data must be read from the FMPI2C_RXDR register before the 9th
SCL pulse (ACK pulse) of the next data byte occurs. If not an overrun occurs, the OVR
flag is set in the FMPI2C_ISR register and an interrupt is generated if the ERRIE bit is
set in the FMPI2C_CR1 register.

RM0402 Rev 6 669/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Slave byte control mode

In order to allow byte ACK control in slave reception mode, The Slave byte control mode
must be enabled by setting the SBC bit in the FMPI2C_CR1 register. This is required to be
compliant with SMBus standards.
The Reload mode must be selected in order to allow byte ACK control in slave reception
mode (RELOAD=1). To get control of each byte, NBYTES must be initialized to 0x1 in the
ADDR interrupt subroutine, and reloaded to 0x1 after each received byte. When the byte is
received, the TCR bit is set, stretching the SCL signal low between the 8th and 9th SCL
pulses. The user can read the data from the FMPI2C_RXDR register, and then decide to
acknowledge it or not by configuring the ACK bit in the FMPI2C_CR2 register. The SCL
stretch is released by programming NBYTES to a non-zero value: the acknowledge or not-
acknowledge is sent and next byte can be received.
NBYTES can be loaded with a value greater than 0x1, and in this case, the reception flow is
continuous during NBYTES data reception.
Note: The SBC bit must be configured when the FMPI2C is disabled, or when the slave is not
addressed, or when ADDR=1.
The RELOAD bit value can be changed when ADDR=1, or when TCR=1.
Caution: The Slave byte control mode is not compatible with NOSTRETCH mode. Setting SBC when
NOSTRETCH=1 is not allowed.

Figure 214. Slave initialization flowchart

Slave
initialization

Initial settings

Clear {OA1EN, OA2EN} in FMPI2C_OAR1 and FMPI2C_OAR2

Configure {OA1[9:0], OA1MODE, OA1EN,

OA2[6:0], OA2MSK[2:0], OA2EN, GCEN}

Configure SBC in FMPI2C_CR1*

Enable interrupts and/or

DMA in FMPI2C_CR1

End

*SBC must be set to support SMBus features

MSv35963V1

670/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Slave transmitter
A transmit interrupt status (TXIS) is generated when the FMPI2C_TXDR register becomes
empty. An interrupt is generated if the TXIE bit is set in the FMPI2C_CR1 register.
The TXIS bit is cleared when the FMPI2C_TXDR register is written with the next data byte
to be transmitted.
When a NACK is received, the NACKF bit is set in the FMPI2C_ISR register and an
interrupt is generated if the NACKIE bit is set in the FMPI2C_CR1 register. The slave
automatically releases the SCL and SDA lines in order to let the master perform a STOP or
a RESTART condition. The TXIS bit is not set when a NACK is received.
When a STOP is received and the STOPIE bit is set in the FMPI2C_CR1 register, the
STOPF flag is set in the FMPI2C_ISR register and an interrupt is generated. In most
applications, the SBC bit is usually programmed to ‘0’. In this case, If TXE = 0 when the
slave address is received (ADDR=1), the user can choose either to send the content of the
FMPI2C_TXDR register as the first data byte, or to flush the FMPI2C_TXDR register by
setting the TXE bit in order to program a new data byte.
In Slave byte control mode (SBC=1), the number of bytes to be transmitted must be
programmed in NBYTES in the address match interrupt subroutine (ADDR=1). In this case,
the number of TXIS events during the transfer corresponds to the value programmed in
NBYTES.
Caution: When NOSTRETCH=1, the SCL clock is not stretched while the ADDR flag is set, so the
user cannot flush the FMPI2C_TXDR register content in the ADDR subroutine, in order to
program the first data byte. The first data byte to be sent must be previously programmed in
the FMPI2C_TXDR register:
• This data can be the data written in the last TXIS event of the previous transmission
message.
• If this data byte is not the one to be sent, the FMPI2C_TXDR register can be flushed by
setting the TXE bit in order to program a new data byte. The STOPF bit must be
cleared only after these actions, in order to guarantee that they are executed before the
first data transmission starts, following the address acknowledge.
If STOPF is still set when the first data transmission starts, an underrun error is
generated (the OVR flag is set).
If a TXIS event is needed, (transmit interrupt or transmit DMA request), the user must
set the TXIS bit in addition to the TXE bit, in order to generate a TXIS event.

RM0402 Rev 6 671/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 215. Transfer sequence flowchart for FMPI2C slave transmitter,

NOSTRETCH= 0

Slave
transmission

Slave initialization

No

FMPI2C_ISR.ADDR
=1?

Yes
SCL
stretched
Read ADDCODE and DIR in FMPI2C_ISR
Optional: Set FMPI2C_ISR.TXE = 1
Set FMPI2C_ICR.ADDRCF

No
FMPI2C_ISR.TXIS
=1?

Yes

Write FMPI2C_TXDR.TXDATA

MSv35964V1

672/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Figure 216. Transfer sequence flowchart for FMPI2C slave transmitter,

NOSTRETCH= 1

Slave
transmission

Slave initialization

No
No
FMPI2C_ISR.TXIS FMPI2C_ISR.STOPF
=1? =1?

Yes Yes

Write FMPI2C_TXDR.TXDATA Optional: Set FMPI2C_ISR.TXE = 1

and FMPI2C_ISR.TXIS=1

Set FMPI2C_ICR.STOPCF

MSv35965V1

RM0402 Rev 6 673/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 217. Transfer bus diagrams for FMPI2C slave transmitter

legend:
Example FMPI2C slave transmitter 3 bytes with 1st data flushed
NOSTRETCH=0: transmission
ADDR TXIS TXIS TXIS TXIS reception

S Address A A A data3 NA P
SCL stretch
data1 data2

EV1 EV2 EV3 EV4 EV5

TXE

EV1: ADDR ISR: check ADDCODE and DIR, set TXE, set ADDRCF
EV2: TXIS ISR: wr data1
EV3: TXIS ISR: wr data2
EV4: TXIS ISR: wr data3
EV5: TXIS ISR: wr data4 (not sent)

legend :
Example FMPI2C slave transmitter 3 bytes without 1st data flush,
NOSTRETCH=0: transmission
ADDR TXIS TXIS TXIS reception

SCL stretch
S Address A data1 A data2 A data3 NA P

EV1 EV2 EV3 EV4

TXE

EV1: ADDR ISR: check ADDCODE and DIR, set ADDRCF

EV2: TXIS ISR: wr data2
EV3: TXIS ISR: wr data3
EV4: TXIS ISR: wr data4 (not sent)

legend:
Example FMPI2C slave transmitter 3 bytes, NOSTRETCH=1:
transmission
TXIS TXIS TXIS STOPF
reception

S Address A data1 A data2 A data3 NA P SCL stretch

EV1 EV2 EV3 EV4 EV5

TXE

EV1: wr data1
EV2: TXIS ISR: wr data2
EV3: TXIS ISR: wr data3
EV4: TXIS ISR: wr data4 (not sent)
EV5: STOPF ISR: (optional: set TXE and TXIS), set STOPCF

MS35975V1

674/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Slave receiver
RXNE is set in FMPI2C_ISR when the FMPI2C_RXDR is full, and generates an interrupt if
RXIE is set in FMPI2C_CR1. RXNE is cleared when FMPI2C_RXDR is read.
When a STOP is received and STOPIE is set in FMPI2C_CR1, STOPF is set in
FMPI2C_ISR and an interrupt is generated.

Figure 218. Transfer sequence flowchart for slave receiver with NOSTRETCH=0

Slave reception

Slave initialization

No

FMPI2C_ISR.ADDR
=1?

Yes
SCL
stretched
Read ADDCODE and DIR in FMPI2C_ISR
Set FMPI2C_ICR.ADDRCF

No
FMPI2C_ISR.RXNE
=1?

Yes

Write FMPI2C_RXDR.RXDATA

MSv35966V1

RM0402 Rev 6 675/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 219. Transfer sequence flowchart for slave receiver with NOSTRETCH=1

Slave reception

Slave initialization

No
No
FMPI2C_ISR.RXNE FMPI2C_ISR.STOPF
=1? =1?

Yes Yes

Read FMPI2C_RXDR.RXDATA Set FMPI2C_ICR.STOPCF

MSv35967V1

Figure 220. Transfer bus diagrams for FMPI2C slave receiver

legend:
Example FMPI2C slave receiver 3 bytes, NOSTRETCH=0:
transmission
ADDR RXNE RXNE RXNE reception

SCL stretch
S Address A data1 A data2 A data3 A

EV1 EV2 EV3 EV4

RXNE

EV1: ADDR ISR: check ADDCODE and DIR, set ADDRCF

EV2: RXNE ISR: rd data1
EV3 : RXNE ISR: rd data2
EV4: RXNE ISR: rd data3

Example FMPI2C slave receiver 3 bytes, NOSTRETCH=1: legend:

transmission
RXNE RXNE RXNE reception

S Address A data 1 A data 2 A data 3 A P SCL stretch

EV1 EV2 EV3

RXNE

EV1: RXNE ISR: rd data1

EV2: RXNE ISR: rd data2
EV3: RXNE ISR: rd data3
EV4: STOPF ISR: set STOPCF
MS35978V1

676/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

23.4.9 FMPI2C master mode

FMPI2C master initialization
Before enabling the peripheral, the FMPI2C master clock must be configured by setting the
SCLH and SCLL bits in the FMPI2C_TIMINGR register.
The STM32CubeMX tool calculates and provides the I2C_TIMINGR content in the I2C
Configuration window.
A clock synchronization mechanism is implemented in order to support multi-master
environment and slave clock stretching.
In order to allow clock synchronization:
• The low level of the clock is counted using the SCLL counter, starting from the SCL low
level internal detection.
• The high level of the clock is counted using the SCLH counter, starting from the SCL
high level internal detection.
The FMPI2C detects its own SCL low level after a tSYNC1 delay depending on the SCL falling
edge, SCL input noise filters (analog + digital) and SCL synchronization to the I2CxCLK
clock. The FMPI2C releases SCL to high level once the SCLL counter reaches the value
programmed in the SCLL[7:0] bits in the FMPI2C_TIMINGR register.
The FMPI2C detects its own SCL high level after a tSYNC2 delay depending on the SCL rising
edge, SCL input noise filters (analog + digital) and SCL synchronization to I2CxCLK clock.
The FMPI2C ties SCL to low level once the SCLH counter is reached reaches the value
programmed in the SCLH[7:0] bits in the FMPI2C_TIMINGR register.
Consequently the master clock period is:
tSCL = tSYNC1 + tSYNC2 + {[(SCLH+1) + (SCLL+1)] x (PRESC+1) x tI2CCLK}
The duration of tSYNC1 depends on these parameters:
– SCL falling slope
– When enabled, input delay induced by the analog filter.
– When enabled, input delay induced by the digital filter: DNF x tI2CCLK
– Delay due to SCL synchronization with FMPI2CCLK clock (2 to 3 FMPI2CCLK
periods)
The duration of tSYNC2 depends on these parameters:
– SCL rising slope
– When enabled, input delay induced by the analog filter.
– When enabled, input delay induced by the digital filter: DNF x tI2CCLK
– Delay due to SCL synchronization with FMPI2CCLK clock (2 to 3 FMPI2CCLK
periods)

RM0402 Rev 6 677/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 221. Master clock generation

SCL master clock generation

SCL high level detected

SCLH counter starts

tSYNC2 SCLH
SCLL
tSYNC1
SCL

SCL low level detected

SCL released
SCLL counter starts

SCL driven low

SCL master clock synchronization

SCL high level detected SCL high level detected SCL high level detected
SCLH counter starts SCLH counter starts SCLH counter starts

SCLH SCLH SCLH

SCLL SCLL

SCL driven low by SCL driven low by

another device another device
SCL low level detected
SCLL counter starts

SCL low level detected

SCLL counter starts SCL released

MS19858V1

Caution: In order to be I2C or SMBus compliant, the master clock must respect the timings given the
table below.

678/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Table 121. I2C-SMBus specification clock timings

Standard- Fast-mode Fast-mode
SMBus
mode (Sm) (Fm) Plus (Fm+)
Symbol Parameter Unit
Min Max Min Max Min Max Min Max

fSCL SCL clock frequency - 100 - 400 - 1000 - 100 kHz

tHD:STA Hold time (repeated) START condition 4.0 - 0.6 - 0.26 - 4.0 - µs
Set-up time for a repeated START
tSU:STA 4.7 - 0.6 - 0.26 - 4.7 - µs
condition
tSU:STO Set-up time for STOP condition 4.0 - 0.6 - 0.26 - 4.0 - µs
Bus free time between a STOP and
tBUF 4.7 - 1.3 - 0.5 - 4.7 - µs
START condition
tLOW Low period of the SCL clock 4.7 - 1.3 - 0.5 - 4.7 - µs
tHIGH Period of the SCL clock 4.0 - 0.6 - 0.26 - 4.0 50 µs
tr Rise time of both SDA and SCL signals - 1000 - 300 - 120 - 1000 ns
tf Fall time of both SDA and SCL signals - 300 - 300 - 120 - 300 ns

Note: SCLL is also used to generate the tBUF and tSU:STA timings.
SCLH is also used to generate the tHD:STA and tSU:STO timings.
Refer to Section 23.4.10: FMPI2C_TIMINGR register configuration examples for examples
of FMPI2C_TIMINGR settings vs. FMPI2CCLK frequency.

Master communication initialization (address phase)

In order to initiate the communication, the user must program the following parameters for
the addressed slave in the FMPI2C_CR2 register:
• Addressing mode (7-bit or 10-bit): ADD10
• Slave address to be sent: SADD[9:0]
• Transfer direction: RD_WRN
• In case of 10-bit address read: HEAD10R bit. HEAD10R must be configure to indicate
if the complete address sequence must be sent, or only the header in case of a
direction change.
• The number of bytes to be transferred: NBYTES[7:0]. If the number of bytes is equal to
or greater than 255 bytes, NBYTES[7:0] must initially be filled with 0xFF.
The user must then set the START bit in FMPI2C_CR2 register. Changing all the above bits
is not allowed when START bit is set.
Then the master automatically sends the START condition followed by the slave address as
soon as it detects that the bus is free (BUSY = 0) and after a delay of tBUF.
In case of an arbitration loss, the master automatically switches back to slave mode and can
acknowledge its own address if it is addressed as a slave.
Note: The START bit is reset by hardware when the slave address has been sent on the bus,
whatever the received acknowledge value. The START bit is also reset by hardware if an
arbitration loss occurs.
In 10-bit addressing mode, when the Slave Address first 7 bits is NACKed by the slave, the

RM0402 Rev 6 679/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

master re-launches automatically the slave address transmission until ACK is received. In
this case ADDRCF must be set if a NACK is received from the slave, in order to stop
sending the slave address.
If the FMPI2C is addressed as a slave (ADDR=1) while the START bit is set, the FMPI2C
switches to slave mode and the START bit is cleared, when the ADDRCF bit is set.
Note: The same procedure is applied for a Repeated Start condition. In this case BUSY=1.

Figure 222. Master initialization flowchart

Master
initialization

Initial settings

Enable interrupts and/or DMA in FMPI2C_CR1

End

MSv35968V1

Initialization of a master receiver addressing a 10-bit address slave

• If the slave address is in 10-bit format, the user can choose to send the complete read
sequence by clearing the HEAD10R bit in the FMPI2C_CR2 register. In this case the
master automatically sends the following complete sequence after the START bit is set:
(Re)Start + Slave address 10-bit header Write + Slave address 2nd byte + REStart +
Slave address 10-bit header Read

Figure 223. 10-bit address read access with HEAD10R=0

11110XX 0 11110XX 1

Slave address Slave address Slave address

S R/W A1 A2 Sr R/W A3 DATA A DATA A P
1st 7 bits 2nd byte 1st 7 bits

Write Read

MSv41066V1

680/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

• If the master addresses a 10-bit address slave, transmits data to this slave and then
reads data from the same slave, a master transmission flow must be done first. Then a
repeated start is set with the 10 bit slave address configured with HEAD10R=1. In this
case the master sends this sequence: ReStart + Slave address 10-bit header Read.

Figure 224. 10-bit address read access with HEAD10R=1

11110XX 0

Slave address Slave address

S R/W A A DATA A DATA A/A
1st 7 bits 2nd byte

Write

11110XX 1
Slave address
Sr R/W A DATA A DATA A P
1st 7 bits

Read

MS19823V1

Master transmitter
In the case of a write transfer, the TXIS flag is set after each byte transmission, after the 9th
SCL pulse when an ACK is received.
A TXIS event generates an interrupt if the TXIE bit is set in the FMPI2C_CR1 register. The
flag is cleared when the FMPI2C_TXDR register is written with the next data byte to be
transmitted.
The number of TXIS events during the transfer corresponds to the value programmed in
NBYTES[7:0]. If the total number of data bytes to be sent is greater than 255, reload mode
must be selected by setting the RELOAD bit in the FMPI2C_CR2 register. In this case,
when NBYTES data have been transferred, the TCR flag is set and the SCL line is stretched
low until NBYTES[7:0] is written to a non-zero value.
The TXIS flag is not set when a NACK is received.
• When RELOAD=0 and NBYTES data have been transferred:
– In automatic end mode (AUTOEND=1), a STOP is automatically sent.
– In software end mode (AUTOEND=0), the TC flag is set and the SCL line is
stretched low in order to perform software actions:
A RESTART condition can be requested by setting the START bit in the
FMPI2C_CR2 register with the proper slave address configuration, and number of
bytes to be transferred. Setting the START bit clears the TC flag and the START
condition is sent on the bus.
A STOP condition can be requested by setting the STOP bit in the FMPI2C_CR2
register. Setting the STOP bit clears the TC flag and the STOP condition is sent on
the bus.
• If a NACK is received: the TXIS flag is not set, and a STOP condition is automatically
sent after the NACK reception. the NACKF flag is set in the FMPI2C_ISR register, and
an interrupt is generated if the NACKIE bit is set.

RM0402 Rev 6 681/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 225. Transfer sequence flowchart for FMPI2C master transmitter for N≤255
bytes

Master
transmission

Master initialization

NBYTES = N
AUTOEND = 0 for RESTART; 1 for STOP
Configure slave address
Set FMPI2C_CR2.START

No

No
FMPI2C_ISR.NACKF FMPI2C_ISR.TXIS
= 1? = 1?

Yes Yes

Write FMPI2C_TXDR
End

NBYTES No
transmitted?

Yes

Yes
FMPI2C_ISR.TC
= 1?
Set FMPI2C_CR2.START
No with slave addess NBYTES
...
End

MSv35969V1

682/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Figure 226. Transfer sequence flowchart for FMPI2C master transmitter for N>255
bytes

Master
transmission

Master initialization

NBYTES = 0xFF; N=N-255

RELOAD = 1
Configure slave address
Set FMPI2C_CR2.START

No

No
FMPI2C_ISR.NACKF FMPI2C_ISR.TXIS
= 1? = 1?

Yes Yes

Write FMPI2C_TXDR
End

No
NBYTES
transmitted ?

Yes

Yes
FMPI2C_ISR.TC
= 1?

Set FMPI2C_CR2.START
with slave addess No
NBYTES ...

FMPI2C_ISR.TCR
= 1?

Yes
IF N< 256
NBYTES = N; N = 0; RELOAD = 0
AUTOEND = 0 for RESTART; 1 for STOP
End
ELSE
NBYTES = 0xFF; N = N-255
RELOAD = 1

MSv35970V1

RM0402 Rev 6 683/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 227. Transfer bus diagrams for FMPI2C master transmitter

Example FMPI2C master transmitter 2 bytes, automatic end mode (STOP)

legend:

TXIS TXIS transmission

reception
S Address A data1 A data2 A P
SCL stretch
INIT EV1 EV2

TXE

NBYTES xx 2

INIT: program Slave address, program NBYTES = 2, AUTOEND=1, set START

EV1: TXIS ISR: wr data1
EV2: TXIS ISR: wr data2

Example FMPI2C master transmitter 2 bytes, software end mode (RESTART)

TXIS TXIS TC legend:

transmission
S Address A data1 A data2 A ReS Address
reception

INIT EV1 EV2 EV3 SCL stretch

TXE

NBYTES xx 2

INIT: program Slave address, program NBYTES = 2, AUTOEND=0, set START

EV1: TXIS ISR: wr data1
EV2: TXIS ISR: wr data2
EV3: TC ISR: program Slave address, program NBYTES = N, set START

MS35980V1

684/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Master receiver
In the case of a read transfer, the RXNE flag is set after each byte reception, after the 8th
SCL pulse. An RXNE event generates an interrupt if the RXIE bit is set in the FMPI2C_CR1
register. The flag is cleared when FMPI2C_RXDR is read.
If the total number of data bytes to be received is greater than 255, reload mode must be
selected by setting the RELOAD bit in the FMPI2C_CR2 register. In this case, when
NBYTES[7:0] data have been transferred, the TCR flag is set and the SCL line is stretched
low until NBYTES[7:0] is written to a non-zero value.
• When RELOAD=0 and NBYTES[7:0] data have been transferred:
– In automatic end mode (AUTOEND=1), a NACK and a STOP are automatically
sent after the last received byte.
– In software end mode (AUTOEND=0), a NACK is automatically sent after the last
received byte, the TC flag is set and the SCL line is stretched low in order to allow
software actions:
A RESTART condition can be requested by setting the START bit in the
FMPI2C_CR2 register with the proper slave address configuration, and number of
bytes to be transferred. Setting the START bit clears the TC flag and the START
condition, followed by slave address, are sent on the bus.
A STOP condition can be requested by setting the STOP bit in the FMPI2C_CR2
register. Setting the STOP bit clears the TC flag and the STOP condition is sent on
the bus.

RM0402 Rev 6 685/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 228. Transfer sequence flowchart for FMPI2C master receiver for N≤255 bytes

Master reception

Master initialization

NBYTES = N
AUTOEND = 0 for RESTART; 1 for STOP
Configure slave address
Set FMPI2C_CR2.START

No
FMPI2C_ISR.RXNE
=1?

Yes

Read FMPI2C_RXDR

NBYTES No
received?

Yes

Yes
FMPI2C_ISR.TC
= 1?
Set FMPI2C_CR2.START
No with slave addess NBYTES
...
End

MSv35971V1

686/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Figure 229. Transfer sequence flowchart for FMPI2C master receiver for N >255 bytes

Master reception

Master initialization

NBYTES = 0xFF; N=N-255

RELOAD =1
Configure slave address
Set FMPI2C_CR2.START

No
FMPI2C_ISR.RXNE
= 1?

Yes

Read FMPI2C_RXDR

NBYTES No
received?

Yes

Yes
FMPI2C_ISR.TC
= 1?
Set FMPI2C_CR2.START
with slave addess No
NBYTES ...
No
FMPI2C_ISR.TCR
= 1?

Yes
IF N< 256
NBYTES =N; N=0;RELOAD=0
AUTOEND=0 for RESTART; 1 for STOP
ELSE
NBYTES =0xFF;N=N-255
RELOAD=1

End

MSv35972V1

RM0402 Rev 6 687/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 230. Transfer bus diagrams for FMPI2C master receiver

Example FMPI2C master receiver 2 bytes, automatic end mode (STOP)

RXNE RXNE
legend:

S Address A data1 A data2 NA P transmission

reception
INIT EV1 EV2
SCL stretch
NBYTES xx 2

INIT: program Slave address, program NBYTES = 2, AUTOEND=1, set START

EV1: RXNE ISR: rd data1
EV2: RXNE ISR: rd data2

Example FMPI2C master receiver 2 bytes,software end mode (RESTART)

RXNE RXNE TC legend:

transmission
S Address A data1 A data2 NA ReS Address
reception

INIT EV1 EV2 SCL stretch

NBYTES

xx 2 N

INIT: program Slave address, program NBYTES = 2, AUTOEND=0, set START

EV1: RXNE ISR: rd data1
EV2: RXNE ISR: read data2
EV3: TC ISR: program Slave address, program NBYTES = N, set START

MS35979V1

688/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

23.4.10 FMPI2C_TIMINGR register configuration examples

The tables below provide examples of how to program the FMPI2C_TIMINGR to obtain
timings compliant with the I2C specification. In order to get more accurate configuration
values, the STM32CubeMX tool (I2C Configuration window) must be used.

Table 122. Examples of timing settings for fI2CCLK = 8 MHz

Standard-mode (Sm) Fast-mode (Fm) Fast-mode Plus (Fm+)
Parameter
10 kHz 100 kHz 400 kHz 500 kHz

PRESC 1 1 0 0
SCLL 0xC7 0x13 0x9 0x6
tSCLL 200x250 ns = 50 µs 20x250 ns = 5.0 µs 10x125 ns = 1250 ns 7x125 ns = 875 ns
SCLH 0xC3 0xF 0x3 0x3
tSCLH 196x250 ns = 49 µs 16x250 ns = 4.0µs 4x125 ns = 500 ns 4x125 ns = 500 ns
(1)
tSCL ~100 µs(2) ~10 µs(2) ~2500 ns(3) ~2000 ns(4)
SDADEL 0x2 0x2 0x1 0x0
tSDADEL 2x250 ns = 500 ns 2x250 ns = 500 ns 1x125 ns = 125 ns 0 ns
SCLDEL 0x4 0x4 0x3 0x1
tSCLDEL 5x250 ns = 1250 ns 5x250 ns = 1250 ns 4x125 ns = 500 ns 2x125 ns = 250 ns
1. SCL period tSCL is greater than tSCLL + tSCLH due to SCL internal detection delay. Values provided for tSCL are examples
only.
2. tSYNC1 + tSYNC2 minimum value is 4 x tI2CCLK = 500 ns. Example with tSYNC1 + tSYNC2 = 1000 ns.
3. tSYNC1 + tSYNC2 minimum value is 4 x tI2CCLK = 500 ns. Example with tSYNC1 + tSYNC2 = 750 ns.
4. tSYNC1 + tSYNC2 minimum value is 4 x tI2CCLK = 500 ns. Example with tSYNC1 + tSYNC2 = 655 ns.

Table 123. Examples of timings settings for fI2CCLK = 16 MHz

Standard-mode (Sm) Fast-mode (Fm) Fast-mode Plus (Fm+)
Parameter
10 kHz 100 kHz 400 kHz 1000 kHz

PRESC 3 3 1 0
SCLL 0xC7 0x13 0x9 0x4
tSCLL 200 x 250 ns = 50 µs 20 x 250 ns = 5.0 µs 10 x 125 ns = 1250 ns 5 x 62.5 ns = 312.5 ns
SCLH 0xC3 0xF 0x3 0x2
tSCLH 196 x 250 ns = 49 µs 16 x 250 ns = 4.0 µs 4 x 125 ns = 500 ns 3 x 62.5 ns = 187.5 ns
(1)
tSCL ~100 µs(2) ~10 µs(2) ~2500 ns(3) ~1000 ns(4)
SDADEL 0x2 0x2 0x2 0x0
tSDADEL 2 x 250 ns = 500 ns 2 x 250 ns = 500 ns 2 x 125 ns = 250 ns 0 ns
SCLDEL 0x4 0x4 0x3 0x2
tSCLDEL 5 x 250 ns = 1250 ns 5 x 250 ns = 1250 ns 4 x 125 ns = 500 ns 3 x 62.5 ns = 187.5 ns
1. SCL period tSCL is greater than tSCLL + tSCLH due to SCL internal detection delay. Values provided for tSCL are examples
only.

RM0402 Rev 6 689/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

2. tSYNC1 + tSYNC2 minimum value is 4 x tI2CCLK = 250 ns. Example with tSYNC1 + tSYNC2 = 1000 ns.
3. tSYNC1 + tSYNC2 minimum value is 4 x tI2CCLK = 250 ns. Example with tSYNC1 + tSYNC2 = 750 ns.
4. tSYNC1 + tSYNC2 minimum value is 4 x tI2CCLK = 250 ns. Example with tSYNC1 + tSYNC2 = 500 ns.

23.4.11 SMBus specific features

This section is relevant only when SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.

Introduction
The system management bus (SMBus) is a two-wire interface through which various
devices can communicate with each other and with the rest of the system. It is based on I2C
principles of operation. The SMBus provides a control bus for system and power
management related tasks.
This peripheral is compatible with the SMBus specification (http://smbus.org).
The System Management Bus Specification refers to three types of devices.
• A slave is a device that receives or responds to a command.
• A master is a device that issues commands, generates the clocks and terminates the
transfer.
• A host is a specialized master that provides the main interface to the system’s CPU. A
host must be a master-slave and must support the SMBus host notify protocol. Only
one host is allowed in a system.
This peripheral can be configured as master or slave device, and also as a host.

Bus protocols
There are eleven possible command protocols for any given device. A device may use any
or all of the eleven protocols to communicate. The protocols are Quick Command, Send
Byte, Receive Byte, Write Byte, Write Word, Read Byte, Read Word, Process Call, Block
Read, Block Write and Block Write-Block Read Process Call. These protocols should be
implemented by the user software.
For more details of these protocols, refer to SMBus specification (http://smbus.org).

Address resolution protocol (ARP)

SMBus slave address conflicts can be resolved by dynamically assigning a new unique
address to each slave device. In order to provide a mechanism to isolate each device for the
purpose of address assignment each device must implement a unique device identifier
(UDID). This 128-bit number is implemented by software.
This peripheral supports the Address Resolution Protocol (ARP). The SMBus Device
Default Address (0b1100 001) is enabled by setting SMBDEN bit in FMPI2C_CR1 register.
The ARP commands should be implemented by the user software.
Arbitration is also performed in slave mode for ARP support.
For more details of the SMBus address resolution protocol, refer to SMBus specification
(http://smbus.org).

690/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Received command and data acknowledge control

A SMBus receiver must be able to NACK each received command or data. In order to allow
the ACK control in slave mode, the Slave Byte Control mode must be enabled by setting
SBC bit in FMPI2C_CR1 register. Refer to Slave byte control mode on page 670 for more
details.

Host notify protocol

This peripheral supports the host notify protocol by setting the SMBHEN bit in the
FMPI2C_CR1 register. In this case the host acknowledges the SMBus host address
(0b0001 000).
When this protocol is used, the device acts as a master and the host as a slave.

SMBus alert
The SMBus ALERT optional signal is supported. A slave-only device can signal the host
through the SMBALERT# pin that it wants to talk. The host processes the interrupt and
simultaneously accesses all SMBALERT# devices through the alert response address
(0b0001 100). Only the device(s) which pulled SMBALERT# low acknowledges the alert
response address.
When configured as a slave device(SMBHEN=0), the SMBA pin is pulled low by setting the
ALERTEN bit in the FMPI2C_CR1 register. The Alert Response Address is enabled at the
same time.
When configured as a host (SMBHEN=1), the ALERT flag is set in the FMPI2C_ISR register
when a falling edge is detected on the SMBA pin and ALERTEN=1. An interrupt is
generated if the ERRIE bit is set in the FMPI2C_CR1 register. When ALERTEN=0, the
ALERT line is considered high even if the external SMBA pin is low.
If the SMBus ALERT pin is not needed, the SMBA pin can be used as a standard GPIO if
ALERTEN=0.

Packet error checking

A packet error checking mechanism has been introduced in the SMBus specification to
improve reliability and communication robustness. The packet error checking is
implemented by appending a packet error code (PEC) at the end of each message transfer.
The PEC is calculated by using the C(x) = x8 + x2 + x + 1 CRC-8 polynomial on all the
message bytes (including addresses and read/write bits).
The peripheral embeds a hardware PEC calculator and allows a not acknowledge to be sent
automatically when the received byte does not match with the hardware calculated PEC.

RM0402 Rev 6 691/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Timeouts
This peripheral embeds hardware timers in order to be compliant with the 3 timeouts defined
in SMBus specification.

Table 124. SMBus timeout specifications

Limits
Symbol Parameter Unit
Min Max

tTIMEOUT Detect clock low timeout 25 35 ms

tLOW:SEXT(1) Cumulative clock low extend time (slave device) - 25 ms
tLOW:MEXT(2) Cumulative clock low extend time (master device) - 10 ms
1. tLOW:SEXT is the cumulative time a given slave device is allowed to extend the clock cycles in one message
from the initial START to the STOP. It is possible that, another slave device or the master also extends the
clock causing the combined clock low extend time to be greater than tLOW:SEXT. Therefore, this parameter is
measured with the slave device as the sole target of a full-speed master.
2. tLOW:MEXT is the cumulative time a master device is allowed to extend its clock cycles within each byte of a
message as defined from START-to-ACK, ACK-to-ACK, or ACK-to-STOP. It is possible that a slave device
or another master also extends the clock causing the combined clock low time to be greater than tLOW:MEXT
on a given byte. Therefore, this parameter is measured with a full speed slave device as the sole target of
the master.

Figure 231. Timeout intervals for tLOW:SEXT, tLOW:MEXT.

Start Stop
tLOW:SEXT

ClkAck ClkAck
tLOW:MEXT tLOW:MEXT tLOW:MEXT

SMBCLK

SMBDAT

MS19866V1

692/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Bus idle detection

A master can assume that the bus is free if it detects that the clock and data signals have
been high for tIDLE greater than tHIGH,MAX. (refer to Table 119: I2C-SMBus specification data
setup and hold times)
This timing parameter covers the condition where a master has been dynamically added to
the bus and may not have detected a state transition on the SMBCLK or SMBDAT lines. In
this case, the master must wait long enough to ensure that a transfer is not currently in
progress. The peripheral supports a hardware bus idle detection.

23.4.12 SMBus initialization

This section is relevant only when SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
In addition to FMPI2C initialization, some other specific initialization must be done in order
to perform SMBus communication:

Received command and data acknowledge control (Slave mode)

A SMBus receiver must be able to NACK each received command or data. In order to allow
ACK control in slave mode, the Slave byte control mode must be enabled by setting the
SBC bit in the FMPI2C_CR1 register. Refer to Slave byte control mode on page 670 for
more details.

Specific address (Slave mode)

The specific SMBus addresses must be enabled if needed. Refer to Bus idle detection on
page 693 for more details.
• The SMBus device default address (0b1100 001) is enabled by setting the SMBDEN
bit in the FMPI2C_CR1 register.
• The SMBus host address (0b0001 000) is enabled by setting the SMBHEN bit in the
FMPI2C_CR1 register.
• The alert response address (0b0001100) is enabled by setting the ALERTEN bit in the
FMPI2C_CR1 register.

Packet error checking

PEC calculation is enabled by setting the PECEN bit in the FMPI2C_CR1 register. Then the
PEC transfer is managed with the help of a hardware byte counter: NBYTES[7:0] in the
FMPI2C_CR2 register. The PECEN bit must be configured before enabling the FMPI2C.
The PEC transfer is managed with the hardware byte counter, so the SBC bit must be set
when interfacing the SMBus in slave mode. The PEC is transferred after NBYTES-1 data
have been transferred when the PECBYTE bit is set and the RELOAD bit is cleared. If
RELOAD is set, PECBYTE has no effect.
Caution: Changing the PECEN configuration is not allowed when the FMPI2C is enabled.

RM0402 Rev 6 693/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Table 125. SMBus with PEC configuration

Mode SBC bit RELOAD bit AUTOEND bit PECBYTE bit

Master Tx/Rx NBYTES + PEC+ STOP x 0 1 1

Master Tx/Rx NBYTES + PEC + ReSTART x 0 0 1
Slave Tx/Rx with PEC 1 0 x 1

Timeout detection
The timeout detection is enabled by setting the TIMOUTEN and TEXTEN bits in the
FMPI2C_TIMEOUTR register. The timers must be programmed in such a way that they
detect a timeout before the maximum time given in the SMBus specification.
• tTIMEOUT check
In order to enable the tTIMEOUT check, the 12-bit TIMEOUTA[11:0] bits must be
programmed with the timer reload value in order to check the tTIMEOUT parameter. The
TIDLE bit must be configured to ‘0’ in order to detect the SCL low level timeout.
Then the timer is enabled by setting the TIMOUTEN in the FMPI2C_TIMEOUTR
register.
If SCL is tied low for a time greater than (TIMEOUTA+1) x 2048 x tI2CCLK, the TIMEOUT
flag is set in the FMPI2C_ISR register.
Refer to Table 126: Examples of TIMEOUTA settings for various FMPI2CCLK
frequencies (max tTIMEOUT = 25 ms).
Caution: Changing the TIMEOUTA[11:0] bits and TIDLE bit configuration is not allowed when the
TIMEOUTEN bit is set.
• tLOW:SEXT and tLOW:MEXT check
Depending on if the peripheral is configured as a master or as a slave, The 12-bit
TIMEOUTB timer must be configured in order to check tLOW:SEXT for a slave and
tLOW:MEXT for a master. As the standard specifies only a maximum, the user can choose
the same value for the both.
Then the timer is enabled by setting the TEXTEN bit in the FMPI2C_TIMEOUTR
register.
If the SMBus peripheral performs a cumulative SCL stretch for a time greater than
(TIMEOUTB+1) x 2048 x tI2CCLK, and in the timeout interval described in Bus idle
detection on page 693 section, the TIMEOUT flag is set in the FMPI2C_ISR register.
Refer to Table 127: Examples of TIMEOUTB settings for various FMPI2CCLK
frequencies
Caution: Changing the TIMEOUTB configuration is not allowed when the TEXTEN bit is set.

Bus idle detection

In order to enable the tIDLE check, the 12-bit TIMEOUTA[11:0] field must be programmed
with the timer reload value in order to obtain the tIDLE parameter. The TIDLE bit must be
configured to ‘1 in order to detect both SCL and SDA high level timeout.
Then the timer is enabled by setting the TIMOUTEN bit in the FMPI2C_TIMEOUTR register.
If both the SCL and SDA lines remain high for a time greater than (TIMEOUTA+1) x 4 x
tI2CCLK, the TIMEOUT flag is set in the FMPI2C_ISR register.

694/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Refer to Table 128: Examples of TIMEOUTA settings for various FMPI2CCLK frequencies
(max tIDLE = 50 µs)
Caution: Changing the TIMEOUTA and TIDLE configuration is not allowed when the TIMEOUTEN is
set.

23.4.13 SMBus: FMPI2C_TIMEOUTR register configuration examples

This section is relevant only when SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
• Configuring the maximum duration of tTIMEOUT to 25 ms:

Table 126. Examples of TIMEOUTA settings for various FMPI2CCLK frequencies

(max tTIMEOUT = 25 ms)
fI2CCLK TIMEOUTA[11:0] bits TIDLE bit TIMEOUTEN bit tTIMEOUT
8 MHz 0x61 0 1 98 x 2048 x 125 ns = 25 ms
16 MHz 0xC3 0 1 196 x 2048 x 62.5 ns = 25 ms

• Configuring the maximum duration of tLOW:SEXT and tLOW:MEXT to 8 ms:

Table 127. Examples of TIMEOUTB settings for various FMPI2CCLK frequencies

fI2CCLK TIMEOUTB[11:0] bits TEXTEN bit tLOW:EXT
8 MHz 0x1F 1 32 x 2048 x 125 ns = 8 ms
16 MHz 0x3F 1 64 x 2048 x 62.5 ns = 8 ms

• Configuring the maximum duration of tIDLE to 50 µs

Table 128. Examples of TIMEOUTA settings for various FMPI2CCLK frequencies

(max tIDLE = 50 µs)
fI2CCLK TIMEOUTA[11:0] bits TIDLE bit TIMEOUTEN bit tTIDLE
8 MHz 0x63 1 1 100 x 4 x 125 ns = 50 µs
16 MHz 0xC7 1 1 200 x 4 x 62.5 ns = 50 µs

23.4.14 SMBus slave mode

This section is relevant only when the SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
In addition to FMPI2C slave transfer management (refer to Section 23.4.8: FMPI2C slave
mode) some additional software flowcharts are provided to support the SMBus.

SMBus slave transmitter

When the IP is used in SMBus, SBC must be programmed to ‘1’ in order to allow the PEC
transmission at the end of the programmed number of data bytes. When the PECBYTE bit
is set, the number of bytes programmed in NBYTES[7:0] includes the PEC transmission. In
that case the total number of TXIS interrupts is NBYTES-1 and the content of the

RM0402 Rev 6 695/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

FMPI2C_PECR register is automatically transmitted if the master requests an extra byte

after the NBYTES-1 data transfer.
Caution: The PECBYTE bit has no effect when the RELOAD bit is set.

Figure 232. Transfer sequence flowchart for SMBus slave transmitter N bytes + PEC

SMBus slave
transmission

Slave initialization

No

FMPI2C_ISR.ADDR
= 1?

Yes

Read ADDCODE and DIR in FMPI2C_ISR SCL

FMPI2C_CR2.NBYTES = N + 1 stretched
PECBYTE=1
Set FMPI2C_ICR.ADDRCF

No
FMPI2C_ISR.TXIS
=1?

Yes

Write FMPI2C_TXDR.TXDATA

MSv35973V1

696/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Figure 233. Transfer bus diagrams for SMBus slave transmitter (SBC=1)
legend:
Example SMBus slave transmitter 2 bytes + PEC,
transmission
ADDR TXIS TXIS reception

S Address A data1 A A PEC NA P

SCL stretch
data2

EV1 EV2 EV3

NBYTES 3

EV1: ADDR ISR: check ADDCODE, program NBYTES=3, set PECBYTE, set ADDRCF
EV2: TXIS ISR: wr data1
EV3: TXIS ISR: wr data2

MS19869V2

SMBus Slave receiver

When the FMPI2C is used in SMBus mode, SBC must be programmed to ‘1’ in order to
allow the PEC checking at the end of the programmed number of data bytes. In order to
allow the ACK control of each byte, the reload mode must be selected (RELOAD=1). Refer
to Slave byte control mode on page 670 for more details.
In order to check the PEC byte, the RELOAD bit must be cleared and the PECBYTE bit
must be set. In this case, after NBYTES-1 data have been received, the next received byte
is compared with the internal FMPI2C_PECR register content. A NACK is automatically
generated if the comparison does not match, and an ACK is automatically generated if the
comparison matches, whatever the ACK bit value. Once the PEC byte is received, it is
copied into the FMPI2C_RXDR register like any other data, and the RXNE flag is set.
In the case of a PEC mismatch, the PECERR flag is set and an interrupt is generated if the
ERRIE bit is set in the FMPI2C_CR1 register.
If no ACK software control is needed, the user can program PECBYTE=1 and, in the same
write operation, program NBYTES with the number of bytes to be received in a continuous
flow. After NBYTES-1 are received, the next received byte is checked as being the PEC.
Caution: The PECBYTE bit has no effect when the RELOAD bit is set.

RM0402 Rev 6 697/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 234. Transfer sequence flowchart for SMBus slave receiver N Bytes + PEC

SMBus slave
reception

Slave initialization

No

FMPI2C_ISR.ADDR
= 1?

Yes

Read ADDCODE and DIR in FMPI2C_ISR SCL

FMPI2C_CR2.NBYTES = 1, RELOAD =1 stretched
PECBYTE=1
Set FMPI2C_ICR.ADDRCF

No
FMPI2C_ISR.RXNE =1?
FMPI2C_ISR.TCR = 1?

Yes
Read FMPI2C_RXDR.RXDATA
Program FMPI2C_CR2.NACK = 0
FMPI2C_CR2.NBYTES = 1
N=N-1

No
N = 1?

Yes
Read FMPI2C_RXDR.RXDATA
Program RELOAD = 0
NACK = 0 and NBYTES = 1

No
FMPI2C_ISR.RXNE
=1?

Yes
Read FMPI2C_RXDR.RXDATA

End
MSv35974V1

698/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Figure 235. Bus transfer diagrams for SMBus slave receiver (SBC=1)
legend:
Example SMBus slave receiver 2 bytes + PEC
transmission
ADDR RXNE RXNE RXNE
reception

S Address A data1 A data2 A PEC A P SCL stretch

EV1 EV2 EV3 EV4

NBYTES 3

EV1: ADDR ISR: check ADDCODE and DIR, program NBYTES = 3, PECBYTE=1, RELOAD=0, set ADDRCF
EV2: RXNE ISR: rd data1
EV3: RXNE ISR: rd data2
EV4: RXNE ISR: rd PEC

Example SMBus slave receiver 2 bytes + PEC, with ACK control legend :

(RELOAD=1/0) transmission
ADDR RXNE,TCR RXNE,TCR RXNE
reception

S Address A data1 A data2 A PEC A P SCL stretch

EV1 EV2 EV3 EV4

NBYTES 1

EV1: ADDR ISR: check ADDCODE and DIR, program NBYTES = 1, PECBYTE=1, RELOAD=1, set ADDRCF
EV2: RXNE-TCR ISR: rd data1, program NACK=0 and NBYTES = 1
EV3: RXNE-TCR ISR: rd data2, program NACK=0, NBYTES = 1 and RELOAD=0
EV4: RXNE-TCR ISR: rd PEC

MS19870V2

This section is relevant only when the SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
In addition to FMPI2C master transfer management (refer to Section 23.4.9: FMPI2C
master mode) some additional software flowcharts are provided to support the SMBus.

SMBus master transmitter

When the SMBus master wants to transmit the PEC, the PECBYTE bit must be set and the
number of bytes must be programmed in the NBYTES[7:0] field, before setting the START
bit. In this case the total number of TXIS interrupts is NBYTES-1. So if the PECBYTE bit is
set when NBYTES=0x1, the content of the FMPI2C_PECR register is automatically
transmitted.
If the SMBus master wants to send a STOP condition after the PEC, automatic end mode
must be selected (AUTOEND=1). In this case, the STOP condition automatically follows the
PEC transmission.

RM0402 Rev 6 699/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

When the SMBus master wants to send a RESTART condition after the PEC, software
mode must be selected (AUTOEND=0). In this case, once NBYTES-1 have been
transmitted, the FMPI2C_PECR register content is transmitted and the TC flag is set after
the PEC transmission, stretching the SCL line low. The RESTART condition must be
programmed in the TC interrupt subroutine.
Caution: The PECBYTE bit has no effect when the RELOAD bit is set.

Figure 236. Bus transfer diagrams for SMBus master transmitter

Example SMBus master transmitter 2 bytes + PEC, automatic end mode (STOP)

TXIS TXIS
legend:

S Address A data1 A data2 A PEC A P transmission

reception
INIT EV1 EV2
SCL stretch
TXE

NBYTES xx 3

INIT: program Slave address, program NBYTES = 3, AUTOEND=1, set PECBYTE, set START
EV1: TXIS ISR: wr data1
EV2: TXIS ISR: wr data2

Example SMBus master transmitter 2 bytes + PEC, software end mode (RESTART)

TC legend:
TXIS TXIS
transmission
S Address A data1 A data2 A PEC A Rstart Address
reception

INIT EV1 EV2 EV3 SCL stretch

xx 3 N

NBYTES

INIT: program Slave address, program NBYTES = 3, AUTOEND=0, set PECBYTE, set START
EV1: TXIS ISR: wr data1
EV2: TXIS ISR: wr data2
EV3: TC ISR: program Slave address, program NBYTES = N, set START

MS19871V2

700/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

SMBus master receiver

When the SMBus master wants to receive the PEC followed by a STOP at the end of the
transfer, automatic end mode can be selected (AUTOEND=1). The PECBYTE bit must be
set and the slave address must be programmed, before setting the START bit. In this case,
after NBYTES-1 data have been received, the next received byte is automatically checked
versus the FMPI2C_PECR register content. A NACK response is given to the PEC byte,
followed by a STOP condition.
When the SMBus master receiver wants to receive the PEC byte followed by a RESTART
condition at the end of the transfer, software mode must be selected (AUTOEND=0). The
PECBYTE bit must be set and the slave address must be programmed, before setting the
START bit. In this case, after NBYTES-1 data have been received, the next received byte is
automatically checked versus the FMPI2C_PECR register content. The TC flag is set after
the PEC byte reception, stretching the SCL line low. The RESTART condition can be
programmed in the TC interrupt subroutine.
Caution: The PECBYTE bit has no effect when the RELOAD bit is set.

RM0402 Rev 6 701/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Figure 237. Bus transfer diagrams for SMBus master receiver

Example SMBus master receiver 2 bytes + PEC, automatic end mode (STOP)

RXNE RXNE RXNE

legend:

S Address A data1 A data2 A PEC NA P transmission

reception
INIT EV1 EV2 EV3
SCL stretch
NBYTES xx 3

INIT: program Slave address, program NBYTES = 3, AUTOEND=1, set PECBYTE, set START
EV1: RXNE ISR: rd data1
EV2: RXNE ISR: rd data2
EV3: RXNE ISR: rd PEC

Example SMBus master receiver 2 bytes + PEC, software end mode (RESTART)

RXNE RXNE RXNE TC legend:

transmission
S Address A data1 A data2 A PEC NA Restart Address
reception

INIT EV1 EV2 EV3 EV4 SCL stretch

NBYTES

xx 3 N

INIT: program Slave address, program NBYTES = 3, AUTOEND=0, set PECBYTE, set START
EV1: RXNE ISR: rd data1
EV2: RXNE ISR: rd data2
EV3: RXNE ISR: read PEC
EV4: TC ISR: program Slave address, program NBYTES = N, set START

MS19872V2

23.4.15 Error conditions

The following errors are the error conditions which may cause communication to fail.

Bus error (BERR)

A bus error is detected when a START or a STOP condition is detected and is not located
after a multiple of 9 SCL clock pulses. A START or a STOP condition is detected when a
SDA edge occurs while SCL is high.
The bus error flag is set only if the FMPI2C is involved in the transfer as master or
addressed slave (i.e not during the address phase in slave mode).
In case of a misplaced START or RESTART detection in slave mode, the FMPI2C enters
address recognition state like for a correct START condition.

702/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

When a bus error is detected, the BERR flag is set in the FMPI2C_ISR register, and an
interrupt is generated if the ERRIE bit is set in the FMPI2C_CR1 register.

Arbitration lost (ARLO)

An arbitration loss is detected when a high level is sent on the SDA line, but a low level is
sampled on the SCL rising edge.
• In master mode, arbitration loss is detected during the address phase, data phase and
data acknowledge phase. In this case, the SDA and SCL lines are released, the
START control bit is cleared by hardware and the master switches automatically to
slave mode.
• In slave mode, arbitration loss is detected during data phase and data acknowledge
phase. In this case, the transfer is stopped, and the SCL and SDA lines are released.
When an arbitration loss is detected, the ARLO flag is set in the FMPI2C_ISR register, and
an interrupt is generated if the ERRIE bit is set in the FMPI2C_CR1 register.

Overrun/underrun error (OVR)

An overrun or underrun error is detected in slave mode when NOSTRETCH=1 and:
• In reception when a new byte is received and the RXDR register has not been read yet.
The new received byte is lost, and a NACK is automatically sent as a response to the
new byte.
• In transmission:
– When STOPF=1 and the first data byte should be sent. The content of the
FMPI2C_TXDR register is sent if TXE=0, 0xFF if not.
– When a new byte must be sent and the FMPI2C_TXDR register has not been
written yet, 0xFF is sent.
When an overrun or underrun error is detected, the OVR flag is set in the FMPI2C_ISR
register, and an interrupt is generated if the ERRIE bit is set in the FMPI2C_CR1 register.

Packet error checking error (PECERR)

This section is relevant only when the SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
A PEC error is detected when the received PEC byte does not match with the
FMPI2C_PECR register content. A NACK is automatically sent after the wrong PEC
reception.
When a PEC error is detected, the PECERR flag is set in the FMPI2C_ISR register, and an
interrupt is generated if the ERRIE bit is set in the FMPI2C_CR1 register.

RM0402 Rev 6 703/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Timeout Error (TIMEOUT)

This section is relevant only when the SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
A timeout error occurs for any of these conditions:
• TIDLE=0 and SCL remained low for the time defined in the TIMEOUTA[11:0] bits: this is
used to detect a SMBus timeout.
• TIDLE=1 and both SDA and SCL remained high for the time defined in the TIMEOUTA
[11:0] bits: this is used to detect a bus idle condition.
• Master cumulative clock low extend time reached the time defined in the
TIMEOUTB[11:0] bits (SMBus tLOW:MEXT parameter)
• Slave cumulative clock low extend time reached the time defined in TIMEOUTB[11:0]
bits (SMBus tLOW:SEXT parameter)
When a timeout violation is detected in master mode, a STOP condition is automatically
sent.
When a timeout violation is detected in slave mode, SDA and SCL lines are automatically
released.
When a timeout error is detected, the TIMEOUT flag is set in the FMPI2C_ISR register, and
an interrupt is generated if the ERRIE bit is set in the FMPI2C_CR1 register.

Alert (ALERT)
This section is relevant only when the SMBus feature is supported. Refer to Section 23.3:
FMPI2C implementation.
The ALERT flag is set when the FMPI2C interface is configured as a Host (SMBHEN=1),
the alert pin detection is enabled (ALERTEN=1) and a falling edge is detected on the SMBA
pin. An interrupt is generated if the ERRIE bit is set in the FMPI2C_CR1 register.

23.4.16 DMA requests

Transmission using DMA
DMA (direct memory access) can be enabled for transmission by setting the TXDMAEN bit
in the FMPI2C_CR1 register. Data is loaded from an SRAM area configured using the DMA
peripheral (see Section 9: Direct memory access controller (DMA)) to the FMPI2C_TXDR
register whenever the TXIS bit is set.
Only the data are transferred with DMA.
• In master mode: the initialization, the slave address, direction, number of bytes and
START bit are programmed by software (the transmitted slave address cannot be
transferred with DMA). When all data are transferred using DMA, the DMA must be

704/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

initialized before setting the START bit. The end of transfer is managed with the
NBYTES counter. Refer to Master transmitter on page 681.
• In slave mode:
– With NOSTRETCH=0, when all data are transferred using DMA, the DMA must be
initialized before the address match event, or in ADDR interrupt subroutine, before
clearing ADDR.
– With NOSTRETCH=1, the DMA must be initialized before the address match
event.
• For instances supporting SMBus: the PEC transfer is managed with NBYTES counter.
Refer to SMBus slave transmitter on page 695 and SMBus master transmitter on
page 699.
Note: If DMA is used for transmission, the TXIE bit does not need to be enabled.

Reception using DMA

DMA (direct memory access) can be enabled for reception by setting the RXDMAEN bit in
the FMPI2C_CR1 register. Data is loaded from the FMPI2C_RXDR register to an SRAM
area configured using the DMA peripheral (refer to Section 9: Direct memory access
controller (DMA)) whenever the RXNE bit is set. Only the data (including PEC) are
transferred with DMA.
• In Master mode, the initialization, the slave address, direction, number of bytes and
START bit are programmed by software. When all data are transferred using DMA, the
DMA must be initialized before setting the START bit. The end of transfer is managed
with the NBYTES counter.
• In Slave mode with NOSTRETCH=0, when all data are transferred using DMA, the
DMA must be initialized before the address match event, or in the ADDR interrupt
subroutine, before clearing the ADDR flag.
• If SMBus is supported (see Section 23.3: FMPI2C implementation): the PEC transfer is
managed with the NBYTES counter. Refer to SMBus Slave receiver on page 697 and
SMBus master receiver on page 701.
Note: If DMA is used for reception, the RXIE bit does not need to be enabled.

23.4.17 Debug mode

When the microcontroller enters debug mode (core halted), the SMBus timeout either
continues to work normally or stops, depending on the DBG_I2Cx_ configuration bits in the
DBG module.

23.5 FMPI2C low-power modes

Table 129. Effect of low-power modes on the FMPI2C
Mode Description

No effect
Sleep
FMPI2C interrupts cause the device to exit the Sleep mode.
Stop The contents of FMPI2C registers are kept.
Standby The FMPI2C peripheral is powered down and must be reinitialized after exiting Standby.

RM0402 Rev 6 705/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.6 FMPI2C interrupts

The table below gives the list of FMPI2C interrupt requests.

Table 130. FMPI2C Interrupt requests

Exit the Exit the Exit the
Interrupt Interrupt Event Enable Interrupt clear
Sleep Stop Standby
acronym event flag control bit method
mode mode modes

Read
Receive buffer
RXNE RXIE FMPI2C_RXDR
not empty
register
Write
Transmit buffer
TXIS TXIE FMPI2C_TXDR
interrupt status
register
Stop detection Write
STOPF STOPIE
interrupt flag STOPCF=1
Write
Transfer No
FMPI2C_CR2
complete TCR
with
FMPI2C reload TCIE Yes
NBYTES[7:0] ≠ 0 No
_EV
Transfer Write START=1
TC
complete or STOP=1
Address Write
ADDR ADDRIE
matched ADDRCF=1
NACK Write
FMP NACKF NACKIE
reception NACKCF=1
I2C
Write
Bus error BERR
BERRCF=1
Write
Arbitration loss ARLO
ARLOCF=1
Overrun/
OVR Write OVRCF=1
FMPI2C Underrun
ERRIE Yes No No
_ER Write
PEC error PECERR
PECERRCF=1
Timeout/ Write
TIMEOUT
tLOW error TIMEOUTCF=1
Write
SMBus alert ALERT
ALERTCF=1

706/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

23.7 FMPI2C registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers are accessed by words (32-bit).

23.7.1 FMPI2C control register 1 (FMPI2C_CR1)

Address offset: 0x00
Reset value: 0x0000 0000
Access: No wait states, except if a write access occurs while a write access to this register is
ongoing. In this case, wait states are inserted in the second write access until the previous
one is completed. The latency of the second write access can be up to
2 x PCLK1 + 6 x FMPI2CCLK.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ALERT SMBD SMBH NOSTR
Res. Res. Res. Res. Res. Res. Res. Res. PECEN GCEN Res. SBC
EN EN EN ETCH

rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RXDMA TXDMA ANF STOP NACK ADDR
Res. DNF[3:0] ERRIE TCIE RXIE TXIE PE
EN EN OFF IE IE IE
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:24 Reserved, must be kept at reset value.

Bit 23 PECEN: PEC enable
0: PEC calculation disabled
1: PEC calculation enabled
Note: If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 22 ALERTEN: SMBus alert enable
0: The SMBus alert pin (SMBA) is not supported in host mode (SMBHEN=1). In device mode
(SMBHEN=0), the SMBA pin is released and the Alert Response Address header is disabled
(0001100x followed by NACK).
1: The SMBus alert pin is supported in host mode (SMBHEN=1). In device mode
(SMBHEN=0), the SMBA pin is driven low and the Alert Response Address header is
enabled (0001100x followed by ACK).
Note: When ALERTEN=0, the SMBA pin can be used as a standard GPIO.
If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 21 SMBDEN: SMBus device default address enable
0: Device default address disabled. Address 0b1100001x is NACKed.
1: Device default address enabled. Address 0b1100001x is ACKed.
Note: If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 20 SMBHEN: SMBus host address enable
0: Host address disabled. Address 0b0001000x is NACKed.
1: Host address enabled. Address 0b0001000x is ACKed.
Note: If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.

RM0402 Rev 6 707/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Bit 19 GCEN: General call enable

0: General call disabled. Address 0b00000000 is NACKed.
1: General call enabled. Address 0b00000000 is ACKed.
Bit 18 Reserved, must be kept at reset value.
Bit 17 NOSTRETCH: Clock stretching disable
This bit is used to disable clock stretching in slave mode. It must be kept cleared in master
mode.
0: Clock stretching enabled
1: Clock stretching disabled
Note: This bit can only be programmed when the I2C is disabled (PE = 0).
Bit 16 SBC: Slave byte control
This bit is used to enable hardware byte control in slave mode.
0: Slave byte control disabled
1: Slave byte control enabled
Bit 15 RXDMAEN: DMA reception requests enable
0: DMA mode disabled for reception
1: DMA mode enabled for reception
Bit 14 TXDMAEN: DMA transmission requests enable
0: DMA mode disabled for transmission
1: DMA mode enabled for transmission
Bit 13 Reserved, must be kept at reset value.
Bit 12 ANFOFF: Analog noise filter OFF
0: Analog noise filter enabled
1: Analog noise filter disabled
Note: This bit can only be programmed when the FMPI2C is disabled (PE = 0).
Bits 11:8 DNF[3:0]: Digital noise filter
These bits are used to configure the digital noise filter on SDA and SCL input. The digital
filter, filters spikes with a length of up to DNF[3:0] * tI2CCLK
0000: Digital filter disabled
0001: Digital filter enabled and filtering capability up to 1 tI2CCLK
...
1111: digital filter enabled and filtering capability up to15 tI2CCLK
Note: If the analog filter is also enabled, the digital filter is added to the analog filter.
This filter can only be programmed when the FMPI2C is disabled (PE = 0).
Bit 7 ERRIE: Error interrupts enable
0: Error detection interrupts disabled
1: Error detection interrupts enabled
Note: Any of these errors generate an interrupt:
Arbitration Loss (ARLO)
Bus Error detection (BERR)
Overrun/Underrun (OVR)
Timeout detection (TIMEOUT)
PEC error detection (PECERR)
Alert pin event detection (ALERT)

708/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Bit 6 TCIE: Transfer Complete interrupt enable

0: Transfer Complete interrupt disabled
1: Transfer Complete interrupt enabled
Note: Any of these events generate an interrupt:
Transfer Complete (TC)
Transfer Complete Reload (TCR)
Bit 5 STOPIE: Stop detection Interrupt enable
0: Stop detection (STOPF) interrupt disabled
1: Stop detection (STOPF) interrupt enabled
Bit 4 NACKIE: Not acknowledge received Interrupt enable
0: Not acknowledge (NACKF) received interrupts disabled
1: Not acknowledge (NACKF) received interrupts enabled
Bit 3 ADDRIE: Address match Interrupt enable (slave only)
0: Address match (ADDR) interrupts disabled
1: Address match (ADDR) interrupts enabled
Bit 2 RXIE: RX Interrupt enable
0: Receive (RXNE) interrupt disabled
1: Receive (RXNE) interrupt enabled
Bit 1 TXIE: TX Interrupt enable
0: Transmit (TXIS) interrupt disabled
1: Transmit (TXIS) interrupt enabled
Bit 0 PE: Peripheral enable
0: Peripheral disable
1: Peripheral enable
Note: When PE=0, the FMPI2C SCL and SDA lines are released. Internal state machines and
status bits are put back to their reset value. When cleared, PE must be kept low for at
least 3 APB clock cycles.

23.7.2 FMPI2C control register 2 (FMPI2C_CR2)

Address offset: 0x04
Reset value: 0x0000 0000
Access: No wait states, except if a write access occurs while a write access to this register is
ongoing. In this case, wait states are inserted in the second write access until the previous
one is completed. The latency of the second write access can be up to 2 x PCLK1 + 6 x
FMPI2CCLK.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PEC AUTOE RE
Res. Res. Res. Res. Res. NBYTES[7:0]
BYTE ND LOAD

rs rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HEAD1 RD_
NACK STOP START ADD10 SADD[9:0]
0R WRN
rs rs rs rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 709/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Bits 31:27 Reserved, must be kept at reset value.

Bit 26 PECBYTE: Packet error checking byte
This bit is set by software, and cleared by hardware when the PEC is transferred, or when a
STOP condition or an Address matched is received, also when PE=0.
0: No PEC transfer.
1: PEC transmission/reception is requested
Note: Writing ‘0’ to this bit has no effect.
This bit has no effect when RELOAD is set.
This bit has no effect is slave mode when SBC=0.
If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 25 AUTOEND: Automatic end mode (master mode)
This bit is set and cleared by software.
0: software end mode: TC flag is set when NBYTES data are transferred, stretching SCL low.
1: Automatic end mode: a STOP condition is automatically sent when NBYTES data are
transferred.
Note: This bit has no effect in slave mode or when the RELOAD bit is set.
Bit 24 RELOAD: NBYTES reload mode
This bit is set and cleared by software.
0: The transfer is completed after the NBYTES data transfer (STOP or RESTART follows).
1: The transfer is not completed after the NBYTES data transfer (NBYTES is reloaded). TCR
flag is set when NBYTES data are transferred, stretching SCL low.
Bits 23:16 NBYTES[7:0]: Number of bytes
The number of bytes to be transmitted/received is programmed there. This field is don’t care
in slave mode with SBC=0.
Note: Changing these bits when the START bit is set is not allowed.
Bit 15 NACK: NACK generation (slave mode)
The bit is set by software, cleared by hardware when the NACK is sent, or when a STOP
condition or an Address matched is received, or when PE=0.
0: an ACK is sent after current received byte.
1: a NACK is sent after current received byte.
Note: Writing ‘0’ to this bit has no effect.
This bit is used in slave mode only: in master receiver mode, NACK is automatically
generated after last byte preceding STOP or RESTART condition, whatever the NACK
bit value.
When an overrun occurs in slave receiver NOSTRETCH mode, a NACK is
automatically generated whatever the NACK bit value.
When hardware PEC checking is enabled (PECBYTE=1), the PEC acknowledge value
does not depend on the NACK value.
Bit 14 STOP: Stop generation (master mode)
The bit is set by software, cleared by hardware when a STOP condition is detected, or when
PE = 0.
In Master Mode:
0: No Stop generation.
1: Stop generation after current byte transfer.
Note: Writing ‘0’ to this bit has no effect.

710/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Bit 13 START: Start generation

This bit is set by software, and cleared by hardware after the Start followed by the address
sequence is sent, by an arbitration loss, by a timeout error detection, or when PE = 0. It can
also be cleared by software by writing ‘1’ to the ADDRCF bit in the FMPI2C_ICR register.
0: No Start generation.
1: Restart/Start generation:
If the FMPI2C is already in master mode with AUTOEND = 0, setting this bit generates a
Repeated Start condition when RELOAD=0, after the end of the NBYTES transfer.
Otherwise setting this bit generates a START condition once the bus is free.
Note: Writing ‘0’ to this bit has no effect.
The START bit can be set even if the bus is BUSY or FMPI2C is in slave mode.
This bit has no effect when RELOAD is set.
Bit 12 HEAD10R: 10-bit address header only read direction (master receiver mode)
0: The master sends the complete 10 bit slave address read sequence: Start + 2 bytes 10bit
address in write direction + Restart + 1st 7 bits of the 10 bit address in read direction.
1: The master only sends the 1st 7 bits of the 10 bit address, followed by Read direction.
Note: Changing this bit when the START bit is set is not allowed.
Bit 11 ADD10: 10-bit addressing mode (master mode)
0: The master operates in 7-bit addressing mode,
1: The master operates in 10-bit addressing mode
Note: Changing this bit when the START bit is set is not allowed.
Bit 10 RD_WRN: Transfer direction (master mode)
0: Master requests a write transfer.
1: Master requests a read transfer.
Note: Changing this bit when the START bit is set is not allowed.
Bits 9:0 SADD[9:0]: Slave address (master mode)
In 7-bit addressing mode (ADD10 = 0):
SADD[7:1] should be written with the 7-bit slave address to be sent. The bits SADD[9],
SADD[8] and SADD[0] are don't care.
In 10-bit addressing mode (ADD10 = 1):
SADD[9:0] should be written with the 10-bit slave address to be sent.
Note: Changing these bits when the START bit is set is not allowed.

RM0402 Rev 6 711/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.7.3 FMPI2C own address 1 register (FMPI2C_OAR1)

Address offset: 0x08
Reset value: 0x0000 0000
Access: No wait states, except if a write access occurs while a write access to this register is
ongoing. In this case, wait states are inserted in the second write access until the previous
one is completed. The latency of the second write access can be up to 2 x PCLK1 + 6 x
FMPI2CCLK.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OA1
OA1EN Res. Res. Res. Res. OA1[9:0]
MODE

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bit 15 OA1EN: Own Address 1 enable
0: Own address 1 disabled. The received slave address OA1 is NACKed.
1: Own address 1 enabled. The received slave address OA1 is ACKed.
Bits 14:11 Reserved, must be kept at reset value.
Bit 10 OA1MODE: Own Address 1 10-bit mode
0: Own address 1 is a 7-bit address.
1: Own address 1 is a 10-bit address.
Note: This bit can be written only when OA1EN=0.
Bits 9:0 OA1[9:0]: Interface own slave address
7-bit addressing mode: OA1[7:1] contains the 7-bit own slave address. The bits OA1[9],
OA1[8] and OA1[0] are don't care.
10-bit addressing mode: OA1[9:0] contains the 10-bit own slave address.
Note: These bits can be written only when OA1EN=0.

712/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

23.7.4 FMPI2C own address 2 register (FMPI2C_OAR2)

Address offset: 0x0C
Reset value: 0x0000 0000
Access: No wait states, except if a write access occurs while a write access to this register is
ongoing. In this case, wait states are inserted in the second write access until the previous
one is completed. The latency of the second write access can be up to 2 x PCLK1 + 6 x
FMPI2CCLK.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OA2EN Res. Res. Res. Res. OA2MSK[2:0] OA2[7:1] Res.

rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bit 15 OA2EN: Own Address 2 enable
0: Own address 2 disabled. The received slave address OA2 is NACKed.
1: Own address 2 enabled. The received slave address OA2 is ACKed.
Bits 14:11 Reserved, must be kept at reset value.
Bits 10:8 OA2MSK[2:0]: Own Address 2 masks
000: No mask
001: OA2[1] is masked and don’t care. Only OA2[7:2] are compared.
010: OA2[2:1] are masked and don’t care. Only OA2[7:3] are compared.
011: OA2[3:1] are masked and don’t care. Only OA2[7:4] are compared.
100: OA2[4:1] are masked and don’t care. Only OA2[7:5] are compared.
101: OA2[5:1] are masked and don’t care. Only OA2[7:6] are compared.
110: OA2[6:1] are masked and don’t care. Only OA2[7] is compared.
111: OA2[7:1] are masked and don’t care. No comparison is done, and all (except reserved)
7-bit received addresses are acknowledged.
Note: These bits can be written only when OA2EN=0.
As soon as OA2MSK is not equal to 0, the reserved FMPI2C addresses (0b0000xxx
and 0b1111xxx) are not acknowledged even if the comparison matches.
Bits 7:1 OA2[7:1]: Interface address
7-bit addressing mode: 7-bit address
Note: These bits can be written only when OA2EN=0.
Bit 0 Reserved, must be kept at reset value.

RM0402 Rev 6 713/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.7.5 FMPI2C timing register (FMPI2C_TIMINGR)

Address offset: 0x10
Reset value: 0x0000 0000
Access: No wait states

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PRESC[3:0] Res. Res. Res. Res. SCLDEL[3:0] SDADEL[3:0]

rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SCLH[7:0] SCLL[7:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:28 PRESC[3:0]: Timing prescaler

This field is used to prescale FMPI2CCLK in order to generate the clock period tPRESC used
for data setup and hold counters (refer to FMPI2C timings on page 662) and for SCL high
and low level counters (refer to FMPI2C master initialization on page 677).
tPRESC = (PRESC+1) x tI2CCLK
Bits 27:24 Reserved, must be kept at reset value.
Bits 23:20 SCLDEL[3:0]: Data setup time
This field is used to generate a delay tSCLDEL between SDA edge and SCL rising edge. In
master mode and in slave mode with NOSTRETCH = 0, the SCL line is stretched low during
tSCLDEL.
tSCLDEL = (SCLDEL+1) x tPRESC
Note: tSCLDEL is used to generate tSU:DAT timing.
Bits 19:16 SDADEL[3:0]: Data hold time
This field is used to generate the delay tSDADEL between SCL falling edge and SDA edge. In
master mode and in slave mode with NOSTRETCH = 0, the SCL line is stretched low during
tSDADEL.
tSDADEL= SDADEL x tPRESC
Note: SDADEL is used to generate tHD:DAT timing.
Bits 15:8 SCLH[7:0]: SCL high period (master mode)
This field is used to generate the SCL high period in master mode.
tSCLH = (SCLH+1) x tPRESC
Note: SCLH is also used to generate tSU:STO and tHD:STA timing.
Bits 7:0 SCLL[7:0]: SCL low period (master mode)
This field is used to generate the SCL low period in master mode.
tSCLL = (SCLL+1) x tPRESC
Note: SCLL is also used to generate tBUF and tSU:STA timings.

Note: This register must be configured when the FMPI2C is disabled (PE = 0).
Note: The STM32CubeMX tool calculates and provides the I2C_TIMINGR content in the I2C
Configuration window.

714/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

23.7.6 FMPI2C timeout register (FMPI2C_TIMEOUTR)

Address offset: 0x14
Reset value: 0x0000 0000
Access: No wait states, except if a write access occurs while a write access to this register is
ongoing. In this case, wait states are inserted in the second write access until the previous
one is completed. The latency of the second write access can be up to 2 x PCLK1 + 6 x
FMPI2CCLK.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
TEXTEN Res. Res. Res. TIMEOUTB[11:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TIMOUTEN Res. Res. TIDLE TIMEOUTA[11:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 TEXTEN: Extended clock timeout enable

0: Extended clock timeout detection is disabled
1: Extended clock timeout detection is enabled. When a cumulative SCL stretch for more
than tLOW:EXT is done by the FMPI2C interface, a timeout error is detected (TIMEOUT=1).
Bits 30:28 Reserved, must be kept at reset value.
Bits 27:16 TIMEOUTB[11:0]: Bus timeout B
This field is used to configure the cumulative clock extension timeout:
In master mode, the master cumulative clock low extend time (tLOW:MEXT) is detected
In slave mode, the slave cumulative clock low extend time (tLOW:SEXT) is detected
tLOW:EXT= (TIMEOUTB+1) x 2048 x tI2CCLK
Note: These bits can be written only when TEXTEN=0.
Bit 15 TIMOUTEN: Clock timeout enable
0: SCL timeout detection is disabled
1: SCL timeout detection is enabled: when SCL is low for more than tTIMEOUT (TIDLE=0) or
high for more than tIDLE (TIDLE=1), a timeout error is detected (TIMEOUT=1).
Bits 14:13 Reserved, must be kept at reset value.
Bit 12 TIDLE: Idle clock timeout detection
0: TIMEOUTA is used to detect SCL low timeout
1: TIMEOUTA is used to detect both SCL and SDA high timeout (bus idle condition)
Note: This bit can be written only when TIMOUTEN=0.
Bits 11:0 TIMEOUTA[11:0]: Bus Timeout A
This field is used to configure:
The SCL low timeout condition tTIMEOUT when TIDLE=0
tTIMEOUT= (TIMEOUTA+1) x 2048 x tI2CCLK
The bus idle condition (both SCL and SDA high) when TIDLE=1
tIDLE= (TIMEOUTA+1) x 4 x tI2CCLK
Note: These bits can be written only when TIMOUTEN=0.

Note: If the SMBus feature is not supported, this register is reserved and forced by hardware to
“0x00000000”. Refer to Section 23.3: FMPI2C implementation.

RM0402 Rev 6 715/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.7.7 FMPI2C interrupt and status register (FMPI2C_ISR)

Address offset: 0x18
Reset value: 0x0000 0001
Access: No wait states

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. ADDCODE[6:0] DIR

r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TIME PEC
BUSY Res. ALERT OVR ARLO BERR TCR TC STOPF NACKF ADDR RXNE TXIS TXE
OUT ERR

r r r r r r r r r r r r r rs rs

Bits 31:24 Reserved, must be kept at reset value.

Bits 23:17 ADDCODE[6:0]: Address match code (Slave mode)
These bits are updated with the received address when an address match event occurs
(ADDR = 1).
In the case of a 10-bit address, ADDCODE provides the 10-bit header followed by the 2
MSBs of the address.
Bit 16 DIR: Transfer direction (Slave mode)
This flag is updated when an address match event occurs (ADDR=1).
0: Write transfer, slave enters receiver mode.
1: Read transfer, slave enters transmitter mode.
Bit 15 BUSY: Bus busy
This flag indicates that a communication is in progress on the bus. It is set by hardware
when a START condition is detected. It is cleared by hardware when a STOP condition is
detected, or when PE=0.
Bit 14 Reserved, must be kept at reset value.
Bit 13 ALERT: SMBus alert
This flag is set by hardware when SMBHEN=1 (SMBus host configuration), ALERTEN=1
and a SMBALERT event (falling edge) is detected on SMBA pin. It is cleared by software by
setting the ALERTCF bit.
Note: This bit is cleared by hardware when PE=0.
If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 12 TIMEOUT: Timeout or tLOW detection flag
This flag is set by hardware when a timeout or extended clock timeout occurred. It is cleared
by software by setting the TIMEOUTCF bit.
Note: This bit is cleared by hardware when PE=0.
If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.

716/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Bit 11 PECERR: PEC Error in reception

This flag is set by hardware when the received PEC does not match with the PEC register
content. A NACK is automatically sent after the wrong PEC reception. It is cleared by
software by setting the PECCF bit.
Note: This bit is cleared by hardware when PE=0.
If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 10 OVR: Overrun/Underrun (slave mode)
This flag is set by hardware in slave mode with NOSTRETCH=1, when an overrun/underrun
error occurs. It is cleared by software by setting the OVRCF bit.
Note: This bit is cleared by hardware when PE=0.
Bit 9 ARLO: Arbitration lost
This flag is set by hardware in case of arbitration loss. It is cleared by software by setting the
ARLOCF bit.
Note: This bit is cleared by hardware when PE=0.
Bit 8 BERR: Bus error
This flag is set by hardware when a misplaced Start or STOP condition is detected whereas
the peripheral is involved in the transfer. The flag is not set during the address phase in slave
mode. It is cleared by software by setting BERRCF bit.
Note: This bit is cleared by hardware when PE=0.
Bit 7 TCR: Transfer Complete Reload
This flag is set by hardware when RELOAD=1 and NBYTES data have been transferred. It is
cleared by software when NBYTES is written to a non-zero value.
Note: This bit is cleared by hardware when PE=0.
This flag is only for master mode, or for slave mode when the SBC bit is set.
Bit 6 TC: Transfer Complete (master mode)
This flag is set by hardware when RELOAD=0, AUTOEND=0 and NBYTES data have been
transferred. It is cleared by software when START bit or STOP bit is set.
Note: This bit is cleared by hardware when PE=0.
Bit 5 STOPF: Stop detection flag
This flag is set by hardware when a STOP condition is detected on the bus and the
peripheral is involved in this transfer:
– either as a master, provided that the STOP condition is generated by the peripheral.
– or as a slave, provided that the peripheral has been addressed previously during
this transfer.
It is cleared by software by setting the STOPCF bit.
Note: This bit is cleared by hardware when PE=0.
Bit 4 NACKF: Not Acknowledge received flag
This flag is set by hardware when a NACK is received after a byte transmission. It is cleared
by software by setting the NACKCF bit.
Note: This bit is cleared by hardware when PE=0.
Bit 3 ADDR: Address matched (slave mode)
This bit is set by hardware as soon as the received slave address matched with one of the
enabled slave addresses. It is cleared by software by setting ADDRCF bit.
Note: This bit is cleared by hardware when PE=0.

RM0402 Rev 6 717/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Bit 2 RXNE: Receive data register not empty (receivers)

This bit is set by hardware when the received data is copied into the FMPI2C_RXDR
register, and is ready to be read. It is cleared when FMPI2C_RXDR is read.
Note: This bit is cleared by hardware when PE=0.
Bit 1 TXIS: Transmit interrupt status (transmitters)
This bit is set by hardware when the FMPI2C_TXDR register is empty and the data to be
transmitted must be written in the FMPI2C_TXDR register. It is cleared when the next data to
be sent is written in the FMPI2C_TXDR register.
This bit can be written to ‘1’ by software when NOSTRETCH=1 only, in order to generate a
TXIS event (interrupt if TXIE=1 or DMA request if TXDMAEN=1).
Note: This bit is cleared by hardware when PE=0.
Bit 0 TXE: Transmit data register empty (transmitters)
This bit is set by hardware when the FMPI2C_TXDR register is empty. It is cleared when the
next data to be sent is written in the FMPI2C_TXDR register.
This bit can be written to ‘1’ by software in order to flush the transmit data register
FMPI2C_TXDR.
Note: This bit is set by hardware when PE=0.

23.7.8 FMPI2C interrupt clear register (FMPI2C_ICR)

Address offset: 0x1C
Reset value: 0x0000 0000
Access: No wait states

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ALERT TIMOU ARLOC BERRC STOPC NACKC ADDR
Res. Res. PECCF OVRCF Res. Res. Res. Res. Res.
CF TCF F F F F CF

w w w w w w w w w

Bits 31:14 Reserved, must be kept at reset value.

Bit 13 ALERTCF: Alert flag clear
Writing 1 to this bit clears the ALERT flag in the FMPI2C_ISR register.
Note: If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 12 TIMOUTCF: Timeout detection flag clear
Writing 1 to this bit clears the TIMEOUT flag in the FMPI2C_ISR register.
Note: If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.
Bit 11 PECCF: PEC Error flag clear
Writing 1 to this bit clears the PECERR flag in the FMPI2C_ISR register.
Note: If the SMBus feature is not supported, this bit is reserved and forced by hardware to ‘0’.
Refer to Section 23.3: FMPI2C implementation.

718/1163 RM0402 Rev 6

RM0402 Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Bit 10 OVRCF: Overrun/Underrun flag clear

Writing 1 to this bit clears the OVR flag in the FMPI2C_ISR register.
Bit 9 ARLOCF: Arbitration lost flag clear
Writing 1 to this bit clears the ARLO flag in the FMPI2C_ISR register.
Bit 8 BERRCF: Bus error flag clear
Writing 1 to this bit clears the BERRF flag in the FMPI2C_ISR register.
Bits 7:6 Reserved, must be kept at reset value.
Bit 5 STOPCF: STOP detection flag clear
Writing 1 to this bit clears the STOPF flag in the FMPI2C_ISR register.
Bit 4 NACKCF: Not Acknowledge flag clear
Writing 1 to this bit clears the NACKF flag in FMPI2C_ISR register.
Bit 3 ADDRCF: Address matched flag clear
Writing 1 to this bit clears the ADDR flag in the FMPI2C_ISR register. Writing 1 to this bit
also clears the START bit in the FMPI2C_CR2 register.
Bits 2:0 Reserved, must be kept at reset value.

23.7.9 FMPI2C PEC register (FMPI2C_PECR)

Address offset: 0x20
Reset value: 0x0000 0000
Access: No wait states

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. PEC[7:0]

r r r r r r r r

Bits 31:8 Reserved, must be kept at reset value.

Bits 7:0 PEC[7:0] Packet error checking register
This field contains the internal PEC when PECEN=1.
The PEC is cleared by hardware when PE=0.

Note: If the SMBus feature is not supported, this register is reserved and forced by hardware to
“0x00000000”. Refer to Section 23.3: FMPI2C implementation.

RM0402 Rev 6 719/1163

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

23.7.10 FMPI2C receive data register (FMPI2C_RXDR)

Address offset: 0x24
Reset value: 0x0000 0000
Access: No wait states

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. RXDATA[7:0]

r r r r r r r r

Bits 31:8 Reserved, must be kept at reset value.

Bits 7:0 RXDATA[7:0] 8-bit receive data
Data byte received from the I2C bus

23.7.11 FMPI2C transmit data register (FMPI2C_TXDR)

Address offset: 0x28
Reset value: 0x0000 0000
Access: No wait states

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. TXDATA[7:0]

rw rw rw rw rw rw rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bits 7:0 TXDATA[7:0] 8-bit transmit data
Data byte to be transmitted to the I2C bus
Note: These bits can be written only when TXE=1.

720/1163 RM0402 Rev 6

 0x8
0x4
0x0

0xC

0x24
0x20
0x18
0x14
0x10

0x1C
Offset
RM0402

23.7.12

name

TIMINGR

FMPI2C_
FMPI2C_
Register

TIMEOUTR

Reset value
Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

FMPI2C_ISR

FMPI2C_ICR
FMPI2C_CR2
FMPI2C_CR1

FMPI2C_PECR
FMPI2C_OAR2
FMPI2C_OAR1

FMPI2C_RXDR
0
0
Res. Res. Res. Res. TEXTEN Res. Res. Res. Res. 31

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 30

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 29

0
PRESC[3:0]
Res. Res. Res. Res. Res. Res. Res. Res. Res. 28

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 27

0
0
Res. Res. Res. Res. Res. Res. Res. PECBYTE Res. 26

0
0
Res. Res. Res. Res. Res. Res. Res. AUTOEND Res. 25
FMPI2C register map

0
0
Res. Res. Res. Res. Res. Res. Res. RELOAD Res. 24

0
0
0
0
0

Res. Res. Res. Res. Res. PECEN 23

0
0
0
0
0

Res. Res. Res. Res. Res. ALERTEN 22

[3:0]

0
0
0
0
0

Res. Res. Res. Res. Res. SMBDEN 21

SCLDEL

0
0
0
0
0

Res. Res. Res. Res. Res. SMBHEN 20

TIMEOUTB[11:0]

0
0
0
0
0

Res. Res. Res. Res. Res. GCEN 19

ADDCODE[6:0]

0
0
0
0
NBYTES[7:0]

Res. Res. Res. Res. Res. Res 18

RM0402 Rev 6
[3:0]

0
0
0
0
0

Res. Res. Res. Res. Res. NOSTRETCH 17

SDADEL

0
0
0
0
0

Res. Res. Res. DIR Res. Res. SBC 16

0
0

0
0
0
0
0

Res. Res. Res. BUSY TIMOUTEN OA2EN OA1EN NACK RXDMAEN 15

0
0
0

Res. Res. Res. Res. Res. Res. Res. STOP TXDMAEN 14

0
0
0

0
Res. Res. ALERTCF ALERT Res. Res. Res. START Res. 13

0
0
0
0

0
0

Res. Res. TIMOUTCF TIMEOUT TIDLE Res. Res. HEAD10R ANFOFF 12

0
0
0
0

0
0

Res. Res. PECCF PECERR Res. Res. ADD10 11

SCLH[7:0]

0
0
0
0
0
0

0
0

Res. Res. OVRCF OVR OA1MODE RD_WRN 10

Table 131. FMPI2C register map and reset values

0
0
0
0
0
0

0
0

Res. Res. ARLOCF ARLO 9

DNF[3:0]

K [2:0]
The table below provides the FMPI2C register map and reset values.

OA2MS

0
0
0
0
0
0

0
0

Res. Res. BERRCF BERR 8

0
0
0
0
0
0
0

0
0
Res. TCR ERRIE 7

0
0
0
0
0
0
0

0
0
Res. TC TCIE 6

0
0
0
0
0
0
0

0
0
0
STOPCF STOPF STOPIE 5

0
0
0
0
0
0
0

0
0
0
NACKCF NACKF NACKIE 4
OA1[9:0]

TIMEOUTA[11:0]
SADD[9:0]

OA2[7:1]

0
0
0
0
0
0
0

0
0
0
ADDRCF ADDR ADDRIE 3

PEC[7:0]
SCLL[7:0]

0
0
0
0
0
0
0

0
0

RXDATA[7:0]
Res. RXNE RXIE 2

0
0
0
0
0
0
0

0
0
Res. TXIS TXIE 1

1
0
0
0
0
0

0
0
0

721/1163
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface

Res. TXE Res. PE

722
Fast-mode Plus Inter-integrated circuit (FMPI2C) interface RM0402

Table 131. FMPI2C register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
FMPI2C_TXDR TXDATA[7:0]
0x28

Reset value 0 0 0 0 0 0 0 0

Refer to Section 2.2 on page 49 for the register boundary addresses.

722/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

24 Inter-integrated circuit (I2C) interface

24.1 I2C introduction

I2C (inter-integrated circuit) bus Interface serves as an interface between the microcontroller
and the serial I2C bus. It provides multimaster capability, and controls all I2C bus-specific
sequencing, protocol, arbitration and timing. It supports the standard mode (Sm, up to 100
kHz) and Fm mode (Fm, up to 400 kHz).
It may be used for a variety of purposes, including CRC generation and verification, SMBus
(system management bus) and PMBus (power management bus).
Depending on specific device implementation DMA capability can be available for reduced
CPU overload.

RM0402 Rev 6 723/1163

757
Inter-integrated circuit (I2C) interface RM0402

24.2 I2C main features

• Parallel-bus/I2C protocol converter
• Multimaster capability: the same interface can act as Master or Slave
• I2C Master features:
– Clock generation
– Start and Stop generation
• I2C Slave features:
– Programmable I2C Address detection
– Dual Addressing Capability to acknowledge 2 slave addresses
– Stop bit detection
• Generation and detection of 7-bit/10-bit addressing and General Call
• Supports different communication speeds:
– Standard Speed (up to 100 kHz)
– Fast Speed (up to 400 kHz)
• Analog noise filter
• Programmable digital noise filter
• Status flags:
– Transmitter/Receiver mode flag
– End-of-Byte transmission flag
– I2C busy flag
• Error flags:
– Arbitration lost condition for master mode
– Acknowledgment failure after address/ data transmission
– Detection of misplaced start or stop condition
– Overrun/Underrun if clock stretching is disabled
• 2 Interrupt vectors:
– 1 Interrupt for successful address/ data communication
– 1 Interrupt for error condition
• Optional clock stretching
• 1-byte buffer with DMA capability
• Configurable PEC (packet error checking) generation or verification:
– PEC value can be transmitted as last byte in Tx mode
– PEC error checking for last received byte
• SMBus 2.0 Compatibility:
– 25 ms clock low timeout delay
– 10 ms master cumulative clock low extend time
– 25 ms slave cumulative clock low extend time
– Hardware PEC generation/verification with ACK control
– Address Resolution Protocol (ARP) supported
• PMBus Compatibility

724/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Note: Some of the above features may not be available in certain products. The user should refer
to the product data sheet, to identify the specific features supported by the I2C interface
implementation.

24.3 I2C functional description

In addition to receiving and transmitting data, this interface converts it from serial to parallel
format and vice versa. The interrupts are enabled or disabled by software. The interface is
connected to the I2C bus by a data pin (SDA) and by a clock pin (SCL). It can be connected
with a standard (up to 100 kHz) or fast (up to 400 kHz) I2C bus.

24.3.1 Mode selection

The interface can operate in one of the four following modes:
• Slave transmitter
• Slave receiver
• Master transmitter
• Master receiver
By default, it operates in slave mode. The interface automatically switches from slave to
master, after it generates a START condition and from master to slave, if an arbitration loss
or a Stop generation occurs, allowing multimaster capability.

Communication flow
In Master mode, the I2C interface initiates a data transfer and generates the clock signal. A
serial data transfer always begins with a start condition and ends with a stop condition. Both
start and stop conditions are generated in master mode by software.
In Slave mode, the interface is capable of recognizing its own addresses (7 or 10-bit), and
the General Call address. The General Call address detection may be enabled or disabled
by software.
Data and addresses are transferred as 8-bit bytes, MSB first. The first byte(s) following the
start condition contain the address (one in 7-bit mode, two in 10-bit mode). The address is
always transmitted in Master mode.
A 9th clock pulse follows the 8 clock cycles of a byte transfer, during which the receiver must
send an acknowledge bit to the transmitter. Refer to Figure 238.

Figure 238. I2C bus protocol

SDA
MSB ACK

SCL
1 2 8 9

Start Stop
condition condition

Acknowledge may be enabled or disabled by software. The I2C interface addresses (dual
addressing 7-bit/ 10-bit and/or general call address) can be selected by software.

RM0402 Rev 6 725/1163

757
Inter-integrated circuit (I2C) interface RM0402

The block diagram of the I2C interface is shown in Figure 239.

Figure 239. I2C block diagram

Data register

Noise Data
SDA Data shift register
filter control

Comparator PEC calculation

Own address register

Dual address register
Clock
Noise PEC register
SCL control
filter

Clock control
Register (CCR)

Control registers
(CR1&CR2)
Control
Status registers logic
(SR1&SR2)
SMBA

Interrupts DMA requests & ACK

MS30035V1

1. SMBA is an optional signal in SMBus mode. This signal is not applicable if SMBus is disabled.

24.3.2 I2C slave mode

By default the I2C interface operates in Slave mode. To switch from default Slave mode to
Master mode a Start condition generation is needed.
The peripheral input clock must be programmed in the I2C_CR2 register in order to
generate correct timings. The peripheral input clock frequency must be at least:
• 2 MHz in Sm mode
• 4 MHz in Fm mode
As soon as a start condition is detected, the address is received from the SDA line and sent
to the shift register. Then it is compared with the address of the interface (OAR1) and with
OAR2 (if ENDUAL=1) or the General Call address (if ENGC = 1).

726/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Note: In 10-bit addressing mode, the comparison includes the header sequence (11110xx0),
where xx denotes the two most significant bits of the address.
Header or address not matched: the interface ignores it and waits for another Start
condition.
Header matched (10-bit mode only): the interface generates an acknowledge pulse if the
ACK bit is set and waits for the 8-bit slave address.
Address matched: the interface generates in sequence:
• An acknowledge pulse if the ACK bit is set
• The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit is
set.
• If ENDUAL=1, the software has to read the DUALF bit to check which slave address
has been acknowledged.
In 10-bit mode, after receiving the address sequence the slave is always in Receiver mode.
It enters Transmitter mode on receiving a repeated Start condition followed by the header
sequence with matching address bits and the least significant bit set (11110xx1).
The TRA bit indicates whether the slave is in Receiver or Transmitter mode.

Slave transmitter
Following the address reception and after clearing ADDR, the slave sends bytes from the
DR register to the SDA line via the internal shift register.
The slave stretches SCL low until ADDR is cleared and DR filled with the data to be sent
(see Figure 240 Transfer sequencing EV1 EV3).
When the acknowledge pulse is received:
• The TxE bit is set by hardware with an interrupt if the ITEVFEN and the ITBUFEN bits
are set.
If TxE is set and some data were not written in the I2C_DR register before the end of the
next data transmission, the BTF bit is set and the interface waits until BTF is cleared by a
read to I2C_SR1 followed by a write to the I2C_DR register, stretching SCL low.

RM0402 Rev 6 727/1163

757
Inter-integrated circuit (I2C) interface RM0402

Figure 240. Transfer sequence diagram for slave transmitter

7-bit slave transmitter
S A ddress A Data1 A Data2 A DataN NA P
.....
EV1 EV3-1 EV3 EV3 EV3 EV3-2

10-bit slave transmitter

S Header A Address A
EV1

Sr Header A Data1 A .... DataN NA P

EV1 EV3_1 EV3 EV3 EV3-2

Legend: S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge, NA= Non-acknowledge ,

EVx= Event (with interrupt if ITEVFEN=1)
AV1: ADDR=1, cleared by reading SR1 followed by reading SR2
EV3-1: TxE=1, shift register empty, data register empty, write Data1 in DR.
EV3-1: TxE=1, shift register not empty, data register empty, cleared by writing DR.
EV3-2: AF=1, AF is cleared by writing ‘0’ in AF bit of SR1 register.
ai18209V2

1. The EV1 and EV3_1 events stretch SCL low until the end of the corresponding software sequence.
2. The EV3 event stretches SCL low if the software sequence is not completed before the end of the next byte
transmission

Slave receiver
Following the address reception and after clearing ADDR, the slave receives bytes from the
SDA line into the DR register via the internal shift register. After each byte the interface
generates in sequence:
• An acknowledge pulse if the ACK bit is set
• The RxNE bit is set by hardware and an interrupt is generated if the ITEVFEN and
ITBUFEN bit is set.
If RxNE is set and the data in the DR register is not read before the end of the next data
reception, the BTF bit is set and the interface waits until BTF is cleared by a read from the
I2C_DR register, stretching SCL low (see Figure 241).

728/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Figure 241. Transfer sequence diagram for slave receiver

7-bit slave receiver
S Address A Data1 A Data2 A DataN A P
.....
EV1 EV2 EV2 EV2 EV4

10-bit slav e receiver

S Header A Address A Data1 A DataN A P
.....
EV2 EV2 EV4

Legend: S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge,

EVx= Event (with interrupt if ITEVFEN=1)

EV1: ADDR=1, cleared by reading SR1 followed by reading SR2

EV2: RxNE=1 cleared by reading DR register.
EV4: STOPF=1, cleared by reading SR1 register followed by writing to the CR1 register
ai18208V2

1. The EV1 event stretches SCL low until the end of the corresponding software sequence.
2. The EV2 event stretches SCL low if the software sequence is not completed before the end of the next byte
reception.
3. After checking the SR1 register content, the user should perform the complete clearing sequence for each
flag found set.
Thus, for ADDR and STOPF flags, the following sequence is required inside the I2C interrupt routine:
READ SR1
if (ADDR == 1) {READ SR1; READ SR2}
if (STOPF == 1) {READ SR1; WRITE CR1}
The purpose is to make sure that both ADDR and STOPF flags are cleared if both are found set.

Closing slave communication

After the last data byte is transferred a Stop Condition is generated by the master. The
interface detects this condition and sets:
• The STOPF bit and generates an interrupt if the ITEVFEN bit is set.
The STOPF bit is cleared by a read of the SR1 register followed by a write to the CR1
register (see Figure 241: Transfer sequence diagram for slave receiver EV4).

24.3.3 I2C master mode

In Master mode, the I2C interface initiates a data transfer and generates the clock signal. A
serial data transfer always begins with a Start condition and ends with a Stop condition.
Master mode is selected as soon as the Start condition is generated on the bus with a
START bit.
The following is the required sequence in master mode.
• Program the peripheral input clock in I2C_CR2 Register in order to generate correct
timings
• Configure the clock control registers
• Configure the rise time register
• Program the I2C_CR1 register to enable the peripheral
• Set the START bit in the I2C_CR1 register to generate a Start condition
The peripheral input clock frequency must be at least:
• 2 MHz in Sm mode
• 4 MHz in Fm mode

RM0402 Rev 6 729/1163

757
Inter-integrated circuit (I2C) interface RM0402

SCL master clock generation

The CCR bits are used to generate the high and low level of the SCL clock, starting from the
generation of the rising and falling edge (respectively). As a slave may stretch the SCL line,
the peripheral checks the SCL input from the bus at the end of the time programmed in
TRISE bits after rising edge generation.
• If the SCL line is low, it means that a slave is stretching the bus, and the high level
counter stops until the SCL line is detected high. This allows to guarantee the minimum
HIGH period of the SCL clock parameter.
• If the SCL line is high, the high level counter keeps on counting.
Indeed, the feedback loop from the SCL rising edge generation by the peripheral to the SCL
rising edge detection by the peripheral takes time even if no slave stretches the clock. This
loopback duration is linked to the SCL rising time (impacting SCL VIH input detection), plus
delay due to the noise filter present on the SCL input path, plus delay due to internal SCL
input synchronization with APB clock. The maximum time used by the feedback loop is
programmed in the TRISE bits, so that the SCL frequency remains stable whatever the SCL
rising time.

Start condition
Setting the START bit causes the interface to generate a Start condition and to switch to
Master mode (MSL bit set) when the BUSY bit is cleared.
Note: In master mode, setting the START bit causes the interface to generate a ReStart condition
at the end of the current byte transfer.
Once the Start condition is sent:
• The SB bit is set by hardware and an interrupt is generated if the ITEVFEN bit is set.
Then the master waits for a read of the SR1 register followed by a write in the DR register
with the Slave address (see Figure 242 and Figure 243 Transfer sequencing EV5).

Slave address transmission

Then the slave address is sent to the SDA line via the internal shift register.
• In 10-bit addressing mode, sending the header sequence causes the following event:
– The ADD10 bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.
Then the master waits for a read of the SR1 register followed by a write in the DR
register with the second address byte (see Figure 242 and Figure 243 Transfer
sequencing).
– The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.
Then the master waits for a read of the SR1 register followed by a read of the SR2
register (see Figure 242 and Figure 243 Transfer sequencing).
• In 7-bit addressing mode, one address byte is sent.
As soon as the address byte is sent,
– The ADDR bit is set by hardware and an interrupt is generated if the ITEVFEN bit
is set.
Then the master waits for a read of the SR1 register followed by a read of the SR2
register (see Figure 242 and Figure 243 Transfer sequencing).

730/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

The master can decide to enter Transmitter or Receiver mode depending on the LSB of the
slave address sent.
• In 7-bit addressing mode,
– To enter Transmitter mode, a master sends the slave address with LSB reset.
– To enter Receiver mode, a master sends the slave address with LSB set.
• In 10-bit addressing mode,
– To enter Transmitter mode, a master sends the header (11110xx0) and then the
slave address, (where xx denotes the two most significant bits of the address).
– To enter Receiver mode, a master sends the header (11110xx0) and then the
slave address. Then it should send a repeated Start condition followed by the
header (11110xx1), (where xx denotes the two most significant bits of the
address).
The TRA bit indicates whether the master is in Receiver or Transmitter mode.

Master transmitter
Following the address transmission and after clearing ADDR, the master sends bytes from
the DR register to the SDA line via the internal shift register.
The master waits until the first data byte is written into I2C_DR (see Figure 242 Transfer
sequencing EV8_1).
When the acknowledge pulse is received, the TxE bit is set by hardware and an interrupt is
generated if the ITEVFEN and ITBUFEN bits are set.
If TxE is set and a data byte was not written in the DR register before the end of the last data
transmission, BTF is set and the interface waits until BTF is cleared by a write to I2C_DR,
stretching SCL low.
Closing the communication
After the last byte is written to the DR register, the STOP bit is set by software to generate a
Stop condition (see Figure 242 Transfer sequencing EV8_2). The interface automatically
goes back to slave mode (MSL bit cleared).
Note: Stop condition should be programmed during EV8_2 event, when either TxE or BTF is set.

RM0402 Rev 6 731/1163

757
Inter-integrated circuit (I2C) interface RM0402

Figure 242. Transfer sequence diagram for master transmitter

7-bit master transmitter

S Address A Data1 A Data2 A DataN A P
.....
EV5 EV6 EV8_1 EV8 EV8 EV8 EV8_2

10-bit master transmitter

S Header A Address A Data1 A DataN A P
.....
EV5 EV9 EV6 EV8_1 EV8 EV8 EV8_2

Legend: S = Start, SR = Repeated start, P = stop, A = Acknowledge

EVx = Event (with interrupt if ITEVFEN = 1)
EV5: SB=1, cleared by reading SR1 register followed by writing DR register with address.
EV6: ADDR=1, cleared by reading SR1 register followed by reading SR2.
EV8_1: TxE=1, shift register empty, data register empty, write Data1 in DR.
EV8: TxE=1, shift register not empty, data register empty, cleared by writing DR register.
EV_2: TxE=1, BTF=1, Program stop request, TxE and BTF are cleared by hardware by the stop condition.
EV9: ADD10=1, cleared by reading SR1 register followed by writing DR register.
ai18210V2

1. The EV5, EV6, EV9, EV8_1 and EV8_2 events stretch SCL low until the end of the corresponding software sequence.
2. The EV8 event stretches SCL low if the software sequence is not complete before the end of the next byte transmission.

732/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Master receiver
Following the address transmission and after clearing ADDR, the I2C interface enters
Master Receiver mode. In this mode the interface receives bytes from the SDA line into the
DR register via the internal shift register. After each byte the interface generates in
sequence:
1. An acknowledge pulse if the ACK bit is set
2. The RxNE bit is set and an interrupt is generated if the ITEVFEN and ITBUFEN bits are
set (see Figure 243 Transfer sequencing EV7).
If the RxNE bit is set and the data in the DR register is not read before the end of the last
data reception, the BTF bit is set by hardware and the interface waits until BTF is cleared by
a read in the DR register, stretching SCL low.
Closing the communication
The master sends a NACK for the last byte received from the slave. After receiving this
NACK, the slave releases the control of the SCL and SDA lines. Then the master can send
a Stop/Restart condition.
1. To generate the nonacknowledge pulse after the last received data byte, the ACK bit
must be cleared just after reading the second last data byte (after second last RxNE
event).
2. In order to generate the Stop/Restart condition, software must set the STOP/START bit
after reading the second last data byte (after the second last RxNE event).
3. In case a single byte has to be received, the Acknowledge disable is made during EV6
(before ADDR flag is cleared) and the STOP condition generation is made after EV6.
After the Stop condition generation, the interface goes automatically back to slave mode
(MSL bit cleared).

RM0402 Rev 6 733/1163

757
Inter-integrated circuit (I2C) interface RM0402

Figure 243. Transfer sequence diagram for master receiver

7-bit master receiver

S Address A Data1 A(1) Data2 A DataN NA P
.....
EV5 EV6 EV7 EV7 EV7_1 EV7

10-bit master receiver

S Header A Address A
EV5 EV9 EV6

Sr Header A Data1 A(1) Data2 A DataN NA P

.....
EV5 EV6 EV7 EV7 EV7_1 EV7

Leg end : S= Start, Sr = Repeated Start, P= Stop, A= Acknowledge, NA= Non-acknowledge,

EVx= Event (with interrupt if ITEVFEN=1)
EV5: SB=1, cleared by reading SR1 register followed by writing DR register.
EV6: ADDR=1, cleared by reading SR1 register followed by reading SR2. In 10-bit master receiver mode, this se-
quence should be followed by writing CR2 with START = 1.
In case of the reception of 1 byte, the Acknowledge disable must be performed during EV6 event, i.e. before clearing ADDR flag.
EV7: RxNE=1 cleared by reading DR register.
EV7_1: RxNE=1 cleared by reading DR register, program AC K=0 and STOP request
EV9: ADD10=1, cleared by reading SR1 register followed by writing DR register.

ai17540d

1. If a single byte is received, it is NA.

2. The EV5, EV6 and EV9 events stretch SCL low until the end of the corresponding software sequence.
3. The EV7 event stretches SCL low if the software sequence is not completed before the end of the next byte reception.
4. The EV7_1 software sequence must be completed before the ACK pulse of the current byte transfer.
The procedures described below are recommended if the EV7-1 software sequence is not
completed before the ACK pulse of the current byte transfer.
These procedures must be followed to make sure:
• The ACK bit is set low on time before the end of the last data reception
• The STOP bit is set high after the last data reception without reception of
supplementary data.
For 2-byte reception:
• Wait until ADDR = 1 (SCL stretched low until the ADDR flag is cleared)
• Set ACK low, set POS high
• Clear ADDR flag
• Wait until BTF = 1 (Data 1 in DR, Data2 in shift register, SCL stretched low until a data
1 is read)
• Set STOP high
• Read data 1 and 2

734/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

For N >2 -byte reception, from N-2 data reception

• Wait until BTF = 1 (data N-2 in DR, data N-1 in shift register, SCL stretched low until
data N-2 is read)
• Set ACK low
• Read data N-2
• Wait until BTF = 1 (data N-1 in DR, data N in shift register, SCL stretched low until a
data N-1 is read)
• Set STOP high
• Read data N-1 and N

24.3.4 Error conditions

The following are the error conditions which may cause communication to fail.

Bus error (BERR)

This error occurs when the I2C interface detects an external Stop or Start condition during
an address or a data transfer. In this case:
• the BERR bit is set and an interrupt is generated if the ITERREN bit is set
• in Slave mode: data are discarded and the lines are released by hardware:
– in case of a misplaced Start, the slave considers it is a restart and waits for an
address, or a Stop condition
– in case of a misplaced Stop, the slave behaves like for a Stop condition and the
lines are released by hardware
• In Master mode: the lines are not released and the state of the current transmission is
not affected. It is up to the software to abort or not the current transmission

Acknowledge failure (AF)

This error occurs when the interface detects a nonacknowledge bit. In this case:
• the AF bit is set and an interrupt is generated if the ITERREN bit is set
• a transmitter which receives a NACK must reset the communication:
– If Slave: lines are released by hardware
– If Master: a Stop or repeated Start condition must be generated by software

Arbitration lost (ARLO)

This error occurs when the I2C interface detects an arbitration lost condition. In this case,
• the ARLO bit is set by hardware (and an interrupt is generated if the ITERREN bit is
set)
• the I2C Interface goes automatically back to slave mode (the MSL bit is cleared). When
the I2C loses the arbitration, it is not able to acknowledge its slave address in the same
transfer, but it can acknowledge it after a repeated Start from the winning master.
• lines are released by hardware

RM0402 Rev 6 735/1163

757
Inter-integrated circuit (I2C) interface RM0402

Overrun/underrun error (OVR)

An overrun error can occur in slave mode when clock stretching is disabled and the I2C
interface is receiving data. The interface has received a byte (RxNE=1) and the data in DR
has not been read, before the next byte is received by the interface. In this case,
• The last received byte is lost.
• In case of Overrun error, software should clear the RxNE bit and the transmitter should
re-transmit the last received byte.
Underrun error can occur in slave mode when clock stretching is disabled and the I2C
interface is transmitting data. The interface has not updated the DR with the next byte
(TxE=1), before the clock comes for the next byte. In this case,
• The same byte in the DR register is sent again
• The user should make sure that data received on the receiver side during an underrun
error are discarded and that the next bytes are written within the clock low time
specified in the I2C bus standard.
For the first byte to be transmitted, the DR must be written after ADDR is cleared and before
the first SCL rising edge. If not possible, the receiver must discard the first data.

24.3.5 Programmable noise filter

In Fm mode, the I2C standard requires that spikes are suppressed to a length of 50 ns on
SDA and SCL lines.
An analog noise filter is implemented in the SDA and SCL I/Os. This filter is enabled by
default and can be disabled by setting the ANOFF bit in the I2C_FLTR register.
A digital noise filter can be enabled by configuring the DNF[3:0] bits to a non-zero value.
This suppresses the spikes on SDA and SCL inputs with a length of up to DNF[3:0] *
TPCLK1.
Enabling the digital noise filter increases the SDA hold time by (DNF[3:0] +1)* TPCLK.
To be compliant with the maximum hold time of the I2C-bus specification version 2.1
(Thd:dat), the DNF bits must be programmed using the constraints shown in Table 132, and
assuming that the analog filter is disabled.
Note: DNF[3:0] must only be configured when the I2C is disabled (PE = 0). If the analog filter is
also enabled, the digital filter is added to the analog filter.

Table 132. Maximum DNF[3:0] value to be compliant with Thd:dat(max)

Maximum DNF value
PCLK1 frequency
Sm mode Fm mode

2 <= FPCLK1 <= 5 2 0

5 < FPCLK1 <= 10 12 0
10 < FPCLK1 <= 20 15 1
20 < FPCLK1 <= 30 15 7
30 < FPCLK1 <= 40 15 13
40 < FPCLK1 <= 50 15 15

736/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Note: For each frequency range, the constraint is given based on the worst case which is the
minimum frequency of the range. Greater DNF values can be used if the system can
support maximum hold time violation.

24.3.6 SDA/SCL line control

• If clock stretching is enabled:
– Transmitter mode: If TxE=1 and BTF=1: the interface holds the clock line low
before transmission to wait for the microcontroller to write the byte in the Data
Register (both buffer and shift register are empty).
– Receiver mode: If RxNE=1 and BTF=1: the interface holds the clock line low after
reception to wait for the microcontroller to read the byte in the Data Register (both
buffer and shift register are full).
• If clock stretching is disabled in Slave mode:
– Overrun Error in case of RxNE=1 and no read of DR has been done before the
next byte is received. The last received byte is lost.
– Underrun Error in case TxE=1 and no write into DR has been done before the next
byte must be transmitted. The same byte is sent again.
– Write Collision not managed.

24.3.7 SMBus
Introduction
The System Management Bus (SMBus) is a two-wire interface through which various
devices can communicate with each other and with the rest of the system. It is based on I2C
principles of operation. SMBus provides a control bus for system and power management
related tasks. A system may use SMBus to pass messages to and from devices instead of
toggling individual control lines.
The System Management Bus Specification refers to three types of devices. A slave is a
device that is receiving or responding to a command. A master is a device that issues
commands, generates the clocks, and terminates the transfer. A host is a specialized
master that provides the main interface to the system's CPU. A host must be a master-slave
and must support the SMBus host notify protocol. Only one host is allowed in a system.

Similarities between SMBus and I2C

• 2 wire bus protocol (1 Clk, 1 Data) + SMBus Alert line optional
• Master-slave communication, Master provides clock
• Multi master capability
• SMBus data format similar to I2C 7-bit addressing format (Figure 238).

Differences between SMBus and I2C

The following table describes the differences between SMBus and I2C.

RM0402 Rev 6 737/1163

757
Inter-integrated circuit (I2C) interface RM0402

Table 133. SMBus vs. I2C

SMBus I2C

Max. speed 100 kHz Max. speed 400 kHz

Min. clock speed 10 kHz No minimum clock speed
35 ms clock low timeout No timeout
Logic levels are fixed Logic levels are VDD dependent
Different address types (reserved, dynamic etc.) 7-bit, 10-bit and general call slave address types
Different bus protocols (quick command, process
No bus protocols
call etc.)

SMBus application usage

With System Management Bus, a device can provide manufacturer information, tell the
system what its model/part number is, save its state for a suspend event, report different
types of errors, accept control parameters, and return its status. SMBus provides a control
bus for system and power management related tasks.

Device identification
Any device that exists on the System Management Bus as a slave has a unique address
called the Slave Address. For the list of reserved slave addresses, refer to the SMBus
specification version. 2.0 (http://smbus.org/).

Bus protocols
The SMBus specification supports up to 9 bus protocols. For more details of these protocols
and SMBus address types, refer to SMBus specification version. 2.0. These protocols
should be implemented by the user software.

Address resolution protocol (ARP)

SMBus slave address conflicts can be resolved by dynamically assigning a new unique
address to each slave device. The Address Resolution Protocol (ARP) has the following
attributes:
• Address assignment uses the standard SMBus physical layer arbitration mechanism
• Assigned addresses remain constant while device power is applied; address retention
through device power loss is also allowed
• No additional SMBus packet overhead is incurred after address assignment. (i.e.
subsequent accesses to assigned slave addresses have the same overhead as
accesses to fixed address devices.)
• Any SMBus master can enumerate the bus

Unique device identifier (UDID)

In order to provide a mechanism to isolate each device for the purpose of address
assignment, each device must implement a unique device identifier (UDID).
For the details on 128 bit UDID and more information on ARP, refer to SMBus specification
version 2.0.

738/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

SMBus alert mode

SMBus Alert is an optional signal with an interrupt line for devices that want to trade their
ability to master for a pin. SMBA is a wired-AND signal just as the SCL and SDA signals are.
SMBA is used in conjunction with the SMBus General Call Address. Messages invoked with
the SMBus are 2 bytes long.
A slave-only device can signal the host through SMBA that it wants to talk by setting ALERT
bit in I2C_CR1 register. The host processes the interrupt and simultaneously accesses all
SMBA devices through the Alert Response Address (known as ARA having a value 0001
100X). Only the device(s) which pulled SMBA low acknowledges the alert Response
address. This status is identified using SMBALERT Status flag in I2C_SR1 register. The
host performs a modified Receive Byte operation. The 7 bit device address provided by the
slave transmit device is placed in the 7 most significant bits of the byte. The eighth bit can
be a zero or one.
If more than one device pulls SMBA low, the highest priority (lowest address) device wins
communication rights via standard arbitration during the slave address transfer. After
acknowledging the slave address the device must disengage its SMBA pull-down. If the
host still sees SMBA low when the message transfer is complete, it knows to read the ARA
again.
A host which does not implement the SMBA signal may periodically access the ARA.
For more details on SMBus Alert mode, refer to SMBus specification version 2.0
(http://smbus.org/).

Timeout error
There are differences in the timing specifications between I2C and SMBus.
SMBus defines a clock low timeout, TIMEOUT of 35 ms. Also SMBus specifies TLOW:
SEXT as the cumulative clock low extend time for a slave device. SMBus specifies TLOW:
MEXT as the cumulative clock low extend time for a master device. For more details on
these timeouts, refer to SMBus specification version 2.0.
The status flag Timeout or Tlow Error in I2C_SR1 shows the status of this feature.

How to use the interface in SMBus mode

To switch from I2C mode to SMBus mode, the following sequence should be performed.
• Set the SMBus bit in the I2C_CR1 register
• Configure the SMBTYPE and ENARP bits in the I2C_CR1 register as required for the
application
If you want to configure the device as a master, follow the Start condition generation
procedure in Section 24.3.3: I2C master mode. Otherwise, follow the sequence in
Section 24.3.2: I2C slave mode.
The application has to control the various SMBus protocols by software.
• SMB Device Default Address acknowledged if ENARP=1 and SMBTYPE=0
• SMB Host Header acknowledged if ENARP=1 and SMBTYPE=1
• SMB Alert Response Address acknowledged if SMBALERT=1

RM0402 Rev 6 739/1163

757
Inter-integrated circuit (I2C) interface RM0402

24.3.8 DMA requests

DMA requests (when enabled) are generated only for data transfer. DMA requests are
generated by Data Register becoming empty in transmission and Data Register becoming
full in reception. The DMA must be initialized and enabled before the I2C data transfer. The
DMAEN bit must be set in the I2C_CR2 register before the ADDR event. In master mode or
in slave mode when clock stretching is enabled, the DMAEN bit can also be set during the
ADDR event, before clearing the ADDR flag. The DMA request must be served before the
end of the current byte transfer. When the number of data transfers which has been
programmed for the corresponding DMA stream is reached, the DMA controller sends an
End of Transfer EOT signal to the I2C interface and generates a Transfer Complete interrupt
if enabled:
• Master transmitter: In the interrupt routine after the EOT interrupt, disable DMA
requests then wait for a BTF event before programming the Stop condition.
• Master receiver
– When the number of bytes to be received is equal to or greater than two, the DMA
controller sends a hardware signal, EOT_1, corresponding to the last but one data
byte (number_of_bytes – 1). If, in the I2C_CR2 register, the LAST bit is set, I2C
automatically sends a NACK after the next byte following EOT_1. The user can
generate a Stop condition in the DMA Transfer Complete interrupt routine if
enabled.
– When a single byte must be received: the NACK must be programmed during EV6
event, i.e. program ACK=0 when ADDR=1, before clearing ADDR flag. Then the
user can program the STOP condition either after clearing ADDR flag, or in the
DMA Transfer Complete interrupt routine.

Transmission using DMA

DMA mode can be enabled for transmission by setting the DMAEN bit in the I2C_CR2
register. Data are loaded from a Memory area configured using the DMA peripheral (refer to
the DMA specification) to the I2C_DR register whenever the TxE bit is set. To map a DMA
stream x for I2C transmission (where x is the stream number), perform the following
sequence:
1. Set the I2C_DR register address in the DMA_SxPAR register. The data are moved to
this address from the memory after each TxE event.
2. Set the memory address in the DMA_SxMA0R register (and in DMA_SxMA1R register
in the case of a bouble buffer mode). The data are loaded into I2C_DR from this
memory after each TxE event.
3. Configure the total number of bytes to be transferred in the DMA_SxNDTR register.
After each TxE event, this value is decremented.
4. Configure the DMA stream priority using the PL[0:1] bits in the DMA_SxCR register
5. Set the DIR bit in the DMA_SxCR register and configure interrupts after half transfer or
full transfer depending on application requirements.
6. Activate the stream by setting the EN bit in the DMA_SxCR register.
When the number of data transfers which has been programmed in the DMA Controller
registers is reached, the DMA controller sends an End of Transfer EOT/ EOT_1 signal to the
I2C interface and the DMA generates an interrupt, if enabled, on the DMA stream interrupt
vector.
Note: Do not enable the ITBUFEN bit in the I2C_CR2 register if DMA is used for transmission.

740/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Reception using DMA

DMA mode can be enabled for reception by setting the DMAEN bit in the I2C_CR2 register.
Data are loaded from the I2C_DR register to a Memory area configured using the DMA
peripheral (refer to the DMA specification) whenever a data byte is received. To map a DMA
stream x for I2C reception (where x is the stream number), perform the following sequence:
1. Set the I2C_DR register address in DMA_SxPAR register. The data are moved from
this address to the memory after each RxNE event.
2. Set the memory address in the DMA_SxMA0R register (and in DMA_SxMA1R register
in the case of a bouble buffer mode). The data are loaded from the I2C_DR register to
this memory area after each RxNE event.
3. Configure the total number of bytes to be transferred in the DMA_SxNDTR register.
After each RxNE event, this value is decremented.
4. Configure the stream priority using the PL[0:1] bits in the DMA_SxCR register
5. Reset the DIR bit and configure interrupts in the DMA_SxCR register after half transfer
or full transfer depending on application requirements.
6. Activate the stream by setting the EN bit in the DMA_SxCR register.
When the number of data transfers which has been programmed in the DMA Controller
registers is reached, the DMA controller sends an End of Transfer EOT/ EOT_1 signal to the
I2C interface and DMA generates an interrupt, if enabled, on the DMA stream interrupt
vector.
Note: Do not enable the ITBUFEN bit in the I2C_CR2 register if DMA is used for reception.

24.3.9 Packet error checking

A PEC calculator has been implemented to improve the reliability of communication. The
PEC is calculated by using the C(x) = x8 + x2 + x + 1 CRC-8 polynomial serially on each bit.
• PEC calculation is enabled by setting the ENPEC bit in the I2C_CR1 register. PEC is a
CRC-8 calculated on all message bytes including addresses and R/W bits.
– In transmission: set the PEC transfer bit in the I2C_CR1 register after the TxE
event corresponding to the last byte. The PEC is transferred after the last
transmitted byte.
– In reception: set the PEC bit in the I2C_CR1 register after the RxNE event
corresponding to the last byte so that the receiver sends a NACK if the next
received byte is not equal to the internally calculated PEC. In case of Master-
Receiver, a NACK must follow the PEC whatever the check result. The PEC must

RM0402 Rev 6 741/1163

757
Inter-integrated circuit (I2C) interface RM0402

be set before the ACK of the CRC reception in slave mode. It must be set when
the ACK is set low in master mode.
• A PECERR error flag/interrupt is also available in the I2C_SR1 register.
• If DMA and PEC calculation are both enabled:-
– In transmission: when the I2C interface receives an EOT signal from the DMA
controller, it automatically sends a PEC after the last byte.
– In reception: when the I2C interface receives an EOT_1 signal from the DMA
controller, it automatically considers the next byte as a PEC and checks it. A DMA
request is generated after PEC reception.
• To allow intermediate PEC transfers, a control bit is available in the I2C_CR2 register
(LAST bit) to determine if it is really the last DMA transfer or not. If it is the last DMA
request for a master receiver, a NACK is automatically sent after the last received byte.
• PEC calculation is corrupted by an arbitration loss.

24.4 I2C interrupts

The table below gives the list of I2C interrupt requests.

Table 134. I2C Interrupt requests

Interrupt event Event flag Enable control bit

Start bit sent (Master) SB

Address sent (Master) or Address matched (Slave) ADDR
10-bit header sent (Master) ADD10 ITEVFEN
Stop received (Slave) STOPF
Data byte transfer finished BTF
Receive buffer not empty RxNE
ITEVFEN and ITBUFEN
Transmit buffer empty TxE
Bus error BERR
Arbitration loss (Master) ARLO
Acknowledge failure AF
Overrun/Underrun OVR ITERREN
PEC error PECERR
Timeout/Tlow error TIMEOUT
SMBus Alert SMBALERT

Note: SB, ADDR, ADD10, STOPF, BTF, RxNE and TxE are logically ORed on the same interrupt
channel.
BERR, ARLO, AF, OVR, PECERR, TIMEOUT and SMBALERT are logically ORed on the
same interrupt channel.

742/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Figure 244. I2C interrupt mapping diagram

ITEVFEN
SB
ADDR
ADD10

STOPF
it_event
BTF

TxE

ITBUFEN

RxNE

ITERREN

BERR
ARLO
it_error
AF
OVR
PECERR
TIMEOUT
SMBALERT

RM0402 Rev 6 743/1163

757
Inter-integrated circuit (I2C) interface RM0402

24.5 I2C debug mode

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core halted), the
SMBUS timeout either continues to work normally or stops, depending on the
DBG_I2Cx_SMBUS_TIMEOUT configuration bits in the DBG module. For more details,
refer to Section 30.16.2: Debug support for timers, watchdog, bxCAN and I2C.

24.6 I2C registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by half-words (16 bits) or words (32 bits).

24.6.1 I2C control register 1 (I2C_CR1)

Address offset: 0x00
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
NO
SW SMB SM
Res. ALERT PEC POS ACK STOP START STRET ENGC ENPEC ENARP Res. PE
RST TYPE BUS
CH
rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 SWRST: Software reset

When set, the I2C is under reset state. Before resetting this bit, make sure the I2C lines are
released and the bus is free.
0: I2C Peripheral not under reset
1: I2C Peripheral under reset state
Note: This bit can be used to reinitialize the peripheral after an error or a locked state. As an
example, if the BUSY bit is set and remains locked due to a glitch on the bus, the
SWRST bit can be used to exit from this state.
Bit 14 Reserved, must be kept at reset value
Bit 13 ALERT: SMBus alert
This bit is set and cleared by software, and cleared by hardware when PE=0.
0: Releases SMBA pin high. Alert Response Address Header followed by NACK.
1: Drives SMBA pin low. Alert Response Address Header followed by ACK.
Bit 12 PEC: Packet error checking
This bit is set and cleared by software, and cleared by hardware when PEC is transferred or
by a START or Stop condition or when PE=0.
0: No PEC transfer
1: PEC transfer (in Tx or Rx mode)
Note: PEC calculation is corrupted by an arbitration loss.

744/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Bit 11 POS: Acknowledge/PEC Position (for data reception)

This bit is set and cleared by software and cleared by hardware when PE=0.
0: ACK bit controls the (N)ACK of the current byte being received in the shift register. The
PEC bit indicates that current byte in shift register is a PEC.
1: ACK bit controls the (N)ACK of the next byte which is received in the shift register. The
PEC bit indicates that the next byte in the shift register is a PEC
Note: The POS bit must be used only in 2-byte reception configuration in master mode. It
must be configured before data reception starts, as described in the 2-byte reception
procedure recommended in Master receiver.
Bit 10 ACK: Acknowledge enable
This bit is set and cleared by software and cleared by hardware when PE=0.
0: No acknowledge returned
1: Acknowledge returned after a byte is received (matched address or data)
Bit 9 STOP: Stop generation
The bit is set and cleared by software, cleared by hardware when a Stop condition is
detected, set by hardware when a timeout error is detected.
In Master Mode:
0: No Stop generation.
1: Stop generation after the current byte transfer or after the current Start condition is sent.
In Slave mode:
0: No Stop generation.
1: Release the SCL and SDA lines after the current byte transfer.
Bit 8 START: Start generation
This bit is set and cleared by software and cleared by hardware when start is sent or PE=0.
In Master Mode:
0: No Start generation
1: Repeated start generation
In Slave mode:
0: No Start generation
1: Start generation when the bus is free
Bit 7 NOSTRETCH: Clock stretching disable (Slave mode)
This bit is used to disable clock stretching in slave mode when ADDR or BTF flag is set, until
it is reset by software.
0: Clock stretching enabled
1: Clock stretching disabled
Bit 6 ENGC: General call enable
0: General call disabled. Address 00h is NACKed.
1: General call enabled. Address 00h is ACKed.
Bit 5 ENPEC: PEC enable
0: PEC calculation disabled
1: PEC calculation enabled
Bit 4 ENARP: ARP enable
0: ARP disable
1: ARP enable
SMBus Device default address recognized if SMBTYPE=0
SMBus Host address recognized if SMBTYPE=1
Bit 3 SMBTYPE: SMBus type
0: SMBus Device
1: SMBus Host

RM0402 Rev 6 745/1163

757
Inter-integrated circuit (I2C) interface RM0402

Bit 2 Reserved, must be kept at reset value

Bit 1 SMBUS: SMBus mode
0: I2C mode
1: SMBus mode
Bit 0 PE: Peripheral enable
0: Peripheral disable
1: Peripheral enable
Note: If this bit is reset while a communication is on going, the peripheral is disabled at the
end of the current communication, when back to IDLE state.
All bit resets due to PE=0 occur at the end of the communication.
In master mode, this bit must not be reset before the end of the communication.

Note: When the STOP, START or PEC bit is set, the software must not perform any write access
to I2C_CR1 before this bit is cleared by hardware. Otherwise there is a risk of setting a
second STOP, START or PEC request.

24.6.2 I2C control register 2 (I2C_CR2)

Address offset: 0x04
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DMA ITBUF ITEVT ITERR
Res. Res. Res. LAST Res. Res. FREQ[5:0]
EN EN EN EN
rw rw rw rw rw rw rw rw rw rw rw

Bits 15:13 Reserved, must be kept at reset value

Bit 12 LAST: DMA last transfer
0: Next DMA EOT is not the last transfer
1: Next DMA EOT is the last transfer
Note: This bit is used in master receiver mode to permit the generation of a NACK on the last
received data.
Bit 11 DMAEN: DMA requests enable
0: DMA requests disabled
1: DMA request enabled when TxE=1 or RxNE =1
Bit 10 ITBUFEN: Buffer interrupt enable
0: TxE = 1 or RxNE = 1 does not generate any interrupt.
1: TxE = 1 or RxNE = 1 generates Event Interrupt (whatever the state of DMAEN)
Bit 9 ITEVTEN: Event interrupt enable
0: Event interrupt disabled
1: Event interrupt enabled
This interrupt is generated when:
– SB = 1 (Master)
– ADDR = 1 (Master/Slave)
– ADD10= 1 (Master)
– STOPF = 1 (Slave)
– BTF = 1 with no TxE or RxNE event
– TxE event to 1 if ITBUFEN = 1
– RxNE event to 1if ITBUFEN = 1

746/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

ITERREN: Error interrupt enable

0: Error interrupt disabled
1: Error interrupt enabled
This interrupt is generated when:
– BERR = 1
– ARLO = 1
– AF = 1
– OVR = 1
– PECERR = 1
– TIMEOUT = 1
– SMBALERT = 1
Bits 7:6 Reserved, must be kept at reset value
Bits 5:0 FREQ[5:0]: Peripheral clock frequency
The FREQ bits must be configured with the APB clock frequency value (I2C peripheral
connected to APB). The FREQ field is used by the peripheral to generate data setup and
hold times compliant with the I2C specifications. The minimum allowed frequency is 2 MHz,
the maximum frequency is limited by the maximum APB frequency and cannot exceed
50 MHz (peripheral intrinsic maximum limit).
0b000000: Not allowed
0b000001: Not allowed
0b000010: 2 MHz
...
0b110010: 50 MHz
Higher than 0b101010: Not allowed

RM0402 Rev 6 747/1163

757
Inter-integrated circuit (I2C) interface RM0402

24.6.3 I2C own address register 1 (I2C_OAR1)

Address offset: 0x08
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ADD
Res. Res. Res. Res. Res. ADD[9:8] ADD[7:1] ADD0
MODE
rw rw rw rw rw rw rw rw rw rw rw

Bit 15 ADDMODE Addressing mode (slave mode)

0: 7-bit slave address (10-bit address not acknowledged)
1: 10-bit slave address (7-bit address not acknowledged)
Bit 14 Should always be kept at 1 by software.
Bits 13:10 Reserved, must be kept at reset value
Bits 9:8 ADD[9:8]: Interface address
7-bit addressing mode: don’t care
10-bit addressing mode: bits9:8 of address
Bits 7:1 ADD[7:1]: Interface address
bits 7:1 of address
Bit 0 ADD0: Interface address
7-bit addressing mode: don’t care
10-bit addressing mode: bit 0 of address

24.6.4 I2C own address register 2 (I2C_OAR2)

Address offset: 0x0C
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EN
Res. Res. Res. Res. Res. Res. Res. Res. ADD2[7:1]
DUAL
rw rw rw rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value

Bits 7:1 ADD2[7:1]: Interface address
bits 7:1 of address in dual addressing mode
Bit 0 ENDUAL: Dual addressing mode enable
0: Only OAR1 is recognized in 7-bit addressing mode
1: Both OAR1 and OAR2 are recognized in 7-bit addressing mode

748/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

24.6.5 I2C data register (I2C_DR)

Address offset: 0x10
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. DR[7:0]
rw rw rw rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value

Bits 7:0 DR[7:0] 8-bit data register
Byte received or to be transmitted to the bus.
– Transmitter mode: Byte transmission starts automatically when a byte is written in the DR
register. A continuous transmit stream can be maintained if the next data to be transmitted is
put in DR once the transmission is started (TxE=1)
– Receiver mode: Received byte is copied into DR (RxNE=1). A continuous transmit stream
can be maintained if DR is read before the next data byte is received (RxNE=1).
Note: In slave mode, the address is not copied into DR.
Write collision is not managed (DR can be written if TxE=0).
If an ARLO event occurs on ACK pulse, the received byte is not copied into DR
and so cannot be read.

24.6.6 I2C status register 1 (I2C_SR1)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMB TIMEO PEC
Res. OVR AF ARLO BERR TxE RxNE Res. STOPF ADD10 BTF ADDR SB
ALERT UT ERR
rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 rc_w0 r r r r r r r

RM0402 Rev 6 749/1163

757
Inter-integrated circuit (I2C) interface RM0402

Bit 15 SMBALERT: SMBus alert

In SMBus host mode:
0: no SMBALERT
1: SMBALERT event occurred on pin
In SMBus slave mode:
0: no SMBALERT response address header
1: SMBALERT response address header to SMBALERT LOW received
– Cleared by software writing 0, or by hardware when PE=0.
Bit 14 TIMEOUT: Timeout or Tlow error
0: No timeout error
1: SCL remained LOW for 25 ms (Timeout)
or
Master cumulative clock low extend time more than 10 ms (Tlow:mext)
or
Slave cumulative clock low extend time more than 25 ms (Tlow:sext)
– When set in slave mode: slave resets the communication and lines are released by
hardware
– When set in master mode: Stop condition sent by hardware
– Cleared by software writing 0, or by hardware when PE=0.
Note: This functionality is available only in SMBus mode.
Bit 13 Reserved, must be kept at reset value
Bit 12 PECERR: PEC Error in reception
0: no PEC error: receiver returns ACK after PEC reception (if ACK=1)
1: PEC error: receiver returns NACK after PEC reception (whatever ACK)
– Cleared by software writing 0, or by hardware when PE=0.
– Note: When the received CRC is wrong, PECERR is not set in slave mode if the PEC control
bit is not set before the end of the CRC reception. Nevertheless, reading the PEC value
determines whether the received CRC is right or wrong.
Bit 11 OVR: Overrun/Underrun
0: No overrun/underrun
1: Overrun or underrun
– Set by hardware in slave mode when NOSTRETCH=1 and:
– In reception when a new byte is received (including ACK pulse) and the DR register has not
been read yet. New received byte is lost.
– In transmission when a new byte should be sent and the DR register has not been written
yet. The same byte is sent twice.
– Cleared by software writing 0, or by hardware when PE=0.
Note: If the DR write occurs very close to SCL rising edge, the sent data is unspecified and a
hold timing error occurs
Bit 10 AF: Acknowledge failure
0: No acknowledge failure
1: Acknowledge failure
– Set by hardware when no acknowledge is returned.
– Cleared by software writing 0, or by hardware when PE=0.

750/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Bit 9 ARLO: Arbitration lost (master mode)

0: No Arbitration Lost detected
1: Arbitration Lost detected
Set by hardware when the interface loses the arbitration of the bus to another master
– Cleared by software writing 0, or by hardware when PE=0.
After an ARLO event the interface switches back automatically to Slave mode (MSL=0).
Note: In SMBUS, the arbitration on the data in slave mode occurs only during the data phase,
or the acknowledge transmission (not on the address acknowledge).
Bit 8 BERR: Bus error
0: No misplaced Start or Stop condition
1: Misplaced Start or Stop condition
– Set by hardware when the interface detects an SDA rising or falling edge while SCL is high,
occurring in a non-valid position during a byte transfer.
– Cleared by software writing 0, or by hardware when PE=0.
Bit 7 TxE: Data register empty (transmitters)
0: Data register not empty
1: Data register empty
– Set when DR is empty in transmission. TxE is not set during address phase.
– Cleared by software writing to the DR register or by hardware after a start or a stop condition
or when PE=0.
TxE is not set if either a NACK is received, or if next byte to be transmitted is PEC (PEC=1)
Note: TxE is not cleared by writing the first data being transmitted, or by writing data when
BTF is set, as in both cases the data register is still empty.
Bit 6 RxNE: Data register not empty (receivers)
0: Data register empty
1: Data register not empty
– Set when data register is not empty in receiver mode. RxNE is not set during address phase.
– Cleared by software reading or writing the DR register or by hardware when PE=0.
RxNE is not set in case of ARLO event.
Note: RxNE is not cleared by reading data when BTF is set, as the data register is still full.
Bit 5 Reserved, must be kept at reset value
Bit 4 STOPF: Stop detection (slave mode)
0: No Stop condition detected
1: Stop condition detected
– Set by hardware when a Stop condition is detected on the bus by the slave after an
acknowledge (if ACK=1).
– Cleared by software reading the SR1 register followed by a write in the CR1 register, or by
hardware when PE=0
Note: The STOPF bit is not set after a NACK reception.
It is recommended to perform the complete clearing sequence (READ SR1 then
WRITE CR1) after the STOPF is set. Refer to Figure 241: Transfer sequence diagram
for slave receiver on page 729.

RM0402 Rev 6 751/1163

757
Inter-integrated circuit (I2C) interface RM0402

Bit 3 ADD10: 10-bit header sent (Master mode)

0: No ADD10 event occurred.
1: Master has sent first address byte (header).
– Set by hardware when the master has sent the first byte in 10-bit address mode.
– Cleared by software reading the SR1 register followed by a write in the DR register of the
second address byte, or by hardware when PE=0.
Note: ADD10 bit is not set after a NACK reception
Bit 2 BTF: Byte transfer finished
0: Data byte transfer not done
1: Data byte transfer succeeded
– Set by hardware when NOSTRETCH=0 and:
– In reception when a new byte is received (including ACK pulse) and DR has not been read
yet (RxNE=1).
– In transmission when a new byte should be sent and DR has not been written yet (TxE=1).
– Cleared by software by either a read or write in the DR register or by hardware after a start or
a stop condition in transmission or when PE=0.
Note: The BTF bit is not set after a NACK reception
The BTF bit is not set if next byte to be transmitted is the PEC (TRA=1 in I2C_SR2
register and PEC=1 in I2C_CR1 register)
Bit 1 ADDR: Address sent (master mode)/matched (slave mode)
This bit is cleared by software reading SR1 register followed reading SR2, or by hardware
when PE=0.
Address matched (Slave)
0: Address mismatched or not received.
1: Received address matched.
– Set by hardware as soon as the received slave address matched with the OAR registers
content or a general call or a SMBus Device Default Address or SMBus Host or SMBus Alert
is recognized. (when enabled depending on configuration).
Note: In slave mode, it is recommended to perform the complete clearing sequence (READ
SR1 then READ SR2) after ADDR is set. Refer to Figure 241: Transfer sequence
diagram for slave receiver on page 729.
Address sent (Master)
0: No end of address transmission
1: End of address transmission
– For 10-bit addressing, the bit is set after the ACK of the 2nd byte.
– For 7-bit addressing, the bit is set after the ACK of the byte.
Note: ADDR is not set after a NACK reception
Bit 0 SB: Start bit (Master mode)
0: No Start condition
1: Start condition generated.
– Set when a Start condition generated.
– Cleared by software by reading the SR1 register followed by writing the DR register, or by
hardware when PE=0

752/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

24.6.7 I2C status register 2 (I2C_SR2)

Address offset: 0x18
Reset value: 0x0000
Note: Reading I2C_SR2 after reading I2C_SR1 clears the ADDR flag, even if the ADDR flag was
set after reading I2C_SR1. Consequently, I2C_SR2 must be read only when ADDR is found
set in I2C_SR1 or when the STOPF bit is cleared.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SMB
SMB GEN
PEC[7:0] DUALF DEFAU Res. TRA BUSY MSL
HOST CALL
LT
r r r r r r r r r r r r r r r

Bits 15:8 PEC[7:0] Packet error checking register

This register contains the internal PEC when ENPEC=1.
Bit 7 DUALF: Dual flag (Slave mode)
0: Received address matched with OAR1
1: Received address matched with OAR2
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 6 SMBHOST: SMBus host header (Slave mode)
0: No SMBus Host address
1: SMBus Host address received when SMBTYPE=1 and ENARP=1.
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 5 SMBDEFAULT: SMBus device default address (Slave mode)
0: No SMBus Device Default address
1: SMBus Device Default address received when ENARP=1
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 4 GENCALL: General call address (Slave mode)
0: No General Call
1: General Call Address received when ENGC=1
– Cleared by hardware after a Stop condition or repeated Start condition, or when PE=0.
Bit 3 Reserved, must be kept at reset value

RM0402 Rev 6 753/1163

757
Inter-integrated circuit (I2C) interface RM0402

Bit 2 TRA: Transmitter/receiver

0: Data bytes received
1: Data bytes transmitted
This bit is set depending on the R/W bit of the address byte, at the end of total address
phase.
It is also cleared by hardware after detection of Stop condition (STOPF=1), repeated Start
condition, loss of bus arbitration (ARLO=1), or when PE=0.
Bit 1 BUSY: Bus busy
0: No communication on the bus
1: Communication ongoing on the bus
– Set by hardware on detection of SDA or SCL low
– cleared by hardware on detection of a Stop condition.
It indicates a communication in progress on the bus. This information is still updated when
the interface is disabled (PE=0).
Bit 0 MSL: Master/slave
0: Slave Mode
1: Master Mode
– Set by hardware as soon as the interface is in Master mode (SB=1).
– Cleared by hardware after detecting a Stop condition on the bus or a loss of arbitration
(ARLO=1), or by hardware when PE=0.

Note: Reading I2C_SR2 after reading I2C_SR1 clears the ADDR flag, even if the ADDR flag was
set after reading I2C_SR1. Consequently, I2C_SR2 must be read only when ADDR is found
set in I2C_SR1 or when the STOPF bit is cleared.

24.6.8 I2C clock control register (I2C_CCR)

Address offset: 0x1C
Reset value: 0x0000
Note: fPCLK1 must be at least 2 MHz to achieve Sm mode I²C frequencies. It must be at least 4
MHz to achieve Fm mode I²C frequencies
The CCR register must be configured only when the I2C is disabled (PE = 0).

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
F/S DUTY Res. Res. CCR[11:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 F/S: I2C master mode selection

0: Sm mode I2C
1: Fm mode I2C

754/1163 RM0402 Rev 6

RM0402 Inter-integrated circuit (I2C) interface

Bit 14 DUTY: Fm mode duty cycle

0: Fm mode tlow/thigh = 2
1: Fm mode tlow/thigh = 16/9 (see CCR)
Note: When the PCLK frequency is a multiple of 10 MHz, the DUTY bit must be set in order to
reach the 400 kHz maximum I2C frequency.
Bits 13:12 Reserved, must be kept at reset value
Bits 11:0 CCR[11:0]: Clock control register in Fm/Sm mode (Master mode)
Controls the SCL clock in master mode.
Sm mode or SMBus:
Thigh = CCR * TPCLK1
Tlow = CCR * TPCLK1
Fm mode:
If DUTY = 0:
Thigh = CCR * TPCLK1
Tlow = 2 * CCR * TPCLK1
If DUTY = 1:
Thigh = 9 * CCR * TPCLK1
Tlow = 16 * CCR * TPCLK1
For instance: in Sm mode, to generate a 100 kHz SCL frequency:
If FREQ = 08, TPCLK1 = 125 ns so CCR must be programmed with 0x28
(0x28 <=> 40d x 125 ns = 5000 ns.)
Note: The minimum allowed value is 0x04, except in FAST DUTY mode where the minimum
allowed value is 0x01
thigh = tr(SCL) + tw(SCLH). See device datasheet for the definitions of parameters.
tlow = tf(SCL) + tw(SCLL). See device datasheet for the definitions of parameters.
I2C communication speed, fSCL ~ 1/(thigh + tlow). The real frequency may differ due to
the analog noise filter input delay.
The CCR register must be configured only when the I2C is disabled (PE = 0).

24.6.9 I2C TRISE register (I2C_TRISE)

Address offset: 0x20
Reset value: 0x0002
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. TRISE[5:0]
rw rw rw rw rw rw

RM0402 Rev 6 755/1163

757
Inter-integrated circuit (I2C) interface RM0402

Bits 15:6 Reserved, must be kept at reset value

Bits 5:0 TRISE[5:0]: Maximum rise time in Fm/Sm mode (Master mode)
These bits should provide the maximum duration of the SCL feedback loop in master mode.
The purpose is to keep a stable SCL frequency whatever the SCL rising edge duration.
These bits must be programmed with the maximum SCL rise time given in the I2C bus
specification, incremented by 1.
For instance: in Sm mode, the maximum allowed SCL rise time is 1000 ns.
If, in the I2C_CR2 register, the value of FREQ[5:0] bits is equal to 0x08 and TPCLK1 = 125 ns
therefore the TRISE[5:0] bits must be programmed with 09h.
(1000 ns / 125 ns = 8 + 1)
The filter value can also be added to TRISE[5:0].
If the result is not an integer, TRISE[5:0] must be programmed with the integer part, in order
to respect the tHIGH parameter.
Note: TRISE[5:0] must be configured only when the I2C is disabled (PE = 0).

24.6.10 I2C FLTR register (I2C_FLTR)

Address offset: 0x24
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. ANOFF DNF[3:0]
rw rw rw rw rw

Bits 15:5 Reserved, must be kept at reset value

Bit 4 ANOFF: Analog noise filter OFF
0: Analog noise filter enable
1: Analog noise filter disable
Note: ANOFF must be configured only when the I2C is disabled (PE = 0).
Bits 3:0 DNF[3:0]: Digital noise filter
These bits are used to configure the digital noise filter on SDA and SCL inputs. The digital filter
suppresses the spikes with a length of up to DNF[3:0] * TPCLK1.
0000: Digital noise filter disable
0001: Digital noise filter enabled and filtering capability up to 1* TPCLK1.
...
1111: Digital noise filter enabled and filtering capability up to 15* TPCLK1.
Note: DNF[3:0] must be configured only when the I2C is disabled (PE = 0). If the analog filter
is also enabled, the digital filter is added to the analog filter.

756/1163 RM0402 Rev 6

 0x24
0x20
0x18
0x14
0x10
0x08
0x04
0x00

0x1C
0x0C
Offset
RM0402

24.6.11

I2C_DR

I2C_SR2
I2C_SR1
I2C_CR2
I2C_CR1

I2C_CCR

I2C_FLTR
I2C_OAR2
I2C_OAR1

I2C_TRISE
Register

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 28
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 27
I2C register map

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 25
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 24
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 23
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 21
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 18

RM0402 Rev 6
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 17
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
0
0

Res. Res. F/S SMBALERT Res. Res. ADDMODE Res. SWRST 15

0
0
0
Res. Res. DUTY TIMEOUT Res. Res. Res. Res. Res. 14

0
0

Res. Res. Res. Res. Res. Res. Res. Res. ALERT 13

0
0
0
0

Res. Res. Res. PECERR Res. Res. Res. LAST PEC 12

0
0
0
0
0
Table 135. I2C register map and reset values

Res. Res. OVR Res. Res. Res. DMAEN POS 11

PEC[7:0]
The table below provides the I2C register map and reset values.

0
0
0
0
0

Res. Res. AF Res. Res. Res. ITBUFEN ACK 10

0
0
0
0
0
0

Res. Res. ARLO Res. Res. ITEVTEN STOP 9

Refer to Section 2.2 on page 49 for the register boundary addresses.

9:8]

0
0
0
0
0
0

ADD[

Res. Res. BERR Res. Res. ITERREN START 8

0
0
0
0

0
0
0

Res. Res. DUALF TxE Res. NOSTRETCH 7

0
0
0
0

0
0
0

Res. Res. SMBHOST RxNE Res. ENGC 6

0
0
0
0
0

0
0
0

Res. SMBDEFAUL Res. ENPEC 5

CCR[11:0]

0
0
0
0
0
0
0

0
0
0

ANOFF GENCALL STOPF ENARP 4

0
0

0
0
0
0
0
0
0

ADD[7:1]

Res.
ADD2[7:1]

ADD10 SMBTYPE 3
DR[7:0]

0
0
0

0
0
0
0
0
0

TRA BTF Res. 2

FREQ[5:0]

TRISE[5:0]
0
0
0

0
1
0
0
0
0
0

BUSY ADDR SMBUS 1

DNF[3:0]
0
0
0
0

0
0
0
0
0
0

MSL SB ENDUAL ADD0 PE 0

Inter-integrated circuit (I2C) interface

757/1163
757
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

25 Universal synchronous receiver transmitter (USART)

/universal asynchronous receiver transmitter (UART)

25.1 USART introduction

The universal synchronous asynchronous receiver transmitter (USART) offers a flexible
means of full-duplex data exchange with external equipment requiring an industry standard
NRZ asynchronous serial data format. The USART offers a very wide range of baud rates
using a fractional baud rate generator.
It supports synchronous one-way communication and half-duplex single wire
communication. It also supports the LIN (local interconnection network), Smartcard Protocol
and IrDA (infrared data association) SIR ENDEC specifications, and modem operations
(CTS/RTS). It allows multiprocessor communication.
High speed data communication is possible by using the DMA for multibuffer configuration.

758/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

25.2 USART main features

• Full duplex, asynchronous communications
• NRZ standard format (Mark/Space)
• Configurable oversampling method by 16 or by 8 to give flexibility between speed and
clock tolerance
• Fractional baud rate generator systems
– Common programmable transmit and receive baud rate (refer to the datasheets
for the value of the baud rate at the maximum APB frequency.
• Programmable data word length (8 or 9 bits)
• Configurable stop bits - support for 1 or 2 stop bits
• LIN Master Synchronous Break send capability and LIN slave break detection
capability
– 13-bit break generation and 10/11 bit break detection when USART is hardware
configured for LIN
• Transmitter clock output for synchronous transmission
• IrDA SIR encoder decoder
– Support for 3/16 bit duration for normal mode
• Smartcard emulation capability
– The Smartcard interface supports the asynchronous protocol Smartcards as
defined in the ISO 7816-3 standards
– 0.5, 1.5 stop bits for Smartcard operation
• Single-wire half-duplex communication
• Configurable multibuffer communication using DMA (direct memory access)
– Buffering of received/transmitted bytes in reserved SRAM using centralized DMA
• Separate enable bits for transmitter and receiver
• Transfer detection flags:
– Receive buffer full
– Transmit buffer empty
– End of transmission flags
• Parity control:
– Transmits parity bit
– Checks parity of received data byte
• Four error detection flags:
– Overrun error
– Noise detection
– Frame error
– Parity error
• Ten interrupt sources with flags:
– CTS changes
– LIN break detection
– Transmit data register empty
– Transmission complete

RM0402 Rev 6 759/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

– Receive data register full

– Idle line received
– Overrun error
– Framing error
– Noise error
– Parity error
• Multiprocessor communication - enter into mute mode if address match does not occur
• Wake up from mute mode (by idle line detection or address mark detection)
• Two receiver wakeup modes: Address bit (MSB, 9th bit), Idle line

25.3 USART implementation

This section describes the full set of features implemented in USART1. Refer to Table 136:
USART features for the differences between USART instances.

Table 136. USART features

USART1, USART2, USART3,
USART modes/features(1)
USART6

Hardware flow control for modem(2) X

Continuous communication using DMA X
Multiprocessor communication X
Synchronous mode(2) X
Smartcard mode X
Single-wire half-duplex communication X
IrDA SIR ENDEC block X
LIN mode X
USART data length 8 or 9 bits
1. X = supported.
2. This feature may be available or not depending on the package. Refer to the product datasheet for more
information.

25.4 USART functional description

The interface is externally connected to another device by three pins (see Figure 245). Any
USART bidirectional communication requires a minimum of two pins: Receive Data In (RX)
and Transmit Data Out (TX):
RX: Receive Data Input is the serial data input. Oversampling techniques are used for data
recovery by discriminating between valid incoming data and noise.
TX: Transmit Data Output. When the transmitter is disabled, the output pin returns to its I/O
port configuration. When the transmitter is enabled and nothing is to be transmitted, the TX
pin is at high level. In single-wire and smartcard modes, this I/O is used to transmit and
receive the data (at USART level, data are then received on SW_RX).

760/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Through these pins, serial data is transmitted and received in normal USART mode as
frames comprising:
• An Idle Line prior to transmission or reception
• A start bit
• A data word (8 or 9 bits) least significant bit first
• 0.5,1, 1.5, 2 Stop bits indicating that the frame is complete
• This interface uses a fractional baud rate generator - with a 12-bit mantissa and 4-bit
fraction
• A status register (USART_SR)
• Data Register (USART_DR)
• A baud rate register (USART_BRR) - 12-bit mantissa and 4-bit fraction.
• A Guardtime Register (USART_GTPR) in case of Smartcard mode.
Refer to Section 25.6: USART registers for the definition of each bit.
The following pin is required to interface in synchronous mode:
• SCLK: Transmitter clock output. This pin outputs the transmitter data clock for
synchronous transmission corresponding to SPI master mode (no clock pulses on start
bit and stop bit, and a software option to send a clock pulse on the last data bit). In
parallel data can be received synchronously on RX. This can be used to control
peripherals that have shift registers (e.g. LCD drivers). The clock phase and polarity
are software programmable. In smartcard mode, SCLK can provide the clock to the
smartcard.
The following pins are required in Hardware flow control mode:
• nCTS: Clear To Send blocks the data transmission at the end of the current transfer
when high
• nRTS: Request to send indicates that the USART is ready to receive a data (when
low).

RM0402 Rev 6 761/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Figure 245. USART block diagram

PWDATA PRDATA
Write Read (Data register) DR

(CPU or DMA) (CPU or DMA)

Transmit data register (TDR) Receive data register (RDR)

TX

RX IrDA
SIR
SW_RX Receive Shift Register
ENDEC Transmit Shift Register
block
IRDA_OUT
IRDA_IN GTPR
GT PSC SCLK control SCLK
CR3 CR2
DMAT DMAR SCEN NACK HD IRLP IREN LINE STOP[1:0] CKEN CPOL CPHA LBCL

CR2 CR1
USART Address UE M WAKE PCE PS PEIE

nRTS Hardware
flow
nCTS controller

Wakeup Receiver
Transmit Receiver clock
control unit
control

CR1 SR
TXEIE TCIE RXNE
IE
IDLE TE RE RWU SBK CTS LBD TXE TC RXNE IDLE ORE NF FE PE
IE

USART
interrupt
control

CR1 USART_BRR
OVER8
TE Transmitter rate
Transmitter control
clock
/ [8 x (2 - OVER8)] /USARTDIV
SAMPLING
DIVIDER DIV_Mantissa DIV_Fraction
15 4 0
fPCLKx(x=1,2)

Receiver rate
RE control

Conventional baud rate generator

USARTDIV = DIV_Mantissa + (DIV_Fraction / 8 × (2 – OVER8)) ai16099

762/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

25.4.1 USART character description

Word length may be selected as being either 8 or 9 bits by programming the M bit in the
USART_CR1 register (see Figure 246).
The TX pin is in low state during the start bit. It is in high state during the stop bit.
An Idle character is interpreted as an entire frame of “1”s followed by the start bit of the
next frame that contains data (The number of “1” ‘s will include the number of stop bits).
A Break character is interpreted on receiving “0”s for a frame period. At the end of the
break frame the transmitter inserts either 1 or 2 stop bits (logic “1” bit) to acknowledge the
start bit.
Transmission and reception are driven by a common baud rate generator, the clock for each
is generated when the enable bit is set respectively for the transmitter and receiver.
The details of each block is given below.

Figure 246.gWord
( length programming
), p
Data frame Next data frame
Possible
parity bit Next
Start Stop Start
Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8
bit bit bit

Clock **

Start
Idle frame bit

Break frame Stop Start

bit bit

8-bit word length (M bit is reset), 1 Stop bit

Data frame Possible Next data frame
parity bit Next
Start Stop Start
Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7
bit bit bit

Clock **

Start
Idle frame bit

Break frame Stop Start

bit bit

** LBCL bit controls last data clock pulse

MS37358V1

RM0402 Rev 6 763/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

25.4.2 Transmitter
The transmitter can send data words of either 8 or 9 bits depending on the M bit status.
When the transmit enable bit (TE) is set, the data in the transmit shift register is output on
the TX pin and the corresponding clock pulses are output on the SCLK pin.

Character transmission
During an USART transmission, data shifts out least significant bit first on the TX pin. In this
mode, the USART_DR register consists of a buffer (TDR) between the internal bus and the
transmit shift register (see Figure 245).
Every character is preceded by a start bit that is a logic level low for one bit period. The
character is terminated by a configurable number of stop bits.
The following stop bits are supported by USART: 0.5, 1, 1.5 and 2 stop bits.
Note: The TE bit should not be reset during transmission of data. Resetting the TE bit during the
transmission will corrupt the data on the TX pin as the baud rate counters will get frozen.
The current data being transmitted will be lost.
An idle frame will be sent after the TE bit is enabled.

Configurable stop bits

The number of stop bits to be transmitted with every character can be programmed in
Control register 2, bits 13,12.
• 1 stop bit: This is the default value of number of stop bits.
• 2 Stop bits: This will be supported by normal USART, single-wire and modem modes.
• 0.5 stop bit: To be used when receiving data in Smartcard mode.
• 1.5 stop bits: To be used when transmitting and receiving data in Smartcard mode.
An idle frame transmission will include the stop bits.
A break transmission will be 10 low bits followed by the configured number of stop bits
(when m = 0) and 11 low bits followed by the configured number of stop bits (when m = 1). It
is not possible to transmit long breaks (break of length greater than 10/11 low bits).

764/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Figure 247. Configurable stop bits

8-bit Word length (M bit is reset)
Possible Next data frame
Data frame parity
bit Next
Start Stop start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 bit
bit
CLOCK ****
**
** LBCL bit controls last data clock pulse

a) 1 Stop Bit
Possible Next data frame
Parity
Data frame
Bit Next
Start start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 bit
1 1/2 stop bits
b) 1 1/2 stop Bits
Possible Next data frame
parity
Data frame
bit Next
Start 2 Stop Start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bits Bit

Possible Next data frame

c) 2 Stop Bits Parity
Data frame
Bit Next
Start start
Bit Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 bit
1/2 stop bit
d) 1/2 Stop Bit

Procedure:
1. Enable the USART by writing the UE bit in USART_CR1 register to 1.
2. Program the M bit in USART_CR1 to define the word length.
3. Program the number of stop bits in USART_CR2.
4. Select DMA enable (DMAT) in USART_CR3 if Multi buffer Communication is to take
place. Configure the DMA register as explained in multibuffer communication.
5. Select the desired baud rate using the USART_BRR register.
6. Set the TE bit in USART_CR1 to send an idle frame as first transmission.
7. Write the data to send in the USART_DR register (this clears the TXE bit). Repeat this
for each data to be transmitted in case of single buffer.
8. After writing the last data into the USART_DR register, wait until TC=1. This indicates
that the transmission of the last frame is complete. This is required for instance when
the USART is disabled or enters the Halt mode to avoid corrupting the last
transmission.

Single byte communication

Clearing the TXE bit is always performed by a write to the data register.
The TXE bit is set by hardware and it indicates:
• The data has been moved from TDR to the shift register and the data transmission has
started.
• The TDR register is empty.
• The next data can be written in the USART_DR register without overwriting the
previous data.
This flag generates an interrupt if the TXEIE bit is set.

RM0402 Rev 6 765/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

When a transmission is taking place, a write instruction to the USART_DR register stores
the data in the TDR register and which is copied in the shift register at the end of the current
transmission.
When no transmission is taking place, a write instruction to the USART_DR register places
the data directly in the shift register, the data transmission starts, and the TXE bit is
immediately set.
If a frame is transmitted (after the stop bit) and the TXE bit is set, the TC bit goes high. An
interrupt is generated if the TCIE bit is set in the USART_CR1 register.
After writing the last data into the USART_DR register, it is mandatory to wait for TC=1
before disabling the USART or causing the microcontroller to enter the low power mode
(see Figure 248: TC/TXE behavior when transmitting).
The TC bit is cleared by the following software sequence:
1. A read from the USART_SR register
2. A write to the USART_DR register
Note: The TC bit can also be cleared by writing a ‘0 to it. This clearing sequence is recommended
only for Multibuffer communication.

Figure 248. TC/TXE behavior when transmitting

Break characters
Setting the SBK bit transmits a break character. The break frame length depends on the M
bit (see Figure 246).
If the SBK bit is set to ‘1 a break character is sent on the TX line after completing the current
character transmission. This bit is reset by hardware when the break character is completed
(during the stop bit of the break character). The USART inserts a logic 1 bit at the end of the
last break frame to guarantee the recognition of the start bit of the next frame.
Note: If the software resets the SBK bit before the commencement of break transmission, the
break character will not be transmitted. For two consecutive breaks, the SBK bit should be
set after the stop bit of the previous break.

Idle characters
Setting the TE bit drives the USART to send an idle frame before the first data frame.

766/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

25.4.3 Receiver
The USART can receive data words of either 8 or 9 bits depending on the M bit in the
USART_CR1 register.

Start bit detection

The start bit detection sequence is the same when oversampling by 16 or by 8.
In the USART, the start bit is detected when a specific sequence of samples is recognized.
This sequence is: 1 1 1 0 X 0 X 0 X 0 0 0 0.

Figure 249. Start bit detection when oversampling by 16 or 8

RX state
Idle Start bit

RX line

Ideal
sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
clock
Sampled values

Real
sample X X X X X X X X 9 10 11 12 13 14 15 16
clock
6/16

7/16 7/16

One-bit time

Conditions
to validate 1 1 1 0 X 0 X 0 X 0 0 0 0 X X X X X X
the start bit
Falling edge At least 2 bits At least 2 bits
detection out of 3 at 0 out of 3 at 0
ai15471b

Note: If the sequence is not complete, the start bit detection aborts and the receiver returns to the
idle state (no flag is set) where it waits for a falling edge.
The start bit is confirmed (RXNE flag set, interrupt generated if RXNEIE=1) if the 3 sampled
bits are at 0 (first sampling on the 3rd, 5th and 7th bits finds the 3 bits at 0 and second
sampling on the 8th, 9th and 10th bits also finds the 3 bits at 0).
The start bit is validated (RXNE flag set, interrupt generated if RXNEIE=1) but the NE noise
flag is set if, for both samplings, at least 2 out of the 3 sampled bits are at 0 (sampling on the
3rd, 5th and 7th bits and sampling on the 8th, 9th and 10th bits). If this condition is not met,
the start detection aborts and the receiver returns to the idle state (no flag is set).
If, for one of the samplings (sampling on the 3rd, 5th and 7th bits or sampling on the 8th, 9th
and 10th bits), 2 out of the 3 bits are found at 0, the start bit is validated but the NE noise
flag bit is set.

Character reception
During an USART reception, data shifts in least significant bit first through the RX pin. In this
mode, the USART_DR register consists of a buffer (RDR) between the internal bus and the
received shift register.

RM0402 Rev 6 767/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Procedure:
1. Enable the USART by writing the UE bit in USART_CR1 register to 1.
2. Program the M bit in USART_CR1 to define the word length.
3. Program the number of stop bits in USART_CR2.
4. Select DMA enable (DMAR) in USART_CR3 if multibuffer communication is to take
place. Configure the DMA register as explained in multibuffer communication. STEP 3
5. Select the desired baud rate using the baud rate register USART_BRR
6. Set the RE bit USART_CR1. This enables the receiver that begins searching for a start
bit.
When a character is received
• The RXNE bit is set. It indicates that the content of the shift register is transferred to the
RDR. In other words, data has been received and can be read (as well as its
associated error flags).
• An interrupt is generated if the RXNEIE bit is set.
• The error flags can be set if a frame error, noise or an overrun error has been detected
during reception.
• In multibuffer, RXNE is set after every byte received and is cleared by the DMA read to
the Data Register.
• In single buffer mode, clearing the RXNE bit is performed by a software read to the
USART_DR register. The RXNE flag can also be cleared by writing a zero to it. The
RXNE bit must be cleared before the end of the reception of the next character to avoid
an overrun error.
Note: The RE bit should not be reset while receiving data. If the RE bit is disabled during
reception, the reception of the current byte will be aborted.

Break character
When a break character is received, the USART handles it as a framing error.

Idle character
When an idle frame is detected, there is the same procedure as a data received character
plus an interrupt if the IDLEIE bit is set.

Overrun error
An overrun error occurs when a character is received when RXNE has not been reset. Data
can not be transferred from the shift register to the RDR register until the RXNE bit is
cleared.

768/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

The RXNE flag is set after every byte received. An overrun error occurs if RXNE flag is set
when the next data is received or the previous DMA request has not been serviced. When
an overrun error occurs:
• The ORE bit is set.
• The RDR content will not be lost. The previous data is available when a read to
USART_DR is performed.
• The shift register will be overwritten. After that point, any data received during overrun
is lost.
• An interrupt is generated if either the RXNEIE bit is set or both the EIE and DMAR bits
are set.
• The ORE bit is reset by a read to the USART_SR register followed by a USART_DR
register read operation.
Note: The ORE bit, when set, indicates that at least 1 data has been lost. There are two
possibilities:
• if RXNE=1, then the last valid data is stored in the receive register RDR and can be
read,
• if RXNE=0, then it means that the last valid data has already been read and thus there
is nothing to be read in the RDR. This case can occur when the last valid data is read in
the RDR at the same time as the new (and lost) data is received. It may also occur
when the new data is received during the reading sequence (between the USART_SR
register read access and the USART_DR read access).

Selecting the proper oversampling method

The receiver implements different user-configurable oversampling techniques (except in
synchronous mode) for data recovery by discriminating between valid incoming data and
noise.
The oversampling method can be selected by programming the OVER8 bit in the
USART_CR1 register and can be either 16 or 8 times the baud rate clock (Figure 250 and
Figure 251).
Depending on the application:
• select oversampling by 8 (OVER8=1) to achieve higher speed (up to fPCLK/8). In this
case the maximum receiver tolerance to clock deviation is reduced (refer to
Section 25.4.5: USART receiver tolerance to clock deviation)
• select oversampling by 16 (OVER8=0) to increase the tolerance of the receiver to clock
deviations. In this case, the maximum speed is limited to maximum fPCLK/16

RM0402 Rev 6 769/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Programming the ONEBIT bit in the USART_CR3 register selects the method used to
evaluate the logic level. There are two options:
• the majority vote of the three samples in the center of the received bit. In this case,
when the 3 samples used for the majority vote are not equal, the NF bit is set
• a single sample in the center of the received bit
Depending on the application:
– select the three samples’ majority vote method (ONEBIT=0) when operating in a
noisy environment and reject the data when a noise is detected (refer to
Figure 137) because this indicates that a glitch occurred during the sampling.
– select the single sample method (ONEBIT=1) when the line is noise-free to
increase the receiver tolerance to clock deviations (see Section 25.4.5: USART
receiver tolerance to clock deviation). In this case the NF bit will never be set.
When noise is detected in a frame:
• The NF bit is set at the rising edge of the RXNE bit.
• The invalid data is transferred from the Shift register to the USART_DR register.
• No interrupt is generated in case of single byte communication. However this bit rises
at the same time as the RXNE bit that itself generates an interrupt. In case of
multibuffer communication an interrupt will be issued if the EIE bit is set in the
USART_CR3 register.
The NF bit is reset by a USART_SR register read operation followed by a USART_DR
register read operation.
Note: Oversampling by 8 is not available in the Smartcard, IrDA and LIN modes. In those modes,
the OVER8 bit is forced to ‘0 by hardware.

Figure 250. Data sampling when oversampling by 16

RX line

sampled values
Sample clock 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

6/16
7/16 7/16
One bit time

MSv31152V1

770/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Figure 251. Data sampling when oversampling by 8

RX line

sampled values
Sample
clock (x8) 1 2 3 4 5 6 7 8

2/8
3/8 3/8
One bit time

MSv31153V1

Table 137. Noise detection from sampled data

Sampled value NE status Received bit value
000 0 0
001 1 0
010 1 0
011 1 1
100 1 0
101 1 1
110 1 1
111 0 1

Framing error
A framing error is detected when:
The stop bit is not recognized on reception at the expected time, following either a de-
synchronization or excessive noise.
When the framing error is detected:
• The FE bit is set by hardware
• The invalid data is transferred from the Shift register to the USART_DR register.
• No interrupt is generated in case of single byte communication. However this bit rises
at the same time as the RXNE bit that itself generates an interrupt. In case of
multibuffer communication an interrupt will be issued if the EIE bit is set in the
USART_CR3 register.
The FE bit is reset by a USART_SR register read operation followed by a USART_DR
register read operation.

Configurable stop bits during reception

The number of stop bits to be received can be configured through the control bits of Control
Register 2 - it can be either 1 or 2 in normal mode and 0.5 or 1.5 in Smartcard mode.

RM0402 Rev 6 771/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

1. 0.5 stop bit (reception in Smartcard mode): No sampling is done for 0.5 stop bit. As
a consequence, no framing error and no break frame can be detected when 0.5 stop bit
is selected.
2. 1 stop bit: Sampling for 1 stop Bit is done on the 8th, 9th and 10th samples.
3. 1.5 stop bits (Smartcard mode): When transmitting in smartcard mode, the device
must check that the data is correctly sent. Thus the receiver block must be enabled (RE
=1 in the USART_CR1 register) and the stop bit is checked to test if the smartcard has
detected a parity error. In the event of a parity error, the smartcard forces the data
signal low during the sampling - NACK signal-, which is flagged as a framing error.
Then, the FE flag is set with the RXNE at the end of the 1.5 stop bit. Sampling for 1.5
stop bits is done on the 16th, 17th and 18th samples (1 baud clock period after the
beginning of the stop bit). The 1.5 stop bit can be decomposed into two parts: one 0.5
baud clock period during which nothing happens, followed by 1 normal stop bit period
during which sampling occurs halfway through. Refer to Section 25.4.11 for more
details.
4. 2 stop bits: Sampling for 2 stop bits is done on the 8th, 9th and 10th samples of the first
stop bit. If a framing error is detected during the first stop bit the framing error flag will
be set. The second stop bit is not checked for framing error. The RXNE flag will be set
at the end of the first stop bit.

25.4.4 Fractional baud rate generation

The baud rate for the receiver and transmitter (Rx and Tx) are both set to the same value as
programmed in the Mantissa and Fraction values of USARTDIV.

Equation 1: Baud rate for standard USART (SPI mode included)

f CK
Tx/Rx baud = ----------------------------------------------------------------------------------
-
8 × ( 2 – OVER8 ) × USARTDIV

Equation 2: Baud rate in Smartcard, LIN and IrDA modes

f CK
Tx/Rx baud = ---------------------------------------------
16 × USARTDIV

USARTDIV is an unsigned fixed point number that is coded on the USART_BRR register.
• When OVER8=0, the fractional part is coded on 4 bits and programmed by the
DIV_fraction[3:0] bits in the USART_BRR register
• When OVER8=1, the fractional part is coded on 3 bits and programmed by the
DIV_fraction[2:0] bits in the USART_BRR register, and bit DIV_fraction[3] must be kept
cleared.
Note: The baud counters are updated to the new value in the baud registers after a write operation
to USART_BRR. Hence the baud rate register value should not be changed during
communication.

How to derive USARTDIV from USART_BRR register values when OVER8=0

Example 1:
If DIV_Mantissa = 0d27 and DIV_Fraction = 0d12 (USART_BRR = 0x1BC), then
Mantissa (USARTDIV) = 0d27

772/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Fraction (USARTDIV) = 12/16 = 0d0.75

Therefore USARTDIV = 0d27.75
Example 2:
To program USARTDIV = 0d25.62
This leads to:
DIV_Fraction = 16*0d0.62 = 0d9.92
The nearest real number is 0d10 = 0xA
DIV_Mantissa = mantissa (0d25.620) = 0d25 = 0x19
Then, USART_BRR = 0x19A hence USARTDIV = 0d25.625

Example 3:
To program USARTDIV = 0d50.99
This leads to:
DIV_Fraction = 16*0d0.99 = 0d15.84
The nearest real number is 0d16 = 0x10 => overflow of DIV_frac[3:0] => carry must be
added up to the mantissa
DIV_Mantissa = mantissa (0d50.990 + carry) = 0d51 = 0x33
Then, USART_BRR = 0x330 hence USARTDIV = 0d51.000

How to derive USARTDIV from USART_BRR register values when OVER8=1

Example 1:
If DIV_Mantissa = 0x27 and DIV_Fraction[2:0]= 0d6 (USART_BRR = 0x1B6), then
Mantissa (USARTDIV) = 0d27
Fraction (USARTDIV) = 6/8 = 0d0.75
Therefore USARTDIV = 0d27.75
Example 2:
To program USARTDIV = 0d25.62
This leads to:
DIV_Fraction = 8*0d0.62 = 0d4.96
The nearest real number is 0d5 = 0x5
DIV_Mantissa = mantissa (0d25.620) = 0d25 = 0x19
Then, USART_BRR = 0x195 => USARTDIV = 0d25.625
Example 3:
To program USARTDIV = 0d50.99
This leads to:
DIV_Fraction = 8*0d0.99 = 0d7.92

RM0402 Rev 6 773/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

The nearest real number is 0d8 = 0x8 => overflow of the DIV_frac[2:0] => carry must be
added up to the mantissa
DIV_Mantissa = mantissa (0d50.990 + carry) = 0d51 = 0x33
Then, USART_BRR = 0x0330 => USARTDIV = 0d51.000

Table 138. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 16(1)
Oversampling by 16 (OVER8=0)

Baud rate fPCLK = 8 MHz fPCLK = 12 MHz

Value % Error = Value

programmed (Calculated - programmed
S.No Desired Actual Actual % Error
in the baud Desired) B.rate / in the baud
rate register Desired B.rate rate register

1 1.2 KBps 1.2 KBps 416.6875 0 1.2 KBps 625 0

2 2.4 KBps 2.4 KBps 208.3125 0.01 2.4 KBps 312.5 0
3 9.6 KBps 9.604 KBps 52.0625 0.04 9.6 KBps 78.125 0
4 19.2 KBps 19.185 KBps 26.0625 0.08 19.2 KBps 39.0625 0
5 38.4 KBps 38.462 KBps 13 0.16 38.339 KBps 19.5625 0.16
6 57.6 KBps 57.554 KBps 8.6875 0.08 57.692 KBps 13 0.16
7 115.2 KBps 115.942 KBps 4.3125 0.64 115.385 KBps 6.5 0.16
8 230.4 KBps 228.571 KBps 2.1875 0.79 230.769 KBps 3.25 0.16
9 460.8 KBps 470.588 KBps 1.0625 2.12 461.538 KBps 1.625 0.16
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 139. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 8(1)
Oversampling by 8 (OVER8 = 1)

Baud rate fPCLK = 8 MHz fPCLK = 12 MHz

% Error =
Value (Calculated - Value
programmed Desired) programmed
S.No Desired Actual Actual % Error
in the baud B.rate / in the baud
rate register Desired rate register
B.rate

1 1.2 KBps 1.2 KBps 833.375 0 1.2 KBps 1250 0

2 2.4 KBps 2.4 KBps 416.625 0.01 2.4 KBps 625 0
3 9.6 KBps 9.604 KBps 104.125 0.04 9.6 KBps 156.25 0
4 19.2 KBps 19.185 KBps 52.125 0.08 19.2 KBps 78.125 0
5 38.4 KBps 38.462 KBps 26 0.16 38.339 KBps 39.125 0.16

774/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Table 139. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 12 MHz,
oversampling by 8(1) (continued)
Oversampling by 8 (OVER8 = 1)

Baud rate fPCLK = 8 MHz fPCLK = 12 MHz

% Error =
Value (Calculated - Value
programmed Desired) programmed
S.No Desired Actual Actual % Error
in the baud B.rate / in the baud
rate register Desired rate register
B.rate

6 57.6 KBps 57.554 KBps 17.375 0.08 57.692 KBps 26 0.16

7 115.2 KBps 115.942 KBps 8.625 0.64 115.385 KBps 13 0.16
8 230.4 KBps 228.571 KBps 4.375 0.79 230.769 KBps 6.5 0.16
9 460.8 KBps 470.588 KBps 2.125 2.12 461.538 KBps 3.25 0.16
10 921.6 KBps 888.889 KBps 1.125 3.55 923.077 KBps 1.625 0.16
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 140. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 16(1)
Oversampling by 16 (OVER8 = 0)

Baud rate fPCLK = 16 MHz fPCLK = 24 MHz

Value % Error = Value

programmed (Calculated - programmed
S.No Desired Actual Actual % Error
in the baud Desired) B.rate / in the baud
rate register Desired B.rate rate register

1 1.2 KBps 1.2 KBps 833.3125 0 1.2 1250 0

2 2.4 KBps 2.4 KBps 416.6875 0 2.4 625 0
3 9.6 KBps 9.598 KBps 104.1875 0.02 9.6 156.25 0
4 19.2 KBps 19.208 KBps 52.0625 0.04 19.2 78.125 0
5 38.4 KBps 38.369 KBps 26.0625 0.08 38.4 39.0625 0
6 57.6 KBps 57.554 KBps 17.375 0.08 57.554 26.0625 0.08
7 115.2 KBps 115.108 KBps 8.6875 0.08 115.385 13 0.16
8 230.4 KBps 231.884 KBps 4.3125 0.64 230.769 6.5 0.16
9 460.8 KBps 457.143 KBps 2.1875 0.79 461.538 3.25 0.16
10 921.6 KBps 941.176 KBps 1.0625 2.12 923.077 1.625 0.16
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

RM0402 Rev 6 775/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Table 141. Error calculation for programmed baud rates at fPCLK = 16 MHz or fPCLK = 24 MHz,
oversampling by 8(1)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 16 MHz fPCLK = 24 MHz

Value % Error = Value

programmed (Calculated - programmed
S.No Desired Actual Actual % Error
in the baud Desired) B.rate / in the baud
rate register Desired B.rate rate register

1 1.2 KBps 1.2 KBps 1666.625 0 1.2 KBps 2500 0

2 2.4 KBps 2.4 KBps 833.375 0 2.4 KBps 1250 0
3 9.6 KBps 9.598 KBps 208.375 0.02 9.6 KBps 312.5 0
4 19.2 KBps 19.208 KBps 104.125 0.04 19.2 KBps 156.25 0
5 38.4 KBps 38.369 KBps 52.125 0.08 38.4 KBps 78.125 0
6 57.6 KBps 57.554 KBps 34.75 0.08 57.554 KBps 52.125 0.08
7 115.2 KBps 115.108 KBps 17.375 0.08 115.385 KBps 26 0.16
8 230.4 KBps 231.884 KBps 8.625 0.64 230.769 KBps 13 0.16
9 460.8 KBps 457.143 KBps 4.375 0.79 461.538 KBps 6.5 0.16
10 921.6 KBps 941.176 KBps 2.125 2.12 923.077 KBps 3.25 0.16
11 2 MBps 2000 KBps 1 0 2000 KBps 1.5 0
12 3 MBps NA NA NA 3000 KBps 1 0
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 142. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 16(1)
Oversampling by 16 (OVER8=0)

Baud rate fPCLK = 8 MHz fPCLK = 16 MHz

Value % Error = Value

programmed (Calculated - programmed
S.No Desired Actual Actual % Error
in the baud Desired)B.Rate in the baud
rate register /Desired B.Rate rate register

1 2.4 KBps 2.400 KBps 208.3125 0.00% 2.400 KBps 416.6875 0.00%
2 9.6 KBps 9.604 KBps 52.0625 0.04% 9.598 KBps 104.1875 0.02%
3 19.2 KBps 19.185 KBps 26.0625 0.08% 19.208 KBps 52.0625 0.04%
4 57.6 KBps 57.554 KBps 8.6875 0.08% 57.554 KBps 17.3750 0.08%
5 115.2 KBps 115.942 KBps 4.3125 0.64% 115.108 KBps 8.6875 0.08%
6 230.4 KBps 228.571 KBps 2.1875 0.79% 231.884 KBps 4.3125 0.64%
7 460.8 KBps 470.588 KBps 1.0625 2.12% 457.143 KBps 2.1875 0.79%

776/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Table 142. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 16(1) (continued)
Oversampling by 16 (OVER8=0)

Baud rate fPCLK = 8 MHz fPCLK = 16 MHz

Value % Error = Value

programmed (Calculated - programmed
S.No Desired Actual Actual % Error
in the baud Desired)B.Rate in the baud
rate register /Desired B.Rate rate register

8 896 KBps NA NA NA 888.889 KBps 1.1250 0.79%

9 921.6 KBps NA NA NA 941.176 KBps 1.0625 2.12%
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

Table 143. Error calculation for programmed baud rates at fPCLK = 8 MHz or fPCLK = 16 MHz,
oversampling by 8(1)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 8 MHz fPCLK = 16 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 2.4 KBps 2.400 KBps 416.625 0.01% 2.400 KBps 833.375 0.00%
2 9.6 KBps 9.604 KBps 104.125 0.04% 9.598 KBps 208.375 0.02%
3 19.2 KBps 19.185 KBps 52.125 0.08% 19.208 KBps 104.125 0.04%
4 57.6 KBps 57.557 KBps 17.375 0.08% 57.554 KBps 34.750 0.08%
5 115.2 KBps 115.942 KBps 8.625 0.64% 115.108 KBps 17.375 0.08%
6 230.4 KBps 228.571 KBps 4.375 0.79% 231.884 KBps 8.625 0.64%
7 460.8 KBps 470.588 KBps 2.125 2.12% 457.143 KBps 4.375 0.79%
8 896 KBps 888.889 KBps 1.125 0.79% 888.889 KBps 2.250 0.79%
9 921.6 KBps 888.889 KBps 1.125 3.55% 941.176 KBps 2.125 2.12%
10 1.792 MBps NA NA NA 1.7777 MBps 1.125 0.79%
11 1.8432 MBps NA NA NA 1.7777 MBps 1.125 3.55%
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.

RM0402 Rev 6 777/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Table 144. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 16(1)(2)
Oversampling by 16 (OVER8=0)

Baud rate fPCLK = 30 MHz fPCLK = 60 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 2.4 KBps 2.400 KBps 781.2500 0.00% 2.400 KBps 1562.5000 0.00%
2 9.6 KBps 9.600 KBps 195.3125 0.00% 9.600 KBps 390.6250 0.00%
3 19.2 KBps 19.194 KBps 97.6875 0.03% 19.200 KBps 195.3125 0.00%
4 57.6 KBps 57.582KBps 32.5625 0.03% 57.582 KBps 65.1250 0.03%
5 115.2 KBps 115.385 KBps 16.2500 0.16% 115.163 KBps 32.5625 0.03%
6 230.4 KBps 230.769 KBps 8.1250 0.16% 230.769 KBps 16.2500 0.16%
7 460.8 KBps 461.538 KBps 4.0625 0.16% 461.538 KBps 8.1250 0.16%
8 896 KBps 909.091 KBps 2.0625 1.46% 895.522 KBps 4.1875 0.05%
9 921.6 KBps 909.091 KBps 2.0625 1.36% 923.077 KBps 4.0625 0.16%
10 1.792 MBps 1.1764 MBps 1.0625 1.52% 1.8182 MBps 2.0625 1.36%
11 1.8432 MBps 1.8750 MBps 1.0000 1.73% 1.8182 MBps 2.0625 1.52%
12 3.584 MBps NA NA NA 3.2594 MBps 1.0625 1.52%
13 3.6864 MBps NA NA NA 3.7500 MBps 1.0000 1.73%
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 145. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 8(1) (2)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 30 MHz fPCLK =60 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 2.4 KBps 2.400 KBps 1562.5000 0.00% 2.400 KBps 3125.0000 0.00%
2 9.6 KBps 9.600 KBps 390.6250 0.00% 9.600 KBps 781.2500 0.00%
3 19.2 KBps 19.194 KBps 195.3750 0.03% 19.200 KBps 390.6250 0.00%
4 57.6 KBps 57.582 KBps 65.1250 0.16% 57.582 KBps 130.2500 0.03%
5 115.2 KBps 115.385 KBps 32.5000 0.16% 115.163 KBps 65.1250 0.03%
6 230.4 KBps 230.769 KBps 16.2500 0.16% 230.769 KBps 32.5000 0.16%

778/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Table 145. Error calculation for programmed baud rates at fPCLK = 30 MHz or fPCLK = 60 MHz,
oversampling by 8(1) (2) (continued)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 30 MHz fPCLK =60 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

7 460.8 KBps 461.538 KBps 8.1250 0.16% 461.538 KBps 16.2500 0.16%
8 896 KBps 909.091 KBps 4.1250 1.46% 895.522 KBps 8.3750 0.05%
9 921.6 KBps 909.091 KBps 4.1250 1.36% 923.077 KBps 8.1250 0.16%
10 1.792 MBps 1.7647 MBps 2.1250 1.52% 1.8182 MBps 4.1250 1.46%
11 1.8432 MBps 1.8750 MBps 2.0000 1.73% 1.8182 MBps 4.1250 1.36%
12 3.584 MBps 3.7500 MBps 1.0000 4.63% 3.5294 MBps 2.1250 1.52%
13 3.6864 MBps 3.7500 MBps 1.0000 1.73% 3.7500 MBps 2.0000 1.73%
14 7.168 MBps NA NA NA 7.5000 MBps 1.0000 4.63%
15 7.3728 MBps NA NA NA 7.5000 MBps 1.0000 1.73%
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 146. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16(1)(2)
Oversampling by 16 (OVER8=0)

Baud rate fPCLK = 42 MHz fPCLK = 84 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 1.2 KBps 1.2 KBps 2187.5 0 1.2 KBps 4375 0

2 2.4 KBps 2.4 KBps 1093.75 0 2.4 KBps 2187.5 0
3 9.6 KBps 9.6 KBps 273.4375 0 9.6 KBps 546.875 0
4 19.2 KBps 19.195 KBps 136.75 0.02 19.2 KBps 273.4375 0
5 38.4 KBps 38.391 KBps 68.375 0.02 38.391 KBps 136.75 0.02
6 57.6 KBps 57.613 KBps 45.5625 0.02 57.613 KBps 91.125 0.02
7 115.2 KBps 115.068 KBps 22.8125 0.11 115.226 KBps 45.5625 0.02
8 230.4 KBps 230.769 KBps 11.375 0.16 230.137 KBps 22.8125 0.11
9 460.8 KBps 461.538 KBps 5.6875 0.16 461.538 KBps 11.375 0.16

RM0402 Rev 6 779/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Table 146. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 Hz,
oversampling by 16(1)(2) (continued)
Oversampling by 16 (OVER8=0)

Baud rate fPCLK = 42 MHz fPCLK = 84 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

10 921.6 KBps 913.043 KBps 2.875 0.93 923.076 KBps 5.6875 0.93
11 1.792 MBps 1.826 MBps 1.4375 1.9 1.787 MBps 2.9375 0.27
12 1.8432 MBps 1.826 MBps 1.4375 0.93 1.826 MBps 2.875 0.93
13 3.584 MBps NA NA NA 3.652 MBps 1.4375 1.9
14 3.6864 MBps NA NA NA 3.652 MBps 1.4375 0.93
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 147. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 MHz,
oversampling by 8(1)(2)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 42 MHz fPCLK = 84 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 1.2 KBps 1.2 KBps 4375 0 1.2 KBps 8750 0

2 2.4 KBps 2.4 KBps 2187.5 0 2.4 KBps 4375 0
3 9.6 KBps 9.6 KBps 546.875 0 9.6 KBps 1093.75 0
4 19.2 KBps 19.195 KBps 273.5 0.02 19.2 KBps 546.875 0
5 38.4 KBps 38.391 KBps 136.75 0.02 38.391 KBps 273.5 0.02
6 57.6 KBps 57.613 KBps 91.125 0.02 57.613 KBps 182.25 0.02
7 115.2 KBps 115.068 KBps 45.625 0.11 115.226 KBps 91.125 0.02
8 230.4 KBps 230.769 KBps 22.75 0.11 230.137 KBps 45.625 0.11
9 460.8 KBps 461.538 KBps 11.375 0.16 461.538 KBps 22.75 0.16
10 921.6 KBps 913.043 KBps 5.75 0.93 923.076 KBps 11.375 0.93
11 1.792 MBps 1.826 MBps 2.875 1.9 1.787Mbps 5.875 0.27
12 1.8432 MBps 1.826 MBps 2.875 0.93 1.826 MBps 5.75 0.93
13 3.584 MBps 3.5 MBps 1.5 2.34 3.652 MBps 2.875 1.9
14 3.6864 MBps 3.82 MBps 1.375 3.57 3.652 MBps 2.875 0.93

780/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Table 147. Error calculation for programmed baud rates at fPCLK = 42 MHz or fPCLK = 84 MHz,
oversampling by 8(1)(2) (continued)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 42 MHz fPCLK = 84 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

15 7.168 MBps NA NA NA 7 MBps 1.5 2.34

16 7.3728 MBps NA NA NA 7.636 MBps 1.375 3.57
18 9 MBps NA NA NA 9.333 MBps 1.125 3.7
20 10.5 MBps NA NA NA 10.5 MBps 1 0
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

Table 148. Error calculation for programmed baud rates at fPCLK = 100 MHz or fPCLK = 50 MHz,
oversampling by 16(1)(2)
Oversampling by 16 (OVER16=1)

Baud rate fPCLK = 100 MHz fPCLK = 50 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 9.600 KBps 9.601 KBps 651 0.006 9.601 KBps 325.5 0.006
2 19.200 KBps 19.201 KBps 325 0.006 19.201 KBps 162.75 0.006
3 38.400 KBps 38.402 KBps 162.75 0.006 38.402 KBps 81.375 0.006
4 57.600 KBps 57.603 KBps 108.5 0.006 57.603 KBps 54.25 0.006
5 115.200 KBps 115.207 KBps 54.25 0.006 115.207 KBps 27.125 0.006
6 230.400 KBps 230.414 KBps 27.125 0.006 230.414 KBps 13.5625 0.006
7 460.800 KBps 460.829 KBps 13.5625 0.006 462.962 KBps 6.75 0.47
8 921.600 KBps 925.925 KBps 6.75 0.470 925.925 KBps 3.375 0.47
9 3.125 MBps 3.125 MBps 2 0 3.125 MBps 1 0
10 4.000 MBps 4.000 MBps 1.5625 0 NA NA NA
11 6.250 MBps 6.250 MBps 1 0 NA NA NA
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

RM0402 Rev 6 781/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Table 149. Error calculation for programmed baud rates at fPCLK = 100 MHz or fPCLK = 50 MHz,
oversampling by 8(1)(2)
Oversampling by 8 (OVER8=1)

Baud rate fPCLK = 100 MHz fPCLK = 50 MHz

Value % Error = Value

programmed (Calculated - programmed %
S.No Desired Actual Actual
in the baud Desired)B.Rate in the baud Error
rate register /Desired B.Rate rate register

1 9.600 KBps 9.601 KBps 1302 0.006 9.601 KBps 651 0.006
2 19.200 KBps 19.201 KBps 651 0.006 19.201 KBps 325.5 0.006
3 38.400 KBps 38.402 KBps 325.5 0.006 38.402 KBps 162.75 0.006
4 57.600 KBps 57.603 KBps 217 0.006 57.603 KBps 108.5 0.006
5 115.200 KBps 115.207 KBps 108.5 0.006 115.207 KBps 54.25 0.006
6 230.400 KBps 230.414 KBps 54.25 0.006 230.414 KBps 27.125 0.006
7 460.800 KBps 460.829 KBps 27.125 0.006 462.962 KBps 13.5 0.470
8 921.600 KBps 925.925 KBps 13.5 0.470 925.925 KBps 6.75 0.470
9 4.000 MBps 4 MBps 3.125 0.000 4.167 MBps 1.5 4.170
10 6.250 MBps 6.25 MBps 2 0.000 6.250 MBps 1 0.000
11 12.500 MBps 12.500 MBps 1 0.000 NA NA NA
1. The lower the CPU clock the lower the accuracy for a particular baud rate. The upper limit of the achievable baud rate can
be fixed with these data.
2. Only USART1 and USART6 are clocked with PCLK2. Other USARTs are clocked with PCLK1. Refer to the device
datasheets for the maximum values for PCLK1 and PCLK2.

25.4.5 USART receiver tolerance to clock deviation

The USART asynchronous receiver works correctly only if the total clock system deviation is
smaller than the USART receiver tolerance. The causes that contribute to the total deviation
are:
• DTRA: Deviation due to the transmitter error (also includes the deviation of the
transmitter local oscillator)
• DQUANT: Error due to the baud rate quantization of the receiver
• DREC: Deviation of the receiver local oscillator
• DTCL: Deviation due to the transmission line (generally due to the transceivers that
can introduce an asymmetry between the low-to-high transition timing and the
high-to-low transition timing)
DTRA + DQUANT + DREC + DTCL < USART receiver tolerance

782/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

The USART receiver tolerance to properly receive data is equal to the maximum tolerated
deviation and depends on the following choices:
• 10- or 11-bit character length defined by the M bit in the USART_CR1 register
• oversampling by 8 or 16 defined by the OVER8 bit in the USART_CR1 register
• use of fractional baud rate or not
• use of 1 bit or 3 bits to sample the data, depending on the value of the ONEBIT bit in
the USART_CR3 register

Table 150. USART receiver tolerance when DIV fraction is 0

OVER8 bit = 0 OVER8 bit = 1
M bit
ONEBIT=0 ONEBIT=1 ONEBIT=0 ONEBIT=1
0 3.75% 4.375% 2.50% 3.75%
1 3.41% 3.97% 2.27% 3.41%

Table 151. USART receiver tolerance when DIV_Fraction is different from 0

OVER8 bit = 0 OVER8 bit = 1
M bit
ONEBIT=0 ONEBIT=1 ONEBIT=0 ONEBIT=1
0 3.33% 3.88% 2% 3%
1 3.03% 3.53% 1.82% 2.73%

Note: The figures specified in Table 150 and Table 151 may slightly differ in the special case when
the received frames contain some Idle frames of exactly 10-bit times when M=0 (11-bit times
when M=1).

25.4.6 Multiprocessor communication

There is a possibility of performing multiprocessor communication with the USART (several
USARTs connected in a network). For instance one of the USARTs can be the master, its TX
output is connected to the RX input of the other USART. The others are slaves, their
respective TX outputs are logically ANDed together and connected to the RX input of the
master.
In multiprocessor configurations it is often desirable that only the intended message
recipient should actively receive the full message contents, thus reducing redundant USART
service overhead for all non addressed receivers.
The non addressed devices may be placed in mute mode by means of the muting function.
In mute mode:
• None of the reception status bits can be set.
• All the receive interrupts are inhibited.
• The RWU bit in USART_CR1 register is set to 1. RWU can be controlled automatically
by hardware or written by the software under certain conditions.
The USART can enter or exit from mute mode using one of two methods, depending on the
WAKE bit in the USART_CR1 register:
• Idle Line detection if the WAKE bit is reset,
• Address Mark detection if the WAKE bit is set.

RM0402 Rev 6 783/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Idle line detection (WAKE=0)

The USART enters mute mode when the RWU bit is written to 1.
It wakes up when an Idle frame is detected. Then the RWU bit is cleared by hardware but
the IDLE bit is not set in the USART_SR register. RWU can also be written to 0 by software.
An example of mute mode behavior using Idle line detection is given in Figure 252.

Figure 252. Mute mode using Idle line detection

RXNE RXNE

RX Data 1 Data 2 Data 3 Data 4 IDLE Data 5 Data 6

RWU Mute mode Normal mode

MMRQ written to 1 Idle frame detected

MSv31154V1

Address mark detection (WAKE=1)

In this mode, bytes are recognized as addresses if their MSB is a ‘1 else they are
considered as data. In an address byte, the address of the targeted receiver is put on the 4
LSB. This 4-bit word is compared by the receiver with its own address that is programmed in
the ADD bits in the USART_CR2 register.
The USART enters mute mode when an address character is received that does not match
its programmed address. In this case, the RWU bit is set by hardware. The RXNE flag is not
set for this address byte and no interrupt nor DMA request is issued as the USART would
have entered mute mode.
It exits from mute mode when an address character is received that matches the
programmed address. Then the RWU bit is cleared and subsequent bytes are received
normally. The RXNE bit is set for the address character since the RWU bit has been
cleared.
The RWU bit can be written to as 0 or 1 when the receiver buffer contains no data (RXNE=0
in the USART_SR register). Otherwise the write attempt is ignored.
An example of mute mode behavior using address mark detection is given in Figure 253.

784/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Figure 253. Mute mode using address mark detection

In this example, the current address of the receiver is 1
(programmed in the USART_CR2 register)
RXNE RXNE RXNE

RX IDLE Addr=0 Data 1 Data 2 IDLE Addr=1 Data 3 Data 4 Addr=2 Data 5

RWU Mute mode Normal mode Mute mode

MMRQ written to 1 Matching address Non-matching address

(RXNE was cleared)

Non-matching address
MSv31155V1

25.4.7 Parity control

Parity control (generation of parity bit in transmission and parity checking in reception) can
be enabled by setting the PCE bit in the USART_CR1 register. Depending on the frame
length defined by the M bit, the possible USART frame formats are as listed in Table 152.

Table 152. Frame formats

M bit PCE bit USART frame(1)

0 0 | SB | 8 bit data | STB |

0 1 | SB | 7-bit data | PB | STB |
1 0 | SB | 9-bit data | STB |
1 1 | SB | 8-bit data PB | STB |
1. Legends: SB: start bit, STB: stop bit, PB: parity bit.

Even parity
The parity bit is calculated to obtain an even number of “1s” inside the frame made of the 7
or 8 LSB bits (depending on whether M is equal to 0 or 1) and the parity bit.
E.g.: data=00110101; 4 bits set => parity bit will be 0 if even parity is selected (PS bit in
USART_CR1 = 0).

Odd parity
The parity bit is calculated to obtain an odd number of “1s” inside the frame made of the 7 or
8 LSB bits (depending on whether M is equal to 0 or 1) and the parity bit.
E.g.: data=00110101; 4 bits set => parity bit will be 1 if odd parity is selected (PS bit in
USART_CR1 = 1).

Parity checking in reception

If the parity check fails, the PE flag is set in the USART_SR register and an interrupt is
generated if PEIE is set in the USART_CR1 register. The PE flag is cleared by a software

RM0402 Rev 6 785/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

sequence (a read from the status register followed by a read or write access to the
USART_DR data register).
Note: In case of wakeup by an address mark: the MSB bit of the data is taken into account to
identify an address but not the parity bit. And the receiver does not check the parity of the
address data (PE is not set in case of a parity error).

Parity generation in transmission

If the PCE bit is set in USART_CR1, then the MSB bit of the data written in the data register
is transmitted but is changed by the parity bit (even number of “1s” if even parity is selected
(PS=0) or an odd number of “1s” if odd parity is selected (PS=1)).
Note: The software routine that manages the transmission can activate the software sequence
that clears the PE flag (a read from the status register followed by a read or write access to
the data register). When operating in half-duplex mode, depending on the software, this can
cause the PE flag to be unexpectedly cleared.

25.4.8 LIN (local interconnection network) mode

The LIN mode is selected by setting the LINEN bit in the USART_CR2 register. In LIN
mode, the following bits must be kept cleared:
• STOP[1:0] and CLKEN in the USART_CR2 register
• SCEN, HDSEL and IREN in the USART_CR3 register.

LIN transmission
The same procedure explained in Section 25.4.2 has to be applied for LIN Master
transmission than for normal USART transmission with the following differences:
• Clear the M bit to configure 8-bit word length.
• Set the LINEN bit to enter LIN mode. In this case, setting the SBK bit sends 13 ‘0 bits
as a break character. Then a bit of value ‘1 is sent to allow the next start detection.

LIN reception
A break detection circuit is implemented on the USART interface. The detection is totally
independent from the normal USART receiver. A break can be detected whenever it occurs,
during Idle state or during a frame.
When the receiver is enabled (RE=1 in USART_CR1), the circuit looks at the RX input for a
start signal. The method for detecting start bits is the same when searching break
characters or data. After a start bit has been detected, the circuit samples the next bits
exactly like for the data (on the 8th, 9th and 10th samples). If 10 (when the LBDL = 0 in
USART_CR2) or 11 (when LBDL=1 in USART_CR2) consecutive bits are detected as ‘0,
and are followed by a delimiter character, the LBD flag is set in USART_SR. If the LBDIE
bit=1, an interrupt is generated. Before validating the break, the delimiter is checked for as it
signifies that the RX line has returned to a high level.
If a ‘1 is sampled before the 10 or 11 have occurred, the break detection circuit cancels the
current detection and searches for a start bit again.
If the LIN mode is disabled (LINEN=0), the receiver continues working as normal USART,
without taking into account the break detection.
If the LIN mode is enabled (LINEN=1), as soon as a framing error occurs (i.e. stop bit
detected at ‘0, which will be the case for any break frame), the receiver stops until the break

786/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

detection circuit receives either a ‘1, if the break word was not complete, or a delimiter
character if a break has been detected.
The behavior of the break detector state machine and the break flag is shown in Figure 254.
Examples of break frames are given on Figure 255, where we suppose that LBDL=1 (11-bit
break length), and M=0 (8-bit data).

Figure 254. Break detection in LIN mode (11-bit break length - LBDL bit is set)

Case 1: break signal not long enough => break discarded, LBDF is not set

Break frame
RX line

Capture strobe

Break state
Idle Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 Bit9 Bit10 Idle
machine
Read samples 0 0 0 0 0 0 0 0 0 0 1

Case 2: break signal just long enough => break detected, LBDF is set

Break frame
RX line
Delimiter is immediate
Capture strobe

Break state
Idle Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 Bit9 B10 Idle
machine
Read samples 0 0 0 0 0 0 0 0 0 0 0

LBDF

Case 3: break signal long enough => break detected, LBDF is set

Break frame
RX line

Capture strobe

Break state
Idle Bit0 Bit1 Bit2 Bit3 Bit4 Bit5 Bit6 Bit7 Bit8 Bit9 Bit10 wait delimiter Idle
machine
Read samples 0 0 0 0 0 0 0 0 0 0 0

LBDF

MSv31156V1

RM0402 Rev 6 787/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Figure 255. Break detection in LIN mode vs. Framing error detection
Case 1: break occurring after an Idle

RX line data 1 IDLE BREAK data 2 (0x55) data 3 (header)

1 data time 1 data time

RXNE /FE

LBDF

Case 2: break occurring while data is being received

RX line data 1 data2 BREAK data 2 (0x55) data 3 (header)

1 data time 1 data time

RXNE /FE

LBDF

MSv31157V1

25.4.9 USART synchronous mode

The synchronous mode is selected by writing the CLKEN bit in the USART_CR2 register to
1. In synchronous mode, the following bits must be kept cleared:
• LINEN bit in the USART_CR2 register,
• SCEN, HDSEL and IREN bits in the USART_CR3 register.
The USART allows the user to control a bidirectional synchronous serial communications in
master mode. The SCLK pin is the output of the USART transmitter clock. No clock pulses
are sent to the SCLK pin during start bit and stop bit. Depending on the state of the LBCL bit
in the USART_CR2 register clock pulses will or will not be generated during the last valid
data bit (address mark). The CPOL bit in the USART_CR2 register allows the user to select
the clock polarity, and the CPHA bit in the USART_CR2 register allows the user to select the
phase of the external clock (see Figure 256, Figure 257 and Figure 258).
During the Idle state, preamble and send break, the external SCLK clock is not activated.
In synchronous mode the USART transmitter works exactly like in asynchronous mode. But
as SCLK is synchronized with TX (according to CPOL and CPHA), the data on TX is
synchronous.
In this mode the USART receiver works in a different manner compared to the
asynchronous mode. If RE=1, the data is sampled on SCLK (rising or falling edge,
depending on CPOL and CPHA), without any oversampling. A setup and a hold time (that
depends on the baud rate: 1/16 bit time) must be respected.
Note: The SCLK pin works in conjunction with the TX pin. Thus, the clock is provided only if the
transmitter is enabled (TE=1) and a data is being transmitted (the data register USART_DR

788/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

has been written). This means that it is not possible to receive a synchronous data without
transmitting data.
The LBCL, CPOL and CPHA bits have to be selected when both the transmitter and the
receiver are disabled (TE=RE=0) to ensure that the clock pulses function correctly. These
bits should not be changed while the transmitter or the receiver is enabled.
It is advised that TE and RE are set in the same instruction in order to minimize the setup
and the hold time of the receiver.
The USART supports master mode only: it cannot receive or send data related to an input
clock (SCLK is always an output).

Figure 256. USART example of synchronous transmission

RX Data out
TX Data in
Synchronous device
USART
(e.g. slave SPI)

SCLK Clock

MSv31158V1

Figure 257. USART data clock timing diagram (M=0)

Idle or preceding M bits = 00 (8 data bits) Idle or next

transmission Start Stop transmission

Clock (CPOL=0, CPHA=0) *

Clock (CPOL=0, CPHA=1)

*

Clock (CPOL=1, CPHA=0) *

Clock (CPOL=1, CPHA=1) *

Data on TX
0 1 2 3 4 5 6 7
(from master)
Start LSB MSB Stop
Data on RX
0 1 2 3 4 5 6 7
(from slave)
LSB MSB
*
Capture strobe
*LBCL bit controls last data pulse

MSv34709V2

RM0402 Rev 6 789/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Figure 258. USART data clock timing diagram (M=1)

Idle or
Idle or next
preceding Start M bits =01 (9 data bits) Stop
transmission
transmission

Clock (CPOL=0,
CPHA=0) *
Clock (CPOL=0,
CPHA=1) *

Clock (CPOL=1, *
CPHA=0)
Clock (CPOL=1, *
CPHA=1)

Data on TX
0 1 2 3 4 5 6 7 8
(from master)
Start LSB MSB Stop
Data on RX
0 1 2 3 4 5 6 7 8
(from slave)
LSB MSB
Capture *
strobe
*LBCL bit controls last data pulse

MSv34710V1

Figure 259. RX data setup/hold time

SCLK
(capture strobe on SCLK
rising edge in this example)

Data on RX (from slave) Valid DATA bit

tSETUP tHOLD

tSETUP=tHOLD 1/16 bit time

MSv31161V1

Note: The function of SCLK is different in Smartcard mode. Refer to the Smartcard mode chapter
for more details.

25.4.10 Single-wire half-duplex communication

The single-wire half-duplex mode is selected by setting the HDSEL bit in the USART_CR3
register. In this mode, the following bits must be kept cleared:
• LINEN and CLKEN bits in the USART_CR2 register,
• SCEN and IREN bits in the USART_CR3 register.
The USART can be configured to follow a single-wire half-duplex protocol where the TX and
RX lines are internally connected. The selection between half- and full-duplex
communication is made with a control bit ‘HALF DUPLEX SEL’ (HDSEL in USART_CR3).

790/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

As soon as HDSEL is written to 1:

• the TX and RX lines are internally connected
• the RX pin is no longer used
• the TX pin is always released when no data is transmitted. Thus, it acts as a standard
I/O in idle or in reception. It means that the I/O must be configured so that TX is
configured as floating input (or output high open-drain) when not driven by the USART.
Apart from this, the communications are similar to what is done in normal USART mode.
The conflicts on the line must be managed by the software (by the use of a centralized
arbiter, for instance). In particular, the transmission is never blocked by hardware and
continue to occur as soon as a data is written in the data register while the TE bit is set.

25.4.11 Smartcard
The Smartcard mode is selected by setting the SCEN bit in the USART_CR3 register. In
smartcard mode, the following bits must be kept cleared:
• LINEN bit in the USART_CR2 register,
• HDSEL and IREN bits in the USART_CR3 register.
Moreover, the CLKEN bit may be set in order to provide a clock to the smartcard.
The Smartcard interface is designed to support asynchronous protocol Smartcards as
defined in the ISO 7816-3 standard. The USART should be configured as:
• 8 bits plus parity: where M=1 and PCE=1 in the USART_CR1 register
• 1.5 stop bits when transmitting and receiving: where STOP=11 in the USART_CR2
register.
Note: It is also possible to choose 0.5 stop bit for receiving but it is recommended to use 1.5 stop
bits for both transmitting and receiving to avoid switching between the two configurations.
Figure 260 shows examples of what can be seen on the data line with and without parity
error.

Figure 260. ISO 7816-3 asynchronous protocol

Without Parity error

Guard time
S 0 1 2 3 4 5 6 7 p
Start bit

WithParity error
Guard time
S 0 1 2 3 4 5 6 7 p

Start bit
Line pulled low by receiver
during stop in case of parity error

MSv31162V1

When connected to a Smartcard, the TX output of the USART drives a bidirectional line that
is also driven by the Smartcard. The TX pin must be configured as open-drain.
Smartcard is a single wire half duplex communication protocol.
• Transmission of data from the transmit shift register is guaranteed to be delayed by a
minimum of 1/2 baud clock. In normal operation a full transmit shift register will start

RM0402 Rev 6 791/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

shifting on the next baud clock edge. In Smartcard mode this transmission is further
delayed by a guaranteed 1/2 baud clock.
• If a parity error is detected during reception of a frame programmed with a 0.5 or 1.5
stop bit period, the transmit line is pulled low for a baud clock period after the
completion of the receive frame. This is to indicate to the Smartcard that the data
transmitted to USART has not been correctly received. This NACK signal (pulling
transmit line low for 1 baud clock) will cause a framing error on the transmitter side
(configured with 1.5 stop bits). The application can handle re-sending of data according
to the protocol. A parity error is ‘NACK’ed by the receiver if the NACK control bit is set,
otherwise a NACK is not transmitted.
• The assertion of the TC flag can be delayed by programming the Guard Time register.
In normal operation, TC is asserted when the transmit shift register is empty and no
further transmit requests are outstanding. In Smartcard mode an empty transmit shift
register triggers the guard time counter to count up to the programmed value in the
Guard Time register. TC is forced low during this time. When the guard time counter
reaches the programmed value TC is asserted high.
• The de-assertion of TC flag is unaffected by Smartcard mode.
• If a framing error is detected on the transmitter end (due to a NACK from the receiver),
the NACK will not be detected as a start bit by the receive block of the transmitter.
According to the ISO protocol, the duration of the received NACK can be 1 or 2 baud
clock periods.
• On the receiver side, if a parity error is detected and a NACK is transmitted the receiver
will not detect the NACK as a start bit.
Note: A break character is not significant in Smartcard mode. A 0x00 data with a framing error will
be treated as data and not as a break.
No Idle frame is transmitted when toggling the TE bit. The Idle frame (as defined for the
other configurations) is not defined by the ISO protocol.
Figure 261 details how the NACK signal is sampled by the USART. In this example the
USART is transmitting a data and is configured with 1.5 stop bits. The receiver part of the
USART is enabled in order to check the integrity of the data and the NACK signal.

Figure 261. Parity error detection using the 1.5 stop bits
Bit 7 Parity bit 1.5 Stop bit

1 bit time 1.5 bit time

Sampling at Sampling at
8th, 9th, 10th 16th, 17th, 18th

0.5 bit
time

Sampling at Sampling at
8th, 9th, 10th 8th, 9th, 10th
MS37359V1

The USART can provide a clock to the smartcard through the SCLK output. In smartcard
mode, SCLK is not associated to the communication but is simply derived from the internal
peripheral input clock through a 5-bit prescaler. The division ratio is configured in the

792/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

prescaler register USART_GTPR. SCLK frequency can be programmed from fCK/2 to

fCK/62, where fCK is the peripheral input clock.

25.4.12 IrDA SIR ENDEC block

The IrDA mode is selected by setting the IREN bit in the USART_CR3 register. In IrDA
mode, the following bits must be kept cleared:
• LINEN, STOP and CLKEN bits in the USART_CR2 register,
• SCEN and HDSEL bits in the USART_CR3 register.
The IrDA SIR physical layer specifies use of a Return to Zero, Inverted (RZI) modulation
scheme that represents logic 0 as an infrared light pulse (see Figure 262).
The SIR Transmit encoder modulates the Non Return to Zero (NRZ) transmit bit stream
output from USART. The output pulse stream is transmitted to an external output driver and
infrared LED. USART supports only bit rates up to 115.2Kbps for the SIR ENDEC. In normal
mode the transmitted pulse width is specified as 3/16 of a bit period.
The SIR receive decoder demodulates the return-to-zero bit stream from the infrared
detector and outputs the received NRZ serial bit stream to USART. The decoder input is
normally HIGH (marking state) in the Idle state. The transmit encoder output has the
opposite polarity to the decoder input. A start bit is detected when the decoder input is low.
• IrDA is a half duplex communication protocol. If the Transmitter is busy (i.e. the USART
is sending data to the IrDA encoder), any data on the IrDA receive line will be ignored
by the IrDA decoder and if the Receiver is busy (USART is receiving decoded data
from the USART), data on the TX from the USART to IrDA will not be encoded by IrDA.
While receiving data, transmission should be avoided as the data to be transmitted
could be corrupted.
• A ‘0 is transmitted as a high pulse and a ‘1 is transmitted as a ‘0. The width of the pulse
is specified as 3/16th of the selected bit period in normal mode (see Figure 263).
• The SIR decoder converts the IrDA compliant receive signal into a bit stream for
USART.
• The SIR receive logic interprets a high state as a logic one and low pulses as logic
zeros.
• The transmit encoder output has the opposite polarity to the decoder input. The SIR
output is in low state when Idle.
• The IrDA specification requires the acceptance of pulses greater than 1.41 us. The
acceptable pulse width is programmable. Glitch detection logic on the receiver end
filters out pulses of width less than 2 PSC periods (PSC is the prescaler value
programmed in the IrDA low-power Baud Register, USART_GTPR). Pulses of width
less than 1 PSC period are always rejected, but those of width greater than one and
less than two periods may be accepted or rejected, those greater than 2 periods will be
accepted as a pulse. The IrDA encoder/decoder doesn’t work when PSC=0.
• The receiver can communicate with a low-power transmitter.
• In IrDA mode, the STOP bits in the USART_CR2 register must be configured to “1 stop
bit”.

RM0402 Rev 6 793/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

IrDA low-power mode

Transmitter:
In low-power mode the pulse width is not maintained at 3/16 of the bit period. Instead, the
width of the pulse is 3 times the low-power baud rate that can be a minimum of 1.42 MHz.
Generally this value is 1.8432 MHz (1.42 MHz < PSC< 2.12 MHz). A low-power mode
programmable divisor divides the system clock to achieve this value.
Receiver:
Receiving in low-power mode is similar to receiving in normal mode. For glitch detection the
USART should discard pulses of duration shorter than 1/PSC. A valid low is accepted only if
its duration is greater than 2 periods of the IrDA low-power Baud clock (PSC value in
USART_GTPR).
Note: A pulse of width less than two and greater than one PSC period(s) may or may not be
rejected.
The receiver set up time should be managed by software. The IrDA physical layer
specification specifies a minimum of 10 ms delay between transmission and reception (IrDA
is a half duplex protocol).

Figure 262. IrDA SIR ENDEC- block diagram

SIREN
TX
OR USART_TX

SIR
Transmit IrDA_OUT
Encoder
USART

SIR
RX
Receive IrDA_IN
DEcoder

USART_RX

MSv31164V2

Figure 263. IrDA data modulation (3/16) -Normal mode

Start Stop
bit bit
0 1 0 1 0 0 1 1 0 1
TX

IrDA_OUT
Bit period 3/16
IrDA_IN

RX 0 1 0 1 0 0 1 1 0 1

MSv31165V1

794/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

25.4.13 Continuous communication using DMA

The USART is capable of continuous communication using the DMA. The DMA requests for
Rx buffer and Tx buffer are generated independently.

Transmission using DMA

DMA mode can be enabled for transmission by setting DMAT bit in the USART_CR3
register. Data is loaded from a SRAM area configured using the DMA peripheral (refer to the
DMA specification) to the USART_DR register whenever the TXE bit is set. To map a DMA
channel for USART transmission, use the following procedure (x denotes the channel
number):
1. Write the USART_DR register address in the DMA control register to configure it as the
destination of the transfer. The data will be moved to this address from memory after
each TXE event.
2. Write the memory address in the DMA control register to configure it as the source of
the transfer. The data will be loaded into the USART_DR register from this memory
area after each TXE event.
3. Configure the total number of bytes to be transferred to the DMA control register.
4. Configure the channel priority in the DMA register
5. Configure DMA interrupt generation after half/ full transfer as required by the
application.
6. Clear the TC bit in the SR register by writing 0 to it.
7. Activate the channel in the DMA register.
When the number of data transfers programmed in the DMA Controller is reached, the DMA
controller generates an interrupt on the DMA channel interrupt vector.
In transmission mode, once the DMA has written all the data to be transmitted (the TCIF flag
is set in the DMA_ISR register), the TC flag can be monitored to make sure that the USART
communication is complete. This is required to avoid corrupting the last transmission before
disabling the USART or entering the Stop mode. The software must wait until TC=1. The TC
flag remains cleared during all data transfers and it is set by hardware at the last frame end
of transmission.

RM0402 Rev 6 795/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Figure 264. Transmission using DMA

Idle preamble Frame 1 Frame 2 Frame 3
TX line

Set by hardware Set by hardware

TXE flag cleared by DMA read cleared by DMA read Set by hardware

DMA request Ignored by the DMA because

the transfer is complete

USART_TDR F1 F2 F3

TC flag Set by
hardware
DMA writes
USART_TDR
Cleared
DMA TCIF flag by
Set by hardware software
(transfer
complete)

Software The DMA

configures DMA DMA writes DMA writes DMA writes transfer is
to send 3 data F1 into F2 into F3 into complete Software waits until TC=1
blocks and USART_TDR USART_TDR USART_TDR (TCIF=1 in
enables USART DMA_ISR)

ai17192b

Reception using DMA

DMA mode can be enabled for reception by setting the DMAR bit in USART_CR3 register.
Data is loaded from the USART_DR register to a SRAM area configured using the DMA
peripheral (refer to the DMA specification) whenever a data byte is received. To map a DMA
channel for USART reception, use the following procedure:
1. Write the USART_DR register address in the DMA control register to configure it as the
source of the transfer. The data will be moved from this address to the memory after
each RXNE event.
2. Write the memory address in the DMA control register to configure it as the destination
of the transfer. The data will be loaded from USART_DR to this memory area after each
RXNE event.
3. Configure the total number of bytes to be transferred in the DMA control register.
4. Configure the channel priority in the DMA control register
5. Configure interrupt generation after half/ full transfer as required by the application.
6. Activate the channel in the DMA control register.
When the number of data transfers programmed in the DMA Controller is reached, the DMA
controller generates an interrupt on the DMA channel interrupt vector. The DMAR bit should
be cleared by software in the USART_CR3 register during the interrupt subroutine.

796/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Figure 265. Reception using DMA

Frame 1 Frame 2 Frame 3

TX line

Set by hardware
RXNE flag cleared by DMA read

DMA request

USART_TDR F1 F2 F3

DMA reads
USART_TDR
Cleared
DMA TCIF flag Set by hardware by
(transfer complete) software

Software configures the The DMA transfer

DMA to receive 3 data DMA reads F1 DMA reads F2 DMA reads F3 is complete
blocks and enables from USART_TDR from USART_TDR from USART_TDR (TCIF=1 in
the USART DMA_ISR)
ai17193c

Error flagging and interrupt generation in multibuffer communication

In case of multibuffer communication if any error occurs during the transaction the error flag
will be asserted after the current byte. An interrupt will be generated if the interrupt enable
flag is set. For framing error, overrun error and noise flag that are asserted with RXNE in
case of single byte reception, there will be separate error flag interrupt enable bit (EIE bit in
the USART_CR3 register), which if set will issue an interrupt after the current byte with
either of these errors.

25.4.14 Hardware flow control

It is possible to control the serial data flow between 2 devices by using the nCTS input and
the nRTS output. The Figure 266 shows how to connect 2 devices in this mode:

Figure 266. Hardware flow control between 2 USARTs

USART 1 USART 2
TX RX
TX circuit RX circuit
nCTS nRTS

RX TX
RX circuit TX circuit
nRTS nCTS

MSv31169V1

RTS and CTS flow control can be enabled independently by writing respectively RTSE and
CTSE bits to 1 (in the USART_CR3 register).

RM0402 Rev 6 797/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

RTS flow control

If the RTS flow control is enabled (RTSE=1), then nRTS is asserted (tied low) as long as the
USART receiver is ready to receive a new data. When the receive register is full, nRTS is
deasserted, indicating that the transmission is expected to stop at the end of the current
frame. Figure 267 shows an example of communication with RTS flow control enabled.

Figure 267. RTS flow control

Start Stop Start Stop

RX Data 1 Idle Data 2
bit bit bit bit

nRTS

RXNE Data 1 read RXNE

Data 2 can now be transmitted

MSv31168V1

CTS flow control

If the CTS flow control is enabled (CTSE=1), then the transmitter checks the nCTS input
before transmitting the next frame. If nCTS is asserted (tied low), then the next data is
transmitted (assuming that a data is to be transmitted, in other words, if TXE=0), else the
transmission does not occur. When nCTS is deasserted during a transmission, the current
transmission is completed before the transmitter stops.
When CTSE=1, the CTSIF status bit is automatically set by hardware as soon as the nCTS
input toggles. It indicates when the receiver becomes ready or not ready for communication.
An interrupt is generated if the CTSIE bit in the USART_CR3 register is set. The figure
below shows an example of communication with CTS flow control enabled.

Figure 268. CTS flow control

CTS CTS

nCTS

Transmit data register

TDR Data 2 empty Data 3 empty

Stop Start Stop Start

TX Data 1 Data 2 Idle Data 3
bit bit bit bit

Writing data 3 in TDR Transmission of Data 3 is

delayed until nCTS = 0
MSv31167V1

798/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Note: Special behavior of break frames: when the CTS flow is enabled, the transmitter does not
check the nCTS input state to send a break.

25.5 USART interrupts

Table 153. USART interrupt requests
Interrupt event Event flag Enable control bit

Transmit Data Register Empty TXE TXEIE

CTS flag CTS CTSIE
Transmission Complete TC TCIE
Received Data Ready to be Read RXNE
RXNEIE
Overrun Error Detected ORE
Idle Line Detected IDLE IDLEIE
Parity Error PE PEIE
Break Flag LBD LBDIE
Noise Flag, Overrun error and Framing Error
NF or ORE or FE EIE
in multibuffer communication

The USART interrupt events are connected to the same interrupt vector (see Figure 269).
• During transmission: Transmission Complete, Clear to Send or Transmit Data Register
empty interrupt.
• While receiving: Idle Line detection, Overrun error, Receive Data register not empty,
Parity error, LIN break detection, Noise Flag (only in multi buffer communication) and
Framing Error (only in multi buffer communication).
These events generate an interrupt if the corresponding Enable Control Bit is set.

RM0402 Rev 6 799/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Figure 269. USART interrupt mapping diagram

TC
TCIE

TXE
TXEIE

CTS
CTSIE

IDLE
IDLEIE

RXNEIE
ORE

RXNEIE
RXNE USART
interrupt
PE
PEIE

LBD
LBDIE

FE
NE
ORE EIE
DMAR

MS35853V1

25.6 USART registers

Refer to Section 1.2 on page 45 for a list of abbreviations used in register descriptions.
The peripheral registers have to be accessed by words (32 bits).

25.6.1 Status register (USART_SR)

Address offset: 0x00
Reset value: 0x00C0 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. CTS LBD TXE TC RXNE IDLE ORE NF FE PE
rc_w0 rc_w0 r rc_w0 rc_w0 r r r r r

800/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Bits 31:10 Reserved, must be kept at reset value

Bit 9 CTS: CTS flag
This bit is set by hardware when the nCTS input toggles, if the CTSE bit is set. It is cleared
by software (by writing it to 0). An interrupt is generated if CTSIE=1 in the USART_CR3
register.
0: No change occurred on the nCTS status line
1: A change occurred on the nCTS status line
Note: This bit is not available for UART4 & UART5.
Bit 8 LBD: LIN break detection flag
This bit is set by hardware when the LIN break is detected. It is cleared by software (by
writing it to 0). An interrupt is generated if LBDIE = 1 in the USART_CR2 register.
0: LIN Break not detected
1: LIN break detected
Note: An interrupt is generated when LBD=1 if LBDIE=1
Bit 7 TXE: Transmit data register empty
This bit is set by hardware when the content of the TDR register has been transferred into
the shift register. An interrupt is generated if the TXEIE bit =1 in the USART_CR1 register. It
is cleared by a write to the USART_DR register.
0: Data is not transferred to the shift register
1: Data is transferred to the shift register)
Note: This bit is used during single buffer transmission.
Bit 6 TC: Transmission complete
This bit is set by hardware if the transmission of a frame containing data is complete and if
TXE is set. An interrupt is generated if TCIE=1 in the USART_CR1 register. It is cleared by
a software sequence (a read from the USART_SR register followed by a write to the
USART_DR register). The TC bit can also be cleared by writing a '0' to it. This clearing
sequence is recommended only for multibuffer communication.
0: Transmission is not complete
1: Transmission is complete
Bit 5 RXNE: Read data register not empty
This bit is set by hardware when the content of the RDR shift register has been transferred
to the USART_DR register. An interrupt is generated if RXNEIE=1 in the USART_CR1
register. It is cleared by a read to the USART_DR register. The RXNE flag can also be
cleared by writing a zero to it. This clearing sequence is recommended only for multibuffer
communication.
0: Data is not received
1: Received data is ready to be read.
Bit 4 IDLE: IDLE line detected
This bit is set by hardware when an Idle Line is detected. An interrupt is generated if the
IDLEIE=1 in the USART_CR1 register. It is cleared by a software sequence (an read to the
USART_SR register followed by a read to the USART_DR register).
0: No Idle Line is detected
1: Idle Line is detected
Note: The IDLE bit will not be set again until the RXNE bit has been set itself (i.e. a new idle
line occurs).

RM0402 Rev 6 801/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Bit 3 ORE: Overrun error

This bit is set by hardware when the word currently being received in the shift register is
ready to be transferred into the RDR register while RXNE=1. An interrupt is generated if
RXNEIE=1 in the USART_CR1 register. It is cleared by a software sequence (an read to the
USART_SR register followed by a read to the USART_DR register).
0: No Overrun error
1: Overrun error is detected
Note: When this bit is set, the RDR register content will not be lost but the shift register will be
overwritten. An interrupt is generated on ORE flag in case of Multi Buffer
communication if the EIE bit is set.
Bit 2 NF: Noise detected flag
This bit is set by hardware when noise is detected on a received frame. It is cleared by a
software sequence (an read to the USART_SR register followed by a read to the
USART_DR register).
0: No noise is detected
1: Noise is detected
Note: This bit does not generate interrupt as it appears at the same time as the RXNE bit that
itself generates an interrupting interrupt is generated on NF flag in case of Multi Buffer
communication if the EIE bit is set.
Note: When the line is noise-free, the NF flag can be disabled by programming the ONEBIT
bit to 1 to increase the USART tolerance to deviations (Refer to Section 25.4.5: USART
receiver tolerance to clock deviation on page 782).
Bit 1 FE: Framing error
This bit is set by hardware when a de-synchronization, excessive noise or a break character
is detected. It is cleared by a software sequence (an read to the USART_SR register
followed by a read to the USART_DR register).
0: No Framing error is detected
1: Framing error or break character is detected
Note: This bit does not generate interrupt as it appears at the same time as the RXNE bit that
itself generates an interrupt. If the word currently being transferred causes both frame
error and overrun error, it will be transferred and only the ORE bit will be set.
An interrupt is generated on FE flag in case of Multi Buffer communication if the EIE bit
is set.
Bit 0 PE: Parity error
This bit is set by hardware when a parity error occurs in receiver mode. It is cleared by a
software sequence (a read from the status register followed by a read or write access to the
USART_DR data register). The software must wait for the RXNE flag to be set before
clearing the PE bit.
An interrupt is generated if PEIE = 1 in the USART_CR1 register.
0: No parity error
1: Parity error

802/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

25.6.2 Data register (USART_DR)

Address offset: 0x04
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. DR[8:0]
rw rw rw rw rw rw rw rw rw

Bits 31:9 Reserved, must be kept at reset value

Bits 8:0 DR[8:0]: Data value
Contains the Received or Transmitted data character, depending on whether it is read from
or written to.
The Data register performs a double function (read and write) since it is composed of two
registers, one for transmission (TDR) and one for reception (RDR)
The TDR register provides the parallel interface between the internal bus and the output shift
register (see Figure 1).
The RDR register provides the parallel interface between the input shift register and the
internal bus.
When transmitting with the parity enabled (PCE bit set to 1 in the USART_CR1 register), the
value written in the MSB (bit 7 or bit 8 depending on the data length) has no effect because
it is replaced by the parity.
When receiving with the parity enabled, the value read in the MSB bit is the received parity
bit.

25.6.3 Baud rate register (USART_BRR)

Note: The baud counters stop counting if the TE or RE bits are disabled respectively.
Address offset: 0x08
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DIV_Mantissa[11:0] DIV_Fraction[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value

Bits 15:4 DIV_Mantissa[11:0]: mantissa of USARTDIV
These 12 bits define the mantissa of the USART Divider (USARTDIV)
Bits 3:0 DIV_Fraction[3:0]: fraction of USARTDIV
These 4 bits define the fraction of the USART Divider (USARTDIV). When OVER8=1, the
DIV_Fraction3 bit is not considered and must be kept cleared.

RM0402 Rev 6 803/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

25.6.4 Control register 1 (USART_CR1)

Address offset: 0x0C
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OVER8 Res. UE M WAKE PCE PS PEIE TXEIE TCIE RXNEIE IDLEIE TE RE RWU SBK
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value

Bit 15 OVER8: Oversampling mode
0: oversampling by 16
1: oversampling by 8
Note: Oversampling by 8 is not available in the Smartcard, IrDA and LIN modes: when
SCEN=1,IREN=1 or LINEN=1 then OVER8 is forced to ‘0 by hardware.
Bit 14 Reserved, must be kept at reset value
Bit 13 UE: USART enable
When this bit is cleared the USART prescalers and outputs are stopped and the end of the
current
byte transfer in order to reduce power consumption. This bit is set and cleared by software.
0: USART prescaler and outputs disabled
1: USART enabled
Bit 12 M: Word length
This bit determines the word length. It is set or cleared by software.
0: 1 Start bit, 8 Data bits, n Stop bit
1: 1 Start bit, 9 Data bits, n Stop bit
Note: The M bit must not be modified during a data transfer (both transmission and reception)
Bit 11 WAKE: Wakeup method
This bit determines the USART wakeup method, it is set or cleared by software.
0: Idle Line
1: Address Mark
Bit 10 PCE: Parity control enable
This bit selects the hardware parity control (generation and detection). When the parity
control is enabled, the computed parity is inserted at the MSB position (9th bit if M=1; 8th bit
if M=0) and parity is checked on the received data. This bit is set and cleared by software.
Once it is set, PCE is active after the current byte (in reception and in transmission).
0: Parity control disabled
1: Parity control enabled
Bit 9 PS: Parity selection
This bit selects the odd or even parity when the parity generation/detection is enabled (PCE
bit set). It is set and cleared by software. The parity will be selected after the current byte.
0: Even parity
1: Odd parity

804/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Bit 8 PEIE: PE interrupt enable

This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever PE=1 in the USART_SR register
Bit 7 TXEIE: TXE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever TXE=1 in the USART_SR register
Bit 6 TCIE: Transmission complete interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever TC=1 in the USART_SR register
Bit 5 RXNEIE: RXNE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever ORE=1 or RXNE=1 in the USART_SR
register
Bit 4 IDLEIE: IDLE interrupt enable
This bit is set and cleared by software.
0: Interrupt is inhibited
1: An USART interrupt is generated whenever IDLE=1 in the USART_SR register
Bit 3 TE: Transmitter enable
This bit enables the transmitter. It is set and cleared by software.
0: Transmitter is disabled
1: Transmitter is enabled
Note: 1: During transmission, a “0” pulse on the TE bit (“0” followed by “1”) sends a preamble
(idle line) after the current word, except in smartcard mode.
2: When TE is set there is a 1 bit-time delay before the transmission starts.
Bit 2 RE: Receiver enable
This bit enables the receiver. It is set and cleared by software.
0: Receiver is disabled
1: Receiver is enabled and begins searching for a start bit
Bit 1 RWU: Receiver wakeup
This bit determines if the USART is in mute mode or not. It is set and cleared by software
and can be cleared by hardware when a wakeup sequence is recognized.
0: Receiver in active mode
1: Receiver in mute mode
Note: 1: Before selecting Mute mode (by setting the RWU bit) the USART must first receive a
data byte, otherwise it cannot function in Mute mode with wakeup by Idle line detection.
2: In Address Mark Detection wakeup configuration (WAKE bit=1) the RWU bit cannot
be modified by software while the RXNE bit is set.
Bit 0 SBK: Send break
This bit set is used to send break characters. It can be set and cleared by software. It should
be set by software, and will be reset by hardware during the stop bit of break.
0: No break character is transmitted
1: Break character will be transmitted

RM0402 Rev 6 805/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

25.6.5 Control register 2 (USART_CR2)

Address offset: 0x10
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. LINEN STOP[1:0] CLKEN CPOL CPHA LBCL Res. LBDIE LBDL Res. ADD[3:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:15 Reserved, must be kept at reset value

Bit 14 LINEN: LIN mode enable
This bit is set and cleared by software.
0: LIN mode disabled
1: LIN mode enabled
The LIN mode enables the capability to send LIN Synch Breaks (13 low bits) using the SBK bit in
the USART_CR1 register, and to detect LIN Sync breaks.
Bits 13:12 STOP: STOP bits
These bits are used for programming the stop bits.
00: 1 Stop bit
01: 0.5 Stop bit
10: 2 Stop bits
11: 1.5 Stop bit
Note: The 0.5 Stop bit and 1.5 Stop bit are not available for UART4 & UART5.
Bit 11 CLKEN: Clock enable
This bit allows the user to enable the SCLK pin.
0: SCLK pin disabled
1: SCLK pin enabled
This bit is not available for UART4 & UART5.
Bit 10 CPOL: Clock polarity
This bit allows the user to select the polarity of the clock output on the SCLK pin in synchronous
mode. It works in conjunction with the CPHA bit to produce the desired clock/data relationship
0: Steady low value on SCLK pin outside transmission window.
1: Steady high value on SCLK pin outside transmission window.
This bit is not available for UART4 & UART5.
Bit 9 CPHA: Clock phase
This bit allows the user to select the phase of the clock output on the SCLK pin in synchronous
mode. It works in conjunction with the CPOL bit to produce the desired clock/data relationship (see
figures 257 to 258)
0: The first clock transition is the first data capture edge
1: The second clock transition is the first data capture edge
Note: This bit is not available for UART4 & UART5.

806/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Bit 8 LBCL: Last bit clock pulse

This bit allows the user to select whether the clock pulse associated with the last data bit
transmitted (MSB) has to be output on the SCLK pin in synchronous mode.
0: The clock pulse of the last data bit is not output to the SCLK pin
1: The clock pulse of the last data bit is output to the SCLK pin
Note: 1: The last bit is the 8th or 9th data bit transmitted depending on the 8 or 9 bit format selected
by the M bit in the USART_CR1 register.
2: This bit is not available for UART4 & UART5.
Bit 7 Reserved, must be kept at reset value
Bit 6 LBDIE: LIN break detection interrupt enable
Break interrupt mask (break detection using break delimiter).
0: Interrupt is inhibited
1: An interrupt is generated whenever LBD=1 in the USART_SR register
Bit 5 LBDL: lin break detection length
This bit is for selection between 11 bit or 10 bit break detection.
0: 10-bit break detection
1: 11-bit break detection
Bit 4 Reserved, must be kept at reset value
Bits 3:0 ADD[3:0]: Address of the USART node
This bit-field gives the address of the USART node.
This is used in multiprocessor communication during mute mode, for wake up with address mark
detection.

Note: These 3 bits (CPOL, CPHA, LBCL) should not be written while the transmitter is enabled.

25.6.6 Control register 3 (USART_CR3)

Address offset: 0x14
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. ONEBIT CTSIE CTSE RTSE DMAT DMAR SCEN NACK HDSEL IRLP IREN EIE

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:12 Reserved, must be kept at reset value

Bit 11 ONEBIT: One sample bit method enable
This bit allows the user to select the sample method. When the one sample bit method is
selected the noise detection flag (NF) is disabled.
0: Three sample bit method
1: One sample bit method
Bit 10 CTSIE: CTS interrupt enable
0: Interrupt is inhibited
1: An interrupt is generated whenever CTS=1 in the USART_SR register
Note: This bit is not available for UART4 & UART5.

RM0402 Rev 6 807/1163

810
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

Bit 9 CTSE: CTS enable

0: CTS hardware flow control disabled
1: CTS mode enabled, data is only transmitted when the nCTS input is asserted (tied to 0).
If the nCTS input is deasserted while a data is being transmitted, then the transmission is
completed before stopping. If a data is written into the data register while nCTS is
deasserted, the transmission is postponed until nCTS is asserted.
Note: This bit is not available for UART4 & UART5.
Bit 8 RTSE: RTS enable
0: RTS hardware flow control disabled
1: RTS interrupt enabled, data is only requested when there is space in the receive buffer.
The transmission of data is expected to cease after the current character has been
transmitted. The nRTS output is asserted (tied to 0) when a data can be received.
Note: This bit is not available for UART4 & UART5.
Bit 7 DMAT: DMA enable transmitter
This bit is set/reset by software
1: DMA mode is enabled for transmission.
0: DMA mode is disabled for transmission.
Bit 6 DMAR: DMA enable receiver
This bit is set/reset by software
1: DMA mode is enabled for reception
0: DMA mode is disabled for reception
Bit 5 SCEN: Smartcard mode enable
This bit is used for enabling Smartcard mode.
0: Smartcard Mode disabled
1: Smartcard Mode enabled
Note: This bit is not available for UART4 & UART5.
Bit 4 NACK: Smartcard NACK enable
0: NACK transmission in case of parity error is disabled
1: NACK transmission during parity error is enabled
Note: This bit is not available for UART4 & UART5.
Bit 3 HDSEL: Half-duplex selection
Selection of Single-wire Half-duplex mode
0: Half duplex mode is not selected
1: Half duplex mode is selected
Bit 2 IRLP: IrDA low-power
This bit is used for selecting between normal and low-power IrDA modes
0: Normal mode
1: Low-power mode
Bit 1 IREN: IrDA mode enable
This bit is set and cleared by software.
0: IrDA disabled
1: IrDA enabled

808/1163 RM0402 Rev 6

RM0402 Universal synchronous receiver transmitter (USART) /universal asynchronous receiver

Bit 0 EIE: Error interrupt enable

Error Interrupt Enable Bit is required to enable interrupt generation in case of a framing
error, overrun error or noise flag (FE=1 or ORE=1 or NF=1 in the USART_SR register) in
case of Multi Buffer Communication (DMAR=1 in the USART_CR3 register).
0: Interrupt is inhibited
1: An interrupt is generated whenever DMAR=1 in the USART_CR3 register and FE=1 or
ORE=1 or NF=1 in the USART_SR register.

25.6.7 Guard time and prescaler register (USART_GTPR)

Address offset: 0x18
Reset value: 0x0000 0000
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GT[7:0] PSC[7:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value

Bits 15:8 GT[7:0]: Guard time value
This bit-field gives the Guard time value in terms of number of baud clocks.
This is used in Smartcard mode. The Transmission Complete flag is set after this guard time
value.
Note: This bit is not available for UART4 & UART5.
Bits 7:0 PSC[7:0]: Prescaler value
– In IrDA Low-power mode:
PSC[7:0] = IrDA Low-Power Baud Rate
Used for programming the prescaler for dividing the system clock to achieve the low-power
frequency:
The source clock is divided by the value given in the register (8 significant bits):
00000000: Reserved - do not program this value
00000001: divides the source clock by 1
00000010: divides the source clock by 2
...
– In normal IrDA mode: PSC must be set to 00000001.
– In smartcard mode:
PSC[4:0]: Prescaler value
Used for programming the prescaler for dividing the system clock to provide the smartcard
clock.
The value given in the register (5 significant bits) is multiplied by 2 to give the division factor
of the source clock frequency:
00000: Reserved - do not program this value
00001: divides the source clock by 2
00010: divides the source clock by 4
00011: divides the source clock by 6
...
Note: 1: Bits [7:5] have no effect if Smartcard mode is used.
2: This bit is not available for UART4 & UART5.

RM0402 Rev 6 809/1163

810
 0x18
0x14
0x10
0x08
0x04
0x00

0x0C
Offset
25.6.8

810/1163
Register

Reset value
Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
USART_SR

USART_DR

USART_CR3
USART_CR2
USART_CR1
USART_BRR

USART_GTPR
Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. 28
Res. Res. Res. Res. Res. Res. Res. 27
Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. 25
USART register map

Res. Res. Res. Res. Res. Res. Res. 24

Res. Res. Res. Res. Res. Res. Res. 23
Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. 21
Res. Res. Res. Res. Res. Res. Res. 20
Res. Res. Res. Res. Res. Res. Res. 19
Res. Res. Res. Res. Res. Res. Res. 18
Res. Res. Res. Res. Res. Res. Res. 17

RM0402 Rev 6
Res. Res. Res. Res. Res. Res. Res. 16

0
0
0

Res. Res. OVER8 Res. Res. 15

0
0
0

Res. LINEN Res. Res. Res. 14

0
0
0
0

Res. UE Res. Res. 13

[1:0]

0
0
0
0

STOP

Res. M Res. Res. 12

0
0
0
0
0

GT[7:0]
ONEBI CLKEN WAKE Res. Res. 11

0
0
0
0
0

CTSIE CPOL PCE Res. Res. 10

Table 154. USART register map and reset values
The table below gives the USART register map and reset values.

0
0
0
0
0
0

CTSE CPHA PS Res. CTS 9

Refer to Section 2.2 on page 49 for the register boundary addresses.

0
0
0
0
0
0
0

RTSE LBCL PEIE LBD 8

DIV_Mantissa[15:4]

0
0
0
0
0
1

DMAT Res. TXEIE TXE 7

0
0
0
0
0
0
1

DMAR LBDIE TCIE TC 6

0
0
0
0
0
0
0

SCEN LBDL RXNEIE RXNE 5

0
0
0
0
0
0

NACK Res. IDLEIE IDLE 4

DR[8:0]

0
0
0
0
0
0
0

HDSEL TE ORE 3
PSC[7:0]

0
0
0
0
0
0
0

IRLP RE NF 2
[3:0]

0
0
0
0
0
0
0

IREN RWU FE 1
ADD[3:0]

0
0
0
0
0
0
0

DIV_Fraction
Universal synchronous receiver transmitter (USART) /universal asynchronous receiver transmit-

EIE SBK PE 0
RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.1 Introduction
The SPI/I²S interface can be used to communicate with external devices using the SPI
protocol or the I2S audio protocol. SPI or I2S mode is selectable by software. SPI mode is
selected by default after a device reset.
The serial peripheral interface (SPI) protocol supports half-duplex, full-duplex and simplex
synchronous, serial communication with external devices. The interface can be configured
as master and in this case it provides the communication clock (SCK) to the external slave
device. The interface is also capable of operating in multimaster configuration.
The Inter-IC sound (I2S) protocol is also a synchronous serial communication interface. It
can operate in slave or master mode with full-duplex and half-duplex communication.
It can address four different audio standards including the Philips I2S standard, the MSB-
and LSB-justified standards and the PCM standard.

Warning: Since some SPI1 pins may be mapped onto some pins used
by the JTAG interface, you can either map SPI/I2S onto other
pins, disable the JTAG and use the SWD interface prior to
configuring the pins listed as SPI I/Os (when debugging the
application) or disable both JTAG/SWD interfaces (for
standalone applications). For more information on the
configuration of the JTAG/SWD interface pins, please refer to
Section 7.3.2: I/O pin multiplexer and mapping.

RM0402 Rev 6 811/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.1.1 SPI main features

• Master or slave operation
• Full-duplex synchronous transfers on three lines
• Half-duplex synchronous transfer on two lines (with bidirectional data line)
• Simplex synchronous transfers on two lines (with unidirectional data line)
• 8-bit to 16-bit transfer frame format selection
• Multimaster mode capability
• 8 master mode baud rate prescalers up to fPCLK/2.
• Slave mode frequency up to fPCLK/2.
• NSS management by hardware or software for both master and slave: dynamic change
of master/slave operations
• Programmable clock polarity and phase
• Programmable data order with MSB-first or LSB-first shifting
• Dedicated transmission and reception flags with interrupt capability
• SPI bus busy status flag
• SPI Motorola support
• Hardware CRC feature for reliable communication:
– CRC value can be transmitted as last byte in Tx mode
– Automatic CRC error checking for last received byte
• Master mode fault, overrun flags with interrupt capability
• CRC Error flag
• 1-byte/word transmission and reception buffer with DMA capability: Tx and Rx requests

812/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.1.2 SPI extended features

• SPI TI mode support

26.1.3 I2S features

• Full-duplex communication
• Half-duplex communication (only transmitter or receiver)
• Master or slave operations
• 8-bit programmable linear prescaler to reach accurate audio sample frequencies (from
8 kHz to 192 kHz)
• Data format may be 16-bit, 24-bit or 32-bit
• Packet frame is fixed to 16-bit (16-bit data frame) or 32-bit (16-bit, 24-bit, 32-bit data
frame) by audio channel
• Programmable clock polarity (steady state)
• Underrun flag in slave transmission mode, overrun flag in reception mode (master and
slave) and Frame Error Flag in reception and transmitter mode (slave only)
• 16-bit register for transmission and reception with one data register for both channel
sides
• Supported I2S protocols:
– I2S Philips standard
– MSB-Justified standard (Left-Justified)
– LSB-Justified standard (Right-Justified)
– PCM standard (with short and long frame synchronization on 16-bit channel frame
or 16-bit data frame extended to 32-bit channel frame)
• Data direction is always MSB first
• DMA capability for transmission and reception (16-bit wide)
• Master clock can be output to drive an external audio component. Ratio is fixed at
256 × FS (where FS is the audio sampling frequency)
• I2S (I2S1, I2S2, I2S3, I2S4 and I2S5) clock can be derived from an external clock
mapped on the I2S_CKIN pin.

26.2 SPI/I2S implementation

This manual describes the full set of features implemented in SPI1, SPI2, SPI3, SPI4 and
SPI5.

Table 155. STM32F412xx SPI implementation

SPI Features(1) SPI1 SPI2 SPI3 SPI4 SPI5

Hardware CRC calculation X X X X X

I2S mode X X X X X
TI mode X X X X X
1. X = supported.

RM0402 Rev 6 813/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.3 SPI functional description

26.3.1 General description

The SPI allows synchronous, serial communication between the MCU and external devices.
Application software can manage the communication by polling the status flag or using
dedicated SPI interrupt. The main elements of SPI and their interactions are shown in the
following block diagram Figure 270.

Figure 270. SPI block diagram

Address and data bus

Read
Rx
buffer
CRC controller
MOSI
MISO Shift register
LSBFIRST CRCEN
CPOL CRCNEXT
CPHA
DFF
TX
buffer
Write Communication
BIDIOE controller
BIDIMODE
RXOLNY
Baud rate
SCK Internal NSS
generator BR[2:0]

NSS
NSS logic
MSv33711V1

Four I/O pins are dedicated to SPI communication with external devices.
• MISO: Master In / Slave Out data. In the general case, this pin is used to transmit data
in slave mode and receive data in master mode.
• MOSI: Master Out / Slave In data. In the general case, this pin is used to transmit data
in master mode and receive data in slave mode.
• SCK: Serial Clock output pin for SPI masters and input pin for SPI slaves.
• NSS: Slave select pin. Depending on the SPI and NSS settings, this pin can be used to
either:
– select an individual slave device for communication
– synchronize the data frame or
– detect a conflict between multiple masters
See Section 26.3.5: Slave select (NSS) pin management for details.
The SPI bus allows the communication between one master device and one or more slave
devices. The bus consists of at least two wires - one for the clock signal and the other for
synchronous data transfer. Other signals can be added depending on the data exchange
between SPI nodes and their slave select signal management.

814/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.3.2 Communications between one master and one slave

The SPI allows the MCU to communicate using different configurations, depending on the
device targeted and the application requirements. These configurations use 2 or 3 wires
(with software NSS management) or 3 or 4 wires (with hardware NSS management).
Communication is always initiated by the master.

Full-duplex communication
By default, the SPI is configured for full-duplex communication. In this configuration, the
shift registers of the master and slave are linked using two unidirectional lines between the
MOSI and the MISO pins. During SPI communication, data is shifted synchronously on the
SCK clock edges provided by the master. The master transmits the data to be sent to the
slave via the MOSI line and receives data from the slave via the MISO line. When the data
frame transfer is complete (all the bits are shifted) the information between the master and
slave is exchanged.

Figure 271. Full-duplex single master/ single slave application

MISO MISO
Rx shift register Tx shift register
MOSI MOSI
Tx shift register Rx shift register

SPI clock SCK SCK

generator
NSS(1) NSS (1)
Master Slave

MSv39623V1

1. The NSS pins can be used to provide a hardware control flow between master and slave. Optionally, the
pins can be left unused by the peripheral. Then the flow has to be handled internally for both master and
slave. For more details see Section 26.3.5: Slave select (NSS) pin management.

Half-duplex communication
The SPI can communicate in half-duplex mode by setting the BIDIMODE bit in the
SPIx_CR1 register. In this configuration, one single cross connection line is used to link the
shift registers of the master and slave together. During this communication, the data is
synchronously shifted between the shift registers on the SCK clock edge in the transfer
direction selected reciprocally by both master and slave with the BDIOE bit in their
SPIx_CR1 registers. In this configuration, the master’s MISO pin and the slave’s MOSI pin
are free for other application uses and act as GPIOs.

RM0402 Rev 6 815/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 272. Half-duplex single master/ single slave application

(2)
MISO MISO
Rx shift register Tx shift register
(3)
MOSI 1kΩ (2)
Tx shift register MOSI Rx shift register

SPI clock SCK SCK

generator
NSS (1) NSS(1)
Master Slave

MSv39624V1

1. The NSS pins can be used to provide a hardware control flow between master and slave. Optionally, the
pins can be left unused by the peripheral. Then the flow has to be handled internally for both master and
slave. For more details see Section 26.3.5: Slave select (NSS) pin management.
2. In this configuration, the master’s MISO pin and the slave’s MOSI pin can be used as GPIOs.
3. A critical situation can happen when communication direction is changed not synchronously between two
nodes working at bidirectionnal mode and new transmitter accesses the common data line while former
transmitter still keeps an opposite value on the line (the value depends on SPI configuration and
communication data). Both nodes then fight while providing opposite output levels on the common line
temporary till next node changes its direction settings correspondingly, too. It is suggested to insert a serial
resistance between MISO and MOSI pins at this mode to protect the outputs and limit the current blowing
between them at this situation.

Simplex communications
The SPI can communicate in simplex mode by setting the SPI in transmit-only or in receive-
only using the RXONLY bit in the SPIx_CR2 register. In this configuration, only one line is
used for the transfer between the shift registers of the master and slave. The remaining
MISO and MOSI pins pair is not used for communication and can be used as standard
GPIOs.
• Transmit-only mode (RXONLY=0): The configuration settings are the same as for full-
duplex. The application has to ignore the information captured on the unused input pin.
This pin can be used as a standard GPIO.
• Receive-only mode (RXONLY=1): The application can disable the SPI output function
by setting the RXONLY bit. In slave configuration, the MISO output is disabled and the
pin can be used as a GPIO. The slave continues to receive data from the MOSI pin
while its slave select signal is active (see 26.3.5: Slave select (NSS) pin management).
Received data events appear depending on the data buffer configuration. In the master
configuration, the MOSI output is disabled and the pin can be used as a GPIO. The
clock signal is generated continuously as long as the SPI is enabled. The only way to
stop the clock is to clear the RXONLY bit or the SPE bit and wait until the incoming
pattern from the MISO pin is finished and fills the data buffer structure, depending on its
configuration.

816/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Figure 273. Simplex single master/single slave application (master in transmit-only/

slave in receive-only mode)

(2) MISO MISO

Rx shift register Tx shift register
MOSI MOSI
Tx shift register Rx shift register

SPI clock SCK SCK

generator
NSS(1) NSS(1)
Master Slave

MSv39625V1

1. The NSS pins can be used to provide a hardware control flow between master and slave. Optionally, the
pins can be left unused by the peripheral. Then the flow has to be handled internally for both master and
slave. For more details see Section 26.3.5: Slave select (NSS) pin management.
2. An accidental input information is captured at the input of transmitter Rx shift register. All the events
associated with the transmitter receive flow must be ignored in standard transmit only mode (e.g. OVF
flag).
3. In this configuration, both the MISO pins can be used as GPIOs.
Note: Any simplex communication can be alternatively replaced by a variant of the half-duplex
communication with a constant setting of the transaction direction (bidirectional mode is
enabled while BDIO bit is not changed).

RM0402 Rev 6 817/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.3.3 Standard multi-slave communication

In a configuration with two or more independent slaves, the master uses GPIO pins to
manage the chip select lines for each slave (see Figure 274.). The master must select one
of the slaves individually by pulling low the GPIO connected to the slave NSS input. When
this is done, a standard master and dedicated slave communication is established.

Figure 274. Master and three independent slaves

NSS (1)

MISO MISO
Rx shift register Tx shift register
MOSI MOSI
Tx shift register Rx shift register

SPI clock SCK SCK

generator
IO1 NSS
Master Slave 1
IO2
IO3

MISO
Tx shift register
MOSI
Rx shift register
SCK

NSS
Slave 2

MISO
Tx shift register
MOSI
Rx shift register
SCK

NSS
Slave 3
MSv39626V1

1. NSS pin is not used on master side at this configuration. It has to be managed internally (SSM=1, SSI=1) to
prevent any MODF error.
2. As MISO pins of the slaves are connected together, all slaves must have the GPIO configuration of their
MISO pin set as alternate function open-drain (see Section 7.3.7: I/O alternate function input/output on
page 174).

818/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.3.4 Multi-master communication

Unless SPI bus is not designed for a multi-master capability primarily, the user can use build
in feature which detects a potential conflict between two nodes trying to master the bus at
the same time. For this detection, NSS pin is used configured at hardware input mode.
The connection of more than two SPI nodes working at this mode is impossible as only one
node can apply its output on a common data line at time.
When nodes are non active, both stay at slave mode by default. Once one node wants to
overtake control on the bus, it switches itself into master mode and applies active level on
the slave select input of the other node via dedicated GPIO pin. After the session is
completed, the active slave select signal is released and the node mastering the bus
temporary returns back to passive slave mode waiting for next session start.
If potentially both nodes raised their mastering request at the same time a bus conflict event
appears (see mode fault MODF event). Then the user can apply some simple arbitration
process (e.g. to postpone next attempt by predefined different time-outs applied at both
nodes).

Figure 275. Multi-master application

MISO MISO
Rx (Tx) shift register Rx (Tx) shift register
MOSI MOSI
Tx (Rx) shift register Tx (Rx) shift register

SPI clock SCK SCK SPI clock

generator generator
(1)
GPIO NSS

Master NSS(1) GPIO Master

(Slave) (Slave)

MSv39628V1

1. The NSS pin is configured at hardware input mode at both nodes. Its active level enables the MISO line
output control as the passive node is configured as a slave.

26.3.5 Slave select (NSS) pin management

In slave mode, the NSS works as a standard “chip select” input and lets the slave
communicate with the master. In master mode, NSS can be used either as output or input.
As an input it can prevent multimaster bus collision, and as an output it can drive a slave
select signal of a single slave.
Hardware or software slave select management can be set using the SSM bit in the
SPIx_CR1 register:
• Software NSS management (SSM = 1): in this configuration, slave select information
is driven internally by the SSI bit value in register SPIx_CR1. The external NSS pin is
free for other application uses.
• Hardware NSS management (SSM = 0): in this case, there are two possible
configurations. The configuration used depends on the NSS output configuration
(SSOE bit in register SPIx_CR1).

RM0402 Rev 6 819/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

– NSS output enable (SSM=0,SSOE = 1): this configuration is only used when the
MCU is set as master. The NSS pin is managed by the hardware. The NSS signal
is driven low as soon as the SPI is enabled in master mode (SPE=1), and is kept
low until the SPI is disabled (SPE =0).
– NSS output disable (SSM=0, SSOE = 0): if the microcontroller is acting as the
master on the bus, this configuration allows multimaster capability. If the NSS pin
is pulled low in this mode, the SPI enters master mode fault state and the device is
automatically reconfigured in slave mode. In slave mode, the NSS pin works as a
standard “chip select” input and the slave is selected while NSS line is at low level.

Figure 276. Hardware/software slave select management

SSI control bit

SSM control bit

NSS Master
Slave mode
Inp. mode
Vdd OK Non active
1 Vss Conflict Active

NSS Input
0
NSS GPIO
pin logic

NSS NSS Output

Output
Control (used in Master mode & NSS
HW management only)

SSOE control bit

NSS external logic NSS internal logic

aiv14746e

820/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.3.6 Communication formats

During SPI communication, receive and transmit operations are performed simultaneously.
The serial clock (SCK) synchronizes the shifting and sampling of the information on the data
lines. The communication format depends on the clock phase, the clock polarity and the
data frame format. To be able to communicate together, the master and slaves devices must
follow the same communication format.

Clock phase and polarity controls

Four possible timing relationships may be chosen by software, using the CPOL and CPHA
bits in the SPIx_CR1 register. The CPOL (clock polarity) bit controls the idle state value of
the clock when no data is being transferred. This bit affects both master and slave modes. If
CPOL is reset, the SCK pin has a low-level idle state. If CPOL is set, the SCK pin has a
high-level idle state.
If the CPHA bit is set, the second edge on the SCK pin captures the first data bit transacted
(falling edge if the CPOL bit is reset, rising edge if the CPOL bit is set). Data are latched on
each occurrence of this clock transition type. If the CPHA bit is reset, the first edge on the
SCK pin captures the first data bit transacted (falling edge if the CPOL bit is set, rising edge
if the CPOL bit is reset). Data are latched on each occurrence of this clock transition type.
The combination of CPOL (clock polarity) and CPHA (clock phase) bits selects the data
capture clock edge.
Figure 277, shows an SPI full-duplex transfer with the four combinations of the CPHA and
CPOL bits.
Note: Prior to changing the CPOL/CPHA bits the SPI must be disabled by resetting the SPE bit.
The idle state of SCK must correspond to the polarity selected in the SPIx_CR1 register (by
pulling up SCK if CPOL=1 or pulling down SCK if CPOL=0).

RM0402 Rev 6 821/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 277. Data clock timing diagram

CPHA =1
CPOL = 1

CPOL = 0

MOSI MSBit LSBit

MISO MSBit LSBit

NSS
(to slave)

Capture strobe

CPHA =0

CPOL = 1

CPOL = 0

MOSI MSBit LSBit

MISO MSBit LSBit

NSS
(to slave)

Capture strobe

ai17154d

Note: The order of data bits depends on LSBFIRST bit setting.

Data frame format

The SPI shift register can be set up to shift out MSB-first or LSB-first, depending on the
value of the LSBFIRST bit. Each data frame is 8 or 16 bit long depending on the size of the
data programmed using the DFF bit in the SPI_CR1 register. The selected data frame
format is applicable both for transmission and reception.

822/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.3.7 SPI configuration

The configuration procedure is almost the same for master and slave. For specific mode
setups, follow the dedicated chapters. When a standard communication is to be initialized,
perform these steps:
1. Write proper GPIO registers: Configure GPIO for MOSI, MISO and SCK pins.
2. Write to the SPI_CR1 register:
a) Configure the serial clock baud rate using the BR[2:0] bits (Note: 3).
b) Configure the CPOL and CPHA bits combination to define one of the four
relationships between the data transfer and the serial clock. (Note: 2 - except the
case when CRC is enabled at TI mode).
c) Select simplex or half-duplex mode by configuring RXONLY or BIDIMODE and
BIDIOE (RXONLY and BIDIMODE can't be set at the same time).
d) Configure the LSBFIRST bit to define the frame format (Note: 2).
e) Configure the CRCEN and CRCEN bits if CRC is needed (while SCK clock signal
is at idle state).
f) Configure SSM and SSI (Note: 2).
g) Configure the MSTR bit (in multimaster NSS configuration, avoid conflict state on
NSS if master is configured to prevent MODF error).
h) Set the DFF bit to configure the data frame format (8 or 16 bits).
3. Write to SPI_CR2 register:
a) Configure SSOE (Note: 1 & 2).
b) Set the FRF bit if the TI protocol is required.
4. Write to SPI_CRCPR register: Configure the CRC polynomial if needed.
5. Write proper DMA registers: Configure DMA streams dedicated for SPI Tx and Rx in
DMA registers if the DMA streams are used.
Note: (1) Step is not required in slave mode.
(2) Step is not required in TI mode.
(3) The step is not required in slave mode except slave working at TI mode.

26.3.8 Procedure for enabling SPI

It is recommended to enable the SPI slave before the master sends the clock. Otherwise,
undesired data transmission might occur. The slave data register must already contain data
to be sent before starting communication with the master (either on the first edge of the
communication clock, or before the end of the ongoing communication if the clock signal is
continuous). The SCK signal must be settled at an idle state level corresponding to the
selected polarity before the SPI slave is enabled.
At full-duplex (or in any transmit-only mode), the master starts communicating when the SPI
is enabled and data to be sent is written in the Tx Buffer.
In any master receive-only mode (RXONLY=1 or BIDIMODE=1 & BIDIOE=0), the master
starts communicating and the clock starts running immediately after the SPI is enabled.
The slave starts communicating when it receives a correct clock signal from the master. The
slave software must write the data to be sent before the SPI master initiates the transfer.
Refer to Section 26.3.11: Communication using DMA (direct memory addressing) for details
on how to handle DMA.

RM0402 Rev 6 823/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.3.9 Data transmission and reception procedures

Rx and Tx buffers
In reception, data are received and then stored into an internal Rx buffer while in
transmission, data are first stored into an internal Tx buffer before being transmitted. A read
access to the SPI_DR register returns the Rx buffered value whereas a write access to the
SPI_DR stores the written data into the Tx buffer.

Tx buffer handling
The data frame is loaded from the Tx buffer into the shift register during the first bit
transmission. Bits are then shifted out serially from the shift register to a dedicated output
pin depending on LSBFIRST bit setting.The TXE flag (Tx buffer empty) is set when the data
are transferred from the Tx buffer to the shift register. It indicates that the internal Tx buffer is
ready to be loaded with the next data. An interrupt can be generated if the TXEIE bit of the
SPI_CR2 register is set. Clearing the TXE bit is performed by writing to the SPI_DR register.
A continuous transmit stream can be achieved if the next data to be transmitted are stored
in the Tx buffer while previous frame transmission is still ongoing. When the software writes
to Tx buffer while the TXE flag is not set, the data waiting for transaction is overwritten.

Rx buffer handling
The RXNE flag (Rx buffer not empty) is set on the last sampling clock edge, when the data
are transferred from the shift register to the Rx buffer. It indicates that data are ready to be
read from the SPI_DR register. An interrupt can be generated if the RXNEIE bit in the
SPI_CR2 register is set. Clearing the RXNE bit is performed by reading the SPI_DR
register.
If a device has not cleared the RXNE bit resulting from the previous data byte transmitted,
an overrun condition occurs when the next value is buffered. The OVR bit is set and an
interrupt is generated if the ERRIE bit is set.
Another way to manage the data exchange is to use DMA (see Section 9.2: DMA main
features).

Sequence handling
The BSY bit is set when a current data frame transaction is ongoing. When the clock signal
runs continuously, the BSY flag remains set between data frames on the master side.
However, on the slave side, it becomes low for a minimum duration of one SPI clock cycle
between each data frame transfer.
For some configurations, the BSY flag can be used during the last data transfer to wait until
the completion of the transfer.
When a receive-only mode is configured on the master side, either in half-duplex
(BIDIMODE=1, BIDIOE=0) or simplex configuration (BIDIMODE=0, RXONLY=1), the
master starts the receive sequence as soon as the SPI is enabled. Then the clock signal is
provided by the master and it does not stop until either the SPI or the receive-only mode is
disabled by the master. The master receives data frames continuously up to this moment.
While the master can provide all the transactions in continuous mode (SCK signal is
continuous), it has to respect slave capability to handle data flow and its content at anytime.
When necessary, the master must slow down the communication and provide either a
slower clock or separate frames or data sessions with sufficient delays. Be aware there is no

824/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

underflow error signal for slave operating in SPI mode, and that data from the slave are
always transacted and processed by the master even if the slave cannot not prepare them
correctly in time. It is preferable for the slave to use DMA, especially when data frames are
shorter and bus rate is high.
Each sequence must be encased by the NSS pulse in parallel with the multislave system to
select just one of the slaves for communication. In single slave systems, using NSS to
control the slave is not necessary. However, the NSS pulse can be used to synchronize the
slave with the beginning of each data transfer sequence. NSS can be managed either by
software or by hardware (see Section 26.3.4: Multi-master communication).
Refer to Figure 278 and Figure 279 for a description of continuous transfers in master / full-
duplex and slave full-duplex mode.

Figure 278. TXE/RXNE/BSY behavior in master / full-duplex mode (BIDIMODE=0,

RXONLY=0) in the case of continuous transfers
Example in Master mode with CPOL=1, CPHA=1

SCK

DATA1 = 0xF1 DATA2 = 0xF2 DATA3 = 0xF3

MISO/MOSI (out) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7

set by hardware set by hardware

TXE flag cleared by software cleared by software set by hardware

Tx buffer 0xF1 0xF2 0xF3

(write SPI_DR)

BSY flag set by hardware reset by hardware

DATA 1 = 0xA1 DATA 2 = 0xA2 DATA 3 = 0xA3

MISO/MOSI (in) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7

set by hardware cleared by software

RXNE flag

Rx buffer 0xA1 0xA2 0xA3

(read SPI_DR)

software software waits software waits software waits software waits software waits
writes 0xF1 until TXE=1 and until RXNE=1 until TXE=1 and until RXNE=1 until RXNE=1
into SPI_DR writes 0xF2 into and reads 0xA1 writes 0xF3 into and reads 0xA2 and reads 0xA3
SPI_DR from SPI_DR SPI_DR from SPI_ DR from SPI_DR
ai17343

RM0402 Rev 6 825/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 279. TXE/RXNE/BSY behavior in slave / full-duplex mode (BIDIMODE=0,

RXONLY=0) in the case of continuous transfers
Example in Slave mode with CPOL=1, CPHA=1

SCK

DATA 1 = 0xF1 DATA 2 = 0xF2 DATA 3 = 0xF3

MISO/MOSI (out) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7

set by hardware set by hardware

TXE flag cleared by software cleared by software set by hardware

Tx buffer 0xF1 0xF2 0xF3

(write to SPI_DR)

BSY flag set by cleared by software reset by hardware

DATA 1 = 0xA1 DATA 2 = 0xA2 DATA 3 = 0xA3

MISO/MOSI (in) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7

set by hardware cleared by software

RXNE flag

Rx buffer 0xA1 0xA2 0xA3

(read from SPI_DR)

software software waits software waits software waits software waits software waits
writes 0xF1 until TXE=1 and until RXNE=1 until TXE=1 and until RXNE=1 until RXNE=1
into SPI_DR writes 0xF2 into and reads 0xA1 writes 0xF3 into and reads 0xA2 and reads 0xA3
SPI_DR from SPI_DR SPI_DR from SPI_ DR from SPI_DR
ai17344

26.3.10 Procedure for disabling the SPI

When SPI is disabled, it is mandatory to follow the disable procedures described in this
paragraph. It is important to do this before the system enters a low-power mode when the
peripheral clock is stopped. Ongoing transactions can be corrupted in this case. In some
modes the disable procedure is the only way to stop continuous communication running.
Master in full-duplex or transmit only mode can finish any transaction when it stops
providing data for transmission. In this case, the clock stops after the last data transaction.
Standard disable procedure is based on pulling BSY status together with TXE flag to check
if a transmission session is fully completed. This check can be done in specific cases, too,
when it is necessary to identify the end of ongoing transactions, for example:
• When NSS signal is managed by an arbitrary GPIO toggle and the master has to
provide proper end of NSS pulse for slave, or
• When transactions’ streams from DMA are completed while the last data frame or CRC
frame transaction is still ongoing in the peripheral bus.
The correct disable procedure is (except when receive-only mode is used):
1. Wait until RXNE=1 to receive the last data.
2. Wait until TXE=1 and then wait until BSY=0 before disabling the SPI.
3. Read received data.

826/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Note: During discontinuous communications, there is a 2 APB clock period delay between the
write operation to the SPI_DR register and BSY bit setting. As a consequence it is
mandatory to wait first until TXE is set and then until BSY is cleared after writing the last
data.
The correct disable procedure for certain receive-only modes is:
1. Interrupt the receive flow by disabling SPI (SPE=0) in the specific time window while
the last data frame is ongoing.
2. Wait until BSY=0 (the last data frame is processed).
3. Read received data.
Note: To stop a continuous receive sequence, a specific time window must be respected during
the reception of the last data frame. It starts when the first bit is sampled and ends before
the last bit transfer starts.

26.3.11 Communication using DMA (direct memory addressing)

To operate at its maximum speed and to facilitate the data register read/write process
required to avoid overrun, the SPI features a DMA capability, which implements a simple
request/acknowledge protocol.
A DMA access is requested when the TXE or RXNE enable bit in the SPIx_CR2 register is
set. Separate requests must be issued to the Tx and Rx buffers.
• In transmission, a DMA request is issued each time TXE is set to 1. The DMA then
writes to the SPIx_DR register.
• In reception, a DMA request is issued each time RXNE is set to 1. The DMA then reads
the SPIx_DR register.
Refer to Figure 280 and Figure 281 for a description of the DMA transmission and reception
waveforms.
When the SPI is used only to transmit data, it is possible to enable only the SPI Tx DMA
channel. In this case, the OVR flag is set because the data received is not read. When the
SPI is used only to receive data, it is possible to enable only the SPI Rx DMA channel.
In transmission mode, when the DMA has written all the data to be transmitted (the TCIF
flag is set in the DMA_ISR register), the BSY flag can be monitored to ensure that the SPI
communication is complete. This is required to avoid corrupting the last transmission before
disabling the SPI or entering the Stop mode. The software must first wait until TXE = 1 and
then until BSY = 0.
When starting communication using DMA, to prevent DMA channel management raising
error events, these steps must be followed in order:
1. Enable DMA Rx buffer in the RXDMAEN bit in the SPI_CR2 register, if DMA Rx is
used.
2. Enable DMA streams for Tx and Rx in DMA registers, if the streams are used.
3. Enable DMA Tx buffer in the TXDMAEN bit in the SPI_CR2 register, if DMA Tx is used.
4. Enable the SPI by setting the SPE bit.

RM0402 Rev 6 827/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

To close communication it is mandatory to follow these steps in order:

1. Disable DMA streams for Tx and Rx in the DMA registers, if the streams are used.
2. Disable the SPI by following the SPI disable procedure.
3. Disable DMA Tx and Rx buffers by clearing the TXDMAEN and RXDMAEN bits in the
SPI_CR2 register, if DMA Tx and/or DMA Rx are used.

Figure 280. Transmission using DMA

Example with CPOL=1, CPHA=1

SCK

DATA 1 = 0xF1 DATA 2 = 0xF2 DATA 3 = 0xF3

MISO/MOSI (out) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7

set by hardware set by hardware

TXE flag cleared by DMA write clear by DMA write set by hardware

reset
BSY flag set by hardware by hardware

DMA request ignored by the DMA because

DMA transfer is complete

Tx buffer 0xF1 0xF2 0xF3

(write to SPI_DR)

DMA writes to SPI_DR

set by hardware clear by software

DMA TCIF flag
(DMA transfer complete)

software configures the DMA writes DMA writes DMA writes DMA transfer is software waits software waits until BSY=0
DMA SPI Tx channel DATA1 into DATA2 into DATA3 into complete (TCIF=1 in until TXE=1
to send 3 data items SPI_DR SPI_DR SPI_DR DMA_ISR)
and enables the SPI
ai17349

828/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Figure 281. Reception using DMA

Example with CPOL=1, CPHA=1

SCK

DATA 1 = 0xA1 DATA 2 = 0xA2 DATA 3 = 0xA3

MISO/MOSI (in) b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7 b0 b1 b2 b3 b4 b5 b6 b7

set by hardware clear by DMA read

RXNE flag

DMA request

Rx buffer
0xA1 0xA2 0xA3
(read from SPI_DR)

DMA read from SPI_DR

clear
set by hardware
flag DMA TCIF by software
(DMA transfer complete)

software configures the DMA reads DMA reads DMA reads The DMA transfer is
DMA SPI Rx channel DATA1 from DATA2 from DATA3 from complete (TCIF=1 in
to receive 3 data items SPI_DR SPI_DR SPI_DR DMA_ISR)
and enables the SPI
ai17350

26.3.12 SPI status flags

Three status flags are provided for the application to completely monitor the state of the SPI
bus.

Tx buffer empty flag (TXE)

When it is set, the TXE flag indicates that the Tx buffer is empty and that the next data to be
transmitted can be loaded into the buffer. The TXE flag is cleared by writing to the SPI_DR
register.

Rx buffer not empty (RXNE)

When set, the RXNE flag indicates that there are valid received data in the Rx buffer. It is
cleared by reading from the SPI_DR register.

Busy flag (BSY)

The BSY flag is set and cleared by hardware (writing to this flag has no effect).
When BSY is set, it indicates that a data transfer is in progress on the SPI (the SPI bus is
busy). There is one exception in master bidirectional receive mode (MSTR=1 and BDM=1
and BDOE=0) where the BSY flag is kept low during reception.
The BSY flag can be used in certain modes to detect the end of a transfer, thus preventing
corruption of the last transfer when the SPI peripheral clock is disabled before entering a
low-power mode or an NSS pulse end is handled by software.
The BSY flag is also useful for preventing write collisions in a multimaster system.

RM0402 Rev 6 829/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

The BSY flag is cleared under any one of the following conditions:
• When the SPI is correctly disabled
• When a fault is detected in Master mode (MODF bit set to 1)
• In Master mode, when it finishes a data transmission and no new data is ready to be
sent
• In Slave mode, when the BSY flag is set to '0' for at least one SPI clock cycle between
each data transfer.
Note: It is recommended to use always the TXE and RXNE flags (instead of the BSY flags) to
handle data transmission or reception operations.

26.3.13 SPI error flags

An SPI interrupt is generated if one of the following error flags is set and interrupt is enabled
by setting the ERRIE bit.

Overrun flag (OVR)

An overrun condition occurs when the master or the slave completes the reception of the
next data frame while the read operation of the previous frame from the Rx buffer has not
completed (case RXNE flag is set).
In this case, the content of the Rx buffer is not updated with the new data received. A read
operation from the SPI_DR register returns the frame previously received. All other
subsequently transmitted data are lost.
Clearing the OVR bit is done by a read access to the SPI_DR register followed by a read
access to the SPI_SR register.

Mode fault (MODF)

Mode fault occurs when the master device has its internal NSS signal (NSS pin in NSS
hardware mode, or SSI bit in NSS software mode) pulled low. This automatically sets the
MODF bit. Master mode fault affects the SPI interface in the following ways:
• The MODF bit is set and an SPI interrupt is generated if the ERRIE bit is set.
• The SPE bit is cleared. This blocks all output from the device and disables the SPI
interface.
• The MSTR bit is cleared, thus forcing the device into slave mode.
Use the following software sequence to clear the MODF bit:
1. Make a read or write access to the SPIx_SR register while the MODF bit is set.
2. Then write to the SPIx_CR1 register.
To avoid any multiple slave conflicts in a system comprising several MCUs, the NSS pin
must be pulled high during the MODF bit clearing sequence. The SPE and MSTR bits can
be restored to their original state after this clearing sequence. As a security, hardware does
not allow the SPE and MSTR bits to be set while the MODF bit is set. In a slave device the
MODF bit cannot be set except as the result of a previous multimaster conflict.

830/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

CRC error (CRCERR)

This flag is used to verify the validity of the value received when the CRCEN bit in the
SPIx_CR1 register is set. The CRCERR flag in the SPIx_SR register is set if the value
received in the shift register does not match the receiver SPIx_RXCRC value. The flag is
cleared by the software.

TI mode frame format error (FRE)

A TI mode frame format error is detected when an NSS pulse occurs during an ongoing
communication when the SPI is operating in slave mode and configured to conform to the TI
mode protocol. When this error occurs, the FRE flag is set in the SPIx_SR register. The SPI
is not disabled when an error occurs, the NSS pulse is ignored, and the SPI waits for the
next NSS pulse before starting a new transfer. The data may be corrupted since the error
detection may result in the loss of two data bytes.
The FRE flag is cleared when SPIx_SR register is read. If the ERRIE bit is set, an interrupt
is generated on the NSS error detection. In this case, the SPI should be disabled because
data consistency is no longer guaranteed and communications should be re-initiated by the
master when the slave SPI is enabled again.

26.4 SPI special features

26.4.1 TI mode
TI protocol in master mode
The SPI interface is compatible with the TI protocol. The FRF bit of the SPIx_CR2 register
can be used to configure the SPI to be compliant with this protocol.
The clock polarity and phase are forced to conform to the TI protocol requirements whatever
the values set in the SPIx_CR1 register. NSS management is also specific to the TI protocol
which makes the configuration of NSS management through the SPIx_CR1 and SPIx_CR2
registers (SSM, SSI, SSOE) impossible in this case.
In slave mode, the SPI baud rate prescaler is used to control the moment when the MISO
pin state changes to HiZ when the current transaction finishes (see Figure 282). Any baud
rate can be used, making it possible to determine this moment with optimal flexibility.
However, the baud rate is generally set to the external master clock baud rate. The delay for
the MISO signal to become HiZ (trelease) depends on internal resynchronization and on the
baud rate value set in through the BR[2:0] bits in the SPIx_CR1 register. It is given by the
formula:

t baud_rate t baud_rate
---------------------- + 4 × t pclk < t release < ---------------------
- + 6 × t pclk
2 2
If the slave detects a misplaced NSS pulse during a data frame transaction the TIFRE flag is
set.
This feature is not available for Motorola SPI communications (FRF bit set to 0).

RM0402 Rev 6 831/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Note: To detect TI frame errors in slave transmitter only mode by using the Error interrupt
(ERRIE=1), the SPI must be configured in 2-line unidirectional mode by setting BIDIMODE
and BIDIOE to 1 in the SPI_CR1 register. When BIDIMODE is set to 0, OVR is set to 1
because the data register is never read and error interrupts are always generated, while
when BIDIMODE is set to 1, data are not received and OVR is never set.
Figure 282 shows the SPI communication waveforms when TI mode is selected.

Figure 282. TI mode transfer

NSS
tri ng

g
t RELEASE
in

in
i
er

er

sa r
pl

pl

pl
e
gg

gg

gg
m

m
sa

sa
tri

tr i

SCK

MOSI Do not care MSB LSB MSB LSB

MISO 1 or 0 MSB LSB MSB LSB

FRAME 1 FRAME 2

MS19835V2

26.4.2 CRC calculation

Two separate CRC calculators (on transmission and reception data flows) are implemented
in order to check the reliability of transmitted and received data. The SPI offers CRC8 or
CRC16 calculation depending on the data format selected through the DFF bit. The CRC is
calculated serially using the polynomial programmed in the SPI_CRCPR register.

CRC principle
CRC calculation is enabled by setting the CRCEN bit in the SPIx_CR1 register before the
SPI is enabled (SPE = 1). The CRC value is calculated using an odd programmable
polynomial on each bit. The calculation is processed on the sampling clock edge defined by
the CPHA and CPOL bits in the SPIx_CR1 register. The calculated CRC value is checked
automatically at the end of the data block as well as for transfer managed by CPU or by the
DMA. When a mismatch is detected between the CRC calculated internally on the received
data and the CRC sent by the transmitter, a CRCERR flag is set to indicate a data corruption
error. The right procedure for handling the CRC calculation depends on the SPI
configuration and the chosen transfer management.
Note: The polynomial value should only be odd. No even values are supported.

CRC transfer managed by CPU

Communication starts and continues normally until the last data frame has to be sent or
received in the SPIx_DR register. Then CRCNEXT bit has to be set in the SPIx_CR1
register to indicate that the CRC frame transaction will follow after the transaction of the
currently processed data frame. The CRCNEXT bit must be set before the end of the last
data frame transaction. CRC calculation is frozen during CRC transaction.

832/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

The received CRC is stored in the Rx buffer like any other data frame.
A CRC-format transaction takes one more data frame to communicate at the end of data
sequence.
When the last CRC data is received, an automatic check is performed comparing the
received value and the value in the SPIx_RXCRC register. Software has to check the
CRCERR flag in the SPIx_SR register to determine if the data transfers were corrupted or
not. Software clears the CRCERR flag by writing '0' to it.
After the CRC reception, the CRC value is stored in the Rx buffer and must be read in the
SPIx_DR register in order to clear the RXNE flag.

CRC transfer managed by DMA

When SPI communication is enabled with CRC communication and DMA mode, the
transmission and reception of the CRC at the end of communication is automatic (with the
exception of reading CRC data in receive-only mode). The CRCNEXT bit does not have to
be handled by the software. The counter for the SPI transmission DMA channel has to be
set to the number of data frames to transmit excluding the CRC frame. On the receiver side,
the received CRC value is handled automatically by DMA at the end of the transaction but
user must take care to flush out the CRC frame received from SPI_DR as it is always loaded
into it.
At the end of the data and CRC transfers, the CRCERR flag in the SPIx_SR register is set if
corruption occurred during the transfer.

Resetting the SPIx_TXCRC and SPIx_RXCRC values

The SPIx_TXCRC and SPIx_RXCRC values are cleared automatically when CRC
calculation is enabled.
When the SPI is configured in slave mode with the CRC feature enabled, a CRC calculation
is performed even if a high level is applied on the NSS pin. This may happen for example in
case of a multislave environment where the communication master addresses slaves
alternately.
Between a slave disabling (high level on NSS) and a new slave enabling (low level on NSS),
the CRC value should be cleared on both master and slave sides to resynchronize the
master and slave respective CRC calculation.
To clear the CRC, follow the below sequence:
1. Disable the SPI
2. Clear the CRCEN bit
3. Enable the CRCEN bit
4. Enable the SPI
Note: When the SPI interface is configured as a slave, the NSS internal signal needs to be kept
low during transaction of the CRC phase once the CRCNEXT signal is released, (see more
details at the product errata sheet).
At TI mode, despite the fact that the clock phase and clock polarity setting is fixed and
independent on the SPIx_CR1 register, the corresponding setting CPOL=0 CPHA=1 has to
be kept at the SPIx_CR1 register anyway if CRC is applied. In addition, the CRC calculation
has to be reset between sessions by the SPI disable sequence by re-enabling the CRCEN
bit described above at both master and slave sides, else the CRC calculation can be
corrupted at this specific mode.

RM0402 Rev 6 833/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.5 SPI interrupts

During SPI communication an interrupts can be generated by the following events:
• Transmit Tx buffer ready to be loaded
• Data received in Rx buffer
• Master mode fault
• Overrun error
• TI frame format error
Interrupts can be enabled and disabled separately.

Table 156. SPI interrupt requests

Interrupt event Event flag Enable Control bit

Transmit Tx buffer ready to be loaded TXE TXEIE

Data received in Rx buffer RXNE RXNEIE
Master Mode fault event MODF
Overrun error OVR
ERRIE
CRC error CRCERR
TI frame format error FRE

834/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.6 I2S functional description

26.6.1 I2S general description

The block diagram of the I2S is shown in Figure 283.

Figure 283. I2S block diagram

Address and data bus

Tx buffer

CRC CH
BSY OVR MODF UDR TxE RxNE FRE
ERR SIDE
16-bit

MOSI/SD

MISO/ Shift register

I2S2ext_SD/ LSB first Communication
I2S3ext_SD(1) 16-bit control
Rx buffer

NSS/WS

I2SCFG I2SSTD CK DATLEN CH

[1:0] [1:0] POL [1:0] LEN

I2S
MOD I2SE

Master control logic

Bidi Bidi CRC CRC Rx
DFF SSM SSI
mode OE EN Next only

SPI LSB
SPE BR2 BR1 BR0 MSTR CPOL CPHA
baud rate generator First

CK

I2S clock generator

I2S_CK

I2SMOD

MCK I2SxCLK
MCKOE ODD I2SDIV[7:0]

MS19909V1

1. I2S2ext_SD and I2S3ext_SD are the extended SD pins that control the I2S full-duplex mode.
2. MCK is mapped on the MISO pin.

RM0402 Rev 6 835/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

The SPI can function as an audio I2S interface when the I2S capability is enabled (by setting
the I2SMOD bit in the SPIx_I2SCFGR register). This interface mainly uses the same pins,
flags and interrupts as the SPI.
The I2S shares three common pins with the SPI:
• SD: Serial Data (mapped on the MOSI pin) to transmit or receive the two time-
multiplexed data channels (in half-duplex mode only).
• WS: Word Select (mapped on the NSS pin) is the data control signal output in master
mode and input in slave mode.
• CK: Serial Clock (mapped on the SCK pin) is the serial clock output in master mode
and serial clock input in slave mode.
An additional pin can be used when a master clock output is needed for some external
audio devices:
• MCK: Master Clock (mapped separately) is used, when the I2S is configured in master
mode (and when the MCKOE bit in the SPIx_I2SPR register is set), to output this
additional clock generated at a preconfigured frequency rate equal to 256 × fS, where
fS is the audio sampling frequency.
The I2S uses its own clock generator to produce the communication clock when it is set in
master mode. This clock generator is also the source of the master clock output. Two
additional registers are available in I2S mode. One is linked to the clock generator
configuration SPIx_I2SPR and the other one is a generic I2S configuration register
SPIx_I2SCFGR (audio standard, slave/master mode, data format, packet frame, clock
polarity, etc.).
The SPIx_CR1 register and all CRC registers are not used in the I2S mode. Likewise, the
SSOE bit in the SPIx_CR2 register and the MODF and CRCERR bits in the SPIx_SR are
not used.
The I2S uses the same SPI register for data transfer (SPIx_DR) in 16-bit wide mode.

26.6.2 I2S full-duplex

To support I2S full-duplex mode, two extra I2S instances called extended I2Ss (I2S2_ext,
I2S3_ext) are available in addition to I2S2 and I2S3 (see Figure 284). The first I2S full-
duplex interface is consequently based on I2S2 and I2S2_ext, and the second one on I2S3
and I2S3_ext.
Note: I2S2_ext an I2S3_ext are used only in full-duplex mode.

Figure 284. I2S full-duplex block diagram

SPI/I2Sx SPIx_MOSI/I2Sx_SD(in/out)

I2Sx_SCK

I2S_WS
I2Sx_ext I2Sx_extSD(in/out)

MS19910V2

1. Where x can be 2 or 3.

836/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

I2Sx can operate in master mode. As a result:

• Only I2Sx can output SCK and WS in half-duplex mode
• Only I2Sx can deliver SCK and WS to I2S2_ext and I2S3_ext in full-duplex mode.
The extended I2Ss (I2Sx_ext) can be used only in full-duplex mode. The I2Sx_ext operate
always in slave mode.
Both I2Sx and I2Sx_ext can be configured as transmitters or receivers.

26.6.3 Supported audio protocols

The three-line bus has to handle only audio data generally time-multiplexed on two
channels: the right channel and the left channel. However there is only one 16-bit register
for transmission or reception. So, it is up to the software to write into the data register the
appropriate value corresponding to each channel side, or to read the data from the data
register and to identify the corresponding channel by checking the CHSIDE bit in the
SPIx_SR register. Channel left is always sent first followed by the channel right (CHSIDE
has no meaning for the PCM protocol).
Four data and packet frames are available. Data may be sent with a format of:
• 16-bit data packed in a 16-bit frame
• 16-bit data packed in a 32-bit frame
• 24-bit data packed in a 32-bit frame
• 32-bit data packed in a 32-bit frame
When using 16-bit data extended on 32-bit packet, the first 16 bits (MSB) are the significant
bits, the 16-bit LSB is forced to 0 without any need for software action or DMA request (only
one read/write operation).
The 24-bit and 32-bit data frames need two CPU read or write operations to/from the
SPIx_DR register or two DMA operations if the DMA is preferred for the application. For 24-
bit data frame specifically, the 8 non significant bits are extended to 32 bits with 0-bits (by
hardware).
For all data formats and communication standards, the most significant bit is always sent
first (MSB first).
The I2S interface supports four audio standards, configurable using the I2SSTD[1:0] and
PCMSYNC bits in the SPIx_I2SCFGR register.

I2S Philips standard

For this standard, the WS signal is used to indicate which channel is being transmitted. It is
activated one CK clock cycle before the first bit (MSB) is available.

RM0402 Rev 6 837/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 285. I2S Philips protocol waveforms (16/32-bit full accuracy, CPOL = 0)
CK

WS transmission reception

Can be 16-bit or 32-bit

SD MSB LSB MSB

Channel left
Channel
right
MS19591V1

Data are latched on the falling edge of CK (for the transmitter) and are read on the rising
edge (for the receiver). The WS signal is also latched on the falling edge of CK.

Figure 286. I2S Philips standard waveforms (24-bit frame with CPOL = 0)

CK

WS Transmission Reception

24-bit data 8-bit remaining 0 forced

SD
MSB LSB

Channel left 32-bit

Channel right

MS19592V1

This mode needs two write or read operations to/from the SPIx_DR register.
• In transmission mode:
If 0x8EAA33 has to be sent (24-bit):

Figure 287. Transmitting 0x8EAA33

First write to Data register Second write to Data register

0x8EAA 0x33XX
Only the 8 MSB are sent
to compare the 24 bits
8 LSBs have no meaning
and can be anything

MS19593V1

• In reception mode:
If data 0x8EAA33 is received:

838/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Figure 288. Receiving 0x8EAA33

First read to Data register Second read to Data register
0x8EAA 0x33XX
Only the 8 MSB are sent
to compare the 24 bits
8 LSBs have no meaning
and can be anything

MS19594V1

Figure 289. I2S Philips standard (16-bit extended to 32-bit packet frame with
CPOL = 0)
CK

WS Transmission Reception

16-bit data 16-bit remaining 0 forced

SD
MSB LSB

Channel left 32-bit

Channel right

MS19599V1

2
When 16-bit data frame extended to 32-bit channel frame is selected during the I S
configuration phase, only one access to the SPIx_DR register is required. The 16 remaining
bits are forced by hardware to 0x0000 to extend the data to 32-bit format.
If the data to transmit or the received data are 0x76A3 (0x76A30000 extended to 32-bit), the
operation shown in Figure 290 is required.

Figure 290. Example of 16-bit data frame extended to 32-bit channel frame

Only one access to SPIx_DR

0x76A3

MS19595V1

For transmission, each time an MSB is written to SPIx_DR, the TXE flag is set and its
interrupt, if allowed, is generated to load the SPIx_DR register with the new value to send.
This takes place even if 0x0000 have not yet been sent because it is done by hardware.
For reception, the RXNE flag is set and its interrupt, if allowed, is generated when the first
16 MSB half-word is received.
In this way, more time is provided between two write or read operations, which prevents
underrun or overrun conditions (depending on the direction of the data transfer).

MSB justified standard

For this standard, the WS signal is generated at the same time as the first data bit, which is
the MSBit.

RM0402 Rev 6 839/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 291. MSB Justified 16-bit or 32-bit full-accuracy length with CPOL = 0

CK

WS Transmission Reception

16- or 32 bit data

SD
MSB LSB MSB

Channel left
Channel right

MS30100 V1

Data are latched on the falling edge of CK (for transmitter) and are read on the rising edge
(for the receiver).

Figure 292. MSB justified 24-bit frame length with CPOL = 0

CK

WS Transmission Reception

24 bit data 8-bit remaining

SD 0 forced
MSB LSB
Channel left 32-bit

Channel right

MS30101V1

Figure 293. MSB justified 16-bit extended to 32-bit packet frame with CPOL = 0

CK

WS Transmission Reception

16-bit data 16-bit remaining

SD 0 forced
MSB LSB
Channel left 32-bit

Channel right

MS30102V1

LSB justified standard

This standard is similar to the MSB justified standard (no difference for the 16-bit and 32-bit
full-accuracy frame formats).

840/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Figure 294. LSB justified 16-bit or 32-bit full-accuracy with CPOL = 0

CK

WS
Transmission Reception
16- or 32-bit data
SD
MSB LSB MSB
Channel left

Channel right
MS30103V1

Figure 295. LSB justified 24-bit frame length with CPOL = 0

CK

WS Reception
Transmission
8-bit data 24-bit remaining
SD 0 forced
MSB LSB

Channel left 32-bit

Channel right

MS30104V1

• In transmission mode:
If data 0x3478AE have to be transmitted, two write operations to the SPIx_DR register
are required by software or by DMA. The operations are shown below.

Figure 296. Operations required to transmit 0x3478AE

First write to Data register Second write to Data register

conditioned by TXE=1 conditioned by TXE=1

0xXX34 0x78AE

Only the 8 LSB of the

half-word are significant.
A field of 0x00 is forced
instead of the 8 MSBs. MS19596V1

• In reception mode:
If data 0x3478AE are received, two successive read operations from the SPIx_DR
register are required on each RXNE event.

RM0402 Rev 6 841/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 297. Operations required to receive 0x3478AE

First read from Data register Second read from Data register
conditioned by RXNE=1 conditioned by RXNE=1

0xXX34 0x78AE

Only the 8 LSB of the

half-word are significant.
A field of 0x00 is forced
instead of the 8 MSBs.

MS19597V1

Figure 298. LSB justified 16-bit extended to 32-bit packet frame with CPOL = 0

CK

Reception
WS
Transmission
16-bit data 16-bit remaining
SD 0 forced
MSB LSB

Channel left 32-bit

Channel right
MS30105V1

When 16-bit data frame extended to 32-bit channel frame is selected during the I2S
configuration phase, Only one access to the SPIx_DR register is required. The 16 remaining
bits are forced by hardware to 0x0000 to extend the data to 32-bit format. In this case it
corresponds to the half-word MSB.
If the data to transmit or the received data are 0x76A3 (0x0000 76A3 extended to 32-bit),
the operation shown in Figure 299 is required.

Figure 299. Example of 16-bit data frame extended to 32-bit channel frame
Only one access to the SPIx-DR register

0x76A3

MS19598V1

In transmission mode, when a TXE event occurs, the application has to write the data to be
transmitted (in this case 0x76A3). The 0x000 field is transmitted first (extension on 32-bit).
The TXE flag is set again as soon as the effective data (0x76A3) is sent on SD.
In reception mode, RXNE is asserted as soon as the significant half-word is received (and
not the 0x0000 field).
In this way, more time is provided between two write or read operations to prevent underrun
or overrun conditions.

842/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

PCM standard
For the PCM standard, there is no need to use channel-side information. The two PCM
modes (short and long frame) are available and configurable using the PCMSYNC bit in
SPIx_I2SCFGR register.

Figure 300. PCM standard waveforms (16-bit)

CK

WS
short frame
13-bits
WS
long frame

SD MSB LSB MSB

MS30106V1

For long frame synchronization, the WS signal assertion time is fixed to 13 bits in master
mode.
For short frame synchronization, the WS synchronization signal is only one cycle long.

Figure 301. PCM standard waveforms (16-bit extended to 32-bit packet frame)

CK

WS
short frame
Up to 13-bits
WS
long frame
16 bits

SD MSB LSB

MS30107V1

Note: For both modes (master and slave) and for both synchronizations (short and long), the
number of bits between two consecutive pieces of data (and so two synchronization signals)
needs to be specified (DATLEN and CHLEN bits in the SPIx_I2SCFGR register) even in
slave mode.

26.6.4 Clock generator

The I2S bitrate determines the data flow on the I2S data line and the I2S clock signal
frequency.
I2S bitrate = number of bits per channel × number of channels × sampling audio frequency
For a 16-bit audio, left and right channel, the I2S bitrate is calculated as follows:
I2S bitrate = 16 × 2 × fS
It will be: I2S bitrate = 32 x 2 x fS if the packet length is 32-bit wide.

RM0402 Rev 6 843/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Figure 302. Audio sampling frequency definition

16-or 32-bit left 16-or 32-bit

channel right channel

32- or 64-bits
FS
sampling point sampling point

FS : audio sampling frequency

MS30108V1

When the master mode is configured, a specific action needs to be taken to properly
program the linear divider in order to communicate with the desired audio frequency.
Figure 303 presents the communication clock architecture. The I2Sx clock is always the
system clock.

Figure 303. I2S clock generator architecture

MCK

I²SxCLK 8-bit linear divider

+ reshaping stage 0
Div2 0
Divider by 4 1 CK
1

MCKOE
I²SMOD

CHLEN
MCKOE ODD I²SDIV[7:0]

MS30109V1

1. Where x = 2.
The audio sampling frequency may be 192 KHz, 96 kHz, 48 kHz, 44.1 kHz, 32 kHz,
22.05 kHz, 16 kHz, 11.025 kHz or 8 kHz (or any other value within this range). In order to
reach the desired frequency, the linear divider needs to be programmed according to the
formulas below:
When the master clock is generated (MCKOE in the SPIx_I2SPR register is set):
fS = I2SxCLK / [(16*2)*((2*I2SDIV)+ODD)*8)] when the channel frame is 16-bit wide
fS = I2SxCLK / [(32*2)*((2*I2SDIV)+ODD)*4)] when the channel frame is 32-bit wide
When the master clock is disabled (MCKOE bit cleared):
fS = I2SxCLK / [(16*2)*((2*I2SDIV)+ODD))] when the channel frame is 16-bit wide
fS = I2SxCLK / [(32*2)*((2*I2SDIV)+ODD))] when the channel frame is 32-bit wide
Table 157 provides example precision values for different clock configurations.
Note: Other configurations are possible that allow optimum clock precision.

844/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Table 157. Audio-frequency precision using standard 8 MHz HSE(1)

SYSCLK Data Target fS
I2SDIV I2SODD MCLK Real fS (KHz) Error
(MHz) length (Hz)

48 16 8 0 No 96000 93750 2.3438%

48 32 4 0 No 96000 93750 2.3438%
48 16 15 1 No 48000 48387.0968 0.8065%
48 32 8 0 No 48000 46875 2.3438%
48 16 17 0 No 44100 44117.647 0.0400%
48 32 8 1 No 44100 44117.647 0.0400%
48 16 23 1 No 32000 31914.8936 0.2660%
48 32 11 1 No 32000 32608.696 1.9022%
48 16 34 0 No 22050 22058.8235 0.0400%
48 32 17 0 No 22050 22058.8235 0.0400%
48 16 47 0 No 16000 15957.4468 0.2660%
48 32 23 1 No 16000 15957.447 0.2660%
48 16 68 0 No 11025 11029.4118 0.0400%
48 32 34 0 No 11025 11029.412 0.0400%
48 16 94 0 No 8000 7978.7234 0.2660%
48 32 47 0 No 8000 7978.7234 0.2660%
48 16 2 0 Yes 48000 46875 2.3430%
48 32 2 0 Yes 48000 46875 2.3430%
48 16 2 0 Yes 44100 46875 6.2925%
48 32 2 0 Yes 44100 46875 6.2925%
48 16 3 0 Yes 32000 31250 2.3438%
48 32 3 0 Yes 32000 31250 2.3438%
48 16 4 1 Yes 22050 20833.333 5.5178%
48 32 4 1 Yes 22050 20833.333 5.5178%
48 16 6 0 Yes 16000 15625 2.3438%
48 32 6 0 Yes 16000 15625 2.3438%
48 16 8 1 Yes 11025 11029.4118 0.0400%
48 32 8 1 Yes 11025 11029.4118 0.0400%
48 16 11 1 Yes 8000 8152.17391 1.9022%
48 32 11 1 Yes 8000 8152.17391 1.9022%
1. This table gives only example values for different clock configurations. Other configurations allowing
optimum clock precision are possible.

RM0402 Rev 6 845/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.6.5 I2S master mode

The I2S can be configured as follows:
• In master mode for transmission or reception (half-duplex mode using I2Sx)
• In master mode transmission and reception (full-duplex mode using I2Sx and
I2Sx_ext).
This means that the serial clock is generated on the CK pin as well as the Word Select
signal WS. Master clock (MCK) may be output or not, controlled by the MCKOE bit in the
SPIx_I2SPR register.

Procedure
1. Select the I2SDIV[7:0] bits in the SPIx_I2SPR register to define the serial clock baud
rate to reach the proper audio sample frequency. The ODD bit in the SPIx_I2SPR
register also has to be defined.
2. Select the CKPOL bit to define the steady level for the communication clock. Set the
MCKOE bit in the SPIx_I2SPR register if the master clock MCK needs to be provided
to the external ADC audio component (the I2SDIV and ODD values should be
computed depending on the state of the MCK output, for more details refer to
Section 26.6.4: Clock generator).
3. Set the I2SMOD bit in the SPIx_I2SCFGR register to activate the I2S functions and
choose the I2S standard through the I2SSTD[1:0] and PCMSYNC bits, the data length
through the DATLEN[1:0] bits and the number of bits per channel by configuring the
CHLEN bit. Select also the I2S master mode and direction (Transmitter or Receiver)
through the I2SCFG[1:0] bits in the SPIx_I2SCFGR register.
4. If needed, select all the potential interrupt sources and the DMA capabilities by writing
the SPIx_CR2 register.
5. The I2SE bit in SPIx_I2SCFGR register must be set.
WS and CK are configured in output mode. MCK is also an output, if the MCKOE bit in
SPIx_I2SPR is set.

Transmission sequence
The transmission sequence begins when a half-word is written into the Tx buffer.
Lets assume the first data written into the Tx buffer corresponds to the left channel data.
When data are transferred from the Tx buffer to the shift register, TXE is set and data
corresponding to the right channel have to be written into the Tx buffer. The CHSIDE flag
indicates which channel is to be transmitted. It has a meaning when the TXE flag is set
because the CHSIDE flag is updated when TXE goes high.
A full frame has to be considered as a left channel data transmission followed by a right
channel data transmission. It is not possible to have a partial frame where only the left
channel is sent.
The data half-word is parallel loaded into the 16-bit shift register during the first bit
transmission, and then shifted out, serially, to the MOSI/SD pin, MSB first. The TXE flag is
set after each transfer from the Tx buffer to the shift register and an interrupt is generated if
the TXEIE bit in the SPIx_CR2 register is set.
For more details about the write operations depending on the I2S Standard-mode selected,
refer to Section 26.6.3: Supported audio protocols).

846/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

To ensure a continuous audio data transmission, it is mandatory to write the SPIx_DR

register with the next data to transmit before the end of the current transmission.
To switch off the I2S, by clearing I2SE, it is mandatory to wait for TXE = 1 and BSY = 0.

Reception sequence
The operating mode is the same as for transmission mode except for the point 3 (refer to the
procedure described in Section 26.6.5: I2S master mode), where the configuration should
set the master reception mode through the I2SCFG[1:0] bits.
Whatever the data or channel length, the audio data are received by 16-bit packets. This
means that each time the Rx buffer is full, the RXNE flag is set and an interrupt is generated
if the RXNEIE bit is set in SPIx_CR2 register. Depending on the data and channel length
configuration, the audio value received for a right or left channel may result from one or two
receptions into the Rx buffer.
Clearing the RXNE bit is performed by reading the SPIx_DR register.
CHSIDE is updated after each reception. It is sensitive to the WS signal generated by the
I2S cell.
For more details about the read operations depending on the I2S Standard-mode selected,
refer to Section 26.6.3: Supported audio protocols.
If data are received while the previously received data have not been read yet, an overrun is
generated and the OVR flag is set. If the ERRIE bit is set in the SPIx_CR2 register, an
interrupt is generated to indicate the error.
To switch off the I2S, specific actions are required to ensure that the I2S completes the
transfer cycle properly without initiating a new data transfer. The sequence depends on the
configuration of the data and channel lengths, and on the audio protocol mode selected. In
the case of:
• 16-bit data length extended on 32-bit channel length (DATLEN = 00 and CHLEN = 1)
using the LSB justified mode (I2SSTD = 10)
a) Wait for the second to last RXNE = 1 (n – 1)
b) Then wait 17 I2S clock cycles (using a software loop)
c) Disable the I2S (I2SE = 0)
• 16-bit data length extended on 32-bit channel length (DATLEN = 00 and CHLEN = 1) in
MSB justified, I2S or PCM modes (I2SSTD = 00, I2SSTD = 01 or I2SSTD = 11,
respectively)
a) Wait for the last RXNE
b) Then wait 1 I2S clock cycle (using a software loop)
c) Disable the I2S (I2SE = 0)
• For all other combinations of DATLEN and CHLEN, whatever the audio mode selected
through the I2SSTD bits, carry out the following sequence to switch off the I2S:
a) Wait for the second to last RXNE = 1 (n – 1)
b) Then wait one I2S clock cycle (using a software loop)
c) Disable the I2S (I2SE = 0)
Note: The BSY flag is kept low during transfers.

RM0402 Rev 6 847/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.6.6 I2S slave mode

The I2S can be configured as follows:
• In slave mode for transmission or reception (half-duplex mode using I2Sx)
• In slave mode transmission and reception (full-duplex mode using I2Sx and I2Sx_ext).
The operating mode is following mainly the same rules as described for the I2S master
configuration. In slave mode, there is no clock to be generated by the I2S interface. The
clock and WS signals are input from the external master connected to the I2S interface.
There is then no need, for the user, to configure the clock.
The configuration steps to follow are listed below:
1. Set the I2SMOD bit in the SPIx_I2SCFGR register to select I2S mode and choose the
I2S standard through the I2SSTD[1:0] bits, the data length through the DATLEN[1:0]
bits and the number of bits per channel for the frame configuring the CHLEN bit. Select
also the mode (transmission or reception) for the slave through the I2SCFG[1:0] bits in
SPIx_I2SCFGR register.
2. If needed, select all the potential interrupt sources and the DMA capabilities by writing
the SPIx_CR2 register.
3. The I2SE bit in SPIx_I2SCFGR register must be set.

Transmission sequence
The transmission sequence begins when the external master device sends the clock and
when the NSS_WS signal requests the transfer of data. The slave has to be enabled before
the external master starts the communication. The I2S data register has to be loaded before
the master initiates the communication.
For the I2S, MSB justified and LSB justified modes, the first data item to be written into the
data register corresponds to the data for the left channel. When the communication starts,
the data are transferred from the Tx buffer to the shift register. The TXE flag is then set in
order to request the right channel data to be written into the I2S data register.
The CHSIDE flag indicates which channel is to be transmitted. Compared to the master
transmission mode, in slave mode, CHSIDE is sensitive to the WS signal coming from the
external master. This means that the slave needs to be ready to transmit the first data
before the clock is generated by the master. WS assertion corresponds to left channel
transmitted first.
Note: The I2SE has to be written at least two PCLK cycles before the first clock of the master
comes on the CK line.
The data half-word is parallel-loaded into the 16-bit shift register (from the internal bus)
during the first bit transmission, and then shifted out serially to the MOSI/SD pin MSB first.
The TXE flag is set after each transfer from the Tx buffer to the shift register and an interrupt
is generated if the TXEIE bit in the SPIx_CR2 register is set.
Note that the TXE flag should be checked to be at 1 before attempting to write the Tx buffer.
For more details about the write operations depending on the I2S Standard-mode selected,
refer to Section 26.6.3: Supported audio protocols.

848/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

To secure a continuous audio data transmission, it is mandatory to write the SPIx_DR

register with the next data to transmit before the end of the current transmission. An
underrun flag is set and an interrupt may be generated if the data are not written into the
SPIx_DR register before the first clock edge of the next data communication. This indicates
to the software that the transferred data are wrong. If the ERRIE bit is set into the SPIx_CR2
register, an interrupt is generated when the UDR flag in the SPIx_SR register goes high. In
this case, it is mandatory to switch off the I2S and to restart a data transfer starting from the
left channel.
To switch off the I2S, by clearing the I2SE bit, it is mandatory to wait for TXE = 1 and
BSY = 0.

Reception sequence
The operating mode is the same as for the transmission mode except for the point 1 (refer to
the procedure described in Section 26.6.6: I2S slave mode), where the configuration should
set the master reception mode using the I2SCFG[1:0] bits in the SPIx_I2SCFGR register.
Whatever the data length or the channel length, the audio data are received by 16-bit
packets. This means that each time the RX buffer is full, the RXNE flag in the SPIx_SR
register is set and an interrupt is generated if the RXNEIE bit is set in the SPIx_CR2
register. Depending on the data length and channel length configuration, the audio value
received for a right or left channel may result from one or two receptions into the RX buffer.
The CHSIDE flag is updated each time data are received to be read from the SPIx_DR
register. It is sensitive to the external WS line managed by the external master component.
Clearing the RXNE bit is performed by reading the SPIx_DR register.
For more details about the read operations depending the I2S Standard-mode selected,
refer to Section 26.6.3: Supported audio protocols.
If data are received while the preceding received data have not yet been read, an overrun is
generated and the OVR flag is set. If the bit ERRIE is set in the SPIx_CR2 register, an
interrupt is generated to indicate the error.
To switch off the I2S in reception mode, I2SE has to be cleared immediately after receiving
the last RXNE = 1.
Note: The external master components should have the capability of sending/receiving data in 16-
bit or 32-bit packets via an audio channel.

26.6.7 I2S status flags

Three status flags are provided for the application to fully monitor the state of the I2S bus.

Busy flag (BSY)

The BSY flag is set and cleared by hardware (writing to this flag has no effect). It indicates
the state of the communication layer of the I2S.
When BSY is set, it indicates that the I2S is busy communicating. There is one exception in
master receive mode (I2SCFG = 11) where the BSY flag is kept low during reception.
The BSY flag is useful to detect the end of a transfer if the software needs to disable the I2S.
This avoids corrupting the last transfer. For this, the procedure described below must be
strictly respected.
The BSY flag is set when a transfer starts, except when the I2S is in master receiver mode.

RM0402 Rev 6 849/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

The BSY flag is cleared:

• When a transfer completes (except in master transmit mode, in which the
communication is supposed to be continuous)
• When the I2S is disabled
When communication is continuous:
• In master transmit mode, the BSY flag is kept high during all the transfers
• In slave mode, the BSY flag goes low for one I2S clock cycle between each transfer
Note: Do not use the BSY flag to handle each data transmission or reception. It is better to use the
TXE and RXNE flags instead.

Tx buffer empty flag (TXE)

When set, this flag indicates that the Tx buffer is empty and the next data to be transmitted
can then be loaded into it. The TXE flag is reset when the Tx buffer already contains data to
be transmitted. It is also reset when the I2S is disabled (I2SE bit is reset).

RX buffer not empty (RXNE)

When set, this flag indicates that there are valid received data in the RX Buffer. It is reset
when SPIx_DR register is read.

Channel Side flag (CHSIDE)

In transmission mode, this flag is refreshed when TXE goes high. It indicates the channel
side to which the data to transfer on SD has to belong. In case of an underrun error event in
slave transmission mode, this flag is not reliable and I2S needs to be switched off and
switched on before resuming the communication.
In reception mode, this flag is refreshed when data are received into SPIx_DR. It indicates
from which channel side data have been received. Note that in case of error (like OVR) this
flag becomes meaningless and the I2S should be reset by disabling and then enabling it
(with configuration if it needs changing).
This flag has no meaning in the PCM standard (for both Short and Long frame modes).
When the OVR or UDR flag in the SPIx_SR is set and the ERRIE bit in SPIx_CR2 is also
set, an interrupt is generated. This interrupt can be cleared by reading the SPIx_SR status
register (once the interrupt source has been cleared).

26.6.8 I2S error flags

There are three error flags for the I2S cell.

Underrun flag (UDR)

In slave transmission mode this flag is set when the first clock for data transmission appears
while the software has not yet loaded any value into SPIx_DR. It is available when the
I2SMOD bit in the SPIx_I2SCFGR register is set. An interrupt may be generated if the
ERRIE bit in the SPIx_CR2 register is set.
The UDR bit is cleared by a read operation on the SPIx_SR register.

850/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Overrun flag (OVR)

This flag is set when data are received and the previous data have not yet been read from
the SPIx_DR register. As a result, the incoming data are lost. An interrupt may be generated
if the ERRIE bit is set in the SPIx_CR2 register.
In this case, the receive buffer contents are not updated with the newly received data from
the transmitter device. A read operation to the SPIx_DR register returns the previous
correctly received data. All other subsequently transmitted half-words are lost.
Clearing the OVR bit is done by a read operation on the SPIx_DR register followed by a
read access to the SPIx_SR register.

Frame error flag (FRE)

This flag can be set by hardware only if the I2S is configured in Slave mode. It is set if the
external master is changing the WS line while the slave is not expecting this change. If the
synchronization is lost, the following steps are required to recover from this state and
resynchronize the external master device with the I2S slave device:
1. Disable the I2S.
2. Enable it again when the correct level is detected on the WS line (WS line is high in I2S
mode or low for MSB- or LSB-justified or PCM modes.
Desynchronization between master and slave devices may be due to noisy environment on
the SCK communication clock or on the WS frame synchronization line. An error interrupt
can be generated if the ERRIE bit is set. The desynchronization flag (FRE) is cleared by
software when the status register is read.

26.6.9 I2S interrupts

Table 158 provides the list of I2S interrupts.

Table 158. I2S interrupt requests

Interrupt event Event flag Enable control bit

Transmit buffer empty flag TXE TXEIE

Receive buffer not empty flag RXNE RXNEIE
Overrun error OVR
Underrun error UDR ERRIE
Frame error flag FRE

26.6.10 DMA features

In I2S mode, the DMA works in exactly the same way as it does in SPI mode. There is no
difference except that the CRC feature is not available in I2S mode since there is no data
transfer protection system.

RM0402 Rev 6 851/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.7 SPI and I2S registers

The peripheral registers can be accessed by half-words (16-bit) or words (32-bit). SPI_DR
in addition by can be accessed by 8-bit access.
Refer to Section 1.2 for a list of abbreviations used in register descriptions.
The peripheral registers can be accessed by half-words (16 bits) or words (32 bits).

26.7.1 SPI control register 1 (SPI_CR1) (not used in I2S mode)

Address offset: 0x00
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
BIDI BIDI CRC CRC RX LSB
DFF SSM SSI SPE BR [2:0] MSTR CPOL CPHA
MODE OE EN NEXT ONLY FIRST
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 15 BIDIMODE: Bidirectional data mode enable

This bit enables half-duplex communication using common single bidirectional data line.
Keep RXONLY bit clear when bidirectional mode is active.
0: 2-line unidirectional data mode selected
1: 1-line bidirectional data mode selected
Note: This bit is not used in I2S mode
Bit 14 BIDIOE: Output enable in bidirectional mode
This bit combined with the BIDIMODE bit selects the direction of transfer in bidirectional
mode
0: Output disabled (receive-only mode)
1: Output enabled (transmit-only mode)
Note: In master mode, the MOSI pin is used while the MISO pin is used in slave mode.
This bit is not used in I2S mode.
Bit 13 CRCEN: Hardware CRC calculation enable
0: CRC calculation disabled
1: CRC calculation enabled
Note: This bit should be written only when SPI is disabled (SPE = ‘0’) for correct operation.
It is not used in I2S mode.
Bit 12 CRCNEXT: CRC transfer next
0: Data phase (no CRC phase)
1: Next transfer is CRC (CRC phase)
Note: When the SPI is configured in full-duplex or transmitter only modes, CRCNEXT must be
written as soon as the last data is written to the SPI_DR register.
When the SPI is configured in receiver only mode, CRCNEXT must be set after the
second last data reception.
This bit should be kept cleared when the transfers are managed by DMA.
It is not used in I2S mode.

852/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Bit 11 DFF: Data frame format

0: 8-bit data frame format is selected for transmission/reception
1: 16-bit data frame format is selected for transmission/reception
Note: This bit should be written only when SPI is disabled (SPE = ‘0’) for correct operation.
It is not used in I2S mode.
Bit 10 RXONLY: Receive only mode enable
This bit enables simplex communication using a single unidirectional line to receive data
exclusively. Keep BIDIMODE bit clear when receive only mode is active.
This bit is also useful in a multislave system in which this particular slave is not accessed, the
output from the accessed slave is not corrupted.
0: full-duplex (Transmit and receive)
1: Output disabled (Receive-only mode)
Note: This bit is not used in I2S mode
Bit 9 SSM: Software slave management
When the SSM bit is set, the NSS pin input is replaced with the value from the SSI bit.
0: Software slave management disabled
1: Software slave management enabled
Note: This bit is not used in I2S mode and SPI TI mode
Bit 8 SSI: Internal slave select
This bit has an effect only when the SSM bit is set. The value of this bit is forced onto the
NSS pin and the IO value of the NSS pin is ignored.
Note: This bit is not used in I2S mode and SPI TI mode
Bit 7 LSBFIRST: Frame format
0: MSB transmitted first
1: LSB transmitted first
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode
Bit 6 SPE: SPI enable
0: Peripheral disabled
1: Peripheral enabled
Note: When disabling the SPI, follow the procedure described in Section 26.3.10: Procedure
for disabling the SPI.
This bit is not used in I2S mode.
Bits 5:3 BR[2:0]: Baud rate control
000: fPCLK/2
001: fPCLK/4
010: fPCLK/8
011: fPCLK/16
100: fPCLK/32
101: fPCLK/64
110: fPCLK/128
111: fPCLK/256
Note: These bits should not be changed when communication is ongoing.
They are not used in I2S mode.

RM0402 Rev 6 853/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Bit 2 MSTR: Master selection

0: Slave configuration
1: Master configuration
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode.
Bit1 CPOL: Clock polarity
0: CK to 0 when idle
1: CK to 1 when idle
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode except the case when CRC is applied
at TI mode.
Bit 0 CPHA: Clock phase
0: The first clock transition is the first data capture edge
1: The second clock transition is the first data capture edge
Note: This bit should not be changed when communication is ongoing.
It is not used in I2S mode and SPI TI mode except the case when CRC is applied
at TI mode.

26.7.2 SPI control register 2 (SPI_CR2)

Address offset: 0x04
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. TXEIE RXNEIE ERRIE FRF Res. SSOE TXDMAEN RXDMAEN

rw rw rw rw rw rw rw

Bits 15:8 Reserved, must be kept at reset value.

Bit 7 TXEIE: Tx buffer empty interrupt enable
0: TXE interrupt masked
1: TXE interrupt not masked. Used to generate an interrupt request when the TXE flag is set.
Bit 6 RXNEIE: RX buffer not empty interrupt enable
0: RXNE interrupt masked
1: RXNE interrupt not masked. Used to generate an interrupt request when the RXNE flag is
set.
Bit 5 ERRIE: Error interrupt enable
This bit controls the generation of an interrupt when an error condition occurs (OVR,
CRCERR, MODF, FRE in SPI mode, and UDR, OVR, FRE in I2S mode).
0: Error interrupt is masked
1: Error interrupt is enabled
Bit 4 FRF: Frame format
0: SPI Motorola mode
1 SPI TI mode
Note: This bit is not used in I2S mode.
Bit 3 Reserved. Forced to 0 by hardware.

854/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Bit 2 SSOE: SS output enable

0: SS output is disabled in master mode and the cell can work in multimaster configuration
1: SS output is enabled in master mode and when the cell is enabled. The cell cannot work
in a multimaster environment.
Note: This bit is not used in I2S mode and SPI TI mode.
Bit 1 TXDMAEN: Tx buffer DMA enable
When this bit is set, the DMA request is made whenever the TXE flag is set.
0: Tx buffer DMA disabled
1: Tx buffer DMA enabled
Bit 0 RXDMAEN: Rx buffer DMA enable
When this bit is set, the DMA request is made whenever the RXNE flag is set.
0: Rx buffer DMA disabled
1: Rx buffer DMA enabled

26.7.3 SPI status register (SPI_SR)

Address offset: 0x08
Reset value: 0x0002
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRC
Res. Res. Res. Res. Res. Res. Res. FRE BSY OVR MODF UDR CHSIDE TXE RXNE
ERR
r r r r rc_w0 r r r r

Bits 15:9 Reserved. Forced to 0 by hardware.

Bit 8 FRE: Frame Error
0: No frame error
1: Frame error occurred.
This bit is set by hardware and cleared by software when the SPI_SR register is read.
This bit is used in SPI TI mode or in I2S mode whatever the audio protocol selected. It
detects a change on NSS or WS line which takes place in slave mode at a non expected
time, informing about a desynchronization between the external master device and the
slave.
Bit 7 BSY: Busy flag
0: SPI (or I2S) not busy
1: SPI (or I2S) is busy in communication or Tx buffer is not empty
This flag is set and cleared by hardware.
Note: BSY flag must be used with caution: refer to Section 26.3.12: SPI status flags and
Section 26.3.10: Procedure for disabling the SPI.
Bit 6 OVR: Overrun flag
0: No overrun occurred
1: Overrun occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 26.3.13: SPI
error flags for the software sequence.

RM0402 Rev 6 855/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Bit 5 MODF: Mode fault

0: No mode fault occurred
1: Mode fault occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 26.4 on
page 831 for the software sequence.
Note: This bit is not used in I2S mode
Bit 4 CRCERR: CRC error flag
0: CRC value received matches the SPI_RXCRCR value
1: CRC value received does not match the SPI_RXCRCR value
This flag is set by hardware and cleared by software writing 0.
Note: This bit is not used in I2S mode.
Bit 3 UDR: Underrun flag
0: No underrun occurred
1: Underrun occurred
This flag is set by hardware and reset by a software sequence. Refer to Section 26.6.8: I2S
error flags for the software sequence.
Note: This bit is not used in SPI mode.
Bit 2 CHSIDE: Channel side
0: Channel Left has to be transmitted or has been received
1: Channel Right has to be transmitted or has been received
Note: This bit is not used for SPI mode and is meaningless in PCM mode.
Bit 1 TXE: Transmit buffer empty
0: Tx buffer not empty
1: Tx buffer empty
Bit 0 RXNE: Receive buffer not empty
0: Rx buffer empty
1: Rx buffer not empty

856/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.7.4 SPI data register (SPI_DR)

Address offset: 0x0C
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DR[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 DR[15:0]: Data register

Data received or to be transmitted.
The data register is split into 2 buffers - one for writing (Transmit Buffer) and another one for
reading (Receive buffer). A write to the data register will write into the Tx buffer and a read
from the data register will return the value held in the Rx buffer.
Note: These notes apply to SPI mode:
Depending on the data frame format selection bit (DFF in SPI_CR1 register), the data
sent or received is either 8-bit or 16-bit. This selection has to be made before enabling
the SPI to ensure correct operation.
For an 8-bit data frame, the buffers are 8-bit and only the LSB of the register
(SPI_DR[7:0]) is used for transmission/reception. When in reception mode, the MSB of
the register (SPI_DR[15:8]) is forced to 0.
For a 16-bit data frame, the buffers are 16-bit and the entire register, SPI_DR[15:0] is
used for transmission/reception.

26.7.5 SPI CRC polynomial register (SPI_CRCPR) (not used in I2S

mode)
Address offset: 0x10
Reset value: 0x0007
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CRCPOLY[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:0 CRCPOLY[15:0]: CRC polynomial register

This register contains the polynomial for the CRC calculation.
The CRC polynomial (0007h) is the reset value of this register. Another polynomial can be
configured as required.
Note: These bits are not used for the I2S mode.

RM0402 Rev 6 857/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

26.7.6 SPI RX CRC register (SPI_RXCRCR) (not used in I2S mode)

Address offset: 0x14
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RXCRC[15:0]
r r r r r r r r r r r r r r r r

Bits 15:0 RXCRC[15:0]: Rx CRC register

When CRC calculation is enabled, the RxCRC[15:0] bits contain the computed CRC value of
the subsequently received bytes. This register is reset when the CRCEN bit in SPI_CR1
register is written to 1. The CRC is calculated serially using the polynomial programmed in
the SPI_CRCPR register.
Only the 8 LSB bits are considered when the data frame format is set to be 8-bit data (DFF
bit of SPI_CR1 is cleared). CRC calculation is done based on any CRC8 standard.
The entire 16-bits of this register are considered when a 16-bit data frame format is selected
(DFF bit of the SPI_CR1 register is set). CRC calculation is done based on any CRC16
standard.
Note: A read to this register when the BSY Flag is set could return an incorrect value.These
bits are not used for I2S mode.

26.7.7 SPI TX CRC register (SPI_TXCRCR) (not used in I2S mode)

Address offset: 0x18
Reset value: 0x0000
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXCRC[15:0]
r r r r r r r r r r r r r r r r

Bits 15:0 TXCRC[15:0]: Tx CRC register

When CRC calculation is enabled, the TxCRC[7:0] bits contain the computed CRC value of
the subsequently transmitted bytes. This register is reset when the CRCEN bit of SPI_CR1
is written to 1. The CRC is calculated serially using the polynomial programmed in the
SPI_CRCPR register.
Only the 8 LSB bits are considered when the data frame format is set to be 8-bit data (DFF
bit of SPI_CR1 is cleared). CRC calculation is done based on any CRC8 standard.
The entire 16-bits of this register are considered when a 16-bit data frame format is selected
(DFF bit of the SPI_CR1 register is set). CRC calculation is done based on any CRC16
standard.
Note: A read to this register when the BSY flag is set could return an incorrect value. These
bits are not used for I2S mode.

858/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

26.7.8 SPI_I2S configuration register (SPI_I2SCFGR)

Address offset: 0x1C
Reset value: 0x0000

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ASTRE PCMSY
Res. Res. Res. I2SMOD I2SE I2SCFG Res. I2SSTD CKPOL DATLEN CHLEN
N NC
rw rw rw rw rw rw rw rw rw rw rw rw

Bits 15:13 Reserved, must be kept at reset value.

Bit 12 ASTREN: Asynchronous start enable.
0: The Asynchronous start is disabled. When the I2S is enabled in slave mode, the I2S slave starts
the transfer when the I2S clock is received and an appropriate transition (depending on the protocol
selected) is detected on the WS signal.
1: The Asynchronous start is enabled. When the I2S is enabled in slave mode, the I2S slave starts
immediately the transfer when the I2S clock is received from the master without checking the
expected transition of WS signal.
Note: Note: The appropriate transition is a falling edge on WS signal when I2S Philips Standard is
used, or a rising edge for other standards.
Bit 11 I2SMOD: I2S mode selection
0: SPI mode is selected
1: I2S mode is selected
Note: This bit should be configured when the SPI or I2S is disabled
Bit 10 I2SE: I2S Enable
0: I2S peripheral is disabled
1: I2S peripheral is enabled
Note: This bit is not used in SPI mode.
Bits 9:8 I2SCFG: I2S configuration mode
00: Slave - transmit
01: Slave - receive
10: Master - transmit
11: Master - receive
Note: This bit should be configured when the I2S is disabled.
It is not used in SPI mode.
Bit 7 PCMSYNC: PCM frame synchronization
0: Short frame synchronization
1: Long frame synchronization
Note: This bit has a meaning only if I2SSTD = 11 (PCM standard is used)
It is not used in SPI mode.
Bit 6 Reserved: forced at 0 by hardware

RM0402 Rev 6 859/1163

862
Serial peripheral interface/ inter-IC sound (SPI/I2S) RM0402

Bits 5:4 I2SSTD: I2S standard selection

00: I2S Philips standard.
01: MSB justified standard (left justified)
10: LSB justified standard (right justified)
11: PCM standard
For more details on I2S standards, refer to Section 26.6.3 on page 837. Not used in SPI mode.
Note: For correct operation, these bits should be configured when the I2S is disabled.
Bit 3 CKPOL: Steady state clock polarity
0: I2S clock steady state is low level
1: I2S clock steady state is high level
Note: For correct operation, this bit should be configured when the I2S is disabled.
This bit is not used in SPI mode
Bits 2:1 DATLEN: Data length to be transferred
00: 16-bit data length
01: 24-bit data length
10: 32-bit data length
11: Not allowed
Note: For correct operation, these bits should be configured when the I2S is disabled.
This bit is not used in SPI mode.
Bit 0 CHLEN: Channel length (number of bits per audio channel)
0: 16-bit wide
1: 32-bit wide
The bit write operation has a meaning only if DATLEN = 00 otherwise the channel length is fixed to
32-bit by hardware whatever the value filled in. Not used in SPI mode.
Note: For correct operation, this bit should be configured when the I2S is disabled.

26.7.9 SPI_I2S prescaler register (SPI_I2SPR)

Address offset: 0x20
Reset value: 0000 0010 (0x0002)
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. MCKOE ODD I2SDIV
rw rw rw

Bits 15:10 Reserved, must be kept at reset value.

860/1163 RM0402 Rev 6

RM0402 Serial peripheral interface/ inter-IC sound (SPI/I2S)

Bit 9 MCKOE: Master clock output enable

0: Master clock output is disabled
1: Master clock output is enabled
Note: This bit should be configured when the I2S is disabled. It is used only when the I2S is in master
mode.
This bit is not used in SPI mode.
Bit 8 ODD: Odd factor for the prescaler
0: real divider value is = I2SDIV *2
1: real divider value is = (I2SDIV * 2)+1
Refer to Section 26.6.4 on page 843. Not used in SPI mode.
Note: This bit should be configured when the I2S is disabled. It is used only when the I2S is in master
mode.
Bits 7:0 I2SDIV: I2S Linear prescaler
I2SDIV [7:0] = 0 or I2SDIV [7:0] = 1 are forbidden values.
Refer to Section 26.6.4 on page 843. Not used in SPI mode.
Note: These bits should be configured when the I2S is disabled. It is used only when the I2S is in
master mode.

RM0402 Rev 6 861/1163

862
 0x20
0x18
0x14
0x10
0x08
0x04
0x00

0x1C
0x0C
Offset

862/1163
26.7.10

SPI_SR

SPI_DR
Register

SPI_CR2
SPI_CR1

SPI_I2SPR

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

Reset value

SPI_CRCPR

SPI_TXCRCR
SPI_RXCRCR

SPI_I2SCFGR
Res. Res. Res. Res. Res. Res. Res. Res. Res. 31
Res. Res. Res. Res. Res. Res. Res. Res. Res. 30
Res. Res. Res. Res. Res. Res. Res. Res. Res. 29
Res. Res. Res. Res. Res. Res. Res. Res. Res. 28
SPI register map

Res. Res. Res. Res. Res. Res. Res. Res. Res. 27

Res. Res. Res. Res. Res. Res. Res. Res. Res. 26
Res. Res. Res. Res. Res. Res. Res. Res. Res. 25
Res. Res. Res. Res. Res. Res. Res. Res. Res. 24
Res. Res. Res. Res. Res. Res. Res. Res. Res. 23
Res. Res. Res. Res. Res. Res. Res. Res. Res. 22
Res. Res. Res. Res. Res. Res. Res. Res. Res. 21
Res. Res. Res. Res. Res. Res. Res. Res. Res. 20
Serial peripheral interface/ inter-IC sound (SPI/I2S)

Res. Res. Res. Res. Res. Res. Res. Res. Res. 19

Res. Res. Res. Res. Res. Res. Res. Res. Res. 18

RM0402 Rev 6
Res. Res. Res. Res. Res. Res. Res. Res. Res. 17
Res. Res. Res. Res. Res. Res. Res. Res. Res. 16
Res. Res. Res. Res. BIDIMODE 15
Res. Res. Res. Res. BIDIOE 14
Res. Res. Res. Res. CRCEN 13
Res. ASTREN Res. Res. CRCNEXT 12
Table 159. SPI register map and reset values

Res. I2SMOD Res. Res. DFF 11

The table provides shows the SPI register map and reset values.

Res. I2SE Res. Res. RXONLY 10

MCKOE Res. Res. SSM

Refer to Section 2.2 on page 49 for the register boundary addresses.

I2SCFG 9
ODD FRE Res. SSI 8

0 0 0 0 0 0
PCMSYNC BSY TXEIE LSBFIRST 7
DR[15:0]

Res. OVR RXNEIE SPE 6

TxCRC[15:0]
RxCRC[15:0]
MODF ERRIE
CRCPOLY[15:0]

I2SSTD 5
0 0 0 0

CRCERR FRF
BR

4
[2:0]

CKPOL UDR Res. 3

I2SDIV
CHSIDE SSOE MSTR 2
DATLEN
TXE TXDMAEN CPOL 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 1 0
0 0 0 0 0 0
0 0 0 0 0 0 0 1 0
0 0 0
RM0402

CHLEN RXNE RXDMAEN CPHA 0

RM0402 Secure digital input/output interface (SDIO)

27 Secure digital input/output interface (SDIO)

27.1 SDIO main features

The SD/SDIO MMC card host interface (SDIO) provides an interface between the APB2
peripheral bus and MultiMediaCards (MMCs), SD memory cards and SDIO cards.
The MultiMediaCard system specifications are available through the MultiMediaCard
Association website, published by the MMCA technical committee.
SD memory card and SD I/O card system specifications are available through the SD card
Association website.
The SDIO features include the following:
• Full compliance with MultiMediaCard System Specification Version 4.2. Card support
for three different databus modes: 1-bit (default), 4-bit and 8-bit
• Full compatibility with previous versions of MultiMediaCards (forward compatibility)
• Full compliance with SD Memory Card Specifications Version 2.0
• Full compliance with SD I/O Card Specification Version 2.0: card support for two
different databus modes: 1-bit (default) and 4-bit
• Data transfer up to 50 MHz for the 8 bit mode
• Data and command output enable signals to control external bidirectional drivers.
Note: 1 The SDIO does not have an SPI-compatible communication mode.
2 The SD memory card protocol is a superset of the MultiMediaCard protocol as defined in the
MultiMediaCard system specification V2.11. Several commands required for SD memory
devices are not supported by either SD I/O-only cards or the I/O portion of combo cards.
Some of these commands have no use in SD I/O devices, such as erase commands, and
thus are not supported in the SDIO protocol. In addition, several commands are different
between SD memory cards and SD I/O cards and thus are not supported in the SDIO
protocol. For details refer to SD I/O card Specification Version 1.0.
The MultiMediaCard/SD bus connects cards to the controller.
The current version of the SDIO supports only one SD/SDIO/MMC4.2 card at any one time
and a stack of MMC4.1 or previous.

27.2 SDIO bus topology

Communication over the bus is based on command and data transfers.
The basic transaction on the MultiMediaCard/SD/SD I/O bus is the command/response
transaction. These types of bus transaction transfer their information directly within the
command or response structure. In addition, some operations have a data token.
Data transfers to/from SD/SDIO memory cards are done in data blocks. Data transfers
to/from MMC are done data blocks or streams.

RM0402 Rev 6 863/1163

921
Secure digital input/output interface (SDIO) RM0402

Figure 304. “No response” and “no data” operations

From host to card(s) From host to card From card to host

SDIO_CMD Command Command Response

SDIO_D

Operation (no response) Operation (no data)

MSv36068V1

Figure 305. (Multiple) block read operation

From host to card From card to host
data from card to host Stop command
stops data transfer

SDIO_CMD Command Response Command Response

SDIO_D Data block crc Data block crc Data block crc

Block read operation

Data stop operation
Multiple block read operation
MSv36069V1

Figure 306. (Multiple) block write operation

From host to card From card to host

data from host to card Stop command
stops data transfer

SDIO_CMD Command Response Command Response

SDIO_D Data block crc Busy Data block crc Busy

Data stop operation

Block write operation

Multiple block write operation

MSv36070V1

Note: The SDIO will not send any data as long as the Busy signal is asserted (SDIO_D0 pulled
low).

864/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Figure 307. Sequential read operation

From host to
card(s) From card to host
Stop command
Data from card to host stops data transfer

SDIO_CMD Command Response Command Response

SDIO_D Data stream

Data stop operation

Data transfer operation
MSv36071V1

Figure 308. Sequential write operation

From host to
card(s) From card to host Stop command
stops data transfer
Data from host to card

SDIO_CMD Command Response Command Response

SDIO_D Data stream

Data stop operation

Data transfer operation
MSv36072V1

27.3 SDIO functional description

The SDIO consists of two parts:
• The SDIO adapter block provides all functions specific to the MMC/SD/SD I/O card
such as the clock generation unit, command and data transfer.
• The APB2 interface accesses the SDIO adapter registers, and generates interrupt and
DMA request signals.

Figure 309. SDIO block diagram

SDIO
SDIO_CK
Interrupts and
DMA request SDIO_CMD

APB2 SDIO SDIO_D[7:0]

interface adapter
APB2 bus

PCLK2 SDIOCLK
MSv36073V1

RM0402 Rev 6 865/1163

921
Secure digital input/output interface (SDIO) RM0402

By default SDIO_D0 is used for data transfer. After initialization, the host can change the
databus width.
If a MultiMediaCard is connected to the bus, SDIO_D0, SDIO_D[3:0] or SDIO_D[7:0] can be
used for data transfer. MMC V3.31 or previous, supports only 1 bit of data so only SDIO_D0
can be used.
If an SD or SD I/O card is connected to the bus, data transfer can be configured by the host
to use SDIO_D0 or SDIO_D[3:0]. All data lines are operating in push-pull mode.
SDIO_CMD has two operational modes:
• Open-drain for initialization (only for MMCV3.31 or previous)
• Push-pull for command transfer (SD/SD I/O card MMC4.2 use push-pull drivers also for
initialization)
SDIO_CK is the clock to the card: one bit is transferred on both command and data lines
with each clock cycle.
The SDIO uses two clock signals:
• SDIO adapter clock SDIOCLK = 50 MHz)
• APB2 bus clock (PCLK2)
PCLK2 and SDIO_CK clock frequencies must respect the following condition:

Frequenc ( PCLK2 ) > ( ( 3xWidth ) ⁄ 32 ) × Frequency ( SDIO_CK )

The signals shown in Table 160 are used on the MultiMediaCard/SD/SD I/O card bus.

Table 160. SDIO I/O definitions

Pin Direction Description

MultiMediaCard/SD/SDIO card clock. This pin is the clock from

SDIO_CK Output
host to card.
MultiMediaCard/SD/SDIO card command. This pin is the
SDIO_CMD Bidirectional
bidirectional command/response signal.
MultiMediaCard/SD/SDIO card data. These pins are the
SDIO_D[7:0] Bidirectional
bidirectional databus.

866/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

27.3.1 SDIO adapter

Figure 310 shows a simplified block diagram of an SDIO adapter.

Figure 310. SDIO adapter

SDIO adapter

Control unit SDIO_CK

Card bus
Command
SDIO_CMD
Adapter path
registers

To APB2
Data path SDIO_D[7:0]
interface FIFO

PCLK2 SDIOCLK
MSv36074V1

The SDIO adapter is a multimedia/secure digital memory card bus master that provides an
interface to a multimedia card stack or to a secure digital memory card. It consists of five
subunits:
• Adapter register block
• Control unit
• Command path
• Data path
• Data FIFO
Note: The adapter registers and FIFO use the APB2 bus clock domain (PCLK2). The control unit,
command path and data path use the SDIO adapter clock domain (SDIOCLK).

Adapter register block

The adapter register block contains all system registers. This block also generates the
signals that clear the static flags in the multimedia card. The clear signals are generated
when 1 is written into the corresponding bit location in the SDIO Clear register.

Control unit
The control unit contains the power management functions and the clock divider for the
memory card clock.
There are three power phases:
• power-off
• power-up
• power-on

RM0402 Rev 6 867/1163

921
Secure digital input/output interface (SDIO) RM0402

Figure 311. Control unit

Control unit

Power management

Adapter Clock management SDIO_CK

registers

To command and data path

MSv36075V1

The control unit is illustrated in Figure 311. It consists of a power management subunit and a
clock management subunit.
The power management subunit disables the card bus output signals during the power-off
and power-up phases.
The clock management subunit generates and controls the SDIO_CK signal. The SDIO_CK
output can use either the clock divide or the clock bypass mode. The clock output is
inactive:
• after reset
• during the power-off or power-up phases
• if the power saving mode is enabled and the card bus is in the Idle state (eight clock
periods after both the command and data path subunits enter the Idle phase)
The clock management subunit controls SDIO_CK dephasing. When not in bypass mode
the SDIO command and data output are generated on the SDIOCLK falling edge
succeeding the rising edge of SDIO_CK. (SDIO_CK rising edge occurs on SDIOCLK rising
edge) when SDIO_CLKCR[13] bit is reset (NEGEDGE = 0). When SDIO_CLKCR[13] bit is
set (NEGEDGE = 1) SDIO command and data changed on the SDIO_CK falling edge.
When SDIO_CLKCR[10] is set (BYPASS = 1), SDIO_CK rising edge occurs on SDIOCLK
rising edge. The data and the command change on SDIOCLK falling edge whatever
NEGEDGE value.
The data and command responses are latched using SDIO_CK rising edge.

Figure 312. SDIO_CK clock dephasing (BYPASS = 0)

SDIOCLK

SDIO_CK

CMD / Data
output
NEGEDGE = 0 NEGEDGE = 1
MSv36076V1

868/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Command path
The command path unit sends commands to and receives responses from the cards.

Figure 313. SDIO adapter command path

To control unit Status Control Command

flag logic timer

Adapter registers
SDIO_CMDin
CMD

Argument
CRC SDIO_CMDout
Shift
CMD register

Response
To APB2 interface
registers

MSv36078V1

• Command path state machine (CPSM)

– When the command register is written to and the enable bit is set, command
transfer starts. When the command has been sent, the command path state
machine (CPSM) sets the status flags and enters the Idle state if a response is not
required. If a response is required, it waits for the response (see Figure 314 on
page 870). When the response is received, the received CRC code and the
internally generated code are compared, and the appropriate status flags are set.

RM0402 Rev 6 869/1163

921
Secure digital input/output interface (SDIO) RM0402

Figure 314. Command path state machine (SDIO)

On reset

CPSM enabled and

pending command Idle
CPSM disabled
Response received or
disabled or command
Enabled and
CRC failed
Pend command start

CPSM disabled
or no response

Last data
CPSM disabled or
Send command timeout Receive

Wait for response Response

started
Wait
MS34444V1

When the Wait state is entered, the command timer starts running. If the timeout is reached
before the CPSM moves to the Receive state, the timeout flag is set and the Idle state is
entered.
Note: The command timeout has a fixed value of 64 SDIO_CK clock periods.
If the interrupt bit is set in the command register, the timer is disabled and the CPSM waits
for an interrupt request from one of the cards. If a pending bit is set in the command register,
the CPSM enters the Pend state, and waits for a CmdPend signal from the data path
subunit. When CmdPend is detected, the CPSM moves to the Send state. This enables the
data counter to trigger the stop command transmission.
Note: The CPSM remains in the Idle state for at least eight SDIO_CK periods to meet the NCC and
NRC timing constraints. NCC is the minimum delay between two host commands, and NRC is
the minimum delay between the host command and the card response.

870/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Figure 315. SDIO command transfer

at least 8 SDIO_CK cycles

SDIO_CK Command Response Command

State Idle Send Wait Receive Idle Send

SDIO_CMD Hi-Z Controller drives Hi-Z Card drives Hi-Z Controller drives
MSv36079V1

• Command format
– Command: a command is a token that starts an operation. Command are sent
from the host either to a single card (addressed command) or to all connected
cards (broadcast command are available for MMC V3.31 or previous). Commands
are transferred serially on the CMD line. All commands have a fixed length of 48
bits. The general format for a command token for MultiMediaCards, SD-Memory
cards and SDIO-Cards is shown in Table 161.
The command path operates in a half-duplex mode, so that commands and
responses can either be sent or received. If the CPSM is not in the Send state, the
SDIO_CMD output is in the Hi-Z state, as shown in Figure 315 on page 871. Data
on SDIO_CMD are synchronous with the rising edge of SDIO_CK. Table 161
shows the command format.

Table 161. Command format

Bit position Width Value Description

47 1 0 Start bit
46 1 1 Transmission bit
[45:40] 6 - Command index
[39:8] 32 - Argument
[7:1] 7 - CRC7
0 1 1 End bit

– Response: a response is a token that is sent from an addressed card (or

synchronously from all connected cards for MMC V3.31 or previous), to the host
as an answer to a previously received command. Responses are transferred
serially on the CMD line.
The SDIO supports two response types. Both use CRC error checking:
• 48 bit short response
• 136 bit long response
Note: If the response does not contain a CRC (CMD1 response), the device driver must ignore the
CRC failed status.

RM0402 Rev 6 871/1163

921
Secure digital input/output interface (SDIO) RM0402

Table 162. Short response format

Bit position Width Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 - Command index
[39:8] 32 - Argument
[7:1] 7 - CRC7(or 1111111)
0 1 1 End bit

Table 163. Long response format

Bit position Width Value Description

135 1 0 Start bit

134 1 0 Transmission bit
[133:128] 6 111111 Reserved
[127:1] 127 - CID or CSD (including internal CRC7)
0 1 1 End bit

The command register contains the command index (six bits sent to a card) and the
command type. These determine whether the command requires a response, and whether
the response is 48 or 136 bits long (see Section 27.8.4 on page 907). The command path
implements the status flags shown in Table 164:

Table 164. Command path status flags

Flag Description
CMDREND Set if response CRC is OK.
CCRCFAIL Set if response CRC fails.
CMDSENT Set when command (that does not require response) is sent
CTIMEOUT Response timeout.
CMDACT Command transfer in progress.

The CRC generator calculates the CRC checksum for all bits before the CRC code. This
includes the start bit, transmitter bit, command index, and command argument (or card
status). The CRC checksum is calculated for the first 120 bits of CID or CSD for the long
response format. Note that the start bit, transmitter bit and the six reserved bits are not used
in the CRC calculation.
The CRC checksum is a 7-bit value:
CRC[6:0] = Remainder [(M(x) * x7) / G(x)]
G(x) = x7 + x3 + 1
M(x) = (start bit) * x39 + ... + (last bit before CRC) * x0, or
M(x) = (start bit) * x119 + ... + (last bit before CRC) * x0

872/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Data path
The data path subunit transfers data to and from cards. Figure 316 shows a block diagram
of the data path.

Figure 316. Data path

Data path

To control unit Status Control Data

flag logic timer

Data FIFO SDIO_Din[7:0]

Transmit
CRC SDIO_Dout[7:0]
Shift
register
Receive

MSv36080V1

The card databus width can be programmed using the clock control register. If the 4-bit wide
bus mode is enabled, data is transferred at four bits per clock cycle over all four data signals
(SDIO_D[3:0]). If the 8-bit wide bus mode is enabled, data is transferred at eight bits per
clock cycle over all eight data signals (SDIO_D[7:0]). If the wide bus mode is not enabled,
only one bit per clock cycle is transferred over SDIO_D0.
Depending on the transfer direction (send or receive), the data path state machine (DPSM)
moves to the Wait_S or Wait_R state when it is enabled:
• Send: the DPSM moves to the Wait_S state. If there is data in the transmit FIFO, the
DPSM moves to the Send state, and the data path subunit starts sending data to a
card.
• Receive: the DPSM moves to the Wait_R state and waits for a start bit. When it
receives a start bit, the DPSM moves to the Receive state, and the data path subunit
starts receiving data from a card.
Data path state machine (DPSM)
The DPSM operates at SDIO_CK frequency. Data on the card bus signals is synchronous to
the rising edge of SDIO_CK. The DPSM has six states, as shown in Figure 317: Data path
state machine (DPSM).

RM0402 Rev 6 873/1163

921
Secure digital input/output interface (SDIO) RM0402

Figure 317. Data path state machine (DPSM)

On reset DPSM disabled

DPSM enabled and Read Wait

Read Wait Started
and SD I/O mode enabled
Disabled or FIFO underrun or
end of data or CRC fail Idle
Disabled or CRC fail
or timeout Enable and not send

ReadWait Stop
Disabled or
end of data
Disabled or
Busy Rx FIFO empty or timeout or
start bit error
Not busy
Enable and send Data received and
Wait_R Read Wait Started and
SD I/O mode enabled

End of packet

Wait_S End of packet or

end of data or
FIFO overrun
Disabled or CRC fail

Data ready Start bit

Send

Receive

ai14809b

• Idle: the data path is inactive, and the SDIO_D[7:0] outputs are in Hi-Z. When the data
control register is written and the enable bit is set, the DPSM loads the data counter
with a new value and, depending on the data direction bit, moves to either the Wait_S
or the Wait_R state.
• Wait_R: if the data counter equals zero, the DPSM moves to the Idle state when the
receive FIFO is empty. If the data counter is not zero, the DPSM waits for a start bit on
SDIO_D. The DPSM moves to the Receive state if it receives a start bit before a
timeout, and loads the data block counter. If it reaches a timeout before it detects a
start bit, it moves to the Idle state and sets the timeout status flag.
• Receive: serial data received from a card is packed in bytes and written to the data
FIFO. Depending on the transfer mode bit in the data control register, the data transfer
mode can be either block or stream:
– In block mode, when the data block counter reaches zero, the DPSM waits until it
receives the CRC code. If the received code matches the internally generated
CRC code, the DPSM moves to the Wait_R state. If not, the CRC fail status flag is
set and the DPSM moves to the Idle state.
– In stream mode, the DPSM receives data while the data counter is not zero. When
the counter is zero, the remaining data in the shift register is written to the data
FIFO, and the DPSM moves to the Wait_R state.
If a FIFO overrun error occurs, the DPSM sets the FIFO error flag and moves to the
Idle state:
• Wait_S: the DPSM moves to the Idle state if the data counter is zero. If not, it waits until
the data FIFO empty flag is deasserted, and moves to the Send state.

874/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Note: The DPSM remains in the Wait_S state for at least two clock periods to meet the NWR timing
requirements, where NWR is the number of clock cycles between the reception of the card
response and the start of the data transfer from the host.
• Send: the DPSM starts sending data to a card. Depending on the transfer mode bit in
the data control register, the data transfer mode can be either block or stream:
– In block mode, when the data block counter reaches zero, the DPSM sends an
internally generated CRC code and end bit, and moves to the Busy state.
– In stream mode, the DPSM sends data to a card while the enable bit is high and
the data counter is not zero. It then moves to the Idle state.
If a FIFO underrun error occurs, the DPSM sets the FIFO error flag and moves to the
Idle state.
• Busy: the DPSM waits for the CRC status flag:
– If it does not receive a positive CRC status, it moves to the Idle state and sets the
CRC fail status flag.
– If it receives a positive CRC status, it moves to the Wait_S state if SDIO_D0 is not
low (the card is not busy).
If a timeout occurs while the DPSM is in the Busy state, it sets the data timeout flag and
moves to the Idle state.
The data timer is enabled when the DPSM is in the Wait_R or Busy state, and
generates the data timeout error:
– When transmitting data, the timeout occurs if the DPSM stays in the Busy state for
longer than the programmed timeout period
– When receiving data, the timeout occurs if the end of the data is not true, and if the
DPSM stays in the Wait_R state for longer than the programmed timeout period.
• Data: data can be transferred from the card to the host or vice versa. Data is
transferred via the data lines. They are stored in a FIFO of 32 words, each word is 32
bits wide.

Table 165. Data token format

Description Start bit Data CRC16 End bit

Block Data 0 - yes 1

Stream Data 0 - no 1

RM0402 Rev 6 875/1163

921
Secure digital input/output interface (SDIO) RM0402

DPSM Flags
The status of the data path subunit transfer is reported by several status flags

Table 166. DPSM flags

Flag Description

Set to high when data block send/receive CRC check is passed.

DBCKEND In SDIO multibyte transfer mode this flag is set at the end of the transfer (a
multibyte transfer is considered as a single block transfer by the host).
Set to high when SDIO_DCOUNT register decrements and reaches 0.
DATAEND
DATAEND indicates the end of a transfer on SDIO data line.
Set to high when data timeout period is reached.
When data timer reaches zero while DPSM is in Wait_R or Busy state, timeout is
DTIMEOUT
set. DTIMEOUT can be set after DATAEND if DPSM remains in busy state for
longer than the programmed period.
DCRCFAIL Set to high when data block send/receive CRC check fails.

Data FIFO
The data FIFO (first-in-first-out) subunit is a data buffer with a transmit and receive unit.
The FIFO contains a 32-bit wide, 32-word deep data buffer, and transmit and receive logic.
Because the data FIFO operates in the APB2 clock domain (PCLK2), all signals from the
subunits in the SDIO clock domain (SDIOCLK) are resynchronized.
Depending on the TXACT and RXACT flags, the FIFO can be disabled, transmit enabled, or
receive enabled. TXACT and RXACT are driven by the data path subunit and are mutually
exclusive:
– The transmit FIFO refers to the transmit logic and data buffer when TXACT is
asserted
– The receive FIFO refers to the receive logic and data buffer when RXACT is
asserted
• Transmit FIFO:
Data can be written to the transmit FIFO through the APB2 interface when the SDIO is
enabled for transmission.
The transmit FIFO is accessible via 32 sequential addresses. The transmit FIFO
contains a data output register that holds the data word pointed to by the read pointer.
When the data path subunit has loaded its shift register, it increments the read pointer
and drives new data out.
If the transmit FIFO is disabled, all status flags are deasserted. The data path subunit
asserts TXACT when it transmits data.

876/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Table 167. Transmit FIFO status flags

Flag Description

TXFIFOF Set to high when all 32 transmit FIFO words contain valid data.
TXFIFOE Set to high when the transmit FIFO does not contain valid data.
Set to high when 8 or more transmit FIFO words are empty. This flag can be used
TXFIFOHE
as a DMA request.
Set to high when the transmit FIFO contains valid data. This flag is the inverse of
TXDAVL
the TXFIFOE flag.
Set to high when an underrun error occurs. This flag is cleared by writing to the
SDIO Clear register.
TXUNDERR Note: In case of TXUNDERR, and DMA is used to fill SDIO FIFO, user software
should disable DMA stream, and then write DMAEN bit in SDIO_DCTRL
with ‘0’ (to disable DMA request generation).

• Receive FIFO
When the data path subunit receives a word of data, it drives the data on the write
databus. The write pointer is incremented after the write operation completes. On the
read side, the contents of the FIFO word pointed to by the current value of the read
pointer is driven onto the read databus. If the receive FIFO is disabled, all status flags
are deasserted, and the read and write pointers are reset. The data path subunit
asserts RXACT when it receives data. Table 168 lists the receive FIFO status flags.
The receive FIFO is accessible via 32 sequential addresses.

Table 168. Receive FIFO status flags

Flag Description

RXFIFOF Set to high when all 32 receive FIFO words contain valid data
RXFIFOE Set to high when the receive FIFO does not contain valid data.
Set to high when 8 or more receive FIFO words contain valid data. This flag can be
RXFIFOHF
used as a DMA request.
Set to high when the receive FIFO is not empty. This flag is the inverse of the
RXDAVL
RXFIFOE flag.
Set to high when an overrun error occurs. This flag is cleared by writing to the SDIO
Clear register.
RXOVERR Note: In case of RXOVERR, and DMA is used to read SDIO FIFO, user software
should disable DMA stream, and then write DMAEN bit in SDIO_DCTRL
with ‘0’ (to disable DMA request generation).

RM0402 Rev 6 877/1163

921
Secure digital input/output interface (SDIO) RM0402

27.3.2 SDIO APB2 interface

The APB2 interface generates the interrupt and DMA requests, and accesses the SDIO
adapter registers and the data FIFO. It consists of a data path, register decoder, and
interrupt/DMA logic.

SDIO interrupts
The interrupt logic generates an interrupt request signal that is asserted when at least one
of the selected status flags is high. A mask register is provided to allow selection of the
conditions that will generate an interrupt. A status flag generates the interrupt request if a
corresponding mask flag is set.

SDIO/DMA interface
SDIO APB interface controls all subunit to perform transfers between the host and card

Example of read procedure using DMA

Send CMD17 (READ_BLOCK) as follows:
a) Program the SDIO data length register (SDIO data timer register should be
already programmed before the card identification process)
b) Program DMA channel (refer to DMA configuration for SDIO controller)
c) Program the SDIO data control register: DTEN with ‘1’ (SDIO card host enabled to
send data); DTDIR with ‘1’ (from card to controller); DTMODE with ‘0’ (block data
transfer); DMAEN with ‘1’ (DMA enabled); DBLOCKSIZE with 0x9 (512 bytes).
Other fields are don’t care.
d) Program the SDIO argument register with the address location of the card from
where data is to be transferred
e) Program the SDIO command register: CmdIndex with 17(READ_BLOCK);
WaitResp with ‘1’ (SDIO card host waits for a response); CPSMEN with ‘1’ (SDIO
card host enabled to send a command). Other fields are at their reset value.
f) Wait for SDIO_STA[6] = CMDREND interrupt, (CMDREND is set if there is no
error on command path).
g) Wait for SDIO_STA[10] = DBCKEND, (DBCKEND is set in case of no errors until
the CRC check is passed)
h) Wait until the FIFO is empty, when FIFO is empty the SDIO_STA[5] = RXOVERR
value has to be check to guarantee that read succeeded
Note: When FIFO overrun error occurs with last 1-4 bytes, it may happens that RXOVERR flag is
set 2 APB clock cycles after DATAEND flag is set. To guarantee success of read operation
RXOVERR must be cheked after FIFO is empty.

878/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Example of write procedure using DMA

Send CMD24 (WRITE_BLOCK) as follows:
a) Program the SDIO data length register (SDIO data timer register should be
already programmed before the card identification process)
b) Program DMA channel (refer to DMA configuration for SDIO controller)
c) Program the SDIO argument register with the address location of the card from
where data is to be transferred
d) Program the SDIO command register: CmdIndex with 24(WRITE_BLOCK);
WaitResp with ‘1’ (SDIO card host waits for a response); CPSMEN with ‘1’ (SDIO
card host enabled to send a command). Other fields are at their reset value.
e) Wait for SDIO_STA[6] = CMDREND interrupt, then Program the SDIO data control
register: DTEN with ‘1’ (SDIO card host enabled to send data); DTDIR with ‘0’
(from controller to card); DTMODE with ‘0’ (block data transfer); DMAEN with ‘1’
(DMA enabled); DBLOCKSIZE with 0x9 (512 bytes). Other fields are don’t care.
f) Wait for SDIO_STA[10] = DBCKEND, (DBCKEND is set in case of no errors)

DMA configuration for SDIO controller

a) Enable DMA2 controller and clear any pending interrupts.
b) Program the DMA2_Stream3 (or DMA2_Stream6) Channel4 source address
register with the memory location base address and DMA2_Stream3 (or
DMA2_Stream6) Channel4 destination address register with the SDIO_FIFO
register address.
c) Program DMA2_Stream3 (or DMA2_Stream6) Channel4 control register (memory
increment, not peripheral increment, peripheral and source width is word size).
d) Program DMA2_Stream3 (or DMA2_Stream6) Channel4 to select the peripheral
as flow controller (set PFCTRL bit in DMA_S3CR (or DMA_S6CR) configuration
register).
e) Configure the incremental burst transfer to 4 beats (at least from peripheral side)
in DMA2_Stream3 (or DMA2_Stream6) Channel4.
f) Enable DMA2_Stream3 (or DMA2_Stream6) Channel4
Note: SDIO host allows only to use the DMA in peripheral flow controller mode. DMA stream used
to serve SDIO must be configured in peripheral flow controller mode
SDIO generates only DMA burst requests to DMA controller. DMA must be configured in
incremental burst mode on peripheral side.

27.4 Card functional description

27.4.1 Card identification mode

While in card identification mode the host resets all cards, validates the operation voltage
range, identifies cards and sets a relative card address (RCA) for each card on the bus. All
data communications in the card identification mode use the command line (CMD) only.

RM0402 Rev 6 879/1163

921
Secure digital input/output interface (SDIO) RM0402

27.4.2 Card reset

The GO_IDLE_STATE command (CMD0) is the software reset command and it puts the
MultiMediaCard and SD memory in the Idle state. The IO_RW_DIRECT command (CMD52)
resets the SD I/O card. After power-up or CMD0, all cards output bus drivers are in the high-
impedance state and the cards are initialized with a default relative card address
(RCA=0x0001) and with a default driver stage register setting (lowest speed, highest driving
current capability).

27.4.3 Operating voltage range validation

All cards can communicate with the SDIO card host using any operating voltage within the
specification range. The supported minimum and maximum VDD values are defined in the
operation conditions register (OCR) on the card.
Cards that store the card identification number (CID) and card specific data (CSD) in the
payload memory are able to communicate this information only under data-transfer VDD
conditions. When the SDIO card host module and the card have incompatible VDD ranges,
the card is not able to complete the identification cycle and cannot send CSD data. For this
purpose, the special commands, SEND_OP_COND (CMD1), SD_APP_OP_COND
(ACMD41 for SD Memory), and IO_SEND_OP_COND (CMD5 for SD I/O), are designed to
provide a mechanism to identify and reject cards that do not match the VDD range desired
by the SDIO card host. The SDIO card host sends the required VDD voltage window as the
operand of these commands. Cards that cannot perform data transfer in the specified range
disconnect from the bus and go to the inactive state.
By using these commands without including the voltage range as the operand, the SDIO
card host can query each card and determine the common voltage range before placing out-
of-range cards in the inactive state. This query is used when the SDIO card host is able to
select a common voltage range or when the user requires notification that cards are not
usable.

27.4.4 Card identification process

The card identification process differs for MultiMediaCards and SD cards. For
MultiMediaCard cards, the identification process starts at clock rate Fod. The SDIO_CMD
line output drivers are open-drain and allow parallel card operation during this process. The
registration process is accomplished as follows:
1. The bus is activated.
2. The SDIO card host broadcasts SEND_OP_COND (CMD1) to receive operation
conditions.
3. The response is the wired AND operation of the operation condition registers from all
cards.
4. Incompatible cards are placed in the inactive state.
5. The SDIO card host broadcasts ALL_SEND_CID (CMD2) to all active cards.
6. The active cards simultaneously send their CID numbers serially. Cards with outgoing
CID bits that do not match the bits on the command line stop transmitting and must wait
for the next identification cycle. One card successfully transmits a full CID to the SDIO
card host and enters the Identification state.
7. The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to that card. This new
address is called the relative card address (RCA); it is shorter than the CID and

880/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

addresses the card. The assigned card changes to the Standby state, it does not react
to further identification cycles, and its output switches from open-drain to push-pull.
8. The SDIO card host repeats steps 5 through 7 until it receives a timeout condition.
For the SD card, the identification process starts at clock rate Fod, and the SDIO_CMD line
output drives are push-pull drivers instead of open-drain. The registration process is
accomplished as follows:
1. The bus is activated.
2. The SDIO card host broadcasts SD_APP_OP_COND (ACMD41).
3. The cards respond with the contents of their operation condition registers.
4. The incompatible cards are placed in the inactive state.
5. The SDIO card host broadcasts ALL_SEND_CID (CMD2) to all active cards.
6. The cards send back their unique card identification numbers (CIDs) and enter the
Identification state.
7. The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to an active card with an
address. This new address is called the relative card address (RCA); it is shorter than
the CID and addresses the card. The assigned card changes to the Standby state. The
SDIO card host can reissue this command to change the RCA. The RCA of the card is
the last assigned value.
8. The SDIO card host repeats steps 5 through 7 with all active cards.
For the SD I/O card, the registration process is accomplished as follows:
1. The bus is activated.
2. The SDIO card host sends IO_SEND_OP_COND (CMD5).
3. The cards respond with the contents of their operation condition registers.
4. The incompatible cards are set to the inactive state.
5. The SDIO card host issues SET_RELATIVE_ADDR (CMD3) to an active card with an
address. This new address is called the relative card address (RCA); it is shorter than
the CID and addresses the card. The assigned card changes to the Standby state. The
SDIO card host can reissue this command to change the RCA. The RCA of the card is
the last assigned value.

27.4.5 Block write

During block write (CMD24 - 27) one or more blocks of data are transferred from the host to
the card with a CRC appended to the end of each block by the host. A card supporting block
write is always able to accept a block of data defined by WRITE_BL_LEN. If the CRC fails,
the card indicates the failure on the SDIO_D line and the transferred data are discarded and
not written, and all further transmitted blocks (in multiple block write mode) are ignored.
If the host uses partial blocks whose accumulated length is not block aligned and, block
misalignment is not allowed (CSD parameter WRITE_BLK_MISALIGN is not set), the card
will detect the block misalignment error before the beginning of the first misaligned block.
(ADDRESS_ERROR error bit is set in the status register). The write operation will also be
aborted if the host tries to write over a write-protected area. In this case, however, the card
will set the WP_VIOLATION bit.
Programming of the CID and CSD registers does not require a previous block length setting.
The transferred data is also CRC protected. If a part of the CSD or CID register is stored in
ROM, then this unchangeable part must match the corresponding part of the receive buffer.
If this match fails, then the card reports an error and does not change any register contents.

RM0402 Rev 6 881/1163

921
Secure digital input/output interface (SDIO) RM0402

Some cards may require long and unpredictable times to write a block of data. After
receiving a block of data and completing the CRC check, the card begins writing and holds
the SDIO_D line low if its write buffer is full and unable to accept new data from a new
WRITE_BLOCK command. The host may poll the status of the card with a SEND_STATUS
command (CMD13) at any time, and the card will respond with its status. The
READY_FOR_DATA status bit indicates whether the card can accept new data or whether
the write process is still in progress. The host may deselect the card by issuing CMD7 (to
select a different card), which will place the card in the Disconnect state and release the
SDIO_D line(s) without interrupting the write operation. When reselecting the card, it will
reactivate busy indication by pulling SDIO_D to low if programming is still in progress and
the write buffer is unavailable.

27.4.6 Block read

In Block read mode the basic unit of data transfer is a block whose maximum size is defined
in the CSD (READ_BL_LEN). If READ_BL_PARTIAL is set, smaller blocks whose start and
end addresses are entirely contained within one physical block (as defined by
READ_BL_LEN) may also be transmitted. A CRC is appended to the end of each block,
ensuring data transfer integrity. CMD17 (READ_SINGLE_BLOCK) initiates a block read and
after completing the transfer, the card returns to the Transfer state.
CMD18 (READ_MULTIPLE_BLOCK) starts a transfer of several consecutive blocks.
The host can abort reading at any time, within a multiple block operation, regardless of its
type. Transaction abort is done by sending the stop transmission command.
If the card detects an error (for example, out of range, address misalignment or internal
error) during a multiple block read operation (both types) it stops the data transmission and
remains in the data state. The host must than abort the operation by sending the stop
transmission command. The read error is reported in the response to the stop transmission
command.
If the host sends a stop transmission command after the card transmits the last block of a
multiple block operation with a predefined number of blocks, it is responded to as an illegal
command, since the card is no longer in the data state. If the host uses partial blocks whose
accumulated length is not block-aligned and block misalignment is not allowed, the card
detects a block misalignment error condition at the beginning of the first misaligned block
(ADDRESS_ERROR error bit is set in the status register).

27.4.7 Stream access, stream write and stream read

(MultiMediaCard only)
In stream mode, data is transferred in bytes and no CRC is appended at the end of each
block.

Stream write (MultiMediaCard only)

WRITE_DAT_UNTIL_STOP (CMD20) starts the data transfer from the SDIO card host to
the card, beginning at the specified address and continuing until the SDIO card host issues
a stop command. When partial blocks are allowed (CSD parameter WRITE_BL_PARTIAL is
set), the data stream can start and stop at any address within the card address space,
otherwise it can only start and stop at block boundaries. Because the amount of data to be
transferred is not determined in advance, a CRC cannot be used. When the end of the
memory range is reached while sending data and no stop command is sent by the SDIO
card host, any additional transferred data are discarded.

882/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

The maximum clock frequency for a stream write operation is given by the following
equation fields of the card-specific data register:

8 × 2 writebllen ) ( – NSAC ))
Maximumspeed = MIN (TRANSPEED,(------------------------------------------------------------------------
TAAC × R2WFACTOR

• Maximumspeed = maximum write frequency

• TRANSPEED = maximum data transfer rate
• writebllen = maximum write data block length
• NSAC = data read access time 2 in CLK cycles
• TAAC = data read access time 1
• R2WFACTOR = write speed factor
If the host attempts to use a higher frequency, the card may not be able to process the data
and stop programming, set the OVERRUN error bit in the status register, and while ignoring
all further data transfer, wait (in the receive data state) for a stop command. The write
operation is also aborted if the host tries to write over a write-protected area. In this case,
however, the card sets the WP_VIOLATION bit.

Stream read (MultiMediaCard only)

READ_DAT_UNTIL_STOP (CMD11) controls a stream-oriented data transfer.
This command instructs the card to send its data, starting at a specified address, until the
SDIO card host sends STOP_TRANSMISSION (CMD12). The stop command has an
execution delay due to the serial command transmission and the data transfer stops after
the end bit of the stop command. When the end of the memory range is reached while
sending data and no stop command is sent by the SDIO card host, any subsequent data
sent are considered undefined.
The maximum clock frequency for a stream read operation is given by the following
equation and uses fields of the card specific data register.

8 × 2 readbllen ) ( – NSAC ))
Maximumspeed = MIN (TRANSPEED,(-----------------------------------------------------------------------
TAAC × R2WFACTOR

• Maximumspeed = maximum read frequency

• TRANSPEED = maximum data transfer rate
• readbllen = maximum read data block length
• writebllen = maximum write data block length
• NSAC = data read access time 2 in CLK cycles
• TAAC = data read access time 1
• R2WFACTOR = write speed factor
If the host attempts to use a higher frequency, the card is not able to sustain data transfer. If
this happens, the card sets the UNDERRUN error bit in the status register, aborts the
transmission and waits in the data state for a stop command.

RM0402 Rev 6 883/1163

921
Secure digital input/output interface (SDIO) RM0402

27.4.8 Erase: group erase and sector erase

The erasable unit of the MultiMediaCard is the erase group. The erase group is measured in
write blocks, which are the basic writable units of the card. The size of the erase group is a
card-specific parameter and defined in the CSD.
The host can erase a contiguous range of Erase Groups. Starting the erase process is a
three-step sequence.
First the host defines the start address of the range using the ERASE_GROUP_START
(CMD35) command, next it defines the last address of the range using the
ERASE_GROUP_END (CMD36) command and, finally, it starts the erase process by
issuing the ERASE (CMD38) command. The address field in the erase commands is an
Erase Group address in byte units. The card ignores all LSBs below the Erase Group size,
effectively rounding the address down to the Erase Group boundary.
If an erase command is received out of sequence, the card sets the ERASE_SEQ_ERROR
bit in the status register and resets the whole sequence.
If an out-of-sequence (neither of the erase commands, except SEND_STATUS) command
received, the card sets the ERASE_RESET status bit in the status register, resets the erase
sequence and executes the last command.
If the erase range includes write protected blocks, they are left intact and only nonprotected
blocks are erased. The WP_ERASE_SKIP status bit in the status register is set.
The card indicates that an erase is in progress by holding SDIO_D low. The actual erase
time may be quite long, and the host may issue CMD7 to deselect the card.

27.4.9 Wide bus selection or deselection

Wide bus (4-bit bus width) operation mode is selected or deselected using
SET_BUS_WIDTH (ACMD6). The default bus width after power-up or GO_IDLE_STATE
(CMD0) is 1 bit. SET_BUS_WIDTH (ACMD6) is only valid in a transfer state, which means
that the bus width can be changed only after a card is selected by
SELECT/DESELECT_CARD (CMD7).

27.4.10 Protection management

Three write protection methods for the cards are supported in the SDIO card host module:
1. internal card write protection (card responsibility)
2. mechanical write protection switch (SDIO card host module responsibility only)
3. password-protected card lock operation

Internal card write protection

Card data can be protected against write and erase. By setting the permanent or temporary
write-protect bits in the CSD, the entire card can be permanently write-protected by the
manufacturer or content provider. For cards that support write protection of groups of
sectors by setting the WP_GRP_ENABLE bit in the CSD, portions of the data can be
protected, and the write protection can be changed by the application. The write protection
is in units of WP_GRP_SIZE sectors as specified in the CSD. The SET_WRITE_PROT and
CLR_WRITE_PROT commands control the protection of the addressed group. The
SEND_WRITE_PROT command is similar to a single block read command. The card sends
a data block containing 32 write protection bits (representing 32 write protect groups starting

884/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

at the specified address) followed by 16 CRC bits. The address field in the write protect
commands is a group address in byte units.
The card ignores all LSBs below the group size.

Mechanical write protect switch

A mechanical sliding tab on the side of the card allows the user to set or clear the write
protection on a card. When the sliding tab is positioned with the window open, the card is
write-protected, and when the window is closed, the card contents can be changed. A
matched switch on the socket side indicates to the SDIO card host module that the card is
write-protected. The SDIO card host module is responsible for protecting the card. The
position of the write protect switch is unknown to the internal circuitry of the card.

Password protect
The password protection feature enables the SDIO card host module to lock and unlock a
card with a password. The password is stored in the 128-bit PWD register and its size is set
in the 8-bit PWD_LEN register. These registers are nonvolatile so that a power cycle does
not erase them. Locked cards respond to and execute certain commands. This means that
the SDIO card host module is allowed to reset, initialize, select, and query for status,
however it is not allowed to access data on the card. When the password is set (as indicated
by a nonzero value of PWD_LEN), the card is locked automatically after power-up. As with
the CSD and CID register write commands, the lock/unlock commands are available in the
transfer state only. In this state, the command does not include an address argument and
the card must be selected before using it. The card lock/unlock commands have the
structure and bus transaction types of a regular single-block write command. The
transferred data block includes all of the required information for the command (the
password setting mode, the PWD itself, and card lock/unlock). The command data block
size is defined by the SDIO card host module before it sends the card lock/unlock
command, and has the structure shown in Table 182.
The bit settings are as follows:
• ERASE: setting it forces an erase operation. All other bits must be zero, and only the
command byte is sent
• LOCK_UNLOCK: setting it locks the card. LOCK_UNLOCK can be set simultaneously
with SET_PWD, however not with CLR_PWD
• CLR_PWD: setting it clears the password data
• SET_PWD: setting it saves the password data to memory
• PWD_LEN: it defines the length of the password in bytes
• PWD: the password (new or currently used, depending on the command)
The following sections list the command sequences to set/reset a password, lock/unlock the
card, and force an erase.

Setting the password

1. Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.
2. Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode, the 8-bit PWD_LEN, and the number of bytes of the new password.

RM0402 Rev 6 885/1163

921
Secure digital input/output interface (SDIO) RM0402

When a password replacement is done, the block size must take into account that both
the old and the new passwords are sent with the command.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (SET_PWD = 1), the
length (PWD_LEN), and the password (PWD) itself. When a password replacement is
done, the length value (PWD_LEN) includes the length of both passwords, the old and
the new one, and the PWD field includes the old password (currently used) followed by
the new password.
4. When the password is matched, the new password and its size are saved into the PWD
and PWD_LEN fields, respectively. When the old password sent does not correspond
(in size and/or content) to the expected password, the LOCK_UNLOCK_FAILED error
bit is set in the card status register, and the password is not changed.
The password length field (PWD_LEN) indicates whether a password is currently set. When
this field is nonzero, there is a password set and the card locks itself after power-up. It is
possible to lock the card immediately in the current power session by setting the
LOCK_UNLOCK bit (while setting the password) or sending an additional command for card
locking.

Resetting the password

1. Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.
2. Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode, the 8-bit PWD_LEN, and the number of bytes in the currently used
password.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (CLR_PWD = 1), the
length (PWD_LEN) and the password (PWD) itself. The LOCK_UNLOCK bit is ignored.
4. When the password is matched, the PWD field is cleared and PWD_LEN is set to 0.
When the password sent does not correspond (in size and/or content) to the expected
password, the LOCK_UNLOCK_FAILED error bit is set in the card status register, and
the password is not changed.

Locking a card
1. Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.
2. Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit card
lock/unlock mode (byte 0 in Table 182), the 8-bit PWD_LEN, and the number of bytes
of the current password.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (LOCK_UNLOCK = 1), the
length (PWD_LEN), and the password (PWD) itself.
4. When the password is matched, the card is locked and the CARD_IS_LOCKED status
bit is set in the card status register. When the password sent does not correspond (in
size and/or content) to the expected password, the LOCK_UNLOCK_FAILED error bit
is set in the card status register, and the lock fails.
It is possible to set the password and to lock the card in the same sequence. In this case,
the SDIO card host module performs all the required steps for setting the password (see
Setting the password on page 885), however it is necessary to set the LOCK_UNLOCK bit
in Step 3 when the new password command is sent.

886/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

When the password is previously set (PWD_LEN is not 0), the card is locked automatically
after power on reset. An attempt to lock a locked card or to lock a card that does not have a
password fails and the LOCK_UNLOCK_FAILED error bit is set in the card status register.

Unlocking the card

1. Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.
2. Define the block length (SET_BLOCKLEN, CMD16) to send, given by the 8-bit
cardlock/unlock mode (byte 0 in Table 182), the 8-bit PWD_LEN, and the number of
bytes of the current password.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data block size on the data line
including the 16-bit CRC. The data block indicates the mode (LOCK_UNLOCK = 0), the
length (PWD_LEN), and the password (PWD) itself.
4. When the password is matched, the card is unlocked and the CARD_IS_LOCKED
status bit is cleared in the card status register. When the password sent is not correct in
size and/or content and does not correspond to the expected password, the
LOCK_UNLOCK_FAILED error bit is set in the card status register, and the card
remains locked.
The unlocking function is only valid for the current power session. When the PWD field is not
clear, the card is locked automatically on the next power-up.
An attempt to unlock an unlocked card fails and the LOCK_UNLOCK_FAILED error bit is set
in the card status register.

Forcing erase
If the user has forgotten the password (PWD content), it is possible to access the card after
clearing all the data on the card. This forced erase operation erases all card data and all
password data.
1. Select a card (SELECT/DESELECT_CARD, CMD7), if none is already selected.
2. Set the block length (SET_BLOCKLEN, CMD16) to 1 byte. Only the 8-bit card
lock/unlock byte (byte 0 in Table 182) is sent.
3. Send LOCK/UNLOCK (CMD42) with the appropriate data byte on the data line
including the 16-bit CRC. The data block indicates the mode (ERASE = 1). All other bits
must be zero.
4. When the ERASE bit is the only bit set in the data field, all card contents are erased,
including the PWD and PWD_LEN fields, and the card is no longer locked. When any
other bits are set, the LOCK_UNLOCK_FAILED error bit is set in the card status
register and the card retains all of its data, and remains locked.
An attempt to use a force erase on an unlocked card fails and the LOCK_UNLOCK_FAILED
error bit is set in the card status register.

RM0402 Rev 6 887/1163

921
Secure digital input/output interface (SDIO) RM0402

27.4.11 Card status register

The response format R1 contains a 32-bit field named card status. This field is intended to
transmit the card status information (which may be stored in a local status register) to the
host. If not specified otherwise, the status entries are always related to the previously issued
command.
Table 169 defines the different entries of the status. The type and clear condition fields in
the table are abbreviated as follows:
Type:
• E: error bit
• S: status bit
• R: detected and set for the actual command response
• X: detected and set during command execution. The SDIO card host must poll the card
by issuing the status command to read these bits.
Clear condition:
• A: according to the card current state
• B: always related to the previous command. Reception of a valid command clears it
(with a delay of one command)
• C: clear by read

Table 169. Card status

Clear
Bits Identifier Type Value Description
condition

The command address argument was out

of the allowed range for this card.
ADDRESS_ ’0’= no error A multiple block or stream read/write
31 ERX C
OUT_OF_RANGE ’1’= error operation is (although started in a valid
address) attempting to read or write
beyond the card capacity.
The commands address argument (in
accordance with the currently set block
length) positions the first data block
misaligned to the card physical blocks.
’0’= no error A multiple block read/write operation
30 ADDRESS_MISALIGN - C
’1’= error (although started with a valid
address/block-length combination) is
attempting to read or write a data block
which is not aligned with the physical
blocks of the card.
Either the argument of a
SET_BLOCKLEN command exceeds the
maximum value allowed for the card, or
’0’= no error the previously defined block length is
29 BLOCK_LEN_ERROR - illegal for the current command (e.g. the C
’1’= error
host issues a write command, the current
block length is smaller than the maximum
allowed value for the card and it is not
allowed to write partial blocks)

888/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Table 169. Card status (continued)

Clear
Bits Identifier Type Value Description
condition

’0’= no error An error in the sequence of erase

28 ERASE_SEQ_ERROR - C
’1’= error commands occurred.
’0’= no error An invalid selection of erase groups for
27 ERASE_PARAM EX C
’1’= error erase occurred.

’0’= no error Attempt to program a write-protected

26 WP_VIOLATION EX
’1’= error block. C
‘0’ = card
When set, signals that the card is locked
25 CARD_IS_LOCKED SR unlocked A
by the host
‘1’ = card locked
Set when a sequence or password error
LOCK_UNLOCK_ ’0’= no error
24 EX has been detected in lock/unlock card C
FAILED ’1’= error
command
’0’= no error The CRC check of the previous command
23 COM_CRC_ERROR ER B
’1’= error failed.
’0’= no error
22 ILLEGAL_COMMAND ER Command not legal for the card state B
’1’= error
’0’= success Card internal ECC was applied but failed
21 CARD_ECC_FAILED EX C
’1’= failure to correct the data.
(Undefined by the standard) A card error
’0’= no error
20 CC_ERROR ER occurred, which is not related to the host C
’1’= error
command.
(Undefined by the standard) A generic
’0’= no error card error related to the (and detected
19 ERROR EX C
’1’= error during) execution of the last host
command (e.g. read or write failures).
18 Reserved
17 Reserved
Can be either of the following errors:
– The CID register has already been
written and cannot be overwritten
’0’= no error ‘1’= – The read-only section of the CSD does
16 CID/CSD_OVERWRITE EX C
error not match the card contents
– An attempt to reverse the copy (set as
original) or permanent WP
(unprotected) bits was made

’0’= not protected Set when only partial address space

15 WP_ERASE_SKIP EX C
’1’= protected was erased due to existing write

’0’= enabled The command has been executed without

14 CARD_ECC_DISABLED S X A
’1’= disabled using the internal ECC.

RM0402 Rev 6 889/1163

921
Secure digital input/output interface (SDIO) RM0402

Table 169. Card status (continued)

Clear
Bits Identifier Type Value Description
condition

An erase sequence was cleared before

executing because an out of erase
’0’= cleared
13 ERASE_RESET - sequence command was received C
’1’= set
(commands other than CMD35, CMD36,
CMD38 or CMD13)
0 = Idle
1 = Ready
2 = Ident
The state of the card when receiving the
3 = Stby
command. If the command execution
4 = Tran
causes a state change, it will be visible to
12:9 CURRENT_STATE SR 5 = Data B
the host in the response on the next
6 = Rcv
command. The four bits are interpreted as
7 = Prg
a binary number between 0 and 15.
8 = Dis
9 = Btst
10-15 = reserved
’0’= not ready Corresponds to buffer empty signalling on
8 READY_FOR_DATA SR -
‘1’ = ready the bus
If set, the card did not switch to the
’0’= no error
7 SWITCH_ERROR EX expected mode as requested by the B
’1’= switch error
SWITCH command
6 Reserved
The card will expect ACMD, or an
‘0’ = Disabled
5 APP_CMD SR indication that the command has been C
‘1’ = Enabled
interpreted as ACMD
4 Reserved for SD I/O Card
’0’= no error Error in the sequence of the
3 AKE_SEQ_ERROR ER C
’1’= error authentication process
2 Reserved for application specific commands
1
Reserved for manufacturer test mode
0

890/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

27.4.12 SD status register

The SD status contains status bits that are related to the SD memory card proprietary
features and may be used for future application-specific usage. The size of the SD Status is
one data block of 512 bits. The contents of this register are transmitted to the SDIO card
host if ACMD13 is sent (CMD55 followed with CMD13). ACMD13 can be sent to a card in
transfer state only (card is selected).
Table 170 defines the different entries of the SD status register. The type and clear condition
fields in the table are abbreviated as follows:
Type:
• E: error bit
• S: status bit
• R: detected and set for the actual command response
• X: detected and set during command execution. The SDIO card Host must poll the card
by issuing the status command to read these bits
Clear condition:
• A: according to the card current state
• B: always related to the previous command. Reception of a valid command clears it
(with a delay of one command)
• C: clear by read

Table 170. SD status

Clear
Bits Identifier Type Value Description
condition

Shows the currently defined

’00’= 1 (default)
databus width that was
‘01’= reserved
511: 510 DAT_BUS_WIDTH S R defined by A
‘10’= 4 bit width
SET_BUS_WIDTH
‘11’= reserved
command
Card is in Secured Mode of
’0’= Not in the mode
509 SECURED_MODE S R operation (refer to the “SD A
’1’= In Secured Mode
Security Specification”).
508: 496 Reserved
In the future, the 8 LSBs will
’00xxh’= SD Memory Cards as be used to define different
defined in Physical Spec Ver1.01- variations of an SD memory
2.00 (’x’= don’t care). The card (each bit will define
495: 480 SD_CARD_TYPE SR following cards are currently different SD types). The 8 A
defined: MSBs will be used to define
’0000’= Regular SD RD/WR Card. SD Cards that do not comply
’0001’= SD ROM Card with current SD physical
layer specification.
SIZE_OF_PROTE Size of protected area (See
479: 448 SR (See below) A
CT ED_AREA below)
Speed Class of the card (See
447: 440 SPEED_CLASS SR (See below) A
below)

RM0402 Rev 6 891/1163

921
Secure digital input/output interface (SDIO) RM0402

Table 170. SD status (continued)

Clear
Bits Identifier Type Value Description
condition

Performance of move indicated by

PERFORMANCE_ 1 [MB/s] step.
439: 432 SR (See below) A
MOVE
(See below)
Size of AU
431:428 AU_SIZE SR (See below) A
(See below)
427:424 Reserved
Number of AUs to be erased at a
423:408 ERASE_SIZE SR (See below) A
time
Timeout value for erasing areas
407:402 ERASE_TIMEOUT S R specified by (See below) A
UNIT_OF_ERASE_AU
Fixed offset value added to erase
401:400 ERASE_OFFSET SR (See below) A
time.
399:312 Reserved
311:0 Reserved for Manufacturer

SIZE_OF_PROTECTED_AREA
Setting this field differs between standard- and high-capacity cards. In the case of a
standard-capacity card, the capacity of protected area is calculated as follows:
Protected area = SIZE_OF_PROTECTED_AREA_* MULT * BLOCK_LEN.
SIZE_OF_PROTECTED_AREA is specified by the unit in MULT*BLOCK_LEN.
In the case of a high-capacity card, the capacity of protected area is specified in this field:
Protected area = SIZE_OF_PROTECTED_AREA
SIZE_OF_PROTECTED_AREA is specified by the unit in bytes.

SPEED_CLASS
This 8-bit field indicates the speed class and the value can be calculated by PW/2 (where
PW is the write performance).

Table 171. Speed class code field

SPEED_CLASS Value definition

00h Class 0
01h Class 2
02h Class 4
03h Class 6
04h – FFh Reserved

892/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

PERFORMANCE_MOVE
This 8-bit field indicates Pm (performance move) and the value can be set by 1 [MB/sec]
steps. If the card does not move used RUs (recording units), Pm should be considered as
infinity. Setting the field to FFh means infinity.

Table 172. Performance move field

PERFORMANCE_MOVE Value definition

00h Not defined

01h 1 [MB/sec]
02h 02h 2 [MB/sec]
--------- ---------
FEh 254 [MB/sec]
FFh Infinity

AU_SIZE
This 4-bit field indicates the AU size and the value can be selected in the power of 2 base
from 16 KB.

Table 173. AU_SIZE field

AU_SIZE Value definition

00h Not defined

01h 16 KB
02h 32 KB
03h 64 KB
04h 128 KB
05h 256 KB
06h 512 KB
07h 1 MB
08h 2 MB
09h 4 MB
Ah – Fh Reserved

The maximum AU size, which depends on the card capacity, is defined in Table 174. The
card can be set to any AU size between RU size and maximum AU size.

Table 174. Maximum AU size

Capacity 16 MB-64 MB 128 MB-256 MB 512 MB 1 GB-32 GB

Maximum AU Size 512 KB 1 MB 2 MB 4 MB

RM0402 Rev 6 893/1163

921
Secure digital input/output interface (SDIO) RM0402

ERASE_SIZE
This 16-bit field indicates NERASE. When NERASE numbers of AUs are erased, the timeout
value is specified by ERASE_TIMEOUT (Refer to ERASE_TIMEOUT). The host should
determine the proper number of AUs to be erased in one operation so that the host can
show the progress of the erase operation. If this field is set to 0, the erase timeout
calculation is not supported.

Table 175. Erase size field

ERASE_SIZE Value definition

0000h Erase timeout calculation is not supported.

0001h 1 AU
0002h 2 AU
0003h 3 AU
--------- ---------
FFFFh 65535 AU

ERASE_TIMEOUT
This 6-bit field indicates TERASE and the value indicates the erase timeout from offset when
multiple AUs are being erased as specified by ERASE_SIZE. The range of
ERASE_TIMEOUT can be defined as up to 63 seconds and the card manufacturer can
choose any combination of ERASE_SIZE and ERASE_TIMEOUT depending on the
implementation. Determining ERASE_TIMEOUT determines the ERASE_SIZE.

Table 176. Erase timeout field

ERASE_TIMEOUT Value definition

00 Erase timeout calculation is not supported.

01 1 [sec]
02 2 [sec]
03 3 [sec]
--------- ---------
63 63 [sec]

ERASE_OFFSET
This 2-bit field indicates TOFFSET and one of four values can be selected. This field is
meaningless if the ERASE_SIZE and ERASE_TIMEOUT fields are set to 0.

Table 177. Erase offset field

ERASE_OFFSET Value definition

0h 0 [sec]
1h 1 [sec]

894/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Table 177. Erase offset field (continued)

ERASE_OFFSET Value definition

2h 2 [sec]
3h 3 [sec]

27.4.13 SD I/O mode

SD I/O interrupts
To allow the SD I/O card to interrupt the MultiMediaCard/SD module, an interrupt function is
available on a pin on the SD interface. Pin 8, used as SDIO_D1 when operating in the 4-bit
SD mode, signals the cards interrupt to the MultiMediaCard/SD module. The use of the
interrupt is optional for each card or function within a card. The SD I/O interrupt is level-
sensitive, which means that the interrupt line must be held active (low) until it is either
recognized and acted upon by the MultiMediaCard/SD module or deasserted due to the end
of the interrupt period. After the MultiMediaCard/SD module has serviced the interrupt, the
interrupt status bit is cleared via an I/O write to the appropriate bit in the SD I/O card’s
internal registers. The interrupt output of all SD I/O cards is active low and the application
must provide pull-up resistors externally on all data lines (SDIO_D[3:0]). The
MultiMediaCard/SD module samples the level of pin 8 (SDIO_D/IRQ) into the interrupt
detector only during the interrupt period. At all other times, the MultiMediaCard/SD module
ignores this value.
The interrupt period is applicable for both memory and I/O operations. The definition of the
interrupt period for operations with single blocks is different from the definition for multiple-
block data transfers.

SD I/O suspend and resume

Within a multifunction SD I/O or a card with both I/O and memory functions, there are
multiple devices (I/O and memory) that share access to the MMC/SD bus. To share access
to the MMC/SD module among multiple devices, SD I/O and combo cards optionally
implement the concept of suspend/resume. When a card supports suspend/resume, the
MMC/SD module can temporarily halt a data transfer operation to one function or memory
(suspend) to free the bus for a higher-priority transfer to a different function or memory. After
this higher-priority transfer is complete, the original transfer is resumed (restarted) where it
left off. Support of suspend/resume is optional on a per-card basis. To perform the
suspend/resume operation on the MMC/SD bus, the MMC/SD module performs the
following steps:
1. Determines the function currently using the SDIO_D [3:0] line(s)
2. Requests the lower-priority or slower transaction to suspend
3. Waits for the transaction suspension to complete
4. Begins the higher-priority transaction
5. Waits for the completion of the higher priority transaction
6. Restores the suspended transaction

SD I/O ReadWait
The optional ReadWait (RW) operation is defined only for the SD 1-bit and 4-bit modes. The
ReadWait operation allows the MMC/SD module to signal a card that it is reading multiple

RM0402 Rev 6 895/1163

921
Secure digital input/output interface (SDIO) RM0402

registers (IO_RW_EXTENDED, CMD53) to temporarily stall the data transfer while allowing
the MMC/SD module to send commands to any function within the SD I/O device. To
determine when a card supports the ReadWait protocol, the MMC/SD module must test
capability bits in the internal card registers. The timing for ReadWait is based on the
interrupt period.

27.4.14 Commands and responses

Application-specific and general commands
The SDIO card host module system is designed to provide a standard interface for a variety
of applications types. In this environment, there is a need for specific customer/application
features. To implement these features, two types of generic commands are defined in the
standard: application-specific commands (ACMD) and general commands (GEN_CMD).
When the card receives the APP_CMD (CMD55) command, the card expects the next
command to be an application-specific command. ACMDs have the same structure as
regular MultiMediaCard commands and can have the same CMD number. The card
recognizes it as ACMD because it appears after APP_CMD (CMD55). When the command
immediately following the APP_CMD (CMD55) is not a defined application-specific
command, the standard command is used. For example, when the card has a definition for
SD_STATUS (ACMD13), and receives CMD13 immediately following APP_CMD (CMD55),
this is interpreted as SD_STATUS (ACMD13). However, when the card receives CMD7
immediately following APP_CMD (CMD55) and the card does not have a definition for
ACMD7, this is interpreted as the standard (SELECT/DESELECT_CARD) CMD7.
To use one of the manufacturer-specific ACMDs the SD card Host must perform the
following steps:
1. Send APP_CMD (CMD55)
The card responds to the MultiMediaCard/SD module, indicating that the APP_CMD bit
is set and an ACMD is now expected.
2. Send the required ACMD
The card responds to the MultiMediaCard/SD module, indicating that the APP_CMD bit
is set and that the accepted command is interpreted as an ACMD. When a nonACMD
is sent, it is handled by the card as a normal MultiMediaCard command and the
APP_CMD bit in the card status register stays clear.
When an invalid command is sent (neither ACMD nor CMD) it is handled as a standard
MultiMediaCard illegal command error.
The bus transaction for a GEN_CMD is the same as the single-block read or write
commands (WRITE_BLOCK, CMD24 or READ_SINGLE_BLOCK,CMD17). In this case, the
argument denotes the direction of the data transfer rather than the address, and the data
block has vendor-specific format and meaning.
The card must be selected (in transfer state) before sending GEN_CMD (CMD56). The data
block size is defined by SET_BLOCKLEN (CMD16). The response to GEN_CMD (CMD56)
is in R1b format.

896/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Command types
Both application-specific and general commands are divided into the four following types:
• broadcast command (BC): sent to all cards; no responses returned.
• broadcast command with response (BCR): sent to all cards; responses received
from all cards simultaneously.
• addressed (point-to-point) command (AC): sent to the card that is selected; does
not include a data transfer on the SDIO_D line(s).
• addressed (point-to-point) data transfer command (ADTC): sent to the card that is
selected; includes a data transfer on the SDIO_D line(s).

Command formats
See Table 161 on page 871 for command formats.

Commands for the MultiMediaCard/SD module

Table 178. Block-oriented write commands

CMD Response
Type Argument Abbreviation Description
index format

Defines the number of blocks which

[31:16] set to 0
are going to be transferred in the
CMD23 ac [15:0] number R1 SET_BLOCK_COUNT
multiple-block read or write command
of blocks
that follows.
[31:0] data Writes a block of the size selected by
CMD24 adtc R1 WRITE_BLOCK
address the SET_BLOCKLEN command.
Continuously writes blocks of data
[31:0] data until a STOP_TRANSMISSION
CMD25 adtc R1 WRITE_MULTIPLE_BLOCK
address follows or the requested number of
blocks has been received.
Programming of the card identification
register. This command must be
issued only once per card. The card
CMD26 adtc [31:0] stuff bits R1 PROGRAM_CID contains hardware to prevent this
operation after the first programming.
Normally this command is reserved
for manufacturer.
Programming of the programmable
CMD27 adtc [31:0] stuff bits R1 PROGRAM_CSD
bits of the CSD.

RM0402 Rev 6 897/1163

921
Secure digital input/output interface (SDIO) RM0402

Table 179. Block-oriented write protection commands

CMD Response
Type Argument Abbreviation Description
index format

If the card has write protection features,

this command sets the write protection bit
[31:0] data
CMD28 ac R1b SET_WRITE_PROT of the addressed group. The properties of
address
write protection are coded in the card-
specific data (WP_GRP_SIZE).
If the card provides write protection
[31:0] data
CMD29 ac R1b CLR_WRITE_PROT features, this command clears the write
address
protection bit of the addressed group.
If the card provides write protection
[31:0] write
features, this command asks the card to
CMD30 adtc protect data R1 SEND_WRITE_PROT
send the status of the write protection
address
bits.
CMD31 Reserved

Table 180. Erase commands

CMD Response
Type Argument Abbreviation Description
index format

CMD32
Reserved. These command indexes cannot be used in order to maintain backward compatibility with older
...
versions of the MultiMediaCard.
CMD34
Sets the address of the first erase
CMD35 ac [31:0] data address R1 ERASE_GROUP_START group within a range to be selected
for erase.
Sets the address of the last erase
CMD36 ac [31:0] data address R1 ERASE_GROUP_END group within a continuous range to be
selected for erase.
Reserved. This command index cannot be used in order to maintain backward compatibility with older
CMD37
versions of the MultiMediaCards
Erases all previously selected write
CMD38 ac [31:0] stuff bits R1 ERASE
blocks.

Table 181. I/O mode commands

CMD Response
Type Argument Abbreviation Description
index format

Used to write and read 8-bit (register) data

[31:16] RCA fields. The command addresses a card and a
[15:15] register register and provides the data for writing if
write flag the write flag is set. The R4 response
CMD39 ac R4 FAST_IO
[14:8] register contains data read from the addressed
address register. This command accesses
[7:0] register data application-dependent registers that are not
defined in the MultiMediaCard standard.

898/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Table 181. I/O mode commands (continued)

CMD Response
Type Argument Abbreviation Description
index format

CMD40 bcr [31:0] stuff bits R5 GO_IRQ_STATE Places the system in the interrupt mode.
CMD41 Reserved

Table 182. Lock card

CMD Response
Type Argument Abbreviation Description
index format

Sets/resets the password or locks/unlocks

CMD42 adtc [31:0] stuff bits R1b LOCK_UNLOCK the card. The size of the data block is set
by the SET_BLOCK_LEN command.
CMD43
... Reserved
CMD54

Table 183. Application-specific commands

CMD Response
Type Argument Abbreviation Description
index format

Indicates to the card that the next command

[31:16] RCA
CMD55 ac R1 APP_CMD bits is an application specific command rather
[15:0] stuff bits than a standard command
Used either to transfer a data block to the card
or to get a data block from the card for general
[31:1] stuff bits
CMD56 adtc - - purpose/application-specific commands. The
[0]: RD/WR size of the data block shall be set by the
SET_BLOCK_LEN command.
CMD57
... Reserved.
CMD59
CMD60
... Reserved for manufacturer.
CMD63

27.5 Response formats

All responses are sent via the SDIO command line SDIO_CMD. The response transmission
always starts with the left bit of the bit string corresponding to the response code word. The
code length depends on the response type.
A response always starts with a start bit (always 0), followed by the bit indicating the
direction of transmission (card = 0). A value denoted by x in the tables below indicates a
variable entry. All responses, except for the R3 response type, are protected by a CRC.
Every command code word is terminated by the end bit (always 1).
There are five types of responses. Their formats are defined as follows:

RM0402 Rev 6 899/1163

921
Secure digital input/output interface (SDIO) RM0402

27.5.1 R1 (normal response command)

Code length = 48 bits. The 45:40 bits indicate the index of the command to be responded to,
this value being interpreted as a binary-coded number (between 0 and 63). The status of the
card is coded in 32 bits.

Table 184. R1 response

Bit position Width (bits Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 X Command index
[39:8] 32 X Card status
[7:1] 7 X CRC7
0 1 1 End bit

27.5.2 R1b
It is identical to R1 with an optional busy signal transmitted on the data line. The card may
become busy after receiving these commands based on its state prior to the command
reception.

27.5.3 R2 (CID, CSD register)

Code length = 136 bits. The contents of the CID register are sent as a response to the
CMD2 and CMD10 commands. The contents of the CSD register are sent as a response to
CMD9. Only the bits [127...1] of the CID and CSD are transferred, the reserved bit [0] of
these registers is replaced by the end bit of the response. The card indicates that an erase
is in progress by holding SDIO_D0 low. The actual erase time may be quite long, and the
host may issue CMD7 to deselect the card.

Table 185. R2 response

Bit position Width (bits Value Description

135 1 0 Start bit

134 1 0 Transmission bit
[133:128] 6 ‘111111’ Command index
[127:1] 127 X Card status
0 1 1 End bit

900/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

27.5.4 R3 (OCR register)

Code length: 48 bits. The contents of the OCR register are sent as a response to CMD1.
The level coding is as follows: restricted voltage windows = low, card busy = low.

Table 186. R3 response

Bit position Width (bits Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 ‘111111’ Reserved
[39:8] 32 X OCR register
[7:1] 7 ‘1111111’ Reserved
0 1 1 End bit

27.5.5 R4 (Fast I/O)

Code length: 48 bits. The argument field contains the RCA of the addressed card, the
register address to be read out or written to, and its content.

Table 187. R4 response

Bit position Width (bits Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 ‘100111’ CMD39
[31:16] 16 X RCA
[39:8] Argument field [15:8] 8 X register address
[7:0] 8 X read register contents
[7:1] 7 X CRC7
0 1 1 End bit

27.5.6 R4b
For SD I/O only: an SDIO card receiving the CMD5 will respond with a unique SDIO
response R4. The format is:

Table 188. R4b response

Bit position Width (bits Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 X Reserved

RM0402 Rev 6 901/1163

921
Secure digital input/output interface (SDIO) RM0402

Table 188. R4b response (continued)

Bit position Width (bits Value Description

39 16 X Card is ready
[38:36] 3 X Number of I/O functions
[39:8] Argument field 35 1 X Present memory
[34:32] 3 X Stuff bits
[31:8] 24 X I/O ORC
[7:1] 7 X Reserved
0 1 1 End bit

Once an SD I/O card has received a CMD5, the I/O portion of that card is enabled to
respond normally to all further commands. This I/O enable of the function within the I/O card
will remain set until a reset, power cycle or CMD52 with write to I/O reset is received by the
card. Note that an SD memory-only card may respond to a CMD5. The proper response for
a memory-only card would be Present memory = 1 and Number of I/O functions = 0. A
memory-only card built to meet the SD Memory Card specification version 1.0 would detect
the CMD5 as an illegal command and not respond. The I/O aware host will send CMD5. If
the card responds with response R4, the host determines the card’s configuration based on
the data contained within the R4 response.

27.5.7 R5 (interrupt request)

Only for MultiMediaCard. Code length: 48 bits. If the response is generated by the host, the
RCA field in the argument will be 0x0.

Table 189. R5 response

Bit position Width (bits Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 ‘101000’ CMD40
RCA [31:16] of winning
[31:16] 16 X
card or of the host
[39:8] Argument field
Not defined. May be used
[15:0] 16 X
for IRQ data
[7:1] 7 X CRC7
0 1 1 End bit

27.5.8 R6
Only for SD I/O. The normal response to CMD3 by a memory device. It is shown in
Table 190.

902/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Table 190. R6 response

Bit position Width (bits) Value Description

47 1 0 Start bit
46 1 0 Transmission bit
[45:40] 6 ‘101000’ CMD40

[39:8] Argument [31:16] 16 X RCA [31:16] of winning card or of the host

field [15:0] 16 X Not defined. May be used for IRQ data
[7:1] 7 X CRC7
0 1 1 End bit

The card [23:8] status bits are changed when CMD3 is sent to an I/O-only card. In this case,
the 16 bits of response are the SD I/O-only values:
• Bit [15] COM_CRC_ERROR
• Bit [14] ILLEGAL_COMMAND
• Bit [13] ERROR
• Bits [12:0] Reserved

27.6 SDIO I/O card-specific operations

The following features are SD I/O-specific operations:
• SDIO read wait operation by SDIO_D2 signalling
• SDIO read wait operation by stopping the clock
• SDIO suspend/resume operation (write and read suspend)
• SDIO interrupts
The SDIO supports these operations only if the SDIO_DCTRL[11] bit is set, except for read
suspend that does not need specific hardware implementation.

27.6.1 SDIO I/O read wait operation by SDIO_D2 signalling

It is possible to start the readwait interval before the first block is received: when the data
path is enabled (SDIO_DCTRL[0] bit set), the SDIO-specific operation is enabled
(SDIO_DCTRL[11] bit set), read wait starts (SDIO_DCTRL[10] =0 and SDIO_DCTRL[8] =1)
and data direction is from card to SDIO (SDIO_DCTRL[1] = 1), the DPSM directly moves
from Idle to Readwait. In Readwait the DPSM drives SDIO_D2 to 0 after 2 SDIO_CK clock
cycles. In this state, when you set the RWSTOP bit (SDIO_DCTRL[9]), the DPSM remains
in Wait for two more SDIO_CK clock cycles to drive SDIO_D2 to 1 for one clock cycle (in
accordance with SDIO specification). The DPSM then starts waiting again until it receives
data from the card. The DPSM will not start a readwait interval while receiving a block even
if read wait start is set: the readwait interval will start after the CRC is received. The
RWSTOP bit has to be cleared to start a new read wait operation. During the readwait
interval, the SDIO can detect SDIO interrupts on SDIO_D1.

RM0402 Rev 6 903/1163

921
Secure digital input/output interface (SDIO) RM0402

27.6.2 SDIO read wait operation by stopping SDIO_CK

If the SDIO card does not support the previous read wait method, the SDIO can perform a
read wait by stopping SDIO_CK (SDIO_DCTRL is set just like in the method presented in
Section 27.6.1, but SDIO_DCTRL[10] =1): DSPM stops the clock two SDIO_CK cycles after
the end bit of the current received block and starts the clock again after the read wait start bit
is set.
As SDIO_CK is stopped, any command can be issued to the card. During a read/wait
interval, the SDIO can detect SDIO interrupts on SDIO_D1.

27.6.3 SDIO suspend/resume operation

While sending data to the card, the SDIO can suspend the write operation. the
SDIO_CMD[11] bit is set and indicates to the CPSM that the current command is a suspend
command. The CPSM analyzes the response and when the ACK is received from the card
(suspend accepted), it acknowledges the DPSM that goes Idle after receiving the CRC
token of the current block.
The hardware does not save the number of the remaining block to be sent to complete the
suspended operation (resume).
The write operation can be suspended by software, just by disabling the DPSM
(SDIO_DCTRL[0] =0) when the ACK of the suspend command is received from the card.
The DPSM enters then the Idle state.
To suspend a read: the DPSM waits in the Wait_r state as the function to be suspended
sends a complete packet just before stopping the data transaction. The application
continues reading RxFIFO until the FIF0 is empty, and the DPSM goes Idle automatically.

27.6.4 SDIO interrupts

SDIO interrupts are detected on the SDIO_D1 line once the SDIO_DCTRL[11] bit is set.
When SDIO interrupt is detected, SDIO_STA[22] (SDIOIT) bit is set. This static bit can be
cleared with clear bit SDIO_ICR[22] (SDIOITC). An interrupt can be generated when
SDIOIT status bit is set. Separated interrupt enable SDIO_MASK[22] bit (SDIOITE) is
available to enable and disable interrupt request.
When SD card interrupt occurs (SDIO_STA[22] bit set), host software follows below steps to
handle it.
1. Disable SDIOIT interrupt signaling by clearing SDIOITE bit (SDIO_MASK[22] = ‘0’),
2. Serve card interrupt request, and clear the source of interrupt on the SD card,
3. Clear SDIOIT bit by writing ‘1’ to SDIOITC bit (SDIO_ICR[22] = ‘1’),
4. Enable SDIOIT interrupt signaling by writing ‘1’ to SDIOITE bit (SDIO_MASK[22] = ‘1’).
Steps 2 to 4 can be executed out of the SDIO interrupt service routine.

27.7 HW flow control

The HW flow control functionality is used to avoid FIFO underrun (TX mode) and overrun
(RX mode) errors.
The behavior is to stop SDIO_CK and freeze SDIO state machines. The data transfer is
stalled while the FIFO is unable to transmit or receive data. Only state machines clocked by

904/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

SDIOCLK are frozen, the APB2 interface is still alive. The FIFO can thus be filled or emptied
even if flow control is activated.
To enable HW flow control, the SDIO_CLKCR[14] register bit must be set to 1. After reset
Flow Control is disabled.

27.8 SDIO registers

The device communicates to the system via 32-bit-wide control registers accessible via
APB2.

27.8.1 SDIO power control register (SDIO_POWER)

Address offset: 0x00
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PWRCTRL

rw rw

Bits 31:2 Reserved, must be kept at reset value.

[1:0] PWRCTRL: Power supply control bits.
These bits are used to define the current functional state of the card clock:
00: Power-off: the clock to card is stopped.
01: Reserved
10: Reserved power-up
11: Power-on: the card is clocked.

Note: At least seven PCLK2 clock periods are needed between two write accesses to this register.
Note: After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.

27.8.2 SDIO clock control register (SDIO_CLKCR)

Address offset: 0x04
Reset value: 0x0000 0000
The SDIO_CLKCR register controls the SDIO_CK output clock.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HWFC NEGE WID BYPAS PWRS
Res. CLKEN CLKDIV
_EN DGE BUS S AV

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 905/1163

921
Secure digital input/output interface (SDIO) RM0402

Bits 31:15 Reserved, must be kept at reset value.

Bit 14 HWFC_EN: HW Flow Control enable
0b: HW Flow Control is disabled
1b: HW Flow Control is enabled
When HW Flow Control is enabled, the meaning of the TXFIFOE and RXFIFOF interrupt
signals, see SDIO Status register definition in Section 27.8.11.
Bit 13 NEGEDGE: SDIO_CK dephasing selection bit
0b: Command and Data changed on the SDIOCLK falling edge succeeding the rising edge
of SDIO_CK. (SDIO_CK rising edge occurs on SDIOCLK rising edge).
1b: Command and Data changed on the SDIO_CK falling edge.
When BYPASS is active, the data and the command change on SDIOCLK falling edge
whatever NEGEDGE value.
Bits 12:11 WIDBUS: Wide bus mode enable bit
00: Default bus mode: SDIO_D0 used
01: 4-wide bus mode: SDIO_D[3:0] used
10: 8-wide bus mode: SDIO_D[7:0] used
Bit 10 BYPASS: Clock divider bypass enable bit
0: Disable bypass: SDIOCLK is divided according to the CLKDIV value before driving the
SDIO_CK output signal.
1: Enable bypass: SDIOCLK directly drives the SDIO_CK output signal.
Bit 9 PWRSAV: Power saving configuration bit
For power saving, the SDIO_CK clock output can be disabled when the bus is idle by setting
PWRSAV:
0: SDIO_CK clock is always enabled
1: SDIO_CK is only enabled when the bus is active
Bit 8 CLKEN: Clock enable bit
0: SDIO_CK is disabled
1: SDIO_CK is enabled
Bits 7:0 CLKDIV: Clock divide factor
This field defines the divide factor between the input clock (SDIOCLK) and the output clock
(SDIO_CK): SDIO_CK frequency = SDIOCLK / [CLKDIV + 2].

Note: 1 While the SD/SDIO card or MultiMediaCard is in identification mode, the SDIO_CK
frequency must be less than 400 kHz.
2 The clock frequency can be changed to the maximum card bus frequency when relative
card addresses are assigned to all cards.
3 After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods. SDIO_CK can also be stopped during the read wait interval
for SD I/O cards: in this case the SDIO_CLKCR register does not control SDIO_CK.

906/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

27.8.3 SDIO argument register (SDIO_ARG)

Address offset: 0x08
Reset value: 0x0000 0000
The SDIO_ARG register contains a 32-bit command argument, which is sent to a card as
part of a command message.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CMDARG[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CMDARG[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 CMDARG: Command argument

Command argument sent to a card as part of a command message. If a command contains
an argument, it must be loaded into this register before writing a command to the command
register.

27.8.4 SDIO command register (SDIO_CMD)

Address offset: 0x0C
Reset value: 0x0000 0000
The SDIO_CMD register contains the command index and command type bits. The
command index is sent to a card as part of a command message. The command type bits
control the command path state machine (CPSM).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SDIO CPSM WAIT WAIT
Res. Res. Res. Res. WAITRESP CMDINDEX
Suspend EN PEND INT

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:12 Reserved, must be kept at reset value.

Bit 11 SDIOSuspend: SD I/O suspend command
If this bit is set, the command to be sent is a suspend command (to be used only with SDIO
card).
Bit 10 CPSMEN: Command path state machine (CPSM) Enable bit
If this bit is set, the CPSM is enabled.
Bit 9 WAITPEND: CPSM Waits for ends of data transfer (CmdPend internal signal).
If this bit is set, the CPSM waits for the end of data transfer before it starts sending a
command. This feature is available only with Stream data transfer mode
SDIO_DCTRL[2] = 1.

RM0402 Rev 6 907/1163

921
Secure digital input/output interface (SDIO) RM0402

Bit 8 WAITINT: CPSM waits for interrupt request

If this bit is set, the CPSM disables command timeout and waits for an interrupt request.
Bits 7:6 WAITRESP: Wait for response bits
They are used to configure whether the CPSM is to wait for a response, and if yes, which
kind of response.
00: No response, expect CMDSENT flag
01: Short response, expect CMDREND or CCRCFAIL flag
10: No response, expect CMDSENT flag
11: Long response, expect CMDREND or CCRCFAIL flag
Bits 5:0 CMDINDEX: Command index
The command index is sent to the card as part of a command message.

Note: 1 After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.
2 MultiMediaCards can send two kinds of response: short responses, 48 bits long, or long
responses,136 bits long. SD card and SD I/O card can send only short responses, the
argument can vary according to the type of response: the software will distinguish the type
of response according to the sent command.

27.8.5 SDIO command response register (SDIO_RESPCMD)

Address offset: 0x10
Reset value: 0x0000 0000
The SDIO_RESPCMD register contains the command index field of the last command
response received. If the command response transmission does not contain the command
index field (long or OCR response), the RESPCMD field is unknown, although it must
contain 111111b (the value of the reserved field from the response).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. RESPCMD

r r r r r r

Bits 31:6 Reserved, must be kept at reset value.

Bits 5:0 RESPCMD: Response command index
Read-only bit field. Contains the command index of the last command response received.

27.8.6 SDIO response 1..4 register (SDIO_RESPx)

Address offset: (0x10 + (4 × x)); x = 1..4
Reset value: 0x0000 0000
The SDIO_RESP1/2/3/4 registers contain the status of a card, which is part of the received
response.

908/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CARDSTATUSx[31:16]
r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CARDSTATUSx[15:0]
r r r r r r r r r r r r r r r r

Bits 31:0 CARDSTATUSx: see Table 191.

The Card Status size is 32 or 127 bits, depending on the response type.

Table 191. Response type and SDIO_RESPx registers

Register Short response Long response

SDIO_RESP1 Card Status[31:0] Card Status [127:96]

SDIO_RESP2 Unused Card Status [95:64]
SDIO_RESP3 Unused Card Status [63:32]
SDIO_RESP4 Unused Card Status [31:1]0b

The most significant bit of the card status is received first. The SDIO_RESP4 register LSB is
always 0b.

27.8.7 SDIO data timer register (SDIO_DTIMER)

Address offset: 0x24
Reset value: 0x0000 0000
The SDIO_DTIMER register contains the data timeout period, in card bus clock periods.
A counter loads the value from the SDIO_DTIMER register, and starts decrementing when
the data path state machine (DPSM) enters the Wait_R or Busy state. If the timer reaches 0
while the DPSM is in either of these states, the timeout status flag is set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATATIME[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATATIME[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 DATATIME: Data timeout period

Data timeout period expressed in card bus clock periods.

Note: A data transfer must be written to the data timer register and the data length register before
being written to the data control register.

RM0402 Rev 6 909/1163

921
Secure digital input/output interface (SDIO) RM0402

27.8.8 SDIO data length register (SDIO_DLEN)

Address offset: 0x28
Reset value: 0x0000 0000
The SDIO_DLEN register contains the number of data bytes to be transferred. The value is
loaded into the data counter when data transfer starts.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. DATALENGTH[24:16]

rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATALENGTH[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:25 Reserved, must be kept at reset value.

Bits 24:0 DATALENGTH: Data length value
Number of data bytes to be transferred.

Note: For a block data transfer, the value in the data length register must be a multiple of the block
size (see SDMMC_DCTRL). Before being written to the data control register a timeout must
be written to the data timer register and the data length register.
In case of IO_RW_EXTENDED (CMD53):
- If the Stream or SDIO multibyte data transfer is selected the value in the data length
register must be between 1 and 512.
- If the Block data transfer is selected the value in the data length register must be between
1*Data block size and 512*Data block size.

27.8.9 SDIO data control register (SDIO_DCTRL)

Address offset: 0x2C
Reset value: 0x0000 0000
The SDIO_DCTRL register control the data path state machine (DPSM).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SDIO RW RW RW DMA DT
Res. Res. Res. Res. DBLOCKSIZE DTDIR DTEN
EN MOD STOP START EN MODE

rw rw rw rw rw rw rw rw rw rw rw rw

910/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Bits 31:12 Reserved, must be kept at reset value.

Bit 11 SDIOEN: SD I/O enable functions
If this bit is set, the DPSM performs an SD I/O-card-specific operation.
Bit 10 RWMOD: Read wait mode
0: Read Wait control stopping SDIO_D2
1: Read Wait control using SDIO_CK
Bit 9 RWSTOP: Read wait stop
0: Read wait in progress if RWSTART bit is set
1: Enable for read wait stop if RWSTART bit is set
Bit 8 RWSTART: Read wait start
If this bit is set, read wait operation starts.
Bits 7:4 DBLOCKSIZE: Data block size
Define the data block length when the block data transfer mode is selected:
0000: (0 decimal) lock length = 20 = 1 byte
0001: (1 decimal) lock length = 21 = 2 bytes
0010: (2 decimal) lock length = 22 = 4 bytes
0011: (3 decimal) lock length = 23 = 8 bytes
0100: (4 decimal) lock length = 24 = 16 bytes
0101: (5 decimal) lock length = 25 = 32 bytes
0110: (6 decimal) lock length = 26 = 64 bytes
0111: (7 decimal) lock length = 27 = 128 bytes
1000: (8 decimal) lock length = 28 = 256 bytes
1001: (9 decimal) lock length = 29 = 512 bytes
1010: (10 decimal) lock length = 210 = 1024 bytes
1011: (11 decimal) lock length = 211 = 2048 bytes
1100: (12 decimal) lock length = 212 = 4096 bytes
1101: (13 decimal) lock length = 213 = 8192 bytes
1110: (14 decimal) lock length = 214 = 16384 bytes
1111: (15 decimal) reserved
Bit 3 DMAEN: DMA enable bit
0: DMA disabled.
1: DMA enabled.
Bit 2 DTMODE: Data transfer mode selection 1: Stream or SDIO multibyte data transfer.
0: Block data transfer
1: Stream or SDIO multibyte data transfer
Bit 1 DTDIR: Data transfer direction selection
0: From controller to card.
1: From card to controller.
[0] DTEN: Data transfer enabled bit
Data transfer starts if 1b is written to the DTEN bit. Depending on the direction bit, DTDIR,
the DPSM moves to the Wait_S, Wait_R state or Readwait if RW Start is set immediately at
the beginning of the transfer. It is not necessary to clear the enable bit after the end of a data
transfer but the SDIO_DCTRL must be updated to enable a new data transfer

RM0402 Rev 6 911/1163

921
Secure digital input/output interface (SDIO) RM0402

Note: After a data write, data cannot be written to this register for three SDIOCLK clock periods
plus two PCLK2 clock periods.
The meaning of the DTMODE bit changes according to the value of the SDIOEN bit. When
SDIOEN=0 and DTMODE=1, the MultiMediaCard stream mode is enabled, and when
SDIOEN=1 and DTMODE=1, the peripheral enables an SDIO multibyte transfer.

912/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

27.8.10 SDIO data counter register (SDIO_DCOUNT)

Address offset: 0x30
Reset value: 0x0000 0000
The SDIO_DCOUNT register loads the value from the data length register (see
SDIO_DLEN) when the DPSM moves from the Idle state to the Wait_R or Wait_S state. As
data is transferred, the counter decrements the value until it reaches 0. The DPSM then
moves to the Idle state and the data status end flag, DATAEND, is set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. DATACOUNT[24:16]

r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATACOUNT[15:0]
r r r r r r r r r r r r r r r r

Bits 31:25 Reserved, must be kept at reset value.

Bits 24:0 DATACOUNT: Data count value
When this bit is read, the number of remaining data bytes to be transferred is returned. Write
has no effect.

Note: This register should be read only when the data transfer is complete.

27.8.11 SDIO status register (SDIO_STA)

Address offset: 0x34
Reset value: 0x0000 0000
The SDIO_STA register is a read-only register. It contains two types of flag:
• Static flags (bits [23:22,10:0]): these bits remain asserted until they are cleared by
writing to the SDIO Interrupt Clear register (see SDIO_ICR)
• Dynamic flags (bits [21:11]): these bits change state depending on the state of the
underlying logic (for example, FIFO full and empty flags are asserted and deasserted
as data while written to the FIFO)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RXD TXD RX TX RX TX
Res. Res. Res. Res. Res. Res. Res. Res. Res. SDIOIT
AVL AVL FIFOE FIFOE FIFOF FIFOF
r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RX TX
CMD DBCK DATA CMDS CMDR RX TXUND DTIME CTIME DCRC CCRC
FIFO FIFO RXACT TXACT Res.
ACT END END ENT END OVERR ERR OUT OUT FAIL FAIL
HF HE
r r r r r r r r r r r r r r r

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 SDIOIT: SDIO interrupt received
Bit 21 RXDAVL: Data available in receive FIFO

RM0402 Rev 6 913/1163

921
Secure digital input/output interface (SDIO) RM0402

Bit 20 TXDAVL: Data available in transmit FIFO

Bit 19 RXFIFOE: Receive FIFO empty
Bit 18 TXFIFOE: Transmit FIFO empty
When HW Flow Control is enabled, TXFIFOE signals becomes activated when the FIFO
contains 2 words.
Bit 17 RXFIFOF: Receive FIFO full
When HW Flow Control is enabled, RXFIFOF signals becomes activated 2 words before the
FIFO is full.
Bit 16 TXFIFOF: Transmit FIFO full
Bit 15 RXFIFOHF: Receive FIFO half full: there are at least 8 words in the FIFO
Bit 14 TXFIFOHE: Transmit FIFO half empty: at least 8 words can be written into the FIFO
Bit 13 RXACT: Data receive in progress
Bit 12 TXACT: Data transmit in progress
Bit 11 CMDACT: Command transfer in progress
Bit 10 DBCKEND: Data block sent/received (CRC check passed)
Bit 9 Reserved, must be kept at reset value.
Bit 8 DATAEND: Data end (data counter, SDIDCOUNT, is zero)
Bit 7 CMDSENT: Command sent (no response required)
Bit 6 CMDREND: Command response received (CRC check passed)
Bit 5 RXOVERR: Received FIFO overrun error
Note: If DMA is used to read SDIO FIFO (DMAEN bit is set in SDIO_DCTRL register), user
software should disable DMA stream, and then write with ‘0’ (to disable DMA request
generation).
Bit 4 TXUNDERR: Transmit FIFO underrun error
Note: If DMA is used to fill SDIO FIFO (DMAEN bit is set in SDIO_DCTRL register), user
software should disable DMA stream, and then write DMAEN with ‘0’ (to disable DMA
request generation).
Bit 3 DTIMEOUT: Data timeout
Bit 2 CTIMEOUT: Command response timeout
The Command TimeOut period has a fixed value of 64 SDIO_CK clock periods.
Bit 1 DCRCFAIL: Data block sent/received (CRC check failed)
Bit 0 CCRCFAIL: Command response received (CRC check failed)

27.8.12 SDIO interrupt clear register (SDIO_ICR)

Address offset: 0x38
Reset value: 0x0000 0000
The SDIO_ICR register is a write-only register. Writing a bit with 1b clears the corresponding
bit in the SDIO_STA Status register.

914/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SDIO
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
ITC
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
CMD RX TX
DBCK DATA CMD DTIME CTIME DCRC CCRC
Res. Res. Res. Res. Res. Res. REND OVERR UNDERR
ENDC ENDC SENTC OUTC OUTC FAILC FAILC
C C C
rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 SDIOITC: SDIOIT flag clear bit
Set by software to clear the SDIOIT flag.
0: SDIOIT not cleared
1: SDIOIT cleared
Bits 21:11 Reserved, must be kept at reset value.
Bit 10 DBCKENDC: DBCKEND flag clear bit
Set by software to clear the DBCKEND flag.
0: DBCKEND not cleared
1: DBCKEND cleared
Bit 9 Reserved, must be kept at reset value.
Bit 8 DATAENDC: DATAEND flag clear bit
Set by software to clear the DATAEND flag.
0: DATAEND not cleared
1: DATAEND cleared
Bit 7 CMDSENTC: CMDSENT flag clear bit
Set by software to clear the CMDSENT flag.
0: CMDSENT not cleared
1: CMDSENT cleared
Bit 6 CMDRENDC: CMDREND flag clear bit
Set by software to clear the CMDREND flag.
0: CMDREND not cleared
1: CMDREND cleared
Bit 5 RXOVERRC: RXOVERR flag clear bit
Set by software to clear the RXOVERR flag.
0: RXOVERR not cleared
1: RXOVERR cleared
Bit 4 TXUNDERRC: TXUNDERR flag clear bit
Set by software to clear TXUNDERR flag.
0: TXUNDERR not cleared
1: TXUNDERR cleared
Bit 3 DTIMEOUTC: DTIMEOUT flag clear bit
Set by software to clear the DTIMEOUT flag.
0: DTIMEOUT not cleared
1: DTIMEOUT cleared

RM0402 Rev 6 915/1163

921
Secure digital input/output interface (SDIO) RM0402

Bit 2 CTIMEOUTC: CTIMEOUT flag clear bit

Set by software to clear the CTIMEOUT flag.
0: CTIMEOUT not cleared
1: CTIMEOUT cleared
Bit 1 DCRCFAILC: DCRCFAIL flag clear bit
Set by software to clear the DCRCFAIL flag.
0: DCRCFAIL not cleared
1: DCRCFAIL cleared
Bit 0 CCRCFAILC: CCRCFAIL flag clear bit
Set by software to clear the CCRCFAIL flag.
0: CCRCFAIL not cleared
1: CCRCFAIL cleared

27.8.13 SDIO mask register (SDIO_MASK)

Address offset: 0x3C
Reset value: 0x0000 0000
The interrupt mask register determines which status flags generate an interrupt request by
setting the corresponding bit to 1b.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RX TX RX TX
SDIO RXD TXD
Res. Res. Res. Res. Res. Res. Res. Res. Res. FIFO FIFO FIFO FIFO
ITIE AVLIE AVLIE
EIE EIE FIE FIE
rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RX TX CMD CMD RX TX
RX TX CMD DBCK DATA DTIME CTIME DCRC CCRC
FIFO FIFO Res. SENT REND OVERR UNDERR
ACTIE ACTIE ACTIE ENDIE ENDIE OUTIE OUTIE FAILIE FAILIE
HFIE HEIE IE IE IE IE
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:23 Reserved, must be kept at reset value.

Bit 22 SDIOITIE: SDIO mode interrupt received interrupt enable
Set and cleared by software to enable/disable the interrupt generated when receiving the
SDIO mode interrupt.
0: SDIO Mode Interrupt Received interrupt disabled
1: SDIO Mode Interrupt Received interrupt enabled
Bit 21 RXDAVLIE: Data available in Rx FIFO interrupt enable
Set and cleared by software to enable/disable the interrupt generated by the presence of
data available in Rx FIFO.
0: Data available in Rx FIFO interrupt disabled
1: Data available in Rx FIFO interrupt enabled
Bit 20 TXDAVLIE: Data available in Tx FIFO interrupt enable
Set and cleared by software to enable/disable the interrupt generated by the presence of
data available in Tx FIFO.
0: Data available in Tx FIFO interrupt disabled
1: Data available in Tx FIFO interrupt enabled

916/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

Bit 19 RXFIFOEIE: Rx FIFO empty interrupt enable

Set and cleared by software to enable/disable interrupt caused by Rx FIFO empty.
0: Rx FIFO empty interrupt disabled
1: Rx FIFO empty interrupt enabled
Bit 18 TXFIFOEIE: Tx FIFO empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO empty.
0: Tx FIFO empty interrupt disabled
1: Tx FIFO empty interrupt enabled
Bit 17 RXFIFOFIE: Rx FIFO full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO full.
0: Rx FIFO full interrupt disabled
1: Rx FIFO full interrupt enabled
Bit 16 TXFIFOFIE: Tx FIFO full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO full.
0: Tx FIFO full interrupt disabled
1: Tx FIFO full interrupt enabled
Bit 15 RXFIFOHFIE: Rx FIFO half full interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO half full.
0: Rx FIFO half full interrupt disabled
1: Rx FIFO half full interrupt enabled
Bit 14 TXFIFOHEIE: Tx FIFO half empty interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO half empty.
0: Tx FIFO half empty interrupt disabled
1: Tx FIFO half empty interrupt enabled
Bit 13 RXACTIE: Data receive acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by data being received (data
receive acting).
0: Data receive acting interrupt disabled
1: Data receive acting interrupt enabled
Bit 12 TXACTIE: Data transmit acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by data being transferred
(data transmit acting).
0: Data transmit acting interrupt disabled
1: Data transmit acting interrupt enabled
Bit 11 CMDACTIE: Command acting interrupt enable
Set and cleared by software to enable/disable interrupt caused by a command being
transferred (command acting).
0: Command acting interrupt disabled
1: Command acting interrupt enabled
Bit 10 DBCKENDIE: Data block end interrupt enable
Set and cleared by software to enable/disable interrupt caused by data block end.
0: Data block end interrupt disabled
1: Data block end interrupt enabled
Bit 9 Reserved, must be kept at reset value.

RM0402 Rev 6 917/1163

921
Secure digital input/output interface (SDIO) RM0402

Bit 8 DATAENDIE: Data end interrupt enable

Set and cleared by software to enable/disable interrupt caused by data end.
0: Data end interrupt disabled
1: Data end interrupt enabled
Bit 7 CMDSENTIE: Command sent interrupt enable
Set and cleared by software to enable/disable interrupt caused by sending command.
0: Command sent interrupt disabled
1: Command sent interrupt enabled
Bit 6 CMDRENDIE: Command response received interrupt enable
Set and cleared by software to enable/disable interrupt caused by receiving command
response.
0: Command response received interrupt disabled
1: command Response Received interrupt enabled
Bit 5 RXOVERRIE: Rx FIFO overrun error interrupt enable
Set and cleared by software to enable/disable interrupt caused by Rx FIFO overrun error.
0: Rx FIFO overrun error interrupt disabled
1: Rx FIFO overrun error interrupt enabled
Bit 4 TXUNDERRIE: Tx FIFO underrun error interrupt enable
Set and cleared by software to enable/disable interrupt caused by Tx FIFO underrun error.
0: Tx FIFO underrun error interrupt disabled
1: Tx FIFO underrun error interrupt enabled
Bit 3 DTIMEOUTIE: Data timeout interrupt enable
Set and cleared by software to enable/disable interrupt caused by data timeout.
0: Data timeout interrupt disabled
1: Data timeout interrupt enabled
Bit 2 CTIMEOUTIE: Command timeout interrupt enable
Set and cleared by software to enable/disable interrupt caused by command timeout.
0: Command timeout interrupt disabled
1: Command timeout interrupt enabled
Bit 1 DCRCFAILIE: Data CRC fail interrupt enable
Set and cleared by software to enable/disable interrupt caused by data CRC failure.
0: Data CRC fail interrupt disabled
1: Data CRC fail interrupt enabled
Bit 0 CCRCFAILIE: Command CRC fail interrupt enable
Set and cleared by software to enable/disable interrupt caused by command CRC failure.
0: Command CRC fail interrupt disabled
1: Command CRC fail interrupt enabled

27.8.14 SDIO FIFO counter register (SDIO_FIFOCNT)

Address offset: 0x48
Reset value: 0x0000 0000
The SDIO_FIFOCNT register contains the remaining number of words to be written to or
read from the FIFO. The FIFO counter loads the value from the data length register (see
SDIO_DLEN) when the data transfer enable bit, DTEN, is set in the data control register
(SDIO_DCTRL register) and the DPSM is at the Idle state. If the data length is not word-
aligned (multiple of 4), the remaining 1 to 3 bytes are regarded as a word.

918/1163 RM0402 Rev 6

RM0402 Secure digital input/output interface (SDIO)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. FIFOCOUNT[23:16]

r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FIFOCOUNT[15:0]
r r r r r r r r r r r r r r r r

Bits 31:24 Reserved, must be kept at reset value.

Bits 23:0 FIFOCOUNT: Remaining number of words to be written to or read from the FIFO.

27.8.15 SDIO data FIFO register (SDIO_FIFO)

Address offset: 0x80
Reset value: 0x0000 0000
The receive and transmit FIFOs can be read or written as 32-bit wide registers. The FIFOs
contain 32 entries on 32 sequential addresses. This allows the CPU to use its load and store
multiple operands to read from/write to the FIFO.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FIF0Data[31:16]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FIF0Data[15:0]
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

bits 31:0 FIFOData: Receive and transmit FIFO data

The FIFO data occupies 32 entries of 32-bit words, from address:
SDIO base + 0x080 to SDIO base + 0xFC.

RM0402 Rev 6 919/1163

921
Secure digital input/output interface (SDIO) RM0402

27.8.16 SDIO register map

The following table summarizes the SDIO registers.

Table 192. SDIO register map

Offset Register 31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
PWRCTRL
SDIO_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
POWER
0x00

Reset value 0 0

NEGEDGE
HWFC_EN

PWRSAV
WIDBUS

BYPASS

CLKDIV
CLKEN
SDIO_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
CLKCR
0x04

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

SDIO_ARG CMDARG
0x08
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

SDIOSuspend

CMDINDEX
WAITPEND

WAITRESP
CPSMEN

WAITINT
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
SDIO_CMD
0x0C

Reset value 0 0 0 0 0 0 0 0 0 0 0 0

SDIO_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
RESPCMD
RESPCMD
0x10
Reset value 0 0 0 0 0 0

SDIO_
CARDSTATUS1
0x14 RESP1

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SDIO_
CARDSTATUS2
0x18 RESP2

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SDIO_
CARDSTATUS3
0x1C RESP3

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SDIO_
CARDSTATUS4
0x20 RESP4

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SDIO_
DATATIME
0x24 DTIMER

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SDIO_
Res.
Res.
Res.
Res.
Res.
Res.
Res.

DATALENGTH
DLEN
0x28
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DBLOCKSIZE
RWSTART
RWSTOP

DTMODE
RWMOD
SDIOEN

DMAEN

DTDIR
DTEN

SDIO_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

DCTRL
0x2C

Reset value 0 0 0 0 0 0 0 0 0 0 0 0

920/1163 RM0402 Rev 6

 0x80
0x48
0x38
0x34
0x30

0x3C
Offset
RM0402

MASK

SDIO_
SDIO_
SDIO_
DCOUNT

FIFOCNT
SDIO_ICR
SDIO_STA
Register

Reset value
SDIO_FIFO
Reset value
Reset value
Reset value
Reset value
Reset value

0
Res. Res. Res. Res. Res. 31

0
Res. Res. Res. Res. Res. 30

0
Res. Res. Res. Res. Res. 29

0
Res. Res. Res. Res. Res. 28

0
Res. Res. Res. Res. Res. 27

0
Res. Res. Res. Res. Res. 26

0
Res. Res. Res. Res. Res. 25

0
0

Res. Res. Res. Res. 24

0
0
0

Res. Res. Res. 23

0
0
0
0
0
0

SDIOITIE SDIOITC SDIOIT 22

0
0
0
0
0

RXDAVLIE Res. RXDAVL 21

0
0
0
0
0

TXDAVLIE Res. TXDAVL 20

0
0
0
0
0

RXFIFOEIE Res. RXFIFOE 19

0
0
0
0
0

TXFIFOEIE Res. TXFIFOE 18

0
0
0
0
0

RXFIFOFIE Res. RXFIFOF 17

RM0402 Rev 6
0
0
0
0
0

TXFIFOFIE Res. TXFIFOF 16

0
0
0
0
0

RXFIFOHFIE Res. RXFIFOHF 15

FIF0Data
0
0
0
0
0

TXFIFOHEIE Res. TXFIFOHE 14

0
0
0
0
0

RXACTIE Res. RXACT 13

0
0
0
0
0

TXACTIE Res. TXACT 12

0
0
0
0
0
Table 192. SDIO register map (continued)

CMDACTIE Res. CMDACT 11

DATACOUNT

0
0
0
0
0

FIFOCOUNT

DBCKENDIE DBCKENDC DBCKEND 10

0
0
0

Res. Res. Res. 9

Refer to Section 2.2 on page 49 for the register boundary addresses.

0
0
0
0
0

0 DATAENDIE DATAENDC DATAEND 8

0
0
0
0

CMDSENTIE CMDSENTC CMDSENT 7

0
0
0
0
0

CMDRENDIE CMDRENDC CMDREND 6

0
0
0
0
0

RXOVERRIE RXOVERRC RXOVERR 5

0
0
0
0
0

TXUNDERRIE TXUNDERRC TXUNDERR 4

0
0
0
0
0

DTIMEOUTIE DTIMEOUTC DTIMEOUT 3

0
0
0
0
0

CTIMEOUTIE CTIMEOUTC CTIMEOUT 2

0
0
0
0
0

DCRCFAILIE DCRCFAILC DCRCFAIL 1

0
0
0
0
0

CCRCFAILIE CCRCFAILC CCRCFAIL 0

921/1163
Secure digital input/output interface (SDIO)

921
Controller area network (bxCAN) RM0402

28 Controller area network (bxCAN)

28.1 Introduction
The Basic Extended CAN peripheral, named bxCAN, interfaces the CAN network. It
supports the CAN protocols version 2.0A and B. It has been designed to manage a high
number of incoming messages efficiently with a minimum CPU load. It also meets the
priority requirements for transmit messages.
For safety-critical applications, the CAN controller provides all hardware functions for
supporting the CAN Time Triggered Communication option.

28.2 bxCAN main features

• Supports CAN protocol version 2.0 A, B Active
• Bit rates up to 1 Mbit/s
• Supports the Time Triggered Communication option
Transmission
• Three transmit mailboxes
• Configurable transmit priority
• Time Stamp on SOF transmission
Reception
• Two receive FIFOs with three stages
• Scalable filter banks:
– 28 filter banks shared between CAN1 and CAN2 for dual CAN
• Identifier list feature
• Configurable FIFO overrun
• Time Stamp on SOF reception
Time-triggered communication option
• Disable automatic retransmission mode
• 16-bit free running timer
• Time Stamp sent in last two data bytes
Management
• Maskable interrupts
• Software-efficient mailbox mapping at a unique address space

Dual CAN peripheral configuration

• CAN1: Master bxCAN for managing the communication between a Slave bxCAN and
the 512-byte SRAM memory
• CAN2: Slave bxCAN, with no direct access to the SRAM memory.
• The two bxCAN cells share the 512-byte SRAM memory (see Figure 319: Dual-CAN
block diagram)

922/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

28.3 bxCAN general description

In today CAN applications, the number of nodes in a network is increasing and often several
networks are linked together via gateways. Typically the number of messages in the system
(to be handled by each node) has significantly increased. In addition to the application
messages, Network Management and Diagnostic messages have been introduced.
• An enhanced filtering mechanism is required to handle each type of message.
Furthermore, application tasks require more CPU time, therefore real-time constraints
caused by message reception have to be reduced.
• A receive FIFO scheme allows the CPU to be dedicated to application tasks for a long
time period without losing messages.
The standard HLP (Higher Layer Protocol) based on standard CAN drivers requires an
efficient interface to the CAN controller.

Figure 318. CAN network topology

CAN node 1

CAN node 2

CAN node n
MCU
Application

CAN
Controller

CAN CAN
Rx Tx

CAN
Transceiver

CAN CAN
High Low

CAN Bus

MS30392V1

28.3.1 CAN 2.0B active core

The bxCAN module handles the transmission and the reception of CAN messages fully
autonomously. Standard identifiers (11-bit) and extended identifiers (29-bit) are fully
supported by hardware.

28.3.2 Control, status and configuration registers

The application uses these registers to:
• Configure CAN parameters, e.g. baud rate
• Request transmissions
• Handle receptions
• Manage interrupts
• Get diagnostic information

28.3.3 Tx mailboxes
Three transmit mailboxes are provided to the software for setting up messages. The
transmission Scheduler decides which mailbox has to be transmitted first.

RM0402 Rev 6 923/1163

966
Controller area network (bxCAN) RM0402

28.3.4 Acceptance filters

The bxCAN provides up to 28 scalable/configurable identifier filter banks in dual CAN
configuration, for selecting the incoming messages, that the software needs and discarding
the others.

Receive FIFO
Two receive FIFOs are used by hardware to store the incoming messages. Three complete
messages can be stored in each FIFO. The FIFOs are managed completely by hardware.

Figure 319. Dual-CAN block diagram

CAN1 (Master) with 512 bytes SRAM

Master
Tx Mailboxes Master Master
Receive FIFO 0 Receive FIFO 1
2
1 2 2
Mailbox 0 1 1
Mailbox 0 Mailbox 0
Master Control
Master Status
Tx Status
Rx FIFO 0 Status Transmission
Scheduler
Rx FIFO 1 Status
Control/Status/Configuration

Acceptance Filters
Interrupt Enable
27
CAN 2.0B Active Core .. .. 26
Error Status Memory 2 3
1
Access Filter 0
Bit Timing Controller

Filter Master Master Filters Slave Filters

(0 to 27) (0 to 27)

Filter Mode
Transmission
Filter Scale Scheduler
Slave Slave
Slave Receive FIFO 0 Receive FIFO 1
Filter FIFOAssign Tx Mailboxes 2 2
Filter Activation 2 1 1
1 Mailbox 0 Mailbox 0
Mailbox 0

CAN2 (Slave)

Master Control
Control/Status/Configuration

Master Status
Tx Status
Rx FIFO 0 Status
CAN 2.0B Active Core
Rx FIFO 1 Status
Interrupt Enable Note: CAN2 start filter bank number n is configurable by writing
CAN2SB[5:0] bits in the CAN_FMR register
Error Status
Bit Timing

ai16094b

924/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

28.4 bxCAN operating modes

bxCAN has three main operating modes: initialization, normal and Sleep. After a
hardware reset, bxCAN is in Sleep mode to reduce power consumption and an internal pull-
up is active on CANTX. The software requests bxCAN to enter initialization or Sleep mode
by setting the INRQ or SLEEP bits in the CAN_MCR register. Once the mode has been
entered, bxCAN confirms it by setting the INAK or SLAK bits in the CAN_MSR register and
the internal pull-up is disabled. When neither INAK nor SLAK are set, bxCAN is in normal
mode. Before entering normal mode bxCAN always has to synchronize on the CAN bus.
To synchronize, bxCAN waits until the CAN bus is idle, this means 11 consecutive recessive
bits have been monitored on CANRX.

28.4.1 Initialization mode

The software initialization can be done while the hardware is in Initialization mode. To enter
this mode the software sets the INRQ bit in the CAN_MCR register and waits until the
hardware has confirmed the request by setting the INAK bit in the CAN_MSR register.
To leave Initialization mode, the software clears the INQR bit. bxCAN has left Initialization
mode once the INAK bit has been cleared by hardware.
While in Initialization Mode, all message transfers to and from the CAN bus are stopped and
the status of the CAN bus output CANTX is recessive (high).
Entering Initialization Mode does not change any of the configuration registers.
To initialize the CAN Controller, software has to set up the Bit Timing (CAN_BTR) and CAN
options (CAN_MCR) registers.
To initialize the registers associated with the CAN filter banks (mode, scale, FIFO
assignment, activation and filter values), software has to set the FINIT bit (CAN_FMR). Filter
initialization also can be done outside the initialization mode.
Note: When FINIT=1, CAN reception is deactivated.
The filter values also can be modified by deactivating the associated filter activation bits (in
the CAN_FA1R register).
If a filter bank is not used, it is recommended to leave it non active (leave the corresponding
FACT bit cleared).

28.4.2 Normal mode

Once the initialization is complete, the software must request the hardware to enter Normal
mode to be able to synchronize on the CAN bus and start reception and transmission.
The request to enter Normal mode is issued by clearing the INRQ bit in the CAN_MCR
register. The bxCAN enters Normal mode and is ready to take part in bus activities when it
has synchronized with the data transfer on the CAN bus. This is done by waiting for the
occurrence of a sequence of 11 consecutive recessive bits (Bus Idle state). The switch to
Normal mode is confirmed by the hardware by clearing the INAK bit in the CAN_MSR
register.
The initialization of the filter values is independent from Initialization Mode but must be done
while the filter is not active (corresponding FACTx bit cleared). The filter scale and mode
configuration must be configured before entering Normal Mode.

RM0402 Rev 6 925/1163

966
Controller area network (bxCAN) RM0402

28.4.3 Sleep mode (low-power)

To reduce power consumption, bxCAN has a low-power mode called Sleep mode. This
mode is entered on software request by setting the SLEEP bit in the CAN_MCR register. In
this mode, the bxCAN clock is stopped, however software can still access the bxCAN
mailboxes.
If software requests entry to initialization mode by setting the INRQ bit while bxCAN is in
Sleep mode, it must also clear the SLEEP bit.
bxCAN can be woken up (exit Sleep mode) either by software clearing the SLEEP bit or on
detection of CAN bus activity.
On CAN bus activity detection, hardware automatically performs the wakeup sequence by
clearing the SLEEP bit if the AWUM bit in the CAN_MCR register is set. If the AWUM bit is
cleared, software has to clear the SLEEP bit when a wakeup interrupt occurs, in order to exit
from Sleep mode.
Note: If the wakeup interrupt is enabled (WKUIE bit set in CAN_IER register) a wakeup interrupt is
generated on detection of CAN bus activity, even if the bxCAN automatically performs the
wakeup sequence.
After the SLEEP bit has been cleared, Sleep mode is exited once bxCAN has synchronized
with the CAN bus, refer to Figure 320: bxCAN operating modes. The Sleep mode is exited
once the SLAK bit has been cleared by hardware.

Figure 320. bxCAN operating modes

Reset

Sleep
SLAK = 1
INAK = 0
Q SL
R EE
.IN SL P.
IN
C
YN EE R
Q
.S K P. .A
EP . AC IN
R C
K
E
SL EP Q
.A
SLE C
K

Normal INRQ.ACK
Initialization
SLAK = 0 SLAK = 0
INAK = 0
INAK = 1
INRQ.SYNC.SLEEP
ai15902

1. ACK = The wait state during which hardware confirms a request by setting the INAK or SLAK bits in the
CAN_MSR register
2. SYNC = The state during which bxCAN waits until the CAN bus is idle, meaning 11 consecutive recessive
bits have been monitored on CANRX

926/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

28.5 Test mode

Test mode can be selected by the SILM and LBKM bits in the CAN_BTR register. These bits
must be configured while bxCAN is in Initialization mode. Once test mode has been
selected, the INRQ bit in the CAN_MCR register must be reset to enter Normal mode.

28.5.1 Silent mode

The bxCAN can be put in Silent mode by setting the SILM bit in the CAN_BTR register.
In Silent mode, the bxCAN is able to receive valid data frames and valid remote frames, but
it sends only recessive bits on the CAN bus and it cannot start a transmission. If the bxCAN
has to send a dominant bit (ACK bit, overload flag, active error flag), the bit is rerouted
internally so that the CAN Core monitors this dominant bit, although the CAN bus may
remain in recessive state. Silent mode can be used to analyze the traffic on a CAN bus
without affecting it by the transmission of dominant bits (Acknowledge Bits, Error Frames).

Figure 321. bxCAN in silent mode

bxCAN
Tx Rx

=1

CANTX CANRX

MS30393V2

28.5.2 Loop back mode

The bxCAN can be set in Loop Back Mode by setting the LBKM bit in the CAN_BTR
register. In Loop Back Mode, the bxCAN treats its own transmitted messages as received
messages and stores them (if they pass acceptance filtering) in a Receive mailbox.

Figure 322. bxCAN in loop back mode

bxCAN
Tx Rx

CANTX CANRX
MS30394V2

RM0402 Rev 6 927/1163

966
Controller area network (bxCAN) RM0402

This mode is provided for self-test functions. To be independent of external events, the CAN
Core ignores acknowledge errors (no dominant bit sampled in the acknowledge slot of a
data / remote frame) in Loop Back Mode. In this mode, the bxCAN performs an internal
feedback from its Tx output to its Rx input. The actual value of the CANRX input pin is
disregarded by the bxCAN. The transmitted messages can be monitored on the CANTX pin.

28.5.3 Loop back combined with silent mode

It is also possible to combine Loop Back mode and Silent mode by setting the LBKM and
SILM bits in the CAN_BTR register. This mode can be used for a “Hot Selftest”, meaning the
bxCAN can be tested like in Loop Back mode but without affecting a running CAN system
connected to the CANTX and CANRX pins. In this mode, the CANRX pin is disconnected
from the bxCAN and the CANTX pin is held recessive.

Figure 323. bxCAN in combined mode

bxCAN
Tx Rx

=1

CANTX CANRX

MS30395V2

28.6 Behavior in debug mode

When the microcontroller enters the debug mode (Cortex®-M4 with FPU core halted), the
bxCAN continues to work normally or stops, depending on:
• the DBG_CAN1_STOP bit for CAN1 or the DBG_CAN2_STOP bit for CAN2 in the
DBG module for the dual mode.
• the DBF bit in CAN_MCR. For more details, refer to Section 28.9.2: CAN control and
status registers.

28.7 bxCAN functional description

28.7.1 Transmission handling

In order to transmit a message, the application must select one empty transmit mailbox, set
up the identifier, the data length code (DLC) and the data before requesting the transmission
by setting the corresponding TXRQ bit in the CAN_TIxR register. Once the mailbox has left
empty state, the software no longer has write access to the mailbox registers. Immediately
after the TXRQ bit has been set, the mailbox enters pending state and waits to become the
highest priority mailbox, see Transmit Priority. As soon as the mailbox has the highest
priority it is scheduled for transmission. The transmission of the message of the scheduled

928/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

mailbox starts (enter transmit state) when the CAN bus becomes idle. Once the mailbox
has been successfully transmitted, it becomes empty again. The hardware indicates a
successful transmission by setting the RQCP and TXOK bits in the CAN_TSR register.
If the transmission fails, the cause is indicated by the ALST bit in the CAN_TSR register in
case of an Arbitration Lost, and/or the TERR bit, in case of transmission error detection.

Transmit priority
By identifier
When more than one transmit mailbox is pending, the transmission order is given by the
identifier of the message stored in the mailbox. The message with the lowest identifier value
has the highest priority according to the arbitration of the CAN protocol. If the identifier
values are equal, the lower mailbox number is scheduled first.
By transmit request order
The transmit mailboxes can be configured as a transmit FIFO by setting the TXFP bit in the
CAN_MCR register. In this mode the priority order is given by the transmit request order.
This mode is very useful for segmented transmission.

Abort
A transmission request can be aborted by the user setting the ABRQ bit in the CAN_TSR
register. In pending or scheduled state, the mailbox is aborted immediately. An abort
request while the mailbox is in transmit state can have two results. If the mailbox is
transmitted successfully the mailbox becomes empty with the TXOK bit set in the
CAN_TSR register. If the transmission fails, the mailbox becomes scheduled, the
transmission is aborted and becomes empty with TXOK cleared. In all cases the mailbox
becomes empty again at least at the end of the current transmission.

Non automatic retransmission mode

This mode has been implemented in order to fulfill the requirement of the Time Triggered
Communication option of the CAN standard. To configure the hardware in this mode the
NART bit in the CAN_MCR register must be set.
In this mode, each transmission is started only once. If the first attempt fails, due to an
arbitration loss or an error, the hardware does not automatically restart the message
transmission.
At the end of the first transmission attempt, the hardware considers the request as
completed and sets the RQCP bit in the CAN_TSR register. The result of the transmission is
indicated in the CAN_TSR register by the TXOK, ALST and TERR bits.

RM0402 Rev 6 929/1163

966
Controller area network (bxCAN) RM0402

Figure 324. Transmit mailbox states

EMPTY
RQCP=X
TXOK=X
TXRQ=1
TME = 1

PENDING
RQCP=0 Mailbox has
TXOK=0 highest priority
ABRQ=1
TME = 0

Mailbox does not

have highest priority
EMPTY SCHEDULED
RQCP=1 ABRQ=1 RQCP=0
TXOK=0 TXOK=0
TME = 1 TME = 0
CAN Bus = IDLE

Transmit failed * NART TRANSMIT Transmit failed * NART

RQCP=0
TXOK=0
TME = 0

EMPTY
Transmit succeeded
RQCP=1
TXOK=1
TME = 1
MS30396V2

28.7.2 Time triggered communication mode

In this mode, the internal counter of the CAN hardware is activated and used to generate the
Time Stamp value stored in the CAN_RDTxR/CAN_TDTxR registers, respectively (for Rx
and Tx mailboxes). The internal counter is incremented each CAN bit time (refer to
Section 28.7.7: Bit timing). The internal counter is captured on the sample point of the Start
Of Frame bit in both reception and transmission.

28.7.3 Reception handling

For the reception of CAN messages, three mailboxes organized as a FIFO are provided. In
order to save CPU load, simplify the software and guarantee data consistency, the FIFO is
managed completely by hardware. The application accesses the messages stored in the
FIFO through the FIFO output mailbox.

Valid message
A received message is considered as valid when it has been received correctly according to
the CAN protocol (no error until the last but one bit of the EOF field) and It passed through
the identifier filtering successfully, see Section 28.7.4: Identifier filtering.

930/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Figure 325. Receive FIFO states

EMPTY
FMP=0x00 Valid Message
FOVR=0 Received

PENDING_1
Release FMP=0x01
Mailbox FOVR=0

Release Valid Message

Mailbox Received
RFOM=1

PENDING_2
FMP=0x10
FOVR=0

Release Valid Message

Mailbox Received
RFOM=1

PENDING_3
FMP=0x11 Valid Message
FOVR=0 Received

OVERRUN
Release FMP=0x11
Mailbox FOVR=1
RFOM=1

Valid Message
Received
MS30397V2

FIFO management
Starting from the empty state, the first valid message received is stored in the FIFO which
becomes pending_1. The hardware signals the event setting the FMP[1:0] bits in the
CAN_RFR register to the value 01b. The message is available in the FIFO output mailbox.
The software reads out the mailbox content and releases it by setting the RFOM bit in the
CAN_RFR register. The FIFO becomes empty again. If a new valid message has been
received in the meantime, the FIFO stays in pending_1 state and the new message is
available in the output mailbox.
If the application does not release the mailbox, the next valid message is stored in the FIFO
which enters pending_2 state (FMP[1:0] = 10b). The storage process is repeated for the
next valid message putting the FIFO into pending_3 state (FMP[1:0] = 11b). At this point,
the software must release the output mailbox by setting the RFOM bit, so that a mailbox is
free to store the next valid message. Otherwise the next valid message received causes a
loss of message.
Refer also to Section 28.7.5: Message storage

RM0402 Rev 6 931/1163

966
Controller area network (bxCAN) RM0402

Overrun
Once the FIFO is in pending_3 state (i.e. the three mailboxes are full) the next valid
message reception leads to an overrun and a message is lost. The hardware signals the
overrun condition by setting the FOVR bit in the CAN_RFR register. Which message is lost
depends on the configuration of the FIFO:
• If the FIFO lock function is disabled (RFLM bit in the CAN_MCR register cleared) the
last message stored in the FIFO is overwritten by the new incoming message. In this
case the latest messages are always available to the application.
• If the FIFO lock function is enabled (RFLM bit in the CAN_MCR register set) the most
recent message is discarded and the software has the three oldest messages in the
FIFO available.

Reception related interrupts

Once a message has been stored in the FIFO, the FMP[1:0] bits are updated and an
interrupt request is generated if the FMPIE bit in the CAN_IER register is set.
When the FIFO becomes full (i.e. a third message is stored) the FULL bit in the CAN_RFR
register is set and an interrupt is generated if the FFIE bit in the CAN_IER register is set.
On overrun condition, the FOVR bit is set and an interrupt is generated if the FOVIE bit in
the CAN_IER register is set.

28.7.4 Identifier filtering

In the CAN protocol the identifier of a message is not associated with the address of a node
but related to the content of the message. Consequently a transmitter broadcasts its
message to all receivers. On message reception a receiver node decides - depending on
the identifier value - whether the software needs the message or not. If the message is
needed, it is copied into the SRAM. If not, the message must be discarded without
intervention by the software.
To fulfill this requirement the bxCAN Controller provides 28 configurable and scalable filter
banks (27-0) to the application, in order to receive only the messages the software needs.
This hardware filtering saves CPU resources which would be otherwise needed to perform
filtering by software. Each filter bank x consists of two 32-bit registers, CAN_FxR0 and
CAN_FxR1.

Scalable width
To optimize and adapt the filters to the application needs, each filter bank can be scaled
independently. Depending on the filter scale a filter bank provides:
• One 32-bit filter for the STDID[10:0], EXTID[17:0], IDE and RTR bits.
• Two 16-bit filters for the STDID[10:0], RTR, IDE and EXTID[17:15] bits.
Refer to Figure 326.
Furthermore, the filters can be configured in mask mode or in identifier list mode.

Mask mode
In mask mode the identifier registers are associated with mask registers specifying which
bits of the identifier are handled as “must match” or as “don’t care”.

932/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Identifier list mode

In identifier list mode, the mask registers are used as identifier registers. Thus instead of
defining an identifier and a mask, two identifiers are specified, doubling the number of single
identifiers. All bits of the incoming identifier must match the bits specified in the filter
registers.

Filter bank scale and mode configuration

The filter banks are configured by means of the corresponding CAN_FMR register. To
configure a filter bank it must be deactivated by clearing the FACT bit in the CAN_FAR
register. The filter scale is configured by means of the corresponding FSCx bit in the
CAN_FS1R register, refer to Figure 326. The identifier list or identifier mask mode for the
corresponding Mask/Identifier registers is configured by means of the FBMx bits in the
CAN_FMR register.
To filter a group of identifiers, configure the Mask/Identifier registers in mask mode.
To select single identifiers, configure the Mask/Identifier registers in identifier list mode.
Filters not used by the application should be left deactivated.
Each filter within a filter bank is numbered (called the Filter Number) from 0 to a maximum
dependent on the mode and the scale of each of the filter banks.
Concerning the filter configuration, refer to Figure 326.

RM0402 Rev 6 933/1163

966
Controller area network (bxCAN) RM0402

Figure 326. Filter bank scale configuration - register organization

Filter
One 32-Bit Filter - Identifier Mask Num.

ID CAN_FxR1[31:24] CAN_FxR1[23:16] CAN_FxR1[15:8] CAN_FxR1[7:0]

n
Mask CAN_FxR2[31:24] CAN_FxR2[23:16] CAN_FxR2[15:8] CAN_FxR2[7:0]
FBMx = 0

Mapping STD ID STID[10:3] STID[2:0]

Mapping Ext ID EXTID[28:21] EXID[20:13] EXID[12:5] EXID[4:0] IDE RTR 0
FSCx = 1

Two 32-Bit Filters - Identifier List

ID CAN_FxR1[31:24] CAN_FxR1[23:16] CAN_FxR1[15:8] CAN_FxR1[7:0] n

FBMx = 1

ID CAN_FxR2[31:24] CAN_FxR2[23:16] CAN_FxR2[15:8] CAN_FxR2[7:0] n+1

Mapping STD ID STID[10:3] STID[2:0]
Mapping Ext ID EXTID[28:21] EXID[20:13] EXID[12:5] EXID[4:0] IDE RTR 0

Two 16-Bit Filters - Identifier Mask

ID CAN_FxR1[15:8] CAN_FxR1[7:0]
n
FBMx = 0

Mask CAN_FxR1[31:24] CAN_FxR1[23:16]

ID CAN_FxR2[15:8] CAN_FxR2[7:0]
n+1
Mask CAN_FxR2[31:24] CAN_FxR2[23:16]

Mapping STID[10:3] STID[2:0] RTR IDE EXID[17:15]

FSCx = 0

Four 16-Bit Filters - Identifier List

ID CAN_FxR1[15:8] CAN_FxR1[7:0] n
FBMx = 1

ID CAN_FxR1[31:24] CAN_FxR1[23:16] n+1

ID CAN_FxR2[15:8] CAN_FxR2[7:0] n+2

ID CAN_FxR2[31:24] CAN_FxR2[23:16] n+3
Mapping STID[10:3] STID[2:0] RTR IDE EXID[17:15]
2

x = filter bank number

1
Filter Bank Scale

ID=Identifier
Filter Bank Mode

1 These bits are located in the CAN_FS1R register

Config. Bits

2 These bits are located in the CAN_FM1R register

MSv30398V4

Filter match index

Once a message has been received in the FIFO it is available to the application. Typically,
application data is copied into SRAM locations. To copy the data to the right location the
application has to identify the data by means of the identifier. To avoid this, and to ease the
access to the SRAM locations, the CAN controller provides a Filter Match Index.
This index is stored in the mailbox together with the message according to the filter priority
rules. Thus each received message has its associated filter match index.
The Filter Match index can be used in two ways:
• Compare the Filter Match index with a list of expected values.
• Use the Filter Match Index as an index on an array to access the data destination
location.
For non masked filters, the software no longer has to compare the identifier.
If the filter is masked the software reduces the comparison to the masked bits only.

934/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

The index value of the filter number does not take into account the activation state of the
filter banks. In addition, two independent numbering schemes are used, one for each FIFO.
Refer to Figure 327 for an example.

Figure 327. Example of filter numbering

Filter FIFO0 Filter Filter FIFO1 Filter

Bank Num. Bank Num.
0 0
0 ID List (32-bit) 2 ID Mask (16-bit)
1 1

2
1 ID Mask (32-bit) 2 4 ID List (32-bit) 3

3
4 Deactivated 4
3 ID List (16-bit) 5 7
ID Mask (16-bit) 5
6
Deactivated 7 6
5 8 ID Mask (16-bit)
ID List (32-bit) 8 7
8
9 Deactivated 9
6 ID Mask (16-bit) 10 10
10 ID List (16-bit)
11
11 12
9 ID List (32-bit) 11 ID List (32-bit)
12 13

13 ID Mask (32-bit) 13 12 ID Mask (32-bit) 14

ID=Identifier
MS30399V2

Filter priority rules

Depending on the filter combination it may occur that an identifier passes successfully
through several filters. In this case the filter match value stored in the receive mailbox is
chosen according to the following priority rules:
• A 32-bit filter takes priority over a 16-bit filter.
• For filters of equal scale, priority is given to the Identifier List mode over the Identifier
Mask mode
• For filters of equal scale and mode, priority is given by the filter number (the lower the
number, the higher the priority).

RM0402 Rev 6 935/1163

966
Controller area network (bxCAN) RM0402

Figure 328. Filtering mechanism - example

Example of 3 filter banks in 32-bit unidentified mode and

the remaining in 32-bit identifier mask mode
Message Received
Identifier Ctrl Data

Filter bank
Num Receive FIFO
Identifier 0
0
Identifier 1 Message
Identifier List
Identifier 4 Identifier #4 Match Stored
2

Identifier 5
Identifier & Mask

1
Identifier
Mask 2 Filter number stored in the
FMI
Filter Match Index field
within the CAN_RDTxR
Identifier register
4 3
Mask
No Match
Found

Message Discarded
MS31000V2

The example above shows the filtering principle of the bxCAN. On reception of a message,
the identifier is compared first with the filters configured in identifier list mode. If there is a
match, the message is stored in the associated FIFO and the index of the matching filter is
stored in the Filter Match Index. As shown in the example, the identifier matches with
Identifier #2 thus the message content and FMI 2 is stored in the FIFO.
If there is no match, the incoming identifier is then compared with the filters configured in
mask mode.
If the identifier does not match any of the identifiers configured in the filters, the message is
discarded by hardware without disturbing the software.

28.7.5 Message storage

The interface between the software and the hardware for the CAN messages is
implemented by means of mailboxes. A mailbox contains all information related to a
message; identifier, data, control, status and time stamp information.

Transmit mailbox
The software sets up the message to be transmitted in an empty transmit mailbox. The
status of the transmission is indicated by hardware in the CAN_TSR register.

936/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Table 193. Transmit mailbox mapping

Offset to transmit mailbox base address Register name

0 CAN_TIxR
4 CAN_TDTxR
8 CAN_TDLxR
12 CAN_TDHxR

Receive mailbox
When a message has been received, it is available to the software in the FIFO output
mailbox. Once the software has handled the message (e.g. read it) the software must
release the FIFO output mailbox by means of the RFOM bit in the CAN_RFR register to
make the next incoming message available. The filter match index is stored in the MFMI
field of the CAN_RDTxR register. The 16-bit time stamp value is stored in the TIME[15:0]
field of CAN_RDTxR.

Table 194. Receive mailbox mapping

Offset to receive mailbox base
Register name
address (bytes)

0 CAN_RIxR
4 CAN_RDTxR
8 CAN_RDLxR
12 CAN_RDHxR

Figure 329. CAN error state diagram

When TEC or REC > 127

ERROR ACTIVE ERROR PASSIVE

When TEC and REC < 128

When 128*11 recessive bits occur: When TEC > 255

BUS OFF

ai15903

RM0402 Rev 6 937/1163

966
Controller area network (bxCAN) RM0402

28.7.6 Error management

The error management as described in the CAN protocol is handled entirely by hardware
using a Transmit Error Counter (TEC value, in CAN_ESR register) and a Receive Error
Counter (REC value, in the CAN_ESR register), which get incremented or decremented
according to the error condition. For detailed information about TEC and REC management,
refer to the CAN standard.
Both of them may be read by software to determine the stability of the network.
Furthermore, the CAN hardware provides detailed information on the current error status in
CAN_ESR register. By means of the CAN_IER register (ERRIE bit, etc.), the software can
configure the interrupt generation on error detection in a very flexible way.

Bus-Off recovery
The Bus-Off state is reached when TEC is greater than 255, this state is indicated by BOFF
bit in CAN_ESR register. In Bus-Off state, the bxCAN is no longer able to transmit and
receive messages.
Depending on the ABOM bit in the CAN_MCR register, bxCAN recovers from Bus-Off
(become error active again) either automatically or on software request. But in both cases
the bxCAN has to wait at least for the recovery sequence specified in the CAN standard
(128 occurrences of 11 consecutive recessive bits monitored on CANRX).
If ABOM is set, the bxCAN starts the recovering sequence automatically after it has entered
Bus-Off state.
If ABOM is cleared, the software must initiate the recovering sequence by requesting
bxCAN to enter and to leave initialization mode.
Note: In initialization mode, bxCAN does not monitor the CANRX signal, therefore it cannot
complete the recovery sequence. To recover, bxCAN must be in normal mode.

28.7.7 Bit timing

The bit timing logic monitors the serial bus-line and performs sampling and adjustment of
the sample point by synchronizing on the start-bit edge and resynchronizing on the following
edges.
Its operation may be explained simply by splitting nominal bit time into three segments as
follows:
• Synchronization segment (SYNC_SEG): a bit change is expected to occur within this
time segment. It has a fixed length of one time quantum (1 x tq).
• Bit segment 1 (BS1): defines the location of the sample point. It includes the
PROP_SEG and PHASE_SEG1 of the CAN standard. Its duration is programmable
between 1 and 16 time quanta but may be automatically lengthened to compensate for
positive phase drifts due to differences in the frequency of the various nodes of the
network.
• Bit segment 2 (BS2): defines the location of the transmit point. It represents the
PHASE_SEG2 of the CAN standard. Its duration is programmable between 1 and 8
time quanta but may also be automatically shortened to compensate for negative
phase drifts.
The resynchronization Jump Width (SJW) defines an upper bound to the amount of
lengthening or shortening of the bit segments. It is programmable between 1 and 4 time
quanta.

938/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

A valid edge is defined as the first transition in a bit time from dominant to recessive bus
level provided the controller itself does not send a recessive bit.
If a valid edge is detected in BS1 instead of SYNC_SEG, BS1 is extended by up to SJW so
that the sample point is delayed.
Conversely, if a valid edge is detected in BS2 instead of SYNC_SEG, BS2 is shortened by
up to SJW so that the transmit point is moved earlier.
As a safeguard against programming errors, the configuration of the Bit Timing Register
(CAN_BTR) is only possible while the device is in Standby mode.
Note: For a detailed description of the CAN bit timing and resynchronization mechanism, refer to
the ISO 11898 standard.

Figure 330. Bit timing

NOMINAL BIT TIME

SYNC_SEG BIT SEGMENT 1 (BS1) BIT SEGMENT 2 (BS2)

1 x tq tBS1 tBS2

SAMPLE POINT TRANSMIT POINT

1
Baud Rate = ------------------
-----------------
----------
-
NominalBitTime
NominalBitTime = 1 X t q + t BS 1 + t BS 2
with:
tBS1 = tq x (TS1[3:0] + 1),
tBS2 = tq x (TS2[2:0] + 1),
tq = (BRP[9:0] + 1) x tPCLK
where tq refers to the Time quantum
tPCLK = time period of the APB clock,
BRP[9:0], TS1[3:0] and TS2[2:0] are defined in the CAN_BTR Register.

MS31001V2

RM0402 Rev 6 939/1163

966
Controller area network (bxCAN) RM0402

Figure 331. CAN frames

Inter-Frame Space
Inter-Frame Space Data Frame (Standard Identifier) or Overload Frame
44 + 8 *N
Arbitration Field Ctrl Field Data Field CRC Field ACK Field
2
32 6 8 *N 16 7

ID DLC CRC EOF

ACK
SOF

IDE
r0
RTR

Inter-Frame Space
Inter-Frame Space Data Frame (Extended Identifier) or Overload Frame
64 + 8 *N

Arbitration Field Arbitration Field Ctrl Field Data Field CRC Field ACK Field
2
32 32 8 *N 16 7

ID DLC CRC EOF

SOF

SRR
IDE

r1
r0

ACK
RTR

Inter-Frame Space
Inter-Frame Space Remote Frame
or Overload Frame
44
Arbitration Field Ctrl Field CRC Field ACK Field
2
32 6 16 7

ID DLC CRC EOF

SOF

IDE
r0

ACK
RTR

Data Frame or Inter-Frame Space

Error Frame
Remote Frame or Overload Frame
Error Flag Echo Error
Flag Delimiter
6 6 8

Notes:
0 <= N <= 8
Data Frame or SOF = Start Of Frame
Any Frame Inter-Frame Space Remote Frame ID = Identifier
Suspend RTR = Remote Transmission Request
Intermission Bus Idle IDE = Identifier Extension Bit
3 Transmission
8 r0 = Reserved Bit
DLC = Data Length Code
CRC = Cyclic Redundancy Code
End of Frame or Error flag: 6 dominant bits if node is error
Error Delimiter or Inter-Frame Space active, else 6 recessive bits.
Overload Delimiter Overload Frame or Error Frame Suspend transmission: applies to error
passive nodes only.
Overload Overload Overload EOF = End of Frame
Flag Echo Delimiter ACK = Acknowledge bit
Ctrl = Control
6 6 8
ai15154b

940/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

28.8 bxCAN interrupts

Four interrupt vectors are dedicated to bxCAN. Each interrupt source can be independently
enabled or disabled by means of the CAN Interrupt Enable Register (CAN_IER).

Figure 332. Event flags and interrupt generation

CAN_IER
TRANSMIT
INTERRUPT
RQCP0 TMEIE
CAN_TSR RQCP1 + &
RQCP2

FMPIE0
FMP0
& FIFO 0
INTERRUPT
FFIE0
CAN_RF0R FULL0
& +
FOVIE0
FOVR0
&

FMPIE1
FMP1
& FIFO 1
INTERRUPT
FFIE1
CAN_RF1R FULL1
& +
FOVIE1
FOVR1
&

ERRIE

EWGIE
EWGF &
EPVIE
CAN_ESR EPVF & &
BOFIE
+
ERRI
BOFF & CAN_MSR STATUS CHANGE
ERROR
LECIE
1 LEC 6 & INTERRUPT

WKUIE
WKUI
&
CAN_MSR
SLKIE
SLAKI
&

MS31002V2

RM0402 Rev 6 941/1163

966
Controller area network (bxCAN) RM0402

• The transmit interrupt can be generated by the following events:

– Transmit mailbox 0 becomes empty, RQCP0 bit in the CAN_TSR register set.
– Transmit mailbox 1 becomes empty, RQCP1 bit in the CAN_TSR register set.
– Transmit mailbox 2 becomes empty, RQCP2 bit in the CAN_TSR register set.
• The FIFO 0 interrupt can be generated by the following events:
– Reception of a new message, FMP0 bits in the CAN_RF0R register are not ‘00’.
– FIFO0 full condition, FULL0 bit in the CAN_RF0R register set.
– FIFO0 overrun condition, FOVR0 bit in the CAN_RF0R register set.
• The FIFO 1 interrupt can be generated by the following events:
– Reception of a new message, FMP1 bits in the CAN_RF1R register are not ‘00’.
– FIFO1 full condition, FULL1 bit in the CAN_RF1R register set.
– FIFO1 overrun condition, FOVR1 bit in the CAN_RF1R register set.
• The error and status change interrupt can be generated by the following events:
– Error condition, for more details on error conditions refer to the CAN Error Status
register (CAN_ESR).
– Wakeup condition, SOF monitored on the CAN Rx signal.
– Entry into Sleep mode.

28.9 CAN registers

The peripheral registers have to be accessed by words (32 bits).

28.9.1 Register access protection

Erroneous access to certain configuration registers can cause the hardware to temporarily
disturb the whole CAN network. Therefore the CAN_BTR register can be modified by
software only while the CAN hardware is in initialization mode.
Although the transmission of incorrect data does not cause problems at the CAN network
level, it can severely disturb the application. A transmit mailbox can be only modified by
software while it is in empty state, refer to Figure 324: Transmit mailbox states.
The filter values can be modified either deactivating the associated filter banks or by setting
the FINIT bit. Moreover, the modification of the filter configuration (scale, mode and FIFO
assignment) in CAN_FMxR, CAN_FSxR and CAN_FFAR registers can only be done when
the filter initialization mode is set (FINIT=1) in the CAN_FMR register.

28.9.2 CAN control and status registers

Refer to Section 1.2 for a list of abbreviations used in register descriptions.

CAN master control register (CAN_MCR)

Address offset: 0x00
Reset value: 0x0001 0002

942/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. DBF

rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

RESET Res. Res. Res. Res. Res. Res. Res. TTCM ABOM AWUM NART RFLM TXFP SLEEP INRQ

rs rw rw rw rw rw rw rw rw

Bits 31:17 Reserved, must be kept at reset value.

Bit 16 DBF: Debug freeze
0: CAN working during debug
1: CAN reception/transmission frozen during debug. Reception FIFOs can still be
accessed/controlled normally.
Bit 15 RESET: bxCAN software master reset
0: Normal operation.
1: Force a master reset of the bxCAN -> Sleep mode activated after reset (FMP bits and
CAN_MCR register are initialized to the reset values). This bit is automatically reset to 0.
Bits 14:8 Reserved, must be kept at reset value.
Bit 7 TTCM: Time triggered communication mode
0: Time Triggered Communication mode disabled.
1: Time Triggered Communication mode enabled
Note: For more information on Time Triggered Communication mode, refer to Section 28.7.2:
Time triggered communication mode.
Bit 6 ABOM: Automatic bus-off management
This bit controls the behavior of the CAN hardware on leaving the Bus-Off state.
0: The Bus-Off state is left on software request, once 128 occurrences of 11 recessive bits
have been monitored and the software has first set and cleared the INRQ bit of the
CAN_MCR register.
1: The Bus-Off state is left automatically by hardware once 128 occurrences of 11 recessive
bits have been monitored.
For detailed information on the Bus-Off state refer to Section 28.7.6: Error management.
Bit 5 AWUM: Automatic wakeup mode
This bit controls the behavior of the CAN hardware on message reception during Sleep
mode.
0: The Sleep mode is left on software request by clearing the SLEEP bit of the CAN_MCR
register.
1: The Sleep mode is left automatically by hardware on CAN message detection.
The SLEEP bit of the CAN_MCR register and the SLAK bit of the CAN_MSR register are
cleared by hardware.
Bit 4 NART: No automatic retransmission
0: The CAN hardware automatically retransmits the message until it has been successfully
transmitted according to the CAN standard.
1: A message is transmitted only once, independently of the transmission result (successful,
error or arbitration lost).

RM0402 Rev 6 943/1163

966
Controller area network (bxCAN) RM0402

Bit 3 RFLM: Receive FIFO locked mode

0: Receive FIFO not locked on overrun. Once a receive FIFO is full the next incoming
message overwrites the previous one.
1: Receive FIFO locked against overrun. Once a receive FIFO is full the next incoming
message is discarded.
Bit 2 TXFP: Transmit FIFO priority
This bit controls the transmission order when several mailboxes are pending at the same
time.
0: Priority driven by the identifier of the message
1: Priority driven by the request order (chronologically)
Bit 1 SLEEP: Sleep mode request
This bit is set by software to request the CAN hardware to enter the Sleep mode. Sleep
mode is entered as soon as the current CAN activity (transmission or reception of a CAN
frame) has been completed.
This bit is cleared by software to exit Sleep mode.
This bit is cleared by hardware when the AWUM bit is set and a SOF bit is detected on the
CAN Rx signal.
This bit is set after reset - CAN starts in Sleep mode.
Bit 0 INRQ: Initialization request
The software clears this bit to switch the hardware into normal mode. Once 11 consecutive
recessive bits have been monitored on the Rx signal the CAN hardware is synchronized and
ready for transmission and reception. Hardware signals this event by clearing the INAK bit in
the CAN_MSR register.
Software sets this bit to request the CAN hardware to enter initialization mode. Once
software has set the INRQ bit, the CAN hardware waits until the current CAN activity
(transmission or reception) is completed before entering the initialization mode. Hardware
signals this event by setting the INAK bit in the CAN_MSR register.

CAN master status register (CAN_MSR)

Address offset: 0x04
Reset value: 0x0000 0C02

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. RX SAMP RXM TXM Res. Res. Res. SLAKI WKUI ERRI SLAK INAK

r r r r rc_w1 rc_w1 rc_w1 r r

Bits 31:12 Reserved, must be kept at reset value.

Bit 11 RX: CAN Rx signal
Monitors the actual value of the CAN_RX Pin.
Bit 10 SAMP: Last sample point
The value of RX on the last sample point (current received bit value).
Bit 9 RXM: Receive mode
The CAN hardware is currently receiver.

944/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Bit 8 TXM: Transmit mode

The CAN hardware is currently transmitter.
Bits 7:5 Reserved, must be kept at reset value.
Bit 4 SLAKI: Sleep acknowledge interrupt
When SLKIE=1, this bit is set by hardware to signal that the bxCAN has entered Sleep
Mode. When set, this bit generates a status change interrupt if the SLKIE bit in the
CAN_IER register is set.
This bit is cleared by software or by hardware, when SLAK is cleared.
Note: When SLKIE=0, no polling on SLAKI is possible. In this case the SLAK bit can be
polled.
Bit 3 WKUI: Wakeup interrupt
This bit is set by hardware to signal that a SOF bit has been detected while the CAN
hardware was in Sleep mode. Setting this bit generates a status change interrupt if the
WKUIE bit in the CAN_IER register is set.
This bit is cleared by software.
Bit 2 ERRI: Error interrupt
This bit is set by hardware when a bit of the CAN_ESR has been set on error detection and
the corresponding interrupt in the CAN_IER is enabled. Setting this bit generates a status
change interrupt if the ERRIE bit in the CAN_IER register is set.
This bit is cleared by software.
Bit 1 SLAK: Sleep acknowledge
This bit is set by hardware and indicates to the software that the CAN hardware is now in
Sleep mode. This bit acknowledges the Sleep mode request from the software (set SLEEP
bit in CAN_MCR register).
This bit is cleared by hardware when the CAN hardware has left Sleep mode (to be
synchronized on the CAN bus). To be synchronized the hardware has to monitor a
sequence of 11 consecutive recessive bits on the CAN RX signal.
Note: The process of leaving Sleep mode is triggered when the SLEEP bit in the CAN_MCR
register is cleared. Refer to the AWUM bit of the CAN_MCR register description for
detailed information for clearing SLEEP bit
Bit 0 INAK: Initialization acknowledge
This bit is set by hardware and indicates to the software that the CAN hardware is now in
initialization mode. This bit acknowledges the initialization request from the software (set
INRQ bit in CAN_MCR register).
This bit is cleared by hardware when the CAN hardware has left the initialization mode (to
be synchronized on the CAN bus). To be synchronized the hardware has to monitor a
sequence of 11 consecutive recessive bits on the CAN RX signal.

CAN transmit status register (CAN_TSR)

Address offset: 0x08
Reset value: 0x1C00 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

LOW2 LOW1 LOW0 TME2 TME1 TME0 CODE[1:0] ABRQ2 Res. Res. Res. TERR2 ALST2 TXOK2 RQCP2

r r r r r r r r rs rc_w1 rc_w1 rc_w1 rc_w1

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

ABRQ1 Res. Res. Res. TERR1 ALST1 TXOK1 RQCP1 ABRQ0 Res. Res. Res. TERR0 ALST0 TXOK0 RQCP0

rs rc_w1 rc_w1 rc_w1 rc_w1 rs rc_w1 rc_w1 rc_w1 rc_w1

RM0402 Rev 6 945/1163

966
Controller area network (bxCAN) RM0402

Bit 31 LOW2: Lowest priority flag for mailbox 2

This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 2 has the lowest priority.
Bit 30 LOW1: Lowest priority flag for mailbox 1
This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 1 has the lowest priority.
Bit 29 LOW0: Lowest priority flag for mailbox 0
This bit is set by hardware when more than one mailbox are pending for transmission and
mailbox 0 has the lowest priority.
Note: The LOW[2:0] bits are set to zero when only one mailbox is pending.
Bit 28 TME2: Transmit mailbox 2 empty
This bit is set by hardware when no transmit request is pending for mailbox 2.
Bit 27 TME1: Transmit mailbox 1 empty
This bit is set by hardware when no transmit request is pending for mailbox 1.
Bit 26 TME0: Transmit mailbox 0 empty
This bit is set by hardware when no transmit request is pending for mailbox 0.
Bits 25:24 CODE[1:0]: Mailbox code
In case at least one transmit mailbox is free, the code value is equal to the number of the
next transmit mailbox free.
In case all transmit mailboxes are pending, the code value is equal to the number of the
transmit mailbox with the lowest priority.
Bit 23 ABRQ2: Abort request for mailbox 2
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bits 22:20 Reserved, must be kept at reset value.
Bit 19 TERR2: Transmission error of mailbox 2
This bit is set when the previous TX failed due to an error.
Bit 18 ALST2: Arbitration lost for mailbox 2
This bit is set when the previous TX failed due to an arbitration lost.
Bit 17 TXOK2: Transmission OK of mailbox 2
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 2 has been completed
successfully. Refer to Figure 324.
Bit 16 RQCP2: Request completed mailbox2
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ2 set in
CAN_TMID2R register).
Clearing this bit clears all the status bits (TXOK2, ALST2 and TERR2) for Mailbox 2.
Bit 15 ABRQ1: Abort request for mailbox 1
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bits 14:12 Reserved, must be kept at reset value.

946/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Bit 11 TERR1: Transmission error of mailbox1

This bit is set when the previous TX failed due to an error.
Bit 10 ALST1: Arbitration lost for mailbox1
This bit is set when the previous TX failed due to an arbitration lost.
Bit 9 TXOK1: Transmission OK of mailbox1
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 1 has been completed
successfully. Refer to Figure 324
Bit 8 RQCP1: Request completed mailbox1
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ1 set in
CAN_TI1R register).
Clearing this bit clears all the status bits (TXOK1, ALST1 and TERR1) for Mailbox 1.
Bit 7 ABRQ0: Abort request for mailbox0
Set by software to abort the transmission request for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.
Setting this bit has no effect when the mailbox is not pending for transmission.
Bits 6:4 Reserved, must be kept at reset value.
Bit 3 TERR0: Transmission error of mailbox0
This bit is set when the previous TX failed due to an error.
Bit 2 ALST0: Arbitration lost for mailbox0
This bit is set when the previous TX failed due to an arbitration lost.
Bit 1 TXOK0: Transmission OK of mailbox0
The hardware updates this bit after each transmission attempt.
0: The previous transmission failed
1: The previous transmission was successful
This bit is set by hardware when the transmission request on mailbox 1 has been completed
successfully. Refer to Figure 324
Bit 0 RQCP0: Request completed mailbox0
Set by hardware when the last request (transmit or abort) has been performed.
Cleared by software writing a “1” or by hardware on transmission request (TXRQ0 set in
CAN_TI0R register).
Clearing this bit clears all the status bits (TXOK0, ALST0 and TERR0) for Mailbox 0.

CAN receive FIFO 0 register (CAN_RF0R)

Address offset: 0x0C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. RFOM0 FOVR0 FULL0 Res. FMP0[1:0]

rs rc_w1 rc_w1 r r

RM0402 Rev 6 947/1163

966
Controller area network (bxCAN) RM0402

Bits 31:6 Reserved, must be kept at reset value.

Bit 5 RFOM0: Release FIFO 0 output mailbox
Set by software to release the output mailbox of the FIFO. The output mailbox can only be
released when at least one message is pending in the FIFO. Setting this bit when the FIFO
is empty has no effect. If at least two messages are pending in the FIFO, the software has to
release the output mailbox to access the next message.
Cleared by hardware when the output mailbox has been released.
Bit 4 FOVR0: FIFO 0 overrun
This bit is set by hardware when a new message has been received and passed the filter
while the FIFO was full.
This bit is cleared by software.
Bit 3 FULL0: FIFO 0 full
Set by hardware when three messages are stored in the FIFO.
This bit is cleared by software.
Bit 2 Reserved, must be kept at reset value.
Bits 1:0 FMP0[1:0]: FIFO 0 message pending
These bits indicate how many messages are pending in the receive FIFO.
FMP is increased each time the hardware stores a new message in to the FIFO. FMP is
decreased each time the software releases the output mailbox by setting the RFOM0 bit.

CAN receive FIFO 1 register (CAN_RF1R)

Address offset: 0x10
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. RFOM1 FOVR1 FULL1 Res. FMP1[1:0]

rs rc_w1 rc_w1 r r

Bits 31:6 Reserved, must be kept at reset value.

Bit 5 RFOM1: Release FIFO 1 output mailbox
Set by software to release the output mailbox of the FIFO. The output mailbox can only be
released when at least one message is pending in the FIFO. Setting this bit when the FIFO
is empty has no effect. If at least two messages are pending in the FIFO, the software has to
release the output mailbox to access the next message.
Cleared by hardware when the output mailbox has been released.
Bit 4 FOVR1: FIFO 1 overrun
This bit is set by hardware when a new message has been received and passed the filter
while the FIFO was full.
This bit is cleared by software.

948/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Bit 3 FULL1: FIFO 1 full

Set by hardware when three messages are stored in the FIFO.
This bit is cleared by software.
Bit 2 Reserved, must be kept at reset value.
Bits 1:0 FMP1[1:0]: FIFO 1 message pending
These bits indicate how many messages are pending in the receive FIFO1.
FMP1 is increased each time the hardware stores a new message in to the FIFO1. FMP is
decreased each time the software releases the output mailbox by setting the RFOM1 bit.

CAN interrupt enable register (CAN_IER)

Address offset: 0x14
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. SLKIE WKUIE

rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

LEC BOF EPV EWG FOV FF FMP FOV FF FMP TME

ERRIE Res. Res. Res. Res.
IE IE IE IE IE1 IE1 IE1 IE0 IE0 IE0 IE

rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:18 Reserved, must be kept at reset value.

Bit 17 SLKIE: Sleep interrupt enable
0: No interrupt when SLAKI bit is set.
1: Interrupt generated when SLAKI bit is set.
Bit 16 WKUIE: Wakeup interrupt enable
0: No interrupt when WKUI is set.
1: Interrupt generated when WKUI bit is set.
Bit 15 ERRIE: Error interrupt enable
0: No interrupt is generated when an error condition is pending in the CAN_ESR.
1: An interrupt is generation when an error condition is pending in the CAN_ESR.
Bits 14:12 Reserved, must be kept at reset value.
Bit 11 LECIE: Last error code interrupt enable
0: ERRI bit is not set when the error code in LEC[2:0] is set by hardware on error detection.
1: ERRI bit is set when the error code in LEC[2:0] is set by hardware on error detection.
Bit 10 BOFIE: Bus-off interrupt enable
0: ERRI bit is not set when BOFF is set.
1: ERRI bit is set when BOFF is set.
Bit 9 EPVIE: Error passive interrupt enable
0: ERRI bit is not set when EPVF is set.
1: ERRI bit is set when EPVF is set.
Bit 8 EWGIE: Error warning interrupt enable
0: ERRI bit is not set when EWGF is set.
1: ERRI bit is set when EWGF is set.

RM0402 Rev 6 949/1163

966
Controller area network (bxCAN) RM0402

Bit 7 Reserved, must be kept at reset value.

Bit 6 FOVIE1: FIFO overrun interrupt enable
0: No interrupt when FOVR is set.
1: Interrupt generation when FOVR is set.
Bit 5 FFIE1: FIFO full interrupt enable
0: No interrupt when FULL bit is set.
1: Interrupt generated when FULL bit is set.
Bit 4 FMPIE1: FIFO message pending interrupt enable
0: No interrupt generated when state of FMP[1:0] bits are not 00b.
1: Interrupt generated when state of FMP[1:0] bits are not 00b.
Bit 3 FOVIE0: FIFO overrun interrupt enable
0: No interrupt when FOVR bit is set.
1: Interrupt generated when FOVR bit is set.
Bit 2 FFIE0: FIFO full interrupt enable
0: No interrupt when FULL bit is set.
1: Interrupt generated when FULL bit is set.
Bit 1 FMPIE0: FIFO message pending interrupt enable
0: No interrupt generated when state of FMP[1:0] bits are not 00b.
1: Interrupt generated when state of FMP[1:0] bits are not 00b.
Bit 0 TMEIE: Transmit mailbox empty interrupt enable
0: No interrupt when RQCPx bit is set.
1: Interrupt generated when RQCPx bit is set.
Note: Refer to Section 28.8: bxCAN interrupts.

CAN error status register (CAN_ESR)

Address offset: 0x18
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

REC[7:0] TEC[7:0]

r r r r r r r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. LEC[2:0] Res. BOFF EPVF EWGF

rw rw rw r r r

Bits 31:24 REC[7:0]: Receive error counter

The implementing part of the fault confinement mechanism of the CAN protocol. In case of
an error during reception, this counter is incremented by 1 or by 8 depending on the error
condition as defined by the CAN standard. After every successful reception the counter is
decremented by 1 or reset to 120 if its value was higher than 128. When the counter value
exceeds 127, the CAN controller enters the error passive state.
Bits 23:16 TEC[7:0]: Least significant byte of the 9-bit transmit error counter
The implementing part of the fault confinement mechanism of the CAN protocol.
Bits 15:7 Reserved, must be kept at reset value.

950/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Bits 6:4 LEC[2:0]: Last error code

This field is set by hardware and holds a code which indicates the error condition of the last
error detected on the CAN bus. If a message has been transferred (reception or
transmission) without error, this field is cleared to 0.
The LEC[2:0] bits can be set to value 0b111 by software. They are updated by hardware to
indicate the current communication status.
000: No Error
001: Stuff Error
010: Form Error
011: Acknowledgment Error
100: Bit recessive Error
101: Bit dominant Error
110: CRC Error
111: Set by software
Bit 3 Reserved, must be kept at reset value.
Bit 2 BOFF: Bus-off flag
This bit is set by hardware when it enters the bus-off state. The bus-off state is entered on
TEC overflow, greater than 255, refer to Section 28.7.6 on page 938.
Bit 1 EPVF: Error passive flag
This bit is set by hardware when the Error Passive limit has been reached (Receive Error
Counter or Transmit Error Counter>127).
Bit 0 EWGF: Error warning flag
This bit is set by hardware when the warning limit has been reached
(Receive Error Counter or Transmit Error Counter≥96).

CAN bit timing register (CAN_BTR)

Address offset: 0x1C
Reset value: 0x0123 0000
This register can only be accessed by the software when the CAN hardware is in
initialization mode.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

SILM LBKM Res. Res. Res. Res. SJW[1:0] Res. TS2[2:0] TS1[3:0]

rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. BRP[9:0]

rw rw rw rw rw rw rw rw rw rw

Bit 31 SILM: Silent mode (debug)

0: Normal operation
1: Silent Mode
Bit 30 LBKM: Loop back mode (debug)
0: Loop Back Mode disabled
1: Loop Back Mode enabled
Bits 29:26 Reserved, must be kept at reset value.

RM0402 Rev 6 951/1163

966
Controller area network (bxCAN) RM0402

Bits 25:24 SJW[1:0]: Resynchronization jump width

These bits define the maximum number of time quanta the CAN hardware is allowed to
lengthen or shorten a bit to perform the resynchronization.
tRJW = tq x (SJW[1:0] + 1)
Bit 23 Reserved, must be kept at reset value.
Bits 22:20 TS2[2:0]: Time segment 2
These bits define the number of time quanta in Time Segment 2.
tBS2 = tq x (TS2[2:0] + 1)
Bits 19:16 TS1[3:0]: Time segment 1
These bits define the number of time quanta in Time Segment 1
tBS1 = tq x (TS1[3:0] + 1)
For more information on bit timing, refer to Section 28.7.7: Bit timing on page 938.
Bits 15:10 Reserved, must be kept at reset value.
Bits 9:0 BRP[9:0]: Baud rate prescaler
These bits define the length of a time quanta.
tq = (BRP[9:0]+1) x tPCLK

28.9.3 CAN mailbox registers

This chapter describes the registers of the transmit and receive mailboxes. Refer to
Section 28.7.5: Message storage on page 936 for detailed register mapping.
Transmit and receive mailboxes have the same registers except:
• The FMI field in the CAN_RDTxR register.
• A receive mailbox is always write protected.
• A transmit mailbox is write-enabled only while empty, corresponding TME bit in the
CAN_TSR register set.
There are 3 TX Mailboxes and 2 RX Mailboxes. Each RX Mailbox allows access to a 3 level
depth FIFO, the access being offered only to the oldest received message in the FIFO.
Each mailbox consist of 4 registers.

Figure 333. CAN mailbox registers

CAN_RI0R CAN_RI1R CAN_TI0R CAN_TI1R CAN_TI2R

CAN_RDT0R CAN_RDT1R CAN_TDT0R CAN_TDT1R CAN_TDT2R
CAN_RL0R CAN_RL1R CAN_TDL0R CAN_TDL1R CAN_TDL2R
CAN_RH0R CAN_RH1R CAN_TDH0R CAN_TDH1R CAN_TDH2R

FIFO0 FIFO1 Three Tx Mailboxes

MS31003V2

CAN TX mailbox identifier register (CAN_TIxR) (x = 0..2)

Address offsets: 0x180, 0x190, 0x1A0
Reset value: 0xXXXX XXXX (except bit 0, TXRQ = 0)
All TX registers are write protected when the mailbox is pending transmission (TMEx reset).

952/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

This register also implements the TX request control (bit 0) - reset value 0.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

STID[10:0]/EXID[28:18] EXID[17:13]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

EXID[12:0] IDE RTR TXRQ

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:21 STID[10:0]/EXID[28:18]: Standard identifier or extended identifier

The standard identifier or the MSBs of the extended identifier (depending on the IDE bit
value).
Bit 20:3 EXID[17:0]: Extended identifier
The LSBs of the extended identifier.
Bit 2 IDE: Identifier extension
This bit defines the identifier type of message in the mailbox.
0: Standard identifier.
1: Extended identifier.
Bit 1 RTR: Remote transmission request
0: Data frame
1: Remote frame
Bit 0 TXRQ: Transmit mailbox request
Set by software to request the transmission for the corresponding mailbox.
Cleared by hardware when the mailbox becomes empty.

RM0402 Rev 6 953/1163

966
Controller area network (bxCAN) RM0402

CAN mailbox data length control and time stamp register

(CAN_TDTxR) (x = 0..2)
All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x184, 0x194, 0x1A4
Reset value: 0xXXXX XXXX

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

TIME[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. DLC[3:0]

rw rw rw rw

Bits 31:16 TIME[15:0]: Message time stamp

This field contains the 16-bit timer value captured at the SOF transmission.
Bits 15:9 Reserved, must be kept at reset value.
Bit 8 TGT: Transmit global time
This bit is active only when the hardware is in the Time Trigger Communication mode,
TTCM bit of the CAN_MCR register is set.
0: Time stamp TIME[15:0] is not sent.
1: Time stamp TIME[15:0] value is sent in the last two data bytes of the 8-byte message:
TIME[7:0] in data byte 7 and TIME[15:8] in data byte 6, replacing the data written in
CAN_TDHxR[31:16] register (DATA6[7:0] and DATA7[7:0]). DLC must be programmed as 8
in order these two bytes to be sent over the CAN bus.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 DLC[3:0]: Data length code
This field defines the number of data bytes a data frame contains or a remote frame request.
A message can contain from 0 to 8 data bytes, depending on the value in the DLC field.

954/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

CAN mailbox data low register (CAN_TDLxR) (x = 0..2)

All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x188, 0x198, 0x1A8
Reset value: 0xXXXX XXXX

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DATA3[7:0] DATA2[7:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DATA1[7:0] DATA0[7:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:24 DATA3[7:0]: Data byte 3

Data byte 3 of the message.
Bits 23:16 DATA2[7:0]: Data byte 2
Data byte 2 of the message.
Bits 15:8 DATA1[7:0]: Data byte 1
Data byte 1 of the message.
Bits 7:0 DATA0[7:0]: Data byte 0
Data byte 0 of the message.
A message can contain from 0 to 8 data bytes and starts with byte 0.

CAN mailbox data high register (CAN_TDHxR) (x = 0..2)

All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x18C, 0x19C, 0x1AC
Reset value: 0xXXXX XXXX

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DATA7[7:0] DATA6[7:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DATA5[7:0] DATA4[7:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 955/1163

966
Controller area network (bxCAN) RM0402

Bits 31:24 DATA7[7:0]: Data byte 7

Data byte 7 of the message.
Note: If TGT of this message and TTCM are active, DATA7 and DATA6 are replaced by the
TIME stamp value.
Bits 23:16 DATA6[7:0]: Data byte 6
Data byte 6 of the message.
Bits 15:8 DATA5[7:0]: Data byte 5
Data byte 5 of the message.
Bits 7:0 DATA4[7:0]: Data byte 4
Data byte 4 of the message.

CAN receive FIFO mailbox identifier register (CAN_RIxR) (x = 0..1)

Address offsets: 0x1B0, 0x1C0
Reset value: 0xXXXX XXXX
All RX registers are write protected.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

STID[10:0]/EXID[28:18] EXID[17:13]

r r r r r r r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

EXID[12:0] IDE RTR Res

r r r r r r r r r r r r r r r

Bits 31:21 STID[10:0]/EXID[28:18]: Standard identifier or extended identifier

The standard identifier or the MSBs of the extended identifier (depending on the IDE bit
value).
Bits 20:3 EXID[17:0]: Extended identifier
The LSBs of the extended identifier.
Bit 2 IDE: Identifier extension
This bit defines the identifier type of message in the mailbox.
0: Standard identifier.
1: Extended identifier.
Bit 1 RTR: Remote transmission request
0: Data frame
1: Remote frame
Bit 0 Reserved, must be kept at reset value.

956/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

CAN receive FIFO mailbox data length control and time stamp register
(CAN_RDTxR) (x = 0..1)
Address offsets: 0x1B4, 0x1C4
Reset value: 0xXXXX XXXX
All RX registers are write protected.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

TIME[15:0]

r r r r r r r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

FMI[7:0] Res. Res. Res. Res. DLC[3:0]

r r r r r r r r r r r r

Bits 31:16 TIME[15:0]: Message time stamp

This field contains the 16-bit timer value captured at the SOF detection.
Bits 15:8 FMI[7:0]: Filter match index
This register contains the index of the filter the message stored in the mailbox passed
through. For more details on identifier filtering refer to Section 28.7.4: Identifier filtering on
page 932 - Filter Match Index paragraph.
Bits 7:4 Reserved, must be kept at reset value.
Bits 3:0 DLC[3:0]: Data length code
This field defines the number of data bytes a data frame contains (0 to 8). It is 0 in the case
of a remote frame request.

RM0402 Rev 6 957/1163

966
Controller area network (bxCAN) RM0402

CAN receive FIFO mailbox data low register (CAN_RDLxR) (x = 0..1)

All bits of this register are write protected when the mailbox is not in empty state.
Address offsets: 0x1B8, 0x1C8
Reset value: 0xXXXX XXXX
All RX registers are write protected.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DATA3[7:0] DATA2[7:0]
r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DATA1[7:0] DATA0[7:0]
r r r r r r r r r r r r r r r r

Bits 31:24 DATA3[7:0]: Data Byte 3

Data byte 3 of the message.
Bits 23:16 DATA2[7:0]: Data Byte 2
Data byte 2 of the message.
Bits 15:8 DATA1[7:0]: Data Byte 1
Data byte 1 of the message.
Bits 7:0 DATA0[7:0]: Data Byte 0
Data byte 0 of the message.
A message can contain from 0 to 8 data bytes and starts with byte 0.

CAN receive FIFO mailbox data high register (CAN_RDHxR) (x = 0..1)

Address offsets: 0x1BC, 0x1CC
Reset value: 0xXXXX XXXX
All RX registers are write protected.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

DATA7[7:0] DATA6[7:0]

r r r r r r r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DATA5[7:0] DATA4[7:0]

r r r r r r r r r r r r r r r r

Bits 31:24 DATA7[7:0]: Data Byte 7

Data byte 3 of the message.

958/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Bits 23:16 DATA6[7:0]: Data Byte 6

Data byte 2 of the message.
Bits 15:8 DATA5[7:0]: Data Byte 5
Data byte 1 of the message.
Bits 7:0 DATA4[7:0]: Data Byte 4
Data byte 0 of the message.

28.9.4 CAN filter registers

CAN filter master register (CAN_FMR)
Address offset: 0x200
Reset value: 0x2A1C 0E01
All bits of this register are set and cleared by software.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Res. Res. CANSB[5:0] Res. Res. Res. Res. Res. Res. Res. FINIT
rw rw rw rw rw rw rw

Bits 31:14 Reserved, must be kept at reset value.

Bits 13:8 CANSB[5:0]: CAN start bank
These bits are set and cleared by software. When both CAN are used, they define the start
bank of each CAN interface:
000001 = 1 filter assigned to CAN1 and 27 assigned to CAN2
011011 = 27 filters assigned to CAN1 and 1 filter assigned to CAN2
– to assign all filters to one CAN set CANSB value to zero and deactivate the non
used CAN
– to use CAN1 only: stop the clock on CAN2 and/or set the CAN_MCR.INRQ on
CAN2
– to use CAN2 only: set the CAN_MCR.INRQ on CAN1 or deactivate the interupt
register CAN_IER on CAN1
Bits 7:1 Reserved, must be kept at reset value.
Bit 0 FINIT: Filter initialization mode
Initialization mode for filter banks
0: Active filters mode.
1: Initialization mode for the filters.

CAN filter mode register (CAN_FM1R)

Address offset: 0x204
Reset value: 0x0000 0000
This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.

RM0402 Rev 6 959/1163

966
Controller area network (bxCAN) RM0402

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. FBM27 FBM26 FBM25 FBM24 FBM23 FBM22 FBM21 FBM20 FBM19 FBM18 FBM17 FBM16
rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FBM15 FBM14 FBM13 FBM12 FBM11 FBM10 FBM9 FBM8 FBM7 FBM6 FBM5 FBM4 FBM3 FBM2 FBM1 FBM0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Note: Refer to Figure 326: Filter bank scale configuration - register organization on page 934.

Bits 31: Reserved, must be kept at reset value.

Bits 27:0 FBMx: Filter mode
Mode of the registers of Filter x.
0: Two 32-bit registers of filter bank x are in Identifier Mask mode.
1: Two 32-bit registers of filter bank x are in Identifier List mode.

CAN filter scale register (CAN_FS1R)

Address offset: 0x20C
Reset value: 0x0000 0000
This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. FSC27 FSC26 FSC25 FSC24 FSC23 FSC22 FSC21 FSC20 FSC19 FSC18 FSC17 FSC16
rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FSC15 FSC14 FSC13 FSC12 FSC11 FSC10 FSC9 FSC8 FSC7 FSC6 FSC5 FSC4 FSC3 FSC2 FSC1 FSC0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:28 Reserved, must be kept at reset value.

Bits 27:0 FSCx: Filter scale configuration
These bits define the scale configuration of Filters 27-0.
0: Dual 16-bit scale configuration
1: Single 32-bit scale configuration

Note: Refer to Figure 326: Filter bank scale configuration - register organization on page 934.

CAN filter FIFO assignment register (CAN_FFA1R)

Address offset: 0x214
Reset value: 0x0000 0000
This register can be written only when the filter initialization mode is set (FINIT=1) in the
CAN_FMR register.

960/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Res. Res. Res. Res. FFA27 FFA26 FFA25 FFA24 FFA23 FFA22 FFA21 FFA20 FFA19 FFA18 FFA17 FFA16
rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FFA15 FFA14 FFA13 FFA12 FFA11 FFA10 FFA9 FFA8 FFA7 FFA6 FFA5 FFA4 FFA3 FFA2 FFA1 FFA0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:28 Reserved, must be kept at reset value.

Bits 27:0 FFAx: Filter FIFO assignment for filter x
The message passing through this filter is stored in the specified FIFO.
0: Filter assigned to FIFO 0
1: Filter assigned to FIFO 1

CAN filter activation register (CAN_FA1R)

Address offset: 0x21C
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FACT FACT FACT FACT FACT FACT FACT FACT FACT FACT FACT FACT
Res. Res. Res. Res.
27 26 25 24 23 22 21 20 19 18 17 16
rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FACT FACT FACT FACT FACT FACT
FACT9 FACT8 FACT7 FACT6 FACT5 FACT4 FACT3 FACT2 FACT1 FACT0
15 14 13 12 11 10
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:28 Reserved, must be kept at reset value.

Bits 27:0 FACTx: Filter active
The software sets this bit to activate Filter x. To modify the Filter x registers (CAN_FxR[0:7]),
the FACTx bit must be cleared or the FINIT bit of the CAN_FMR register must be set.
0: Filter x is not active
1: Filter x is active

RM0402 Rev 6 961/1163

966
Controller area network (bxCAN) RM0402

Filter bank i register x (CAN_FiRx) (i = 0..27, x = 1, 2)

Address offsets: 0x240 to 0x31C
Reset value: 0xXXXX XXXX
There are 28 filter banks, i= 0 to 27. Each filter bank i is composed of two 32-bit registers,
CAN_FiR[2:1].
This register can only be modified when the FACTx bit of the CAN_FAxR register is cleared
or when the FINIT bit of the CAN_FMR register is set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FB31 FB30 FB29 FB28 FB27 FB26 FB25 FB24 FB23 FB22 FB21 FB20 FB19 FB18 FB17 FB16
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FB15 FB14 FB13 FB12 FB11 FB10 FB9 FB8 FB7 FB6 FB5 FB4 FB3 FB2 FB1 FB0
rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

In all configurations:

Bits 31:0 FB[31:0]: Filter bits

Identifier
Each bit of the register specifies the level of the corresponding bit of the expected identifier.
0: Dominant bit is expected
1: Recessive bit is expected
Mask
Each bit of the register specifies whether the bit of the associated identifier register must
match with the corresponding bit of the expected identifier or not.
0: Do not care, the bit is not used for the comparison
1: Must match, the bit of the incoming identifier must have the same level has specified in
the corresponding identifier register of the filter.

Note: Depending on the scale and mode configuration of the filter the function of each register can
differ. For the filter mapping, functions description and mask registers association, refer to
Section 28.7.4: Identifier filtering on page 932.
A Mask/Identifier register in mask mode has the same bit mapping as in identifier list
mode.
For the register mapping/addresses of the filter banks refer to Table 195 on page 963.

962/1163 RM0402 Rev 6

 0x180
0x018
0x014
0x010
0x008
0x004
0x000

0x17F
0x01C
0x00C

0x020-
Offset
28.9.5
RM0402

-
CAN_IER
CAN_TSR

CAN_BTR
CAN_ESR

CAN_TI0R
CAN_MSR
CAN_MCR
Register

Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

CAN_RF1R
CAN_RF0R

x
0
0
0
Res. SILM Res. Res. Res. Res. Res. 31

x
0
0
0
Res. LBKM Res. Res. Res. LOW[2:0] Res. Res. 30

x
0
0
Res. Res. Res. Res. Res. Res. Res. 29

x
0
1
Res. Res. Res. Res. Res. Res. Res. 28

x
0
1
Res. Res. Res. Res. Res. TME[2:0] Res. Res. 27

REC[7:0]

x
0
1
Res. Res. Res. Res. Res. Res. Res. 26

x
0
0
0
Res. Res. Res. Res. Res. Res. 25
bxCAN register map

SJW[1:0] CODE[1:0]

x
0
0
0
Res. Res. Res. Res. Res. Res. 24

STID[10:0]/EXID[28:18]

x
0
0
Res. Res. Res. Res. Res. ABRQ2 Res. Res. 23

x
0
0
Res. Res. Res. Res. Res. Res. Res. 22

x
1
0
Res. Res. Res. Res. Res. Res. Res. 21

TS2[2:0]

x
0
0
Res. Res. Res. Res. Res. Res. Res. 20

x
0
0
0
Res. Res. Res. Res. TERR2 Res. Res. 19

TEC[7:0]

x
0
0
0
Res. Res. Res. Res. ALST2 Res. Res. 18

x
1
0
0

RM0402 Rev 6
Res. SLKIE Res. Res. TXOK2 Res. Res. 17

TS1[3:0]

x
1
0
0

0
1
offset 0x200 to 0x31C are present only in CAN1.

Res. WKUIE Res. Res. RQCP2 Res. DBF 16

x
0

0
0

Res. Res. Res. ERRIE Res. Res. ABRQ1 Res. RESET 15

x
Res. Res. Res. Res. Res. Res. Res. Res. Res. 14

x
Res. Res. Res. Res. Res. Res. Res. Res. Res. 13

x
Res. Res. Res. Res. Res. Res. Res. Res. Res. 12

x
0

0
1

Res. Res. Res. LECIE Res. Res. TERR1 RX Res. 11

EXID[17:0]

x
0

0
1
Table 195. bxCAN register map and reset values

Res. Res. Res. BOFIE Res. Res. ALST1 SAMP Res. 10

x
0
0

0
0

Res. Res. EPVIE Res. Res. TXOK1 RXM Res. 9

x
0
0

0
0

Res. Res. EWGIE Res. Res. RQCP1 TXM Res. 8

x
0
0
0

Res. Res. Res. Res. Res. ABRQ0 Res. TTCM 7

x
0
0
0
0

Res. FOVIE1 Res. Res. Res. Res. ABOM 6

x
0
0
0
0
0
0

Res. LEC[2:0] FFIE1 RFOM1 RFOM0 Res. Res. AWUM 5

x
0
0
0
0
0
0
0

Res. FMPIE1 FOVR1 FOVR0 Res. SLAKI NART 4

BRP[9:0]

x
0
0
0

0
0
0
0

Res. Res. FOVIE0 FULL1 FULL0 TERR0 WKUI RFLM 3

x
0
0
0
0
0
0

IDE Res. BOFF FFIE0 Res. Res. ALST0 ERRI TXFP 2

x
0
0
0
0

0
0
1
1
Refer to Section 2.2 on page 49 for the register boundary addresses. The registers from

RTR Res. EPVF FMPIE0 TXOK0 SLAK SLEEP 1

FMP1[1:0] FMP0[1:0]

0
0

0
0
0

0
0
0
0

TXRQ Res. EWGF TMEIE RQCP0 INAK INRQ 0

Controller area network (bxCAN)

963/1163
966
Controller area network (bxCAN) RM0402

Table 195. bxCAN register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
TGT
CAN_TDT0R TIME[15:0] DLC[3:0]
0x184

Reset value x x x x x x x x x x x x x x x x - - - - - - - x - - - - x x x x

CAN_TDL0R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x188

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_TDH0R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x18C

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

TXRQ
RTR
IDE
CAN_TI1R STID[10:0]/EXID[28:18] EXID[17:0]
0x190

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
TGT
CAN_TDT1R TIME[15:0] DLC[3:0]
0x194

Reset value x x x x x x x x x x x x x x x x - - - - - - - x - - - - x x x x

CAN_TDL1R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x198

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_TDH1R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x19C

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

TXRQ
RTR
CAN_TI2R STID[10:0]/EXID[28:18] EXID[17:0] IDE
0x1A0

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
TGT

CAN_TDT2R TIME[15:0] DLC[3:0]

0x1A4

Reset value x x x x x x x x x x x x x x x x - - - - - - - x - - - - x x x x

CAN_TDL2R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x1A8

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_TDH2R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x1AC

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x
Res.
RTR
IDE

CAN_RI0R STID[10:0]/EXID[28:18] EXID[17:0]

0x1B0
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x -

964/1163 RM0402 Rev 6

RM0402 Controller area network (bxCAN)

Table 195. bxCAN register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
Res.
Res.
CAN_RDT0R TIME[15:0] FMI[7:0] DLC[3:0]
0x1B4

Reset value x x x x x x x x x x x x x x x x x x x x x x x x - - - - x x x x

CAN_RDL0R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x1B8

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_RDH0R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x1BC

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

Res.
RTR
IDE
CAN_RI1R STID[10:0]/EXID[28:18] EXID[17:0]
0x1C0
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x -

Res.
Res.
Res.
Res.
CAN_RDT1R TIME[15:0] FMI[7:0] DLC[3:0]
0x1C4

Reset value x x x x x x x x x x x x x x x x x x x x x x x x - - - - x x x x

CAN_RDL1R DATA3[7:0] DATA2[7:0] DATA1[7:0] DATA0[7:0]

0x1C8

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_RDH1R DATA7[7:0] DATA6[7:0] DATA5[7:0] DATA4[7:0]

0x1CC

Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

0x1D0-
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
0x1FF
-
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
- Res.
0x208

-
Res.
Res.
Res.
Res.

CAN_FS1R FSC[27:0]
0x20C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

0x210 -
Res.
Res.
Res.
Res.

CAN_FFA1R FFA[27:0]
0x214
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

0x218 -

RM0402 Rev 6 965/1163

966
Controller area network (bxCAN) RM0402

Table 195. bxCAN register map and reset values (continued)

Offset Register

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
Res.
Res.
Res.
Res.
CAN_FA1R FACT[27:0]
0x21C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
0x220 -

0x224-
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
0x23F
-

CAN_F0R1 FB[31:0]
0x240
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_F0R2 FB[31:0]
0x244
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_F1R1 FB[31:0]
0x248
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_F1R2 FB[31:0]
0x24C
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

. . .
. . .
. . .
. . .

CAN_F27R1 FB[31:0]
0x318
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

CAN_F27R2 FB[31:0]
0x31C
Reset value x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

Refer to Section 2.2 on page 49 for the register boundary addresses.

966/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29 USB on-the-go full-speed (OTG_FS)

29.1 Introduction
Portions Copyright (c) Synopsys, Inc. All rights reserved. Used with permission.
This section presents the architecture and the programming model of the OTG_FS
controller.
The following acronyms are used throughout the section:
FS Full-speed
LS Low-speed
MAC Media access controller
OTG On-the-go
PFC Packet FIFO controller
PHY Physical layer
USB Universal serial bus
UTMI USB 2.0 Transceiver Macrocell interface (UTMI)
LPM Link power management
BCD Battery charging detector
HNP Host negotiation protocol
SRP Session request protocol

References are made to the following documents:

• USB On-The-Go Supplement, Revision 2.0
• Universal Serial Bus Revision 2.0 Specification
• USB 2.0 Link Power Management Addendum Engineering Change Notice to the USB
2.0 specification, July 16, 2007
• Errata for USB 2.0 ECN: Link Power Management (LPM) - 7/2007
• Battery Charging Specification, Revision 1.2
The USB OTG is a dual-role device (DRD) controller that supports both device and host
functions and is fully compliant with the On-The-Go Supplement to the USB 2.0
Specification. It can also be configured as a host-only or device-only controller, fully
compliant with the USB 2.0 Specification. OTG_FS supports the speeds defined in the
Table 196: OTG_FS speeds supported below. The USB OTG supports both HNP and SRP.
The only external device required is a charge pump for VBUS in OTG mode.

Table 196. OTG_FS speeds supported

- HS (480 Mb/s) FS (12 Mb/s) LS (1.5 Mb/s)

Host mode - X X
Device mode - X -

RM0402 Rev 6 967/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.2 OTG_FS main features

The main features can be divided into three categories: general, host-mode and device-
mode features.

29.2.1 General features

The OTG_FS interface general features are the following:
• It is USB-IF certified to the Universal Serial Bus Specification Rev 2.0
• OTG_FS supports the following PHY interface:
– An on-chip full-speed PHY
• It includes full support (PHY) for the optional On-The-Go (OTG) protocol detailed in the
On-The-Go Supplement Rev 2.0 specification
– Integrated support for A-B device identification (ID line)
– Integrated support for host Negotiation protocol (HNP) and session request
protocol (SRP)
– It allows host to turn VBUS off to conserve battery power in OTG applications
– It supports OTG monitoring of VBUS levels with internal comparators
– It supports dynamic host-peripheral switch of role
• It is software-configurable to operate as:
– SRP capable USB FS Peripheral (B-device)
– SRP capable USB FS/LS host (A-device)
– USB On-The-Go Full-Speed Dual Role device
• It supports FS SOF and LS Keep-alives with
– SOF pulse PAD connectivity
– SOF pulse internal connection to timer (TIMx)
– Configurable framing period
– Configurable end of frame interrupt
• It includes power saving features such as system stop during USB suspend, switch-off
of clock domains internal to the digital core, PHY and DFIFO power management.
• It features a dedicated RAM of 1.25 Kbytes with advanced FIFO control:
– Configurable partitioning of RAM space into different FIFOs for flexible and
efficient use of RAM
– Each FIFO can hold multiple packets
– Dynamic memory allocation
– Configurable FIFO sizes that are not powers of 2 to allow the use of contiguous
memory locations
• It guarantees max USB bandwidth for up to one frame (1 ms) without system
intervention.
• It supports charging port detection as described in Battery Charging Specification
Revision 1.2 on the FS PHY transceiver only.

968/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.2.2 Host-mode features

The OTG_FS interface main features and requirements in host-mode are the following:
• External charge pump for VBUS voltage generation.
• Up to 12 host channels (pipes): each channel is dynamically reconfigurable to allocate
any type of USB transfer.
• Built-in hardware scheduler holding:
– Up to 12 interrupt plus isochronous transfer requests in the periodic hardware
queue
– Up to 12 control plus bulk transfer requests in the non-periodic hardware queue
• Management of a shared Rx FIFO, a periodic Tx FIFO and a nonperiodic Tx FIFO for
efficient usage of the USB data RAM.

29.2.3 Peripheral-mode features

The OTG_FS interface main features in peripheral-mode are the following:
• 1 bidirectional control endpoint0
• 5 IN endpoints (EPs) configurable to support bulk, interrupt or isochronous transfers
• 5 OUT endpoints configurable to support bulk, interrupt or isochronous transfers
• Management of a shared Rx FIFO and a Tx-OUT FIFO for efficient usage of the USB
data RAM
• Management of up to 6 dedicated Tx-IN FIFOs (one for each active IN EP) to put less
load on the application
• Support for the soft disconnect feature.

29.2.4 Split rail for USB

In some package options it is possible to work with a split rail VDDUSB which has a more
restrictive voltage specification than the VDD which can work at a lower range.
When this is the case, and assuming that VDD is indeed below the minimum functional level
for VDDUSB, please take the following precautions.
For host only applications:
• No special precautions are necessary.
For VBUS powered USB2.0 peripheral device:
• No special precautions are necessary. VBUS detection is not needed, as VBUS state is
reflected directly in the presence of power supplies on the MCU.
All other cases (VBUS detection is required):
• Refer to the datasheet (Electrical Characteristics, and Appendix for application block
diagrams). A different range of the VDD supply will apply this case.

RM0402 Rev 6 969/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.3 OTG_FS implementation

Table 197. OTG_FS implementation(1)
USB features OTG_FS

Device bidirectional endpoints (including EP0) 6

Host mode channels 12
Size of dedicated SRAM 1.2 KBytes
USB 2.0 link power management (LPM) support X
OTG revision supported 2.0
Attach detection protocol (ADP) support -
Battery charging detection (BCD) support X
1. “X” = supported

970/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.4 OTG_FS functional description

29.4.1 OTG_FS block diagram

Figure 334. OTG_FS full-speed block diagram

Cortex® core

OTG_FS_DP
Power
USB2.0 OTG OTG_FS_DM
and
OTG FS UTMIFS FS
clock OTG_FS_ID
core PHY
controller USB suspend
USB clock at 48 MHz System clock domain
USB clock OTG_FS_VBUS
domain
RAM bus

Universal serial bus

OTG_FS_SOF
1.25 Kbyte
USB data
FIFOs

MS19928V4

29.4.2 OTG_FS pin and internal signals

Table 198. OTG_FS input/output pins

Signal name Signal type Description

OTG_FS_DP Digital input/output USB OTG D+ line

OTG_FS_DM Digital input/output USB OTG D- line
OTG_FS_ID Digital input USB OTG ID
OTG_FS_VBUS Analog input USB OTG VBUS
OTG_FS_SOF Digital output USB OTG Start Of Frame (visibility)

RM0402 Rev 6 971/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Table 199. OTG_FS input/output signals

Signal name Signal type Description

usb_sof Digital output USB OTG start-of-frame event for on chip peripherals
usb_wkup Digital output USB OTG wakeup event output
usb_gbl_it Digital output USB OTG global interrupt

29.4.3 OTG_FS core

The USB OTG_FS receives the 48 MHz clock from the reset and clock controller (RCC).
This clock is used for driving the 48 MHz domain at full-speed (12 Mbit/s) and must be
enabled prior to configuring the OTG core.
The CPU reads and writes from/to the OTG core registers through the AHB peripheral bus.
It is informed of USB events through the single USB OTG interrupt line described in
Section 29.13: OTG_FS interrupts.
The CPU submits data over the USB by writing 32-bit words to dedicated OTG locations
(push registers). The data are then automatically stored into Tx-data FIFOs configured
within the USB data RAM. There is one Tx FIFO push register for each in-endpoint
(peripheral mode) or out-channel (host mode).
The CPU receives the data from the USB by reading 32-bit words from dedicated OTG
addresses (pop registers). The data are then automatically retrieved from a shared Rx FIFO
configured within the 1.25-Kbyte USB data RAM. There is one Rx FIFO pop register for
each out-endpoint or in-channel.
The USB protocol layer is driven by the serial interface engine (SIE) and serialized over the
USB by the transceiver module within the on-chip physical layer (PHY).
Caution: To guarantee a correct operation for the USB OTG FS peripheral, the AHB frequency should
be higher than 14.2 MHz.

29.4.4 Embedded full-speed OTG PHY connected to OTG_FS

The embedded full-speed OTG PHY is controlled by the OTG FS core and conveys USB
control & data signals through the full-speed subset of the UTMI+ Bus (UTMIFS). It provides
the physical support to USB connectivity.
The full-speed OTG PHY includes the following components:
• FS/LS transceiver module used by both host and device. It directly drives transmission
and reception on the single-ended USB lines.
• DP/DM integrated pull-up and pull-down resistors controlled by the OTG_FS core
depending on the current role of the device. As a peripheral, it enables the DP pull-up
resistor to signal full-speed peripheral connections as soon as VBUS is sensed to be at
a valid level (B-session valid). In host mode, pull-down resistors are enabled on both
DP/DM. Pull-up and pull-down resistors are dynamically switched when the role of the
device is changed via the host negotiation protocol (HNP).
• Pull-up/pull-down resistor ECN circuit. The DP pull-up consists of two resistors
controlled separately from the OTG_FS as per the resistor Engineering Change Notice
applied to USB Rev2.0. The dynamic trimming of the DP pull-up strength allows for
better noise rejection and Tx/Rx signal quality.

972/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.4.5 OTG detections

Additionally the OTG_FS uses the following functions:
• integrated ID pull-up resistor used to sample the ID line for A/B device identification.
• VBUS sensing comparators with hysteresis used to detect VBUS valid, A-B session valid
and session-end voltage thresholds. They are used to drive the session request
protocol (SRP), detect valid startup and end-of-session conditions, and constantly
monitor the VBUS supply during USB operations.

29.5 OTG_FS dual role device (DRD)

Figure 335. OTG_FS A-B device connection

VDD
5 V to VDD
Voltage
regulator(1)

VDD
STM32 EN STMPS2141STR
GPIO
Current-limited 5 V Pwr
Overcurrent power distribution
GPIO + IRQ switch(2)

VBUS
VBUS

USBmicro-AB connector
DM
DM
OSC_IN DP
DP
ID
ID
OSC_OUT
VSS

MSv36917V2

1. External voltage regulator only needed when building a VBUS powered device.
2. STMPS2141STR needed only if the application has to support a VBUS powered device. A basic power
switch can be used if 5 V are available on the application board.

29.5.1 ID line detection

The host or peripheral (the default) role is assumed depending on the ID input pin. The ID
line status is determined on plugging in the USB cable, depending on whether a MicroA or
MicroB plug is connected to the micro-AB receptacle.
• If the B-side of the USB cable is connected with a floating ID wire, the integrated pull-
up resistor detects a high ID level and the default peripheral role is confirmed. In this
configuration the OTG_FS complies with the standard FSM described in section 4.2.4:
ID pin of the On-the-Go specification Rev2.0, supplement to the USB2.0.
• If the A-side of the USB cable is connected with a grounded ID, the OTG_FS issues an
ID line status change interrupt (CIDSCHG bit in OTG_GINTSTS) for host software
initialization, and automatically switches to the host role. In this configuration the
OTG_FS complies with the standard FSM described by section 4.2.4: ID pin of the On-
the-Go specification Rev2.0, supplement to the USB2.0.

RM0402 Rev 6 973/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.5.2 HNP dual role device

The HNP capable bit in the Global USB configuration register (HNPCAP bit in OTG_
GUSBCFG) enables the OTG_FS core to dynamically change its role from A-host to A-
peripheral and vice-versa, or from B-Peripheral to B-host and vice-versa according to the
host negotiation protocol (HNP). The current device status can be read by the combined
values of the connector ID status bit in the Global OTG control and status register (CIDSTS
bit in OTG_GOTGCTL) and the current mode of operation bit in the global interrupt and
status register (CMOD bit in OTG_GINTSTS).
The HNP program model is described in detail in Section 29.16: OTG_FS programming
model.

29.5.3 SRP dual role device

The SRP capable bit in the global USB configuration register (SRPCAP bit in
OTG_GUSBCFG) enables the OTG_FS core to switch off the generation of VBUS for the A-
device to save power. Note that the A-device is always in charge of driving VBUS regardless
of the host or peripheral role of the OTG_FS.
The SRP A/B-device program model is described in detail in Section 29.16: OTG_FS
programming model.

29.6 OTG_FS as a USB peripheral

This section gives the functional description of the OTG_FS in the USB peripheral mode.
The OTG_FS works as an USB peripheral in the following circumstances:
• OTG B-Peripheral
– OTG B-device default state if B-side of USB cable is plugged in
• OTG A-Peripheral
– OTG A-device state after the HNP switches the OTG_FS to its peripheral role
• B-device
– If the ID line is present, functional and connected to the B-side of the USB cable,
and the HNP-capable bit in the Global USB Configuration register (HNPCAP bit in
OTG_GUSBCFG) is cleared.
• Peripheral only (see Figure 336: OTG_FS peripheral-only connection)
– The force device mode bit (FDMOD) in the Section 29.15.4: OTG USB
configuration register (OTG_GUSBCFG) is set to 1, forcing the OTG_FS core to
work as an USB peripheral-only. In this case, the ID line is ignored even if it is
present on the USB connector.
Note: To build a bus-powered device implementation in case of the B-device or peripheral-only
configuration, an external regulator has to be added, that generates the necessary power-
supply from VBUS.

974/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 336. OTG_FS peripheral-only connection

VDD
5 V to VDD
Voltage
regulator

VDD

GPIO EN STMPS2141STR
Current-limited 5 V Pwr
Overcurrent power distribution
GPIO + IRQ switch(1)

VBUS

DM

USB micro connector

OSC_IN DP

OSC_OUT
VSS

MSv36916V2

1. Use a regulator to build a bus-powered device.

29.6.1 SRP-capable peripheral

The SRP capable bit in the Global USB configuration register (SRPCAP bit in
OTG_GUSBCFG) enables the OTG_FS to support the session request protocol (SRP). In
this way, it allows the remote A-device to save power by switching off VBUS while the USB
session is suspended.
The SRP peripheral mode program model is described in detail in the B-device session
request protocol section.

29.6.2 Peripheral states

Powered state
The VBUS input detects the B-session valid voltage by which the USB peripheral is allowed
to enter the powered state (see USB2.0 section 9.1). The OTG_FS then automatically
connects the DP pull-up resistor to signal full-speed device connection to the host and
generates the session request interrupt (SRQINT bit in OTG_GINTSTS) to notify the
powered state.
The VBUS input also ensures that valid VBUS levels are supplied by the host during USB
operations. If a drop in VBUS below B-session valid happens to be detected (for instance
because of a power disturbance or if the host port has been switched off), the OTG_FS
automatically disconnects and the session end detected (SEDET bit in OTG_GOTGINT)
interrupt is generated to notify that the OTG_FS has exited the powered state.
In the powered state, the OTG_FS expects to receive some reset signaling from the host.
No other USB operation is possible. When a reset signaling is received the reset detected
interrupt (USBRST in OTG_GINTSTS) is generated. When the reset signaling is complete,
the enumeration done interrupt (ENUMDNE bit in OTG_GINTSTS) is generated and the
OTG_FS enters the Default state.

RM0402 Rev 6 975/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Soft disconnect
The powered state can be exited by software with the soft disconnect feature. The DP pull-
up resistor is removed by setting the soft disconnect bit in the device control register (SDIS
bit in OTG_DCTL), causing a device disconnect detection interrupt on the host side even
though the USB cable was not really removed from the host port.

Default state
In the Default state the OTG_FS expects to receive a SET_ADDRESS command from the
host. No other USB operation is possible. When a valid SET_ADDRESS command is
decoded on the USB, the application writes the corresponding number into the device
address field in the device configuration register (DAD bit in OTG_DCFG). The OTG_FS
then enters the address state and is ready to answer host transactions at the configured
USB address.

Suspended state
The OTG_FS peripheral constantly monitors the USB activity. After counting 3 ms of USB
idleness, the early suspend interrupt (ESUSP bit in OTG_GINTSTS) is issued, and
confirmed 3 ms later, if appropriate, by the suspend interrupt (USBSUSP bit in
OTG_GINTSTS). The device suspend bit is then automatically set in the device status
register (SUSPSTS bit in OTG_DSTS) and the OTG_FS enters the suspended state.
The suspended state may optionally be exited by the device itself. In this case the
application sets the remote wakeup signaling bit in the device control register (RWUSIG bit
in OTG_DCTL) and clears it after 1 to 15 ms.
When a resume signaling is detected from the host, the resume interrupt (WKUPINT bit in
OTG_GINTSTS) is generated and the device suspend bit is automatically cleared.

29.6.3 Peripheral endpoints

The OTG_FS core instantiates the following USB endpoints:
• Control endpoint 0:
– Bidirectional and handles control messages only
– Separate set of registers to handle in and out transactions
– Proper control (OTG_DIEPCTL0/OTG_DOEPCTL0), transfer configuration
(OTG_DIEPTSIZ0/OTG_DOEPTSIZ0), and status-interrupt
(OTG_DIEPINT0/)OTG_DOEPINT0) registers. The available set of bits inside the
control and transfer size registers slightly differs from that of other endpoints
• 5 IN endpoints
– Each of them can be configured to support the isochronous, bulk or interrupt
transfer type
– Each of them has proper control (OTG_DIEPCTLx), transfer configuration
(OTG_DIEPTSIZx), and status-interrupt (OTG_DIEPINTx) registers
– The device IN endpoints common interrupt mask register (OTG_DIEPMSK) is
available to enable/disable a single kind of endpoint interrupt source on all of the
IN endpoints (EP0 included)
– Support for incomplete isochronous IN transfer interrupt (IISOIXFR bit in
OTG_GINTSTS), asserted when there is at least one isochronous IN endpoint on

976/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

which the transfer is not completed in the current frame. This interrupt is asserted
along with the end of periodic frame interrupt (OTG_GINTSTS/EOPF).
• 5 OUT endpoints
– Each of them can be configured to support the isochronous, bulk or interrupt
transfer type
– Each of them has a proper control (OTG_DOEPCTLx), transfer configuration
(OTG_DOEPTSIZx) and status-interrupt (OTG_DOEPINTx) register
– Device OUT endpoints common interrupt mask register (OTG_DOEPMSK) is
available to enable/disable a single kind of endpoint interrupt source on all of the
OUT endpoints (EP0 included)
– Support for incomplete isochronous OUT transfer interrupt (INCOMPISOOUT bit
in OTG_GINTSTS), asserted when there is at least one isochronous OUT
endpoint on which the transfer is not completed in the current frame. This interrupt
is asserted along with the end of periodic frame interrupt (OTG_GINTSTS/EOPF).

Endpoint control
• The following endpoint controls are available to the application through the device
endpoint-x IN/OUT control register (OTG_DIEPCTLx/OTG_DOEPCTLx):
– Endpoint enable/disable
– Endpoint activate in current configuration
– Program USB transfer type (isochronous, bulk, interrupt)
– Program supported packet size
– Program Tx FIFO number associated with the IN endpoint
– Program the expected or transmitted data0/data1 PID (bulk/interrupt only)
– Program the even/odd frame during which the transaction is received or
transmitted (isochronous only)
– Optionally program the NAK bit to always negative-acknowledge the host
regardless of the FIFO status
– Optionally program the STALL bit to always stall host tokens to that endpoint
– Optionally program the SNOOP mode for OUT endpoint not to check the CRC
field of received data

Endpoint transfer
The device endpoint-x transfer size registers (OTG_DIEPTSIZx/OTG_DOEPTSIZx) allow
the application to program the transfer size parameters and read the transfer status.
Programming must be done before setting the endpoint enable bit in the endpoint control
register. Once the endpoint is enabled, these fields are read-only as the OTG_FS core
updates them with the current transfer status.
The following transfer parameters can be programmed:
• Transfer size in bytes
• Number of packets that constitute the overall transfer size

Endpoint status/interrupt
The device endpoint-x interrupt registers (OTG_DIEPINTx/OTG_DOPEPINTx) indicate the
status of an endpoint with respect to USB- and AHB-related events. The application must
read these registers when the OUT endpoint interrupt bit or the IN endpoint interrupt bit in

RM0402 Rev 6 977/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

the core interrupt register (OEPINT bit in OTG_GINTSTS or IEPINT bit in OTG_GINTSTS,
respectively) is set. Before the application can read these registers, it must first read the
device all endpoints interrupt (OTG_DAINT) register to get the exact endpoint number for
the device endpoint-x interrupt register. The application must clear the appropriate bit in this
register to clear the corresponding bits in the OTG_DAINT and OTG_GINTSTS registers
The peripheral core provides the following status checks and interrupt generation:
• Transfer completed interrupt, indicating that data transfer was completed on both the
application (AHB) and USB sides
• Setup stage has been done (control-out only)
• Associated transmit FIFO is half or completely empty (in endpoints)
• NAK acknowledge has been transmitted to the host (isochronous-in only)
• IN token received when Tx FIFO was empty (bulk-in/interrupt-in only)
• Out token received when endpoint was not yet enabled
• Babble error condition has been detected
• Endpoint disable by application is effective
• Endpoint NAK by application is effective (isochronous-in only)
• More than 3 back-to-back setup packets were received (control-out only)
• Timeout condition detected (control-in only)
• Isochronous out packet has been dropped, without generating an interrupt

29.7 OTG_FS as a USB host

This section gives the functional description of the OTG_FS in the USB host mode. The
OTG_FS works as a USB host in the following circumstances:
• OTG A-host
– OTG A-device default state when the A-side of the USB cable is plugged in
• OTG B-host
– OTG B-device after HNP switching to the host role
• A-device
– If the ID line is present, functional and connected to the A-side of the USB cable,
and the HNP-capable bit is cleared in the Global USB Configuration register
(HNPCAP bit in OTG_GUSBCFG). Integrated pull-down resistors are
automatically set on the DP/DM lines.
• Host only
– The force host mode bit (FHMOD) in the OTG USB configuration register
(OTG_GUSBCFG) forces the OTG_FS core to work as a USB host-only. In this
case, the ID line is ignored even if present on the USB connector. Integrated pull-
down resistors are automatically set on the DP/DM lines.
Note: On-chip 5 V VBUS generation is not supported. For this reason, a charge pump or, if 5 V are
available on the application board, a basic power switch must be added externally to drive
the 5 V VBUS line. The external charge pump can be driven by any GPIO output. This is
required for the OTG A-host, A-device and host-only configurations.

978/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 337. OTG_FS host-only connection

VDD
5V

EN STMPS2141STR
GPIO
Current-limited 5 V Pwr

GPIO + IRQ Overcurrent power distribution

switch(2)

USB Std-A connector

VBUS
DM
OSC_IN
DP
VSS
OSC_OUT

MSv36915V2

1. VDD range is between 2 V and 3.6 V.

29.7.1 SRP-capable host

SRP support is available through the SRP capable bit in the global USB configuration
register (SRPCAP bit in OTG_GUSBCFG). With the SRP feature enabled, the host can
save power by switching off the VBUS power while the USB session is suspended.
The SRP host mode program model is described in detail in the A-device session request
protocol) section.

29.7.2 USB host states

Host port power
On-chip 5 V VBUS generation is not supported. For this reason, a charge pump or, if 5 V are
available on the application board, a basic power switch, must be added externally to drive
the 5 V VBUS line. The external charge pump can be driven by any GPIO output or via an
I2C interface connected to an external PMIC (power management IC). When the application
decides to power on VBUS, it must also set the port power bit in the host port control and
status register (PPWR bit in OTG_HPRT).

VBUS valid
When HNP or SRP is enabled the VBUS sensing pin should be connected to VBUS. The
VBUS input ensures that valid VBUS levels are supplied by the charge pump during USB
operations. Any unforeseen VBUS voltage drop below the VBUS valid threshold (4.4 V) leads
to an OTG interrupt triggered by the session end detected bit (SEDET bit in
OTG_GOTGINT). The application is then required to remove the VBUS power and clear the
port power bit.
When HNP and SRP are both disabled, the VBUS sensing pin does not need to be
connected to VBUS.
The charge pump overcurrent flag can also be used to prevent electrical damage. Connect
the overcurrent flag output from the charge pump to any GPIO input and configure it to
generate a port interrupt on the active level. The overcurrent ISR must promptly disable the
VBUS generation and clear the port power bit.

RM0402 Rev 6 979/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Host detection of a peripheral connection

If SRP or HNP are enabled, even if USB peripherals or B-devices can be attached at any
time, the OTG_FS does not detect any bus connection until VBUS is no longer sensed at a
valid level (5 V). When VBUS is at a valid level and a remote B-device is attached, the
OTG_FS core issues a host port interrupt triggered by the device connected bit in the host
port control and status register (PCDET bit in OTG_HPRT).
When HNP and SRP are both disabled, USB peripherals or B-device are detected as soon
as they are connected. The OTG_FS core issues a host port interrupt triggered by the
device connected bit in the host port control and status (PCDET bit in OTG_HPRT).

Host detection of peripheral a disconnection

The peripheral disconnection event triggers the disconnect detected interrupt (DISCINT bit
in OTG_GINTSTS).

Host enumeration
After detecting a peripheral connection the host must start the enumeration process by
sending USB reset and configuration commands to the new peripheral.
Before starting to drive a USB reset, the application waits for the OTG interrupt triggered by
the debounce done bit (DBCDNE bit in OTG_GOTGINT), which indicates that the bus is
stable again after the electrical debounce caused by the attachment of a pull-up resistor on
DP (FS) or DM (LS).
The application drives a USB reset signaling (single-ended zero) over the USB by keeping
the port reset bit set in the host port control and status register (PRST bit in OTG_HPRT) for
a minimum of 10 ms and a maximum of 20 ms. The application takes care of the timing
count and then of clearing the port reset bit.
Once the USB reset sequence has completed, the host port interrupt is triggered by the port
enable/disable change bit (PENCHNG bit in OTG_HPRT). This informs the application that
the speed of the enumerated peripheral can be read from the port speed field in the host
port control and status register (PSPD bit in OTG_HPRT) and that the host is starting to
drive SOFs (FS) or Keep alives (LS). The host is now ready to complete the peripheral
enumeration by sending peripheral configuration commands.

Host suspend
The application decides to suspend the USB activity by setting the port suspend bit in the
host port control and status register (PSUSP bit in OTG_HPRT). The OTG_FS core stops
sending SOFs and enters the suspended state.
The suspended state can be optionally exited on the remote device’s initiative (remote
wakeup). In this case the remote wakeup interrupt (WKUPINT bit in OTG_GINTSTS) is
generated upon detection of a remote wakeup signaling, the port resume bit in the host port
control and status register (PRES bit in OTG_HPRT) self-sets, and resume signaling is
automatically driven over the USB. The application must time the resume window and then
clear the port resume bit to exit the suspended state and restart the SOF.
If the suspended state is exited on the host initiative, the application must set the port
resume bit to start resume signaling on the host port, time the resume window and finally
clear the port resume bit.

980/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.7.3 Host channels

The OTG_FS core instantiates 12 host channels. Each host channel supports an USB host
transfer (USB pipe). The host is not able to support more than 12 transfer requests at the
same time. If more than 12 transfer requests are pending from the application, the host
controller driver (HCD) must re-allocate channels when they become available from
previous duty, that is, after receiving the transfer completed and channel halted interrupts.
Each host channel can be configured to support in/out and any type of periodic/nonperiodic
transaction. Each host channel makes us of proper control (OTG_HCCHARx), transfer
configuration (OTG_HCTSIZx) and status/interrupt (OTG_HCINTx) registers with
associated mask (OTG_HCINTMSKx) registers.

Host channel control

• The following host channel controls are available to the application through the host
channel-x characteristics register (OTG_HCCHARx):
– Channel enable/disable
– Program the FS/LS speed of target USB peripheral
– Program the address of target USB peripheral
– Program the endpoint number of target USB peripheral
– Program the transfer IN/OUT direction
– Program the USB transfer type (control, bulk, interrupt, isochronous)
– Program the maximum packet size (MPS)
– Program the periodic transfer to be executed during odd/even frames

Host channel transfer

The host channel transfer size registers (OTG_HCTSIZx) allow the application to program
the transfer size parameters, and read the transfer status. Programming must be done
before setting the channel enable bit in the host channel characteristics register. Once the
endpoint is enabled the packet count field is read-only as the OTG_FS core updates it
according to the current transfer status.
• The following transfer parameters can be programmed:
– transfer size in bytes
– number of packets making up the overall transfer size
– initial data PID

Host channel status/interrupt

The host channel-x interrupt register (OTG_HCINTx) indicates the status of an endpoint
with respect to USB- and AHB-related events. The application must read these register
when the host channels interrupt bit in the core interrupt register (HCINT bit in
OTG_GINTSTS) is set. Before the application can read these registers, it must first read the
host all channels interrupt (OTG_HAINT) register to get the exact channel number for the
host channel-x interrupt register. The application must clear the appropriate bit in this
register to clear the corresponding bits in the OTG_HAINT and OTG_GINTSTS registers.

RM0402 Rev 6 981/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

The mask bits for each interrupt source of each channel are also available in the
OTG_HCINTMSKx register.
• The host core provides the following status checks and interrupt generation:
– Transfer completed interrupt, indicating that the data transfer is complete on both
the application (AHB) and USB sides
– Channel has stopped due to transfer completed, USB transaction error or disable
command from the application
– Associated transmit FIFO is half or completely empty (IN endpoints)
– ACK response received
– NAK response received
– STALL response received
– USB transaction error due to CRC failure, timeout, bit stuff error, false EOP
– Babble error
– frame overrun
– data toggle error

29.7.4 Host scheduler

The host core features a built-in hardware scheduler which is able to autonomously re-order
and manage the USB transaction requests posted by the application. At the beginning of
each frame the host executes the periodic (isochronous and interrupt) transactions first,
followed by the nonperiodic (control and bulk) transactions to achieve the higher level of
priority granted to the isochronous and interrupt transfer types by the USB specification.
The host processes the USB transactions through request queues (one for periodic and one
for nonperiodic). Each request queue can hold up to 8 entries. Each entry represents a
pending transaction request from the application, and holds the IN or OUT channel number
along with other information to perform a transaction on the USB. The order in which the
requests are written to the queue determines the sequence of the transactions on the USB
interface.
At the beginning of each frame, the host processes the periodic request queue first, followed
by the nonperiodic request queue. The host issues an incomplete periodic transfer interrupt
(IPXFR bit in OTG_GINTSTS) if an isochronous or interrupt transaction scheduled for the
current frame is still pending at the end of the current frame. The OTG_FS core is fully
responsible for the management of the periodic and nonperiodic request queues.The
periodic transmit FIFO and queue status register (OTG_HPTXSTS) and nonperiodic
transmit FIFO and queue status register (OTG_HNPTXSTS) are read-only registers which
can be used by the application to read the status of each request queue. They contain:
• The number of free entries currently available in the periodic (nonperiodic) request
queue (8 max)
• Free space currently available in the periodic (nonperiodic) Tx FIFO (out-transactions)
• IN/OUT token, host channel number and other status information.
As request queues can hold a maximum of 8 entries each, the application can push to
schedule host transactions in advance with respect to the moment they physically reach the
SB for a maximum of 8 pending periodic transactions plus 8 pending non-periodic
transactions.
To post a transaction request to the host scheduler (queue) the application must check that
there is at least 1 entry available in the periodic (nonperiodic) request queue by reading the

982/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

PTXQSAV bits in the OTG_HNPTXSTS register or NPTQXSAV bits in the

OTG_HNPTXSTS register.

29.8 OTG_FS SOF trigger

Figure 338. SOF connectivity (SOF trigger output to TIM and ITR1 connection)

STM32

SOF pulse output, to

USB micro-AB connector

external audio control
VBUS

ITR1 SOF pulse

D-

D+

TIM SOFgen ID

VSS

MSv36914V1

The OTG_FS core provides means to monitor, track and configure SOF framing in the host
and peripheral, as well as an SOF pulse output connectivity feature.
Such utilities are especially useful for adaptive audio clock generation techniques, where
the audio peripheral needs to synchronize to the isochronous stream provided by the PC, or
the host needs to trim its framing rate according to the requirements of the audio peripheral.

29.8.1 Host SOFs

In host mode the number of PHY clocks occurring between the generation of two
consecutive SOF (FS) or Keep-alive (LS) tokens is programmable in the host frame interval
register (HFIR), thus providing application control over the SOF framing period. An interrupt
is generated at any start of frame (SOF bit in OTG_GINTSTS). The current frame number
and the time remaining until the next SOF are tracked in the host frame number register
(HFNUM).
A SOF pulse signal, is generated at any SOF starting token and with a width of 20 HCLK
cycles. The SOF pulse is also internally connected to the input trigger of the timer, so that
the input capture feature, the output compare feature and the timer can be triggered by the
SOF pulse.

29.8.2 Peripheral SOFs

In device mode, the start of frame interrupt is generated each time an SOF token is received
on the USB (SOF bit in OTG_GINTSTS). The corresponding frame number can be read
from the device status register (FNSOF bit in OTG_DSTS). A SOF pulse signal with a width
of 20 HCLK cycles is also generated.The SOF pulse signal is also internally connected to
the TIM input trigger, so that the input capture feature, the output compare feature and the
timer can be triggered by the SOF pulse.

RM0402 Rev 6 983/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

The end of periodic frame interrupt (OTG_GINTSTS/EOPF) is used to notify the application
when 80%, 85%, 90% or 95% of the time frame interval elapsed depending on the periodic
frame interval field in the device configuration register (PFIVL bit in OTG_DCFG). This
feature can be used to determine if all of the isochronous traffic for that frame is complete.

29.9 OTG_FS low-power modes

Table 200 below defines the STM32 low power modes and their compatibility with the OTG.

Table 200. Compatibility of STM32 low power modes with the OTG
Mode Description USB compatibility

Required when USB not in

Run MCU fully active
suspend state.
USB suspend exit causes the device to exit Sleep mode. Peripheral Available while USB is in
Sleep
registers content is kept. suspend state.
USB suspend exit causes the device to exit Stop mode. Peripheral Available while USB is in
Stop
registers content is kept(1). suspend state.
Powered-down. The peripheral must be reinitialized after exiting Not compatible with USB
Standby
Standby mode. applications.
1. Within Stop mode there are different possible settings. Some restrictions may also exist, please refer to Section 5: Power
controller (PWR) to understand which (if any) restrictions apply when using OTG.

The following bits and procedures reduce power consumption.

The power consumption of the OTG PHY is controlled by two or three bits in the general
core configuration register, depending on OTG revision supported.
• PHY power down (OTG_GCCFG/PWRDWN)
It switches on/off the full-speed transceiver module of the PHY. It must be preliminarily
set to allow any USB operation
• VBUS detection enable (OTG_GCCFG/VBDEN)
It switches on/off the VBUS sensing comparators associated with OTG operations
Power reduction techniques are available while in the USB suspended state, when the USB
session is not yet valid or the device is disconnected.
• Stop PHY clock (STPPCLK bit in OTG_PCGCCTL)
When setting the stop PHY clock bit in the clock gating control register, most of the
48 MHz clock domain internal to the OTG core is switched off by clock gating. The
dynamic power consumption due to the USB clock switching activity is cut even if the
48 MHz clock input is kept running by the application
Most of the transceiver is also disabled, and only the part in charge of detecting the
asynchronous resume or remote wakeup event is kept alive.
• Gate HCLK (GATEHCLK bit in OTG_PCGCCTL)
When setting the Gate HCLK bit in the clock gating control register, most of the system
clock domain internal to the OTG_FS core is switched off by clock gating. Only the
register read and write interface is kept alive. The dynamic power consumption due to

984/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

the USB clock switching activity is cut even if the system clock is kept running by the
application for other purposes.
• USB system stop
When the OTG_FS is in the USB suspended state, the application may decide to
drastically reduce the overall power consumption by a complete shut down of all the
clock sources in the system. USB System Stop is activated by first setting the Stop
PHY clock bit and then configuring the system deep sleep mode in the power control
system module (PWR).
The OTG_FS core automatically reactivates both system and USB clocks by
asynchronous detection of remote wakeup (as an host) or resume (as a device)
signaling on the USB.
To save dynamic power, the USB data FIFO is clocked only when accessed by the OTG_FS
core.

29.10 OTG_FS Dynamic update of the OTG_HFIR register

The USB core embeds a dynamic trimming capability of SOF framing period in host mode
allowing to synchronize an external device with the SOF frames.
When the OTG_HFIR register is changed within a current SOF frame, the SOF period
correction is applied in the next frame as described in Figure 339.
For a dynamic update, it is required to set RLDCTRL=1.

Figure 339. Updating OTG_HFIR dynamically (RLDCTRL = 1)

SOF reload

OTG_HFIR write

OTG_HFIR value 400 450

... ... ... ...

400
399

400
399

450
449

450
449

Frame timer
1
0

1
0

1
0

1
0

ai18440b

29.11 OTG_FS data FIFOs

The USB system features 1.25 Kbytes of dedicated RAM with a sophisticated FIFO control
mechanism. The packet FIFO controller module in the OTG_FS core organizes RAM space
into Tx FIFOs into which the application pushes the data to be temporarily stored before the
USB transmission, and into a single Rx FIFO where the data received from the USB are
temporarily stored before retrieval (popped) by the application. The number of instructed
FIFOs and how these are organized inside the RAM depends on the device’s role. In
peripheral mode an additional Tx FIFO is instructed for each active IN endpoint. Any FIFO
size is software configured to better meet the application requirements.

RM0402 Rev 6 985/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.11.1 Peripheral FIFO architecture

Figure 340. Device-mode FIFO address mapping and AHB FIFO access mapping
Single data
FIFO
IN endpoint Tx FIFO #x Dedicated Tx OTG_DIEPTXFx[31:16]
Tx FIFO #x
DFIFO push access FIFO #x control
packet
from AHB (optional) OTG_DIEPTXFx[15:0]
MAC pop
. . .
. . .
. . .

IN endpoint Tx FIFO #1 Dedicated Tx OTG_DIEPTXF1[31:16]

Tx FIFO #1
DFIFO push access FIFO #1 control
packet
from AHB (optional) OTG_DIEPTXF1[15:0]
MAC pop

IN endpoint Tx FIFO #0 Dedicated Tx OTG_DIEPTXF0[31:16]

Tx FIFO #0
DFIFO push access FIFO #0 control
packet
from AHB (optional) OTG_DIEPTXF0[15:0]
MAC pop

Any OUT endpoint Dedicated Tx

DFIFO pop access FIFO #1 control Rx packets OTG_GRXFSIZ[15:0]
from AHB (optional)

MAC push
A1=0 (Rx start address fixed
to 0)

MSv36929V1

Peripheral Rx FIFO
The OTG peripheral uses a single receive FIFO that receives the data directed to all OUT
endpoints. Received packets are stacked back-to-back until free space is available in the Rx
FIFO. The status of the received packet (which contains the OUT endpoint destination
number, the byte count, the data PID and the validity of the received data) is also stored by
the core on top of the data payload. When no more space is available, host transactions are
NACKed and an interrupt is received on the addressed endpoint. The size of the receive
FIFO is configured in the receive FIFO size register (OTG_GRXFSIZ).
The single receive FIFO architecture makes it more efficient for the USB peripheral to fill in
the receive RAM buffer:
• All OUT endpoints share the same RAM buffer (shared FIFO)
• The OTG_FS core can fill in the receive FIFO up to the limit for any host sequence of
OUT tokens
The application keeps receiving the Rx FIFO non-empty interrupt (RXFLVL bit in
OTG_GINTSTS) as long as there is at least one packet available for download. It reads the
packet information from the receive status read and pop register (OTG_GRXSTSP) and
finally pops data off the receive FIFO by reading from the endpoint-related pop address.

986/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Peripheral Tx FIFOs
The core has a dedicated FIFO for each IN endpoint. The application configures FIFO sizes
by writing the endpoint 0 transmit FIFO size register (OTG_DIEPTXF0) for IN endpoint0 and
the device IN endpoint transmit FIFOx registers (OTG_DIEPTXFx) for IN endpoint-x.

29.11.2 Host FIFO architecture

Figure 341. Host-mode FIFO address mapping and AHB FIFO access mapping
Single data
FIFO

Periodic Tx
OTG_HPTXFSIZ[31:16]
Any periodic channel packets
Periodic Tx FIFO
DFIFO push access
control (optional)
from AHB OTG_HPTXFSIZ[15:0]
MAC pop
Non-periodic
OTG_HNPTXFSIZ[31:16]
Any non-periodic Tx packets
Non-periodic Tx
channel DFIFO push
FIFO control
access from AHB OTG_HNPTXFSIZ[15:0]
MAC pop

Rx packets OTG_GRXFSIZ[15:0]
Any channel DFIFO pop
Rx FIFO control
access from AHB
Rx start address fixed to 0
A1=0
MAC push

MSv36930V1

Host Rx FIFO
The host uses one receiver FIFO for all periodic and nonperiodic transactions. The FIFO is
used as a receive buffer to hold the received data (payload of the received packet) from the
USB until it is transferred to the system memory. Packets received from any remote IN
endpoint are stacked back-to-back until free space is available. The status of each received
packet with the host channel destination, byte count, data PID and validity of the received
data are also stored into the FIFO. The size of the receive FIFO is configured in the receive
FIFO size register (OTG_GRXFSIZ).
The single receive FIFO architecture makes it highly efficient for the USB host to fill in the
receive data buffer:
• All IN configured host channels share the same RAM buffer (shared FIFO)
• The OTG_FS core can fill in the receive FIFO up to the limit for any sequence of IN
tokens driven by the host software
The application receives the Rx FIFO not-empty interrupt as long as there is at least one
packet available for download. It reads the packet information from the receive status read
and pop register and finally pops the data off the receive FIFO.

RM0402 Rev 6 987/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Host Tx FIFOs
The host uses one transmit FIFO for all non-periodic (control and bulk) OUT transactions
and one transmit FIFO for all periodic (isochronous and interrupt) OUT transactions. FIFOs
are used as transmit buffers to hold the data (payload of the transmit packet) to be
transmitted over the USB. The size of the periodic (nonperiodic) Tx FIFO is configured in the
host periodic (nonperiodic) transmit FIFO size OTG_HPTXFSIZ / OTG_HNPTXFSIZ)
register.
The two Tx FIFO implementation derives from the higher priority granted to the periodic type
of traffic over the USB frame. At the beginning of each frame, the built-in host scheduler
processes the periodic request queue first, followed by the nonperiodic request queue.
The two transmit FIFO architecture provides the USB host with separate optimization for
periodic and nonperiodic transmit data buffer management:
• All host channels configured to support periodic (nonperiodic) transactions in the OUT
direction share the same RAM buffer (shared FIFOs)
• The OTG_FS core can fill in the periodic (nonperiodic) transmit FIFO up to the limit for
any sequence of OUT tokens driven by the host software
The OTG_FS core issues the periodic Tx FIFO empty interrupt (PTXFE bit in
OTG_GINTSTS) as long as the periodic Tx FIFO is half or completely empty, depending on
the value of the periodic Tx FIFO empty level bit in the AHB configuration register
(PTXFELVL bit in OTG_GAHBCFG). The application can push the transmission data in
advance as long as free space is available in both the periodic Tx FIFO and the periodic
request queue. The host periodic transmit FIFO and queue status register
(OTG_HPTXSTS) can be read to know how much space is available in both.
OTG_FS core issues the non periodic Tx FIFO empty interrupt (NPTXFE bit in
OTG_GINTSTS) as long as the nonperiodic Tx FIFO is half or completely empty depending
on the non periodic Tx FIFO empty level bit in the AHB configuration register (TXFELVL bit
in OTG_GAHBCFG). The application can push the transmission data as long as free space
is available in both the nonperiodic Tx FIFO and nonperiodic request queue. The host
nonperiodic transmit FIFO and queue status register (OTG_HNPTXSTS) can be read to
know how much space is available in both.

29.11.3 FIFO RAM allocation

Device mode
Receive FIFO RAM allocation: the application should allocate RAM for SETUP packets:
• 10 locations must be reserved in the receive FIFO to receive SETUP packets on
control endpoint. The core does not use these locations, which are reserved for SETUP
packets, to write any other data.
• One location is to be allocated for Global OUT NAK.
• Status information is written to the FIFO along with each received packet. Therefore, a
minimum space of (largest packet size / 4) + 1 must be allocated to receive packets. If
multiple isochronous endpoints are enabled, then at least two (largest packet size / 4) +
1 spaces must be allocated to receive back-to-back packets. Typically, two (largest
packet size / 4) + 1 spaces are recommended so that when the previous packet is
being transferred to the CPU, the USB can receive the subsequent packet.
• Along with the last packet for each endpoint, transfer complete status information is
also pushed to the FIFO. One location for each OUT endpoint is recommended.

988/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Device RxFIFO =
(5 * number of control endpoints + 8) + ((largest USB packet used / 4) + 1 for status
information) + (2 * number of OUT endpoints) + 1 for Global NAK
Example: The MPS is 1,024 bytes for a periodic USB packet and 512 bytes for a non-
periodic USB packet. There are three OUT endpoints, three IN endpoints, one control
endpoint, and three host channels.
Device RxFIFO = (5 * 1 + 8) + ((1,024 / 4) +1) + (2 * 4) + 1 = 279
Transmit FIFO RAM allocation: the minimum RAM space required for each IN endpoint
Transmit FIFO is the maximum packet size for that particular IN endpoint.
Note: More space allocated in the transmit IN endpoint FIFO results in better performance on the
USB.

Host mode
Receive FIFO RAM allocation:
Status information is written to the FIFO along with each received packet. Therefore, a
minimum space of (largest packet size / 4) + 1 must be allocated to receive packets. If
multiple isochronous channels are enabled, then at least two (largest packet size / 4) + 1
spaces must be allocated to receive back-to-back packets. Typically, two (largest packet
size / 4) + 1 spaces are recommended so that when the previous packet is being transferred
to the CPU, the USB can receive the subsequent packet.
Along with the last packet in the host channel, transfer complete status information is also
pushed to the FIFO. So one location must be allocated for this.
Host RxFIFO = (largest USB packet used / 4) + 1 for status information + 1 transfer
complete
Example: Host RxFIFO = ((1,024 / 4) + 1) + 1 = 258
Transmit FIFO RAM allocation:
The minimum amount of RAM required for the host Non-periodic Transmit FIFO is the
largest maximum packet size among all supported non-periodic OUT channels.
Typically, two largest packet sizes worth of space is recommended, so that when the current
packet is under transfer to the USB, the CPU can get the next packet.
Non-Periodic TxFIFO = largest non-periodic USB packet used / 4
Example: Non-Periodic TxFIFO = (512 / 4) = 128
The minimum amount of RAM required for host periodic Transmit FIFO is the largest
maximum packet size out of all the supported periodic OUT channels. If there is at least one
isochronous OUT endpoint, then the space must be at least two times the maximum packet
size of that channel.
Host Periodic TxFIFO = largest periodic USB packet used / 4
Example: Host Periodic TxFIFO = (1,024 / 4) = 256
Note: More space allocated in the Transmit Non-periodic FIFO results in better performance on
the USB.

RM0402 Rev 6 989/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.12 OTG_FS system performance

Best USB and system performance is achieved owing to the large RAM buffers, the highly
configurable FIFO sizes, the quick 32-bit FIFO access through AHB push/pop registers and,
especially, the advanced FIFO control mechanism. Indeed, this mechanism allows the
OTG_FS to fill in the available RAM space at best regardless of the current USB sequence.
With these features:
• The application gains good margins to calibrate its intervention in order to optimize the
CPU bandwidth usage:
– It can accumulate large amounts of transmission data in advance compared to
when they are effectively sent over the USB
– It benefits of a large time margin to download data from the single receive FIFO
• The USB core is able to maintain its full operating rate, that is to provide maximum full-
speed bandwidth with a great margin of autonomy versus application intervention:
– It has a large reserve of transmission data at its disposal to autonomously manage
the sending of data over the USB
– It has a lot of empty space available in the receive buffer to autonomously fill it in
with the data coming from the USB
As the OTG_FS core is able to fill in the 1.25-Kbyte RAM buffer very efficiently, and as 1.25-
Kbyte of transmit/receive data is more than enough to cover a full speed frame, the USB
system is able to withstand the maximum full-speed data rate for up to one USB frame
(1 ms) without any CPU intervention.

29.13 OTG_FS interrupts

When the OTG_FS controller is operating in one mode, either device or host, the application
must not access registers from the other mode. If an illegal access occurs, a mode
mismatch interrupt is generated and reflected in the core interrupt register (MMIS bit in the
OTG_GINTSTS register). When the core switches from one mode to the other, the registers
in the new mode of operation must be reprogrammed as they would be after a power-on
reset.
Figure 342 shows the interrupt hierarchy.

990/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 342. Interrupt hierarchy

Wakeup interrupt
(1) OTG_FS_WKUP

Global interrupt
OTG_FS
AND

OR

Global interrupt mask (bit 0)

T

IEP P

NT
IN

OTG_AHBCFG

T
PIN
T
RT

IN
IN

GI
AND AHB configuration register
HC

OE
HP

OT
OTG_GINTSTS
Core register interrupt
31:26 25 24 23:20 19 18 17:3 2 1:0
OTG_GINTMSK
Core interrupt mask register

OTG_GOTGINT
OTG interrupt register

(15 + #EP):16 (#EP-1):0

OUT endpoints IN endpoints
OTG_DAINTMSK
Device all endpoints interrupt mask
register
OTG_DAINT
Device all endpoints interrupt register

OTG_DIEPMSK/
OTG_DOEPMSK
Device IN/OUT endpoints common
interrupt mask register

x=0
OTG_DIEPINTx/
...

OTG_DOEPINTx x = #EP-1
Device IN/OUT endpoint interrupt
registers

OTG_HPRT
Host port control and status register

OTG_HAINTMSK
Host all channels interrupt mask register
OTG_HAINT
Host all channels interrupt register

x=0
OTG_HCTINTMSKx
...

Host channels interrupt mask registers x = #HC-1

OTG_HCTINTx
Host channels interrupt registers

MSv36921V4

1. OTG_FS_WKUP becomes active (high state) when resume condition occurs during L1 SLEEP or L2 SUSPEND states.

RM0402 Rev 6 991/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.14 OTG_FS control and status registers

By reading from and writing to the control and status registers (CSRs) through the AHB
slave interface, the application controls the OTG_FS controller. These registers are 32 bits
wide, and the addresses are 32-bit block aligned. The OTG_FS registers must be accessed
by words (32 bits).
CSRs are classified as follows:
• Core global registers
• Host-mode registers
• Host global registers
• Host port CSRs
• Host channel-specific registers
• Device-mode registers
• Device global registers
• Device endpoint-specific registers
• Power and clock-gating registers
• Data FIFO (DFIFO) access registers
Only the core global, power and clock-gating, data FIFO access, and host port control and
status registers can be accessed in both host and device modes. When the OTG_FS
controller is operating in one mode, either device or host, the application must not access
registers from the other mode. If an illegal access occurs, a mode mismatch interrupt is
generated and reflected in the core interrupt register (MMIS bit in the OTG_GINTSTS
register). When the core switches from one mode to the other, the registers in the new mode
of operation must be reprogrammed as they would be after a power-on reset.

29.14.1 CSR memory map

The host and device mode registers occupy different addresses. All registers are
implemented in the AHB clock domain.

Global CSR map

These registers are available in both host and device modes.

Table 201. Core global control and status registers (CSRs)

Address
Acronym Register name
offset

OTG_GOTGCTL 0x000 Section 29.15.1: OTG control and status register (OTG_GOTGCTL)

OTG_GOTGINT 0x004 Section 29.15.2: OTG interrupt register (OTG_GOTGINT)

OTG_GAHBCFG 0x008 Section 29.15.3: OTG AHB configuration register (OTG_GAHBCFG)

OTG_GUSBCFG 0x00C Section 29.15.4: OTG USB configuration register (OTG_GUSBCFG)

OTG_GRSTCTL 0x010 Section 29.15.5: OTG reset register (OTG_GRSTCTL)

OTG_GINTSTS 0x014 Section 29.15.6: OTG core interrupt register (OTG_GINTSTS)

OTG_GINTMSK 0x018 Section 29.15.7: OTG interrupt mask register (OTG_GINTMSK)

992/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Table 201. Core global control and status registers (CSRs) (continued)
Address
Acronym Register name
offset

Section 29.15.8: OTG receive status debug read register (OTG_GRXSTSR)

OTG_GRXSTSR 0x01C Section 29.15.9: OTG receive status debug read [alternate]
(OTG_GRXSTSR)

Section 29.15.10: OTG status read and pop registers (OTG_GRXSTSP)

OTG_GRXSTSP 0x020 Section 29.15.11: OTG status read and pop registers [alternate]
(OTG_GRXSTSP)

OTG_GRXFSIZ 0x024 Section 29.15.12: OTG receive FIFO size register (OTG_GRXFSIZ)

OTG_HNPTXFSIZ/ Section 29.15.13: OTG host non-periodic transmit FIFO size register
0x028
OTG_DIEPTXF0(1) (OTG_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size (OTG_DIEPTXF0)

Section 29.15.14: OTG non-periodic transmit FIFO/queue status register

OTG_HNPTXSTS 0x02C
(OTG_HNPTXSTS)

OTG_GCCFG 0x038 Section 29.15.15: OTG general core configuration register (OTG_GCCFG)

OTG_CID 0x03C Section 29.15.16: OTG core ID register (OTG_CID)

OTG_GLPMCFG 0x54 Section 29.15.17: OTG core LPM configuration register (OTG_GLPMCFG)

Section 29.15.18: OTG host periodic transmit FIFO size register

OTG_HPTXFSIZ 0x100
(OTG_HPTXFSIZ)
0x104
0x108 Section 29.15.19: OTG device IN endpoint transmit FIFO x size register
OTG_DIEPTXFx
... (OTG_DIEPTXFx)
0x114
1. The general rule is to use OTG_HNPTXFSIZ for host mode and OTG_DIEPTXF0 for device mode.

Host-mode CSR map

These registers must be programmed every time the core changes to host mode.

Table 202. Host-mode control and status registers (CSRs)

Offset
Acronym Register name
address

OTG_HCFG 0x400 Section 29.15.21: OTG host configuration register (OTG_HCFG)

OTG_HFIR 0x404 Section 29.15.22: OTG host frame interval register (OTG_HFIR)

Section 29.15.23: OTG host frame number/frame time remaining register

OTG_HFNUM 0x408
(OTG_HFNUM)

Section 29.15.24: OTG_Host periodic transmit FIFO/queue status register

OTG_HPTXSTS 0x410
(OTG_HPTXSTS)

OTG_HAINT 0x414 Section 29.15.25: OTG host all channels interrupt register (OTG_HAINT)

Section 29.15.26: OTG host all channels interrupt mask register

OTG_HAINTMSK 0x418
(OTG_HAINTMSK)

RM0402 Rev 6 993/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Table 202. Host-mode control and status registers (CSRs) (continued)

Offset
Acronym Register name
address

OTG_HPRT 0x440 Section 29.15.27: OTG host port control and status register (OTG_HPRT)

0x500
0x520 Section 29.15.28: OTG host channel x characteristics register
OTG_HCCHARx
... (OTG_HCCHARx)
0x660
0x508
0x528
OTG_HCINTx Section 29.15.29: OTG host channel x interrupt register (OTG_HCINTx)
....
0x668
0x50C
0x52C Section 29.15.30: OTG host channel x interrupt mask register
OTG_HCINTMSKx
.... (OTG_HCINTMSKx)
0x66C
0x510
0x530 Section 29.15.31: OTG host channel x transfer size register
OTG_HCTSIZx
.... (OTG_HCTSIZx)
0x670

Device-mode CSR map

These registers must be programmed every time the core changes to device mode.

Table 203. Device-mode control and status registers

Offset
Acronym Register name
address

OTG_DCFG 0x800 Section 29.15.33: OTG device configuration register (OTG_DCFG)

OTG_DCTL 0x804 Section 29.15.34: OTG device control register (OTG_DCTL)

OTG_DSTS 0x808 Section 29.15.35: OTG device status register (OTG_DSTS)

Section 29.15.36: OTG device IN endpoint common interrupt mask

OTG_DIEPMSK 0x810
register (OTG_DIEPMSK)

Section 29.15.37: OTG device OUT endpoint common interrupt mask

OTG_DOEPMSK 0x814
register (OTG_DOEPMSK)

Section 29.15.38: OTG device all endpoints interrupt register

OTG_DAINT 0x818
(OTG_DAINT)

Section 29.15.39: OTG all endpoints interrupt mask register

OTG_DAINTMSK 0x81C
(OTG_DAINTMSK)

Section 29.15.40: OTG device VBUS discharge time register

OTG_DVBUSDIS 0x828
(OTG_DVBUSDIS)
Section 29.15.41: OTG device VBUS pulsing time register
OTG_DVBUSPULSE 0x82C
(OTG_DVBUSPULSE)

994/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Table 203. Device-mode control and status registers (continued)

Offset
Acronym Register name
address

Section 29.15.42: OTG device IN endpoint FIFO empty interrupt mask

OTG_DIEPEMPMSK 0x834
register (OTG_DIEPEMPMSK)

Section 29.15.43: OTG device control IN endpoint 0 control register

OTG_DIEPCTL0 0x900
(OTG_DIEPCTL0)

0x920
0x940 Section 29.15.44: OTG device IN endpoint x control register
OTG_DIEPCTLx
... (OTG_DIEPCTLx)
0x9A0
0x908
0x928 Section 29.15.45: OTG device IN endpoint x interrupt register
OTG_DIEPINTx
.... (OTG_DIEPINTx)
0x988

Section 29.15.46: OTG device IN endpoint 0 transfer size register

OTG_DIEPTSIZ0 0x910
(OTG_DIEPTSIZ0)

0x918
0x938 Section 29.15.47: OTG device IN endpoint transmit FIFO status register
OTG_DTXFSTSx
.... (OTG_DTXFSTSx)
0x998
0x930
0x950 Section 29.15.48: OTG device IN endpoint x transfer size register
OTG_DIEPTSIZx
... (OTG_DIEPTSIZx)
0x9B0

Section 29.15.49: OTG device control OUT endpoint 0 control register

OTG_DOEPCTL0 0xB00
(OTG_DOEPCTL0)

0xB08
0xB28 Section 29.15.50: OTG device OUT endpoint x interrupt register
OTG_DOEPINTx
... (OTG_DOEPINTx)
0xBA8

Section 29.15.51: OTG device OUT endpoint 0 transfer size register

OTG_DOEPTSIZ0 0xB10
(OTG_DOEPTSIZ0)

0xB20
0xB40 Section 29.15.52: OTG device OUT endpoint x control register
OTG_DOEPCTLx
... (OTG_DOEPCTLx)
0xBA0
0xB30
0xB50 Section 29.15.53: OTG device OUT endpoint x transfer size register
OTG_DOEPTSIZx
... (OTG_DOEPTSIZx)
0xBB0

RM0402 Rev 6 995/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Data FIFO (DFIFO) access register map

These registers, available in both host and device modes, are used to read or write the FIFO
space for a specific endpoint or a channel, in a given direction. If a host channel is of type
IN, the FIFO can only be read on the channel. Similarly, if a host channel is of type OUT, the
FIFO can only be written on the channel.

Table 204. Data FIFO (DFIFO) access register map

FIFO access register section Offset address Access

Device IN endpoint 0/Host OUT Channel 0: DFIFO write access w

0x1000–0x1FFC
Device OUT endpoint 0/Host IN Channel 0: DFIFO read access r

Device IN endpoint 1/Host OUT Channel 1: DFIFO write access w

0x2000–0x2FFC
Device OUT endpoint 1/Host IN Channel 1: DFIFO read access r

... ... ...

Device IN endpoint x(1)/Host OUT Channel x(1): DFIFO write access w

0xX000–0xXFFC
Device OUT endpoint x(1)/Host IN Channel x(1): DFIFO read access r

1. Where x is 5in device mode and 11 in host mode.

Power and clock gating CSR map

There is a single register for power and clock gating. It is available in both host and device
modes.

Table 205. Power and clock gating control and status registers
Acronym Offset address Register name

Section 29.15.54: OTG power and clock gating control

OTG_PCGCCTL 0xE00–0xE04
register (OTG_PCGCCTL)

29.15 OTG_FS registers

These registers are available in both host and device modes, and do not need to be
reprogrammed when switching between these modes.
Bit values in the register descriptions are expressed in binary unless otherwise specified.

996/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.1 OTG control and status register (OTG_GOTGCTL)

Address offset: 0x000
Reset value: 0x0001 0000
The OTG_GOTGCTL register controls the behavior and reflects the status of the OTG
function of the core.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
CUR OTG CID
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. BSVLD ASVLD DBCT
MOD VER STS
r rw r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DHNP HSHNP HNP HNG BVALO BVALO AVALO AVALO VBVAL VBVAL SRQ
Res. Res. Res. EHEN SRQ
EN EN RQ SCS VAL EN VAL EN OVAL OEN SCS
rw rw rw rw r rw rw rw rw rw rw rw r

Bits 31:22 Reserved, must be kept at reset value.

Bit 21 CURMOD: Current mode of operation
Indicates the current mode (host or device).
0: Device mode
1: Host mode
Bit 20 OTGVER: OTG version
Selects the OTG revision.
0:OTG Version 1.3. OTG1.3 is obsolete for new product development.
1:OTG Version 2.0. In this version the core supports only data line pulsing for SRP.
Bit 19 BSVLD: B-session valid
Indicates the device mode transceiver status.
0: B-session is not valid.
1: B-session is valid.
In OTG mode, the user can use this bit to determine if the device is connected or
disconnected.
Note: Only accessible in device mode.
Bit 18 ASVLD: A-session valid
Indicates the host mode transceiver status.
0: A-session is not valid
1: A-session is valid
Note: Only accessible in host mode.
Bit 17 DBCT: Long/short debounce time
Indicates the debounce time of a detected connection.
0: Long debounce time, used for physical connections (100 ms + 2.5 µs)
1: Short debounce time, used for soft connections (2.5 µs)
Note: Only accessible in host mode.
Bit 16 CIDSTS: Connector ID status
Indicates the connector ID status on a connect event.
0: The OTG_FS controller is in A-device mode
1: The OTG_FS controller is in B-device mode
Note: Accessible in both device and host modes.

RM0402 Rev 6 997/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 15:13 Reserved, must be kept at reset value.

Bit 12 EHEN: Embedded host enable
It is used to select between OTG A device state machine and embedded host state machine.
0: OTG A device state machine is selected
1: Embedded host state machine is selected
Bit 11 DHNPEN: Device HNP enabled
The application sets this bit when it successfully receives a SetFeature.SetHNPEnable
command from the connected USB host.
0: HNP is not enabled in the application
1: HNP is enabled in the application
Note: Only accessible in device mode.
Bit 10 HSHNPEN: host set HNP enable
The application sets this bit when it has successfully enabled HNP (using the
SetFeature.SetHNPEnable command) on the connected device.
0: Host Set HNP is not enabled
1: Host Set HNP is enabled
Note: Only accessible in host mode.
Bit 9 HNPRQ: HNP request
The application sets this bit to initiate an HNP request to the connected USB host. The
application can clear this bit by writing a 0 when the host negotiation success status change
bit in the OTG_GOTGINT register (HNSSCHG bit in OTG_GOTGINT) is set. The core clears
this bit when the HNSSCHG bit is cleared.
0: No HNP request
1: HNP request
Note: Only accessible in device mode.
Bit 8 HNGSCS: Host negotiation success
The core sets this bit when host negotiation is successful. The core clears this bit when the
HNP request (HNPRQ) bit in this register is set.
0: Host negotiation failure
1: Host negotiation success
Note: Only accessible in device mode.
Bit 7 BVALOVAL: B-peripheral session valid override value.
This bit is used to set override value for Bvalid signal when BVALOEN bit is set.
0: Bvalid value is '0' when BVALOEN = 1
1: Bvalid value is '1' when BVALOEN = 1
Note: Only accessible in device mode.
Bit 6 BVALOEN: B-peripheral session valid override enable.
This bit is used to enable/disable the software to override the Bvalid signal using the
BVALOVAL bit.
0:Override is disabled and Bvalid signal from the respective PHY selected is used internally
by the core
1:Internally Bvalid received from the PHY is overridden with BVALOVAL bit value
Note: Only accessible in device mode.
Bit 5 AVALOVAL: A-peripheral session valid override value.
This bit is used to set override value for Avalid signal when AVALOEN bit is set.
0: Avalid value is '0' when AVALOEN = 1
1: Avalid value is '1' when AVALOEN = 1
Note: Only accessible in host mode.

998/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 4 AVALOEN: A-peripheral session valid override enable.

This bit is used to enable/disable the software to override the Avalid signal using the
AVALOVAL bit.
0:Override is disabled and Avalid signal from the respective PHY selected is used internally
by the core
1:Internally Avalid received from the PHY is overridden with AVALOVAL bit value
Note: Only accessible in host mode.
Bit 3 VBVALOVAL: VBUS valid override value.
This bit is used to set override value for vbusvalid signal when VBVALOEN bit is set.
0: vbusvalid value is '0' when VBVALOEN = 1
1: vbusvalid value is '1' when VBVALOEN = 1
Note: Only accessible in host mode.
Bit 2 VBVALOEN: VBUS valid override enable.
This bit is used to enable/disable the software to override the vbusvalid signal using the
VBVALOVAL bit.
0: Override is disabled and vbusvalid signal from the respective PHY selected is used
internally by the core
1: Internally vbusvalid received from the PHY is overridden with VBVALOVAL bit value
Note: Only accessible in host mode.
Bit 1 SRQ: Session request
The application sets this bit to initiate a session request on the USB. The application can
clear this bit by writing a 0 when the host negotiation success status change bit in the
OTG_GOTGINT register (HNSSCHG bit in OTG_GOTGINT) is set. The core clears this bit
when the HNSSCHG bit is cleared.
If the user uses the USB 1.1 full-speed serial transceiver interface to initiate the session
request, the application must wait until VBUS discharges to 0.2 V, after the B-session valid bit
in this register (BSVLD bit in OTG_GOTGCTL) is cleared.
0: No session request
1: Session request
Note: Only accessible in device mode.
Bit 0 SRQSCS: Session request success
The core sets this bit when a session request initiation is successful.
0: Session request failure
1: Session request success
Note: Only accessible in device mode.

RM0402 Rev 6 999/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.2 OTG interrupt register (OTG_GOTGINT)

Address offset: 0x04
Reset value: 0x0000 0000
The application reads this register whenever there is an OTG interrupt and clears the bits in
this register to clear the OTG interrupt.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DBC ADTO HNG
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
DNE CHG DET

rc_w1 rc_w1 rc_w1

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HNSS SRSS
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. SEDET Res. Res.
CHG CHG

rc_w1 rc_w1 rc_w1

Bits 31:20 Reserved, must be kept at reset value.

Bit 19 DBCDNE: Debounce done
The core sets this bit when the debounce is completed after the device connect. The
application can start driving USB reset after seeing this interrupt. This bit is only valid when
the HNP Capable or SRP Capable bit is set in the OTG_GUSBCFG register (HNPCAP bit or
SRPCAP bit in OTG_GUSBCFG, respectively).
Note: Only accessible in host mode.
Bit 18 ADTOCHG: A-device timeout change
The core sets this bit to indicate that the A-device has timed out while waiting for the B-device
to connect.
Note: Accessible in both device and host modes.
Bit 17 HNGDET: Host negotiation detected
The core sets this bit when it detects a host negotiation request on the USB.
Note: Accessible in both device and host modes.
Bits 16:10 Reserved, must be kept at reset value.
Bit 9 HNSSCHG: Host negotiation success status change
The core sets this bit on the success or failure of a USB host negotiation request. The
application must read the host negotiation success bit of the OTG_GOTGCTL register
(HNGSCS bit in OTG_GOTGCTL) to check for success or failure.
Note: Accessible in both device and host modes.
Bits 7:3 Reserved, must be kept at reset value.
Bit 8 SRSSCHG: Session request success status change
The core sets this bit on the success or failure of a session request. The application must
read the session request success bit in the OTG_GOTGCTL register (SRQSCS bit in
OTG_GOTGCTL) to check for success or failure.
Note: Accessible in both device and host modes.
Bit 2 SEDET: Session end detected
The core sets this bit to indicate that the level of the voltage on VBUS is no longer valid for a
B-Peripheral session when VBUS < 0.8 V.
Note: Accessible in both device and host modes.
Bits 1:0 Reserved, must be kept at reset value.

1000/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.3 OTG AHB configuration register (OTG_GAHBCFG)

Address offset: 0x008
Reset value: 0x0000 0000
This register can be used to configure the core after power-on or a change in mode. This
register mainly contains AHB system-related configuration parameters. Do not change this
register after the initial programming. The application must program this register before
starting any transactions on either the AHB or the USB.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PTXFE TXFE GINT
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
LVL LVL MSK

rw rw rw

Bits 31:9 Reserved, must be kept at reset value.

Bit 8 PTXFELVL: Periodic Tx FIFO empty level
Indicates when the periodic Tx FIFO empty interrupt bit in the OTG_GINTSTS register
(PTXFE bit in OTG_GINTSTS) is triggered.
0: PTXFE (in OTG_GINTSTS) interrupt indicates that the Periodic Tx FIFO is half empty
1: PTXFE (in OTG_GINTSTS) interrupt indicates that the Periodic Tx FIFO is completely
empty
Note: Only accessible in host mode.
Bit 7 TXFELVL: Tx FIFO empty level
In device mode, this bit indicates when IN endpoint Transmit FIFO empty interrupt (TXFE in
OTG_DIEPINTx) is triggered:
0:The TXFE (in OTG_DIEPINTx) interrupt indicates that the IN endpoint Tx FIFO is half
empty
1:The TXFE (in OTG_DIEPINTx) interrupt indicates that the IN endpoint Tx FIFO is
completely empty
In host mode, this bit indicates when the nonperiodic Tx FIFO empty interrupt (NPTXFE bit in
OTG_GINTSTS) is triggered:
0:The NPTXFE (in OTG_GINTSTS) interrupt indicates that the nonperiodic Tx FIFO is half
empty
1:The NPTXFE (in OTG_GINTSTS) interrupt indicates that the nonperiodic Tx FIFO is
completely empty
Bits 6:1 Reserved, must be kept at reset value.
Bit 0 GINTMSK: Global interrupt mask
The application uses this bit to mask or unmask the interrupt line assertion to itself.
Irrespective of this bit’s setting, the interrupt status registers are updated by the core.
0: Mask the interrupt assertion to the application.
1: Unmask the interrupt assertion to the application.
Note: Accessible in both device and host modes.

RM0402 Rev 6 1001/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.4 OTG USB configuration register (OTG_GUSBCFG)

Address offset: 0x00C
Reset value: 0x0000 1440
This register can be used to configure the core after power-on or a changing to host mode
or device mode. It contains USB and USB-PHY related configuration parameters. The
application must program this register before starting any transactions on either the AHB or
the USB. Do not make changes to this register after the initial programming.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
FD FH
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
MOD MOD

rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
HNP SRP PHY
Res. Res. TRDT Res. Res. Res. Res. TOCAL
CAP CAP SEL

rw rw rw rw rw rw r rw rw rw

Bit 31 Reserved, must be kept at reset value.

Bit 30 FDMOD: Force device mode
Writing a 1 to this bit, forces the core to device mode irrespective of the OTG_ID input pin.
0: Normal mode
1: Force device mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Note: Accessible in both device and host modes.
Bit 29 FHMOD: Force host mode
Writing a 1 to this bit, forces the core to host mode irrespective of the OTG_ID input pin.
0: Normal mode
1: Force host mode
After setting the force bit, the application must wait at least 25 ms before the change takes
effect.
Note: Accessible in both device and host modes.
Bits 28:26 Reserved, must be kept at reset value.
Bit 22 Reserved, must be kept at reset value.
Bit 15 Reserved, must be kept at reset value.
Bit 14 Reserved, must be kept at reset value.
Bits 13:10 TRDT[3:0]: USB turnaround time
These bits allows to set the turnaround time in PHY clocks. They must be configured
according to Table 206: TRDT values (FS), depending on the application AHB frequency.
Higher TRDT values allow stretching the USB response time to IN tokens in order to
compensate for longer AHB read access latency to the data FIFO.
Note: Only accessible in device mode.

1002/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 9 HNPCAP: HNP-capable

The application uses this bit to control the OTG_FS controller’s HNP capabilities.
0: HNP capability is not enabled.
1: HNP capability is enabled.
Note: Accessible in both device and host modes.
Bit 8 SRPCAP: SRP-capable
The application uses this bit to control the OTG_FS controller’s SRP capabilities. If the core
operates as a non-SRP-capable
B-device, it cannot request the connected A-device (host) to activate VBUS and start a
session.
0: SRP capability is not enabled.
1: SRP capability is enabled.
Note: Accessible in both device and host modes.
Bit 7 Reserved, must be kept at reset value.
Bit 6 PHYSEL: Full Speed serial transceiver mode select
This bit is always 1 with read-only access.
Bit 5 Reserved, must be kept at reset value.
Bit 4 Reserved, must be kept at reset value.
Bit 3 Reserved, must be kept at reset value.
Bits 2:0 TOCAL[2:0]: FS timeout calibration
The number of PHY clocks that the application programs in this field is added to the full-
speed interpacket timeout duration in the core to account for any additional delays
introduced by the PHY. This can be required, because the delay introduced by the PHY in
generating the line state condition can vary from one PHY to another.
The USB standard timeout value for full-speed operation is 16 to 18 (inclusive) bit times. The
application must program this field based on the speed of enumeration. The number of bit
times added per PHY clock is 0.25 bit times.

Table 206. TRDT values (FS)

AHB frequency range (MHz)
TRDT minimum value
Min Max

14.2 15 0xF
15 16 0xE
16 17.2 0xD
17.2 18.5 0xC
18.5 20 0xB
20 21.8 0xA
21.8 24 0x9
24 27.5 0x8
27.5 32 0x7
32 - 0x6

RM0402 Rev 6 1003/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.5 OTG reset register (OTG_GRSTCTL)

Address offset: 0x10
Reset value: 0x8000 0000
The application uses this register to reset various hardware features inside the core.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
AHB
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
IDL
r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
TXF RXF
Res. Res. Res. Res. Res. TXFNUM Res. FCRST PSRST CSRST
FLSH FLSH
rw rw rw rw rw rs rs rs rs r

Bit 31 AHBIDL: AHB master idle

Indicates that the AHB master state machine is in the Idle condition.
Note: Accessible in both device and host modes.
Bits 30:11 Reserved, must be kept at reset value.
Bits 10:6 TXFNUM[4:0]: Tx FIFO number
This is the FIFO number that must be flushed using the Tx FIFO Flush bit. This field must not
be changed until the core clears the Tx FIFO Flush bit.
00000:
– Non-periodic Tx FIFO flush in host mode
– Tx FIFO 0 flush in device mode
00001:
– Periodic Tx FIFO flush in host mode
– Tx FIFO 1 flush in device mode
00010: Tx FIFO 2 flush in device mode
...
01111: Tx FIFO 15 flush in device mode
10000: Flush all the transmit FIFOs in device or host mode.
Note: Accessible in both device and host modes.
Bit 5 TXFFLSH: Tx FIFO flush
This bit selectively flushes a single or all transmit FIFOs, but cannot do so if the core is in the
midst of a transaction.
The application must write this bit only after checking that the core is neither writing to the Tx
FIFO nor reading from the Tx FIFO. Verify using these registers:
Read—NAK Effective interrupt ensures the core is not reading from the FIFO
Write—AHBIDL bit in OTG_GRSTCTL ensures the core is not writing anything to the FIFO.
Flushing is normally recommended when FIFOs are reconfigured. FIFO flushing is also
recommended during device endpoint disable. The application must wait until the core clears
this bit before performing any operations. This bit takes eight clocks to clear, using the slower
clock of phy_clk or hclk.
Note: Accessible in both device and host modes.

1004/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 4 RXFFLSH: Rx FIFO flush

The application can flush the entire Rx FIFO using this bit, but must first ensure that the core
is not in the middle of a transaction.
The application must only write to this bit after checking that the core is neither reading from
the Rx FIFO nor writing to the Rx FIFO.
The application must wait until the bit is cleared before performing any other operations. This
bit requires 8 clocks (slowest of PHY or AHB clock) to clear.
Note: Accessible in both device and host modes.
Bit 3 Reserved, must be kept at reset value.
Bit 2 FCRST: Host frame counter reset
The application writes this bit to reset the frame number counter inside the core. When the
frame counter is reset, the subsequent SOF sent out by the core has a frame number of 0.
When application writes '1' to the bit, it might not be able to read back the value as it gets
cleared by the core in a few clock cycles.
Note: Only accessible in host mode.
Bit 1 PSRST: Partial soft reset
Resets the internal state machines but keeps the enumeration info. Could be used to recover
some specific PHY errors.
Note: Accessible in both device and host modes.
Bit 0 CSRST: Core soft reset
Resets the HCLK and PHY clock domains as follows:
Clears the interrupts and all the CSR register bits except for the following bits:
– GATEHCLK bit in OTG_PCGCCTL
– STPPCLK bit in OTG_PCGCCTL
– FSLSPCS bits in OTG_HCFG
– DSPD bit in OTG_DCFG
– SDIS bit in OTG_DCTL
– OTG_GCCFG register
All module state machines (except for the AHB slave unit) are reset to the Idle state, and all
the transmit FIFOs and the receive FIFO are flushed.
Any transactions on the AHB Master are terminated as soon as possible, after completing the
last data phase of an AHB transfer. Any transactions on the USB are terminated immediately.
The application can write to this bit any time it wants to reset the core. This is a self-clearing
bit and the core clears this bit after all the necessary logic is reset in the core, which can take
several clocks, depending on the current state of the core. Once this bit has been cleared,
the software must wait at least 3 PHY clocks before accessing the PHY domain
(synchronization delay). The software must also check that bit 31 in this register is set to 1
(AHB Master is Idle) before starting any operation.
Typically, the software reset is used during software development and also when the user
dynamically changes the PHY selection bits in the above listed USB configuration registers.
When you change the PHY, the corresponding clock for the PHY is selected and used in the
PHY domain. Once a new clock is selected, the PHY domain has to be reset for proper
operation.
Note: Accessible in both device and host modes.

RM0402 Rev 6 1005/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.6 OTG core interrupt register (OTG_GINTSTS)

Address offset: 0x014
Reset value: 0x0400 0020
This register interrupts the application for system-level events in the current mode (device
mode or host mode).
Some of the bits in this register are valid only in host mode, while others are valid in device
mode only. This register also indicates the current mode. To clear the interrupt status bits of
the rc_w1 type, the application must write 1 into the bit.
The FIFO status interrupts are read-only; once software reads from or writes to the FIFO
while servicing these interrupts, FIFO interrupt conditions are cleared automatically.
The application must clear the OTG_GINTSTS register at initialization before unmasking
the interrupt bit to avoid any interrupts generated prior to initialization.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
IPXFR/
IN
WKUP SRQ DISC CIDS LPM HPRT RST IISOI OEP
PTXFE HCINT Res. COMP IEPINT Res. Res.
INT INT INT CHG INT INT DET XFR INT
ISO
OUT
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 r r r rc_w1 rc_w1 rc_w1 r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
GO GI
ISOO ENUM USB USB NPTXF RXF OTG
EOPF ESUSP Res. Res. NAK NAK SOF MMIS CMOD
DRP DNE RST SUSP E LVL INT
EFF EFF
rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 r r r r rc_w1 r rc_w1 r

Bit 31 WKUPINT: Resume/remote wakeup detected interrupt

Wakeup interrupt during suspend(L2) or LPM(L1) state.
– During suspend(L2):
In device mode, this interrupt is asserted when a resume is detected on the USB. In host
mode, this interrupt is asserted when a remote wakeup is detected on the USB.
– During LPM(L1):
This interrupt is asserted for either host initiated resume or device initiated remote wakeup
on USB.
Note: Accessible in both device and host modes.
Bit 30 SRQINT: Session request/new session detected interrupt
In host mode, this interrupt is asserted when a session request is detected from the device.
In device mode, this interrupt is asserted when VBUS is in the valid range for a B-peripheral
device. Accessible in both device and host modes.
Bit 29 DISCINT: Disconnect detected interrupt
Asserted when a device disconnect is detected.
Note: Only accessible in host mode.
Bit 28 CIDSCHG: Connector ID status change
The core sets this bit when there is a change in connector ID status.
Note: Accessible in both device and host modes.

1006/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 27 LPMINT: LPM interrupt

In device mode, this interrupt is asserted when the device receives an LPM transaction and
responds with a non-ERRORed response.
In host mode, this interrupt is asserted when the device responds to an LPM transaction with
a non-ERRORed response or when the host core has completed LPM transactions for the
programmed number of times (RETRYCNT bit in OTG_GLPMCFG).
This field is valid only if the LPMEN bit in OTG_GLPMCFG is set to 1.
Bit 26 PTXFE: Periodic Tx FIFO empty
Asserted when the periodic transmit FIFO is either half or completely empty and there is
space for at least one entry to be written in the periodic request queue. The half or
completely empty status is determined by the periodic Tx FIFO empty level bit in the
OTG_GAHBCFG register (PTXFELVL bit in OTG_GAHBCFG).
Note: Only accessible in host mode.
Bit 25 HCINT: Host channels interrupt
The core sets this bit to indicate that an interrupt is pending on one of the channels of the
core (in host mode). The application must read the OTG_HAINT register to determine the
exact number of the channel on which the interrupt occurred, and then read the
corresponding OTG_HCINTx register to determine the exact cause of the interrupt. The
application must clear the appropriate status bit in the OTG_HCINTx register to clear this bit.
Note: Only accessible in host mode.
Bit 24 HPRTINT: Host port interrupt
The core sets this bit to indicate a change in port status of one of the OTG_FS controller
ports in host mode. The application must read the OTG_HPRT register to determine the
exact event that caused this interrupt. The application must clear the appropriate status bit in
the OTG_HPRT register to clear this bit.
Note: Only accessible in host mode.
Bit 23 RSTDET: Reset detected interrupt
In device mode, this interrupt is asserted when a reset is detected on the USB in partial
power-down mode when the device is in suspend.
Note: Only accessible in device mode.
Bit 22 Reserved, must be kept at reset value.
Bit 21 IPXFR: Incomplete periodic transfer
In host mode, the core sets this interrupt bit when there are incomplete periodic transactions
still pending, which are scheduled for the current frame.
INCOMPISOOUT: Incomplete isochronous OUT transfer
In device mode, the core sets this interrupt to indicate that there is at least one isochronous
OUT endpoint on which the transfer is not completed in the current frame. This interrupt is
asserted along with the End of periodic frame interrupt (EOPF) bit in this register.
Bit 20 IISOIXFR: Incomplete isochronous IN transfer
The core sets this interrupt to indicate that there is at least one isochronous IN endpoint on
which the transfer is not completed in the current frame. This interrupt is asserted along with
the End of periodic frame interrupt (EOPF) bit in this register.
Note: Only accessible in device mode.

RM0402 Rev 6 1007/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 19 OEPINT: OUT endpoint interrupt

The core sets this bit to indicate that an interrupt is pending on one of the OUT endpoints of
the core (in device mode). The application must read the OTG_DAINT register to determine
the exact number of the OUT endpoint on which the interrupt occurred, and then read the
corresponding OTG_DOEPINTx register to determine the exact cause of the interrupt. The
application must clear the appropriate status bit in the corresponding OTG_DOEPINTx
register to clear this bit.
Note: Only accessible in device mode.
Bit 18 IEPINT: IN endpoint interrupt
The core sets this bit to indicate that an interrupt is pending on one of the IN endpoints of the
core (in device mode). The application must read the OTG_DAINT register to determine the
exact number of the IN endpoint on which the interrupt occurred, and then read the
corresponding OTG_DIEPINTx register to determine the exact cause of the interrupt. The
application must clear the appropriate status bit in the corresponding OTG_DIEPINTx
register to clear this bit.
Note: Only accessible in device mode.
Bits 17:16 Reserved, must be kept at reset value.
Bit 15 EOPF: End of periodic frame interrupt
Indicates that the period specified in the periodic frame interval field of the OTG_DCFG
register (PFIVL bit in OTG_DCFG) has been reached in the current frame.
Note: Only accessible in device mode.
Bit 14 ISOODRP: Isochronous OUT packet dropped interrupt
The core sets this bit when it fails to write an isochronous OUT packet into the Rx FIFO
because the Rx FIFO does not have enough space to accommodate a maximum size
packet for the isochronous OUT endpoint.
Note: Only accessible in device mode.
Bit 13 ENUMDNE: Enumeration done
The core sets this bit to indicate that speed enumeration is complete. The application must
read the OTG_DSTS register to obtain the enumerated speed.
Note: Only accessible in device mode.
Bit 12 USBRST: USB reset
The core sets this bit to indicate that a reset is detected on the USB.
Note: Only accessible in device mode.
Bit 11 USBSUSP: USB suspend
The core sets this bit to indicate that a suspend was detected on the USB. The core enters
the suspended state when there is no activity on the data lines for an extended period of
time.
Note: Only accessible in device mode.
Bit 10 ESUSP: Early suspend
The core sets this bit to indicate that an Idle state has been detected on the USB for 3 ms.
Note: Only accessible in device mode.
Bits 9:8 Reserved, must be kept at reset value.
Bit 7 GONAKEFF: Global OUT NAK effective
Indicates that the Set global OUT NAK bit in the OTG_DCTL register (SGONAK bit in
OTG_DCTL), set by the application, has taken effect in the core. This bit can be cleared by
writing the Clear global OUT NAK bit in the OTG_DCTL register (CGONAK bit in
OTG_DCTL).
Note: Only accessible in device mode.

1008/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 6 GINAKEFF: Global IN non-periodic NAK effective

Indicates that the Set global non-periodic IN NAK bit in the OTG_DCTL register (SGINAK bit
in OTG_DCTL), set by the application, has taken effect in the core. That is, the core has
sampled the Global IN NAK bit set by the application. This bit can be cleared by clearing the
Clear global non-periodic IN NAK bit in the OTG_DCTL register (CGINAK bit in
OTG_DCTL).
This interrupt does not necessarily mean that a NAK handshake is sent out on the USB. The
STALL bit takes precedence over the NAK bit.
Note: Only accessible in device mode.
Bit 5 NPTXFE: Non-periodic Tx FIFO empty
This interrupt is asserted when the non-periodic Tx FIFO is either half or completely empty,
and there is space for at least one entry to be written to the non-periodic transmit request
queue. The half or completely empty status is determined by the non-periodic Tx FIFO
empty level bit in the OTG_GAHBCFG register (TXFELVL bit in OTG_GAHBCFG).
Note: Accessible in host mode only.
Bit 4 RXFLVL: Rx FIFO non-empty
Indicates that there is at least one packet pending to be read from the Rx FIFO.
Note: Accessible in both host and device modes.
Bit 3 SOF: Start of frame
In host mode, the core sets this bit to indicate that an SOF (FS), or Keep-Alive (LS) is
transmitted on the USB. The application must write a 1 to this bit to clear the interrupt.
In device mode, in the core sets this bit to indicate that an SOF token has been received on
the USB. The application can read the OTG_DSTS register to get the current frame number.
This interrupt is seen only when the core is operating in FS.
Note: This register may return '1' if read immediately after power on reset. If the register bit
reads '1' immediately after power on reset it does not indicate that an SOF has been
sent (in case of host mode) or SOF has been received (in case of device mode). The
read value of this interrupt is valid only after a valid connection between host and
device is established. If the bit is set after power on reset the application can clear the
bit.
Note: Accessible in both host and device modes.
Bit 2 OTGINT: OTG interrupt
The core sets this bit to indicate an OTG protocol event. The application must read the OTG
interrupt status (OTG_GOTGINT) register to determine the exact event that caused this
interrupt. The application must clear the appropriate status bit in the OTG_GOTGINT
register to clear this bit.
Note: Accessible in both host and device modes.
Bit 1 MMIS: Mode mismatch interrupt
The core sets this bit when the application is trying to access:
– A host mode register, when the core is operating in device mode
– A device mode register, when the core is operating in host mode
The register access is completed on the AHB with an OKAY response, but is ignored by the
core internally and does not affect the operation of the core.
Note: Accessible in both host and device modes.
Bit 0 CMOD: Current mode of operation
Indicates the current mode.
0: Device mode
1: Host mode
Note: Accessible in both host and device modes.

RM0402 Rev 6 1009/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.7 OTG interrupt mask register (OTG_GINTMSK)

Address offset: 0x018
Reset value: 0x0000 0000
This register works with the core interrupt register to interrupt the application. When an
interrupt bit is masked, the interrupt associated with that bit is not generated. However, the
core interrupt (OTG_GINTSTS) register bit corresponding to that interrupt is still set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
IPXFR
DISCIN CIDSC LPMIN PTXFE RSTDE M/IISO IISOIX OEPIN
WUIM SRQIM HCIM PRTIM Res. IEPINT Res. Res.
T HGM TM M TM OXFR FRM T
M

rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
EOPF ISOOD ENUM USBRS USBSU ESUSP GONA GINAK NPTXF RXFLV OTGIN
Res. Res. SOFM MMISM Res.
M RPM DNEM T SPM M KEFFM EFFM EM LM T

rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 WUIM: Resume/remote wakeup detected interrupt mask

0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 30 SRQIM: Session request/new session detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 29 DISCINT: Disconnect detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 28 CIDSCHGM: Connector ID status change mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 27 LPMINTM: LPM interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both host and device modes.
Bit 26 PTXFEM: Periodic Tx FIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 25 HCIM: Host channels interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.

1010/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 24 PRTIM: Host port interrupt mask

0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 23 RSTDETM: Reset detected interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 22 Reserved, must be kept at reset value.
Bit 21 IPXFRM: Incomplete periodic transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
IISOOXFRM: Incomplete isochronous OUT transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 20 IISOIXFRM: Incomplete isochronous IN transfer mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 19 OEPINT: OUT endpoints interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 18 IEPINT: IN endpoints interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bits 17:16 Reserved, must be kept at reset value.
Bit 15 EOPFM: End of periodic frame interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 14 ISOODRPM: Isochronous OUT packet dropped interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 13 ENUMDNEM: Enumeration done mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.

RM0402 Rev 6 1011/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 12 USBRST: USB reset mask

0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 11 USBSUSPM: USB suspend mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 10 ESUSPM: Early suspend mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bits 9:8 Reserved, must be kept at reset value.
Bit 7 GONAKEFFM: Global OUT NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 6 GINAKEFFM: Global non-periodic IN NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in device mode.
Bit 5 NPTXFEM: Non-periodic Tx FIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Only accessible in host mode.
Bit 4 RXFLVLM: Receive FIFO non-empty mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 3 SOFM: Start of frame mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 2 OTGINT: OTG interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 1 MMISM: Mode mismatch interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Note: Accessible in both device and host modes.
Bit 0 Reserved, must be kept at reset value.

1012/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.8 OTG receive status debug read register (OTG_GRXSTSR)

Address offset for read: 0x01C
Reset value: 0x0000 0000
This description is for register OTG_GRXSTSR in Device mode.
A read to the receive status debug read register returns the contents of the top of the
receive FIFO.
The core ignores the receive status read when the receive FIFO is empty and returns a
value of 0x0000 0000.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
STSPH
Res. Res. Res. Res. Res. Res. FRMNUM[3:0] PKTSTS[3:0] DPID[1]
ST

r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DPID[0] BCNT[10:0] EPNUM[3:0]

r r r r r r r r r r r r r r r r

Bits 31:28 Reserved, must be kept at reset value.

Bit 27 STSPHST: Status phase start
Indicates the start of the status phase for a control write transfer. This bit is set along with
the OUT transfer completed PKTSTS pattern.
Bits 26:25 Reserved, must be kept at reset value.
Bits 24:21 FRMNUM[3:0]: Frame number
This is the least significant 4 bits of the frame number in which the packet is received on the
USB. This field is supported only when isochronous OUT endpoints are supported.
Bits 20:17 PKTSTS[3:0]: Packet status
Indicates the status of the received packet
0001: Global OUT NAK (triggers an interrupt)
0010: OUT data packet received
0011: OUT transfer completed (triggers an interrupt)
0100: SETUP transaction completed (triggers an interrupt)
0110: SETUP data packet received
Others: Reserved
Bits 16:15 DPID[1:0]: Data PID
Indicates the data PID of the received OUT data packet
00: DATA0
10: DATA1
Bits 14:4 BCNT[10:0]: Byte count
Indicates the byte count of the received data packet.
Bits 3:0 EPNUM[3:0]: Endpoint number
Indicates the endpoint number to which the current received packet belongs.

RM0402 Rev 6 1013/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.9 OTG receive status debug read [alternate] (OTG_GRXSTSR)

Address offset for read: 0x01C
Reset value: 0x0000 0000
This description is for register OTG_GRXSTSR in Host mode.
A read to the receive status debug read register returns the contents of the top of the
receive FIFO.
The core ignores the receive status read when the receive FIFO is empty and returns a
value of 0x0000 0000.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PKTSTS[3:0] DPID

r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DPID BCNT[10:0] CHNUM[3:0]

r r r r r r r r r r r r r r r r

Bits 31:21 Reserved, must be kept at reset value.

Bits 20:17 PKTSTS[3:0]: Packet status
Indicates the status of the received packet
0010: IN data packet received
0011: IN transfer completed (triggers an interrupt)
0101: Data toggle error (triggers an interrupt)
0111: Channel halted (triggers an interrupt)
Others: Reserved
Bits 16:15 DPID[1:0]: Data PID
Indicates the data PID of the received packet
00: DATA0
10: DATA1
Bits 14:4 BCNT[10:0]: Byte count
Indicates the byte count of the received IN data packet.
Bits 3:0 CHNUM[3:0]: Channel number
Indicates the channel number to which the current received packet belongs.

1014/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.10 OTG status read and pop registers (OTG_GRXSTSP)

Address offset for pop: 0x020
Reset value: 0x0000 0000
This description is for register OTG_GRXSTSP in Device mode.
Similarly to OTG_GRXSTSR (receive status debug read register) where a read returns the
contents of the top of the receive FIFO, a read to OTG_GRXSTSP (receive status read and
pop register) additionally pops the top data entry out of the Rx FIFO.
The core ignores the receive status pop/read when the receive FIFO is empty and returns a
value of 0x0000 0000. The application must only pop the receive status FIFO when the
receive FIFO non-empty bit of the core interrupt register (RXFLVL bit in OTG_GINTSTS) is
asserted.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
STSPH
Res. Res. Res. Res. Res. Res. FRMNUM[3:0] PKTSTS[3:0] DPID[1]
ST

r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DPID[0] BCNT[10:0] EPNUM[3:0]

r r r r r r r r r r r r r r r r

Bits 31:28 Reserved, must be kept at reset value.

Bit 27 STSPHST: Status phase start
Indicates the start of the status phase for a control write transfer. This bit is set along with
the OUT transfer completed PKTSTS pattern.
Bits 26:25 Reserved, must be kept at reset value.
Bits 24:21 FRMNUM[3:0]: Frame number
This is the least significant 4 bits of the frame number in which the packet is received on the
USB. This field is supported only when isochronous OUT endpoints are supported.
Bits 20:17 PKTSTS[3:0]: Packet status
Indicates the status of the received packet
0001: Global OUT NAK (triggers an interrupt)
0010: OUT data packet received
0011: OUT transfer completed (triggers an interrupt)
0100: SETUP transaction completed (triggers an interrupt)
0110: SETUP data packet received
Others: Reserved
Bits 16:15 DPID[1:0]: Data PID
Indicates the data PID of the received OUT data packet
00: DATA0
10: DATA1
Bits 14:4 BCNT[10:0]: Byte count
Indicates the byte count of the received data packet.
Bits 3:0 EPNUM[3:0]: Endpoint number
Indicates the endpoint number to which the current received packet belongs.

RM0402 Rev 6 1015/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.11 OTG status read and pop registers [alternate] (OTG_GRXSTSP)

Address offset for pop: 0x020
Reset value: 0x0000 0000
This description is for register OTG_GRXSTSP in Host mode.
Similarly to OTG_GRXSTSR (receive status debug read register) where a read returns the
contents of the top of the receive FIFO, a read to OTG_GRXSTSP (receive status read and
pop register) additionally pops the top data entry out of the Rx FIFO.
The core ignores the receive status pop/read when the receive FIFO is empty and returns a
value of 0x0000 0000. The application must only pop the receive status FIFO when the
receive FIFO non-empty bit of the core interrupt register (RXFLVL bit in OTG_GINTSTS) is
asserted.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PKTSTS[3:0] DPID

r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DPID BCNT[10:0] CHNUM[3:0]

r r r r r r r r r r r r r r r r

Bits 31:21 Reserved, must be kept at reset value.

Bits 20:17 PKTSTS[3:0]: Packet status
Indicates the status of the received packet
0010: IN data packet received
0011: IN transfer completed (triggers an interrupt)
0101: Data toggle error (triggers an interrupt)
0111: Channel halted (triggers an interrupt)
Others: Reserved
Bits 16:15 DPID[1:0]: Data PID
Indicates the data PID of the received packet
00: DATA0
10: DATA1
Bits 14:4 BCNT[10:0]: Byte count
Indicates the byte count of the received IN data packet.
Bits 3:0 CHNUM[3:0]: Channel number
Indicates the channel number to which the current received packet belongs.

1016/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.12 OTG receive FIFO size register (OTG_GRXFSIZ)

Address offset: 0x024
Reset value: 0x0000 0200
The application can program the RAM size that must be allocated to the Rx FIFO.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

RXFD[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 RXFD[15:0]: Rx FIFO depth
This value is in terms of 32-bit words.
Minimum value is 16
Programmed values must respect the available FIFO memory allocation and must not
exceed the power-on value.

29.15.13 OTG host non-periodic transmit FIFO size register

(OTG_HNPTXFSIZ)/Endpoint 0 Transmit FIFO size
(OTG_DIEPTXF0)
Address offset: 0x028
Reset value: 0x0200 0200

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

NPTXFD/TX0FD[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

NPTXFSA/TX0FSA[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Host mode

Bits 31:16 NPTXFD[15:0]: Non-periodic Tx FIFO depth

This value is in terms of 32-bit words.
Minimum value is 16
Programmed values must respect the available FIFO memory allocation and must not
exceed the power-on value.
Bits 15:0 NPTXFSA[15:0]: Non-periodic transmit RAM start address
This field configures the memory start address for non-periodic transmit FIFO RAM.

RM0402 Rev 6 1017/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Device mode

Bits 31:16 TX0FD: Endpoint 0 Tx FIFO depth

This value is in terms of 32-bit words.
Minimum value is 16
Programmed values must respect the available FIFO memory allocation and must not
exceed the power-on value.
Bits 15:0 TX0FSA: Endpoint 0 transmit RAM start address
This field configures the memory start address for the endpoint 0 transmit FIFO RAM.

29.15.14 OTG non-periodic transmit FIFO/queue status register

(OTG_HNPTXSTS)
Address offset: 0x02C
Reset value: 0x0008 0200
Note: In device mode, this register is not valid.
This read-only register contains the free space information for the non-periodic Tx FIFO and
the non-periodic transmit request queue.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. NPTXQTOP[6:0] NPTQXSAV[7:0]

r r r r r r r r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

NPTXFSAV[15:0]

r r r r r r r r r r r r r r r r

1018/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 31 Reserved, must be kept at reset value.

Bits 30:24 NPTXQTOP[6:0]: Top of the non-periodic transmit request queue
Entry in the non-periodic Tx request queue that is currently being processed by the MAC.
Bits 30:27: Channel/endpoint number
Bits 26:25:
00: IN/OUT token
01: Zero-length transmit packet (device IN/host OUT)
11: Channel halt command
Bit 24: Terminate (last entry for selected channel/endpoint)
Bits 23:16 NPTQXSAV[7:0]: Non-periodic transmit request queue space available
Indicates the amount of free space available in the non-periodic transmit request queue.
This queue holds both IN and OUT requests.
0: Non-periodic transmit request queue is full
1: 1 location available
2: locations available
n: n locations available (0 ≤ n ≤ 8)
Others: Reserved
Bits 15:0 NPTXFSAV[15:0]: Non-periodic Tx FIFO space available
Indicates the amount of free space available in the non-periodic Tx FIFO.
Values are in terms of 32-bit words.
0: Non-periodic Tx FIFO is full
1: 1 word available
2: 2 words available
n: n words available (where 0 ≤ n ≤ 512)
Others: Reserved

29.15.15 OTG general core configuration register (OTG_GCCFG)

Address offset: 0x038
Reset value: 0x0000 XXXX

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DCD PWR
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. VBDEN SDEN PDEN BCDEN
EN DWN

rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PS2
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. SDET PDET DCDET
DET

r r r r

RM0402 Rev 6 1019/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 31:22 Reserved, must be kept at reset value.

Bit 21 VBDEN: USB VBUS detection enable
Enables VBUS sensing comparators to detect VBUS valid levels on the VBUS PAD for USB
host and device operation. If HNP and/or SRP support is enabled, VBUS comparators are
automatically enabled independently of VBDEN value.
0 = VBUS detection disabled
1 = VBUS detection enabled
Bit 20 SDEN: Secondary detection (SD) mode enable
This bit is set by the software to put the BCD into SD mode. Only one detection mode (DCD,
PD, SD or OFF) should be selected to work correctly
Bit 19 PDEN: Primary detection (PD) mode enable
This bit is set by the software to put the BCD into PD mode. Only one detection mode (DCD,
PD, SD or OFF) should be selected to work correctly.
Bit 18 DCDEN: Data contact detection (DCD) mode enable
This bit is set by the software to put the BCD into DCD mode. Only one detection mode
(DCD, PD, SD or OFF) should be selected to work correctly.
Bit 17 BCDEN: Battery charging detector (BCD) enable
This bit is set by the software to enable the BCD support within the USB device. When
enabled, the USB PHY is fully controlled by BCD and cannot be used for normal
communication. Once the BCD discovery is finished, the BCD should be placed in OFF
mode by clearing this bit to ‘0’ in order to allow the normal USB operation.
Bit 16 PWRDWN: Power down control of FS PHY
Used to activate the FS PHY in transmission/reception. When reset, the PHY is kept in
power-down. When set, the BCD function must be off (BCDEN=0).
0 = USB FS PHY disabled
1 = USB FS PHY enabled
Bits 15:4 Reserved, must be kept at reset value.
Bit 3 PS2DET: DM pull-up detection status
This bit is active only during PD and gives the result of comparison between DM voltage
level and VLGC threshold. In normal situation, the DM level should be below this threshold.
If it is above, it means that the DM is externally pulled high. This can be caused by
connection to a PS2 port (which pulls-up both DP and DM lines) or to some proprietary
charger not following the BCD specification.
0: Normal port detected (connected to SDP, CDP or DCP)
1: PS2 port or proprietary charger detected
Bit 2 SDET: Secondary detection (SD) status
This bit gives the result of SD.
0: CDP detected
1: DCP detected
Bit 1 PDET: Primary detection (PD) status
This bit gives the result of PD.
0: no BCD support detected (connected to SDP or proprietary device).
1: BCD support detected (connected to CDP or DCP).
Bit 0 DCDET: Data contact detection (DCD) status
This bit gives the result of DCD.
0: data lines contact not detected
1: data lines contact detected

1020/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.16 OTG core ID register (OTG_CID)

Address offset: 0x03C
Reset value: 0x0000 2000
This is a register containing the Product ID as reset value.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PRODUCT_ID[31:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

PRODUCT_ID[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:0 PRODUCT_ID[31:0]: Product ID field

Application-programmable ID field.

29.15.17 OTG core LPM configuration register (OTG_GLPMCFG)

Address offset: 0x54
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
EN SND L1RSM
Res. Res. Res. LPMRCNTSTS[2:0] LPMRCNT[2:0] LPMCHIDX[3:0]
BESL LPM OK
rw r r r rs rw rw rw rw rw rw rw r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SLP L1DS L1SS REM LPM LPM
LPMRSP[1:0] BESLTHRS[3:0] BESL[3:0]
STS EN EN WAKE ACK EN
r r r rw rw rw rw rw rw rw/r rw/r rw/r rw/r rw/r rw rw

Bits 31:29 Reserved, must be kept at reset value.

Bit 28 ENBESL: Enable best effort service latency
This bit enables the BESL feature as defined in the LPM errata:
0:The core works as described in the following document:
USB 2.0 Link Power Management Addendum Engineering Change Notice to the USB 2.0
specification, July 16, 2007
1:The core works as described in the LPM Errata:
Errata for USB 2.0 ECN: Link Power Management (LPM) - 7/2007
Note: Only the updated behavior (described in LPM Errata) is considered in this document
and so the ENBESL bit should be set to '1' by application SW.
Bits 27:25 LPMRCNTSTS[2:0]: LPM retry count status
Number of LPM host retries still remaining to be transmitted for the current LPM sequence.
Note: Accessible only in host mode.

RM0402 Rev 6 1021/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 24 SNDLPM: Send LPM transaction

When the application software sets this bit, an LPM transaction containing two tokens, EXT
and LPM is sent. The hardware clears this bit once a valid response (STALL, NYET, or
ACK) is received from the device or the core has finished transmitting the programmed
number of LPM retries.
Note: This bit must be set only when the host is connected to a local port.
Note: Accessible only in host mode.
Bits 23:21 LPMRCNT[2:0]: LPM retry count
When the device gives an ERROR response, this is the number of additional LPM retries
that the host performs until a valid device response (STALL, NYET, or ACK) is received.
Note: Accessible only in host mode.
Bits 20:17 LPMCHIDX[3:0]: LPM Channel Index
The channel number on which the LPM transaction has to be applied while sending an LPM
transaction to the local device. Based on the LPM channel index, the core automatically
inserts the device address and endpoint number programmed in the corresponding channel
into the LPM transaction.
Note: Accessible only in host mode.
Bit 16 L1RSMOK: Sleep state resume OK
Indicates that the device or host can start resume from Sleep state. This bit is valid in LPM
sleep (L1) state. It is set in sleep mode after a delay of 50 μs (TL1Residency).
This bit is reset when SLPSTS is 0.
1: The application or host can start resume from Sleep state
0: The application or host cannot start resume from Sleep state
Bit 15 SLPSTS: Port sleep status
Device mode:
This bit is set as long as a Sleep condition is present on the USB bus. The core enters the
Sleep state when an ACK response is sent to an LPM transaction and the TL1TokenRetry
timer has expired. To stop the PHY clock, the application must set the STPPCLK bit in
OTG_PCGCCTL, which asserts the PHY suspend input signal.
The application must rely on SLPSTS and not ACK in LPMRSP to confirm transition into
sleep.
The core comes out of sleep:
– When there is any activity on the USB linestate
– When the application writes to the RWUSIG bit in OTG_DCTL or when the application
resets or soft-disconnects the device.
Host mode:
The host transitions to Sleep (L1) state as a side-effect of a successful LPM transaction by
the core to the local port with ACK response from the device. The read value of this bit
reflects the current Sleep status of the port.
The core clears this bit after:
– The core detects a remote L1 wakeup signal,
– The application sets the PRST bit or the PRES bit in the OTG_HPRT register, or
– The application sets the L1Resume/ remote wakeup detected interrupt bit or disconnect
detected interrupt bit in the core interrupt register (WKUPINT or DISCINT bit in
OTG_GINTSTS, respectively).
0: Core not in L1
1: Core in L1

1022/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 14:13 LPMRSP[1:0]: LPM response

Device mode:
The response of the core to LPM transaction received is reflected in these two bits.
Host mode:
Handshake response received from local device for LPM transaction
11: ACK
10: NYET
01: STALL
00: ERROR (No handshake response)
Bit 12 L1DSEN: L1 deep sleep enable
Enables suspending the PHY in L1 Sleep mode. For maximum power saving during L1
Sleep mode, this bit should be set to '1' by application SW in all the cases.
Bits 11:8 BESLTHRS[3:0]: BESL threshold
Device mode:
The core puts the PHY into deep low power mode in L1 when BESL value is greater than or
equal to the value defined in this field BESL_Thres[3:0].
Host mode:
The core puts the PHY into deep low power mode in L1. BESLTHRS[3:0] specifies the time
for which resume signaling is to be reflected by host (TL1HubDrvResume2) on the USB bus
when it detects device initiated resume.
BESLTHRS must not be programmed with a value greater than 1100b in host mode,
because this exceeds maximum TL1HubDrvResume2.
Thres[3:0] host mode resume signaling time (μs):
0000: 75
0001: 100
0010: 150
0011: 250
0100: 350
0101: 450
0110: 950
All other values: reserved
Bit 7 L1SSEN: L1 Shallow Sleep enable
Enables suspending the PHY in L1 Sleep mode. For maximum power saving during L1
Sleep mode, this bit should be set to '1' by application SW in all the cases.
Bit 6 REMWAKE: bRemoteWake value
Host mode:
The value of remote wake up to be sent in the wIndex field of LPM transaction.
Device mode (read-only):
This field is updated with the received LPM token bRemoteWake bmAttribute when an ACK,
NYET, or STALL response is sent to an LPM transaction.

RM0402 Rev 6 1023/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 5:2 BESL[3:0]: Best effort service latency

Host mode:
The value of BESL to be sent in an LPM transaction. This value is also used to initiate
resume for a duration TL1HubDrvResume1 for host initiated resume.
Device mode (read-only):
This field is updated with the received LPM token BESL bmAttribute when an ACK, NYET,
or STALL response is sent to an LPM transaction.
BESL[3:0]TBESL (μs)
0000: 125
0001: 150
0010: 200
0011: 300
0100: 400
0101: 500
0110: 1000
0111: 2000
1000: 3000
1001: 4000
1010: 5000
1011: 6000
1100: 7000
1101: 8000
1110: 9000
1111: 10000
Bit 1 LPMACK: LPM token acknowledge enable
Handshake response to LPM token preprogrammed by device application software.
1: ACK
Even though ACK is preprogrammed, the core device responds with ACK only on
successful LPM transaction. The LPM transaction is successful if:
– No PID/CRC5 errors in either EXT token or LPM token (else ERROR)
– Valid bLinkState = 0001B (L1) received in LPM transaction (else STALL)
– No data pending in transmit queue (else NYET).
0: NYET
The preprogrammed software bit is over-ridden for response to LPM token when:
– The received bLinkState is not L1 (STALL response), or
– An error is detected in either of the LPM token packets because of corruption (ERROR
response).
Note: Accessible only in device mode.
Bit 0 LPMEN: LPM support enable
The application uses this bit to control the OTG_FS core LPM capabilities.
If the core operates as a non-LPM-capable host, it cannot request the connected device or
hub to activate LPM mode.
If the core operates as a non-LPM-capable device, it cannot respond to any LPM
transactions.
0: LPM capability is not enabled
1: LPM capability is enabled

1024/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.18 OTG host periodic transmit FIFO size register

(OTG_HPTXFSIZ)
Address offset: 0x100
Reset value: 0x0200 0400

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PTXFSIZ[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

PTXSA[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 PTXFSIZ[15:0]: Host periodic Tx FIFO depth

This value is in terms of 32-bit words.
Minimum value is 16
Bits 15:0 PTXSA[15:0]: Host periodic Tx FIFO start address
This field configures the memory start address for periodic transmit FIFO RAM.

29.15.19 OTG device IN endpoint transmit FIFO x size register

(OTG_DIEPTXFx)
Address offset: 0x104 + 0x04 * (x - 1), (x = 1 to 5)
Reset value: Block 1: 0x0200 0400
Reset value: Block 2: 0x0200 0600
Reset value: Block 3: 0x0200 0800
Reset value: Block 4: 0x0200 0A00
Reset value: Block 5: 0x0200 0C00

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

INEPTXFD[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

INEPTXSA[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 INEPTXFD[15:0]: IN endpoint Tx FIFO depth

This value is in terms of 32-bit words.
Minimum value is 16
Bits 15:0 INEPTXSA[15:0]: IN endpoint FIFOx transmit RAM start address
This field contains the memory start address for IN endpoint transmit FIFOx. The address
must be aligned with a 32-bit memory location.

RM0402 Rev 6 1025/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.20 Host-mode registers

Bit values in the register descriptions are expressed in binary unless otherwise specified.
Host-mode registers affect the operation of the core in the host mode. Host mode registers
must not be accessed in device mode, as the results are undefined. Host mode registers
can be categorized as follows:

29.15.21 OTG host configuration register (OTG_HCFG)

Address offset: 0x400
Reset value: 0x0000 0000
This register configures the core after power-on. Do not make changes to this register after
initializing the host.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. FSLSS FSLSPCS[1:0]

r rw rw

Bits 31:3 Reserved, must be kept at reset value.

Bit 2 FSLSS: FS- and LS-only support
The application uses this bit to control the core’s enumeration speed. Using this bit, the
application can make the core enumerate as an FS host, even if the connected device
supports HS traffic. Do not make changes to this field after initial programming.
1: FS/LS-only, even if the connected device can support HS (read-only).
Bits 1:0 FSLSPCS[1:0]: FS/LS PHY clock select
When the core is in FS host mode
01: PHY clock is running at 48 MHz
Others: Reserved
When the core is in LS host mode
00: Reserved
01: Select 48 MHz PHY clock frequency
10: Select 6 MHz PHY clock frequency
11: Reserved
Note: The FSLSPCS must be set on a connection event according to the speed of the
connected device (after changing this bit, a software reset must be performed).

1026/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.22 OTG host frame interval register (OTG_HFIR)

Address offset: 0x404
Reset value: 0x0000 EA60
This register stores the frame interval information for the current speed to which the
OTG_FS controller has enumerated.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RLD
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.
CTRL
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

FRIVL[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:17 Reserved, must be kept at reset value.

Bit 16 RLDCTRL: Reload control
This bit allows dynamic reloading of the HFIR register during run time.
0: The HFIR cannot be reloaded dynamically
1: The HFIR can be dynamically reloaded during run time.
This bit needs to be programmed during initial configuration and its value must not be
changed during run time.
Caution: RLDCTRL = 0 is not recommended.
Bits 15:0 FRIVL[15:0]: Frame interval
The value that the application programs to this field, specifies the interval between two
consecutive SOFs (FS) or Keep-Alive tokens (LS). This field contains the number of PHY
clocks that constitute the required frame interval. The application can write a value to this
register only after the port enable bit of the host port control and status register (PENA bit in
OTG_HPRT) has been set. If no value is programmed, the core calculates the value based
on the PHY clock specified in the FS/LS PHY clock select field of the host configuration
register (FSLSPCS in OTG_HCFG). Do not change the value of this field after the initial
configuration, unless the RLDCTRL bit is set. In such case, the FRIVL is reloaded with each
SOF event.
– Frame interval = 1 ms × (FRIVL - 1)

RM0402 Rev 6 1027/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.23 OTG host frame number/frame time remaining register

(OTG_HFNUM)
Address offset: 0x408
Reset value: 0x0000 3FFF
This register indicates the current frame number. It also indicates the time remaining (in
terms of the number of PHY clocks) in the current frame.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

FTREM[15:0]

r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

FRNUM[15:0]

r r r r r r r r r r r r r r r r

Bits 31:16 FTREM[15:0]: Frame time remaining

Indicates the amount of time remaining in the current frame, in terms of PHY clocks. This
field decrements on each PHY clock. When it reaches zero, this field is reloaded with the
value in the Frame interval register and a new SOF is transmitted on the USB.
Bits 15:0 FRNUM[15:0]: Frame number
This field increments when a new SOF is transmitted on the USB, and is cleared to 0 when
it reaches 0x3FFF.

29.15.24 OTG_Host periodic transmit FIFO/queue status register

(OTG_HPTXSTS)
Address offset: 0x410
Reset value: 0x0008 0100
This read-only register contains the free space information for the periodic Tx FIFO and the
periodic transmit request queue.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PTXQTOP[7:0] PTXQSAV[7:0]

r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

PTXFSAVL[15:0]

r r r r r r r r r r r r r r r r

1028/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 31:24 PTXQTOP[7:0]: Top of the periodic transmit request queue

This indicates the entry in the periodic Tx request queue that is currently being processed by
the MAC.
This register is used for debugging.
Bit 31: Odd/Even frame
0: send in even frame
1: send in odd frame
Bits 30:27: Channel/endpoint number
Bits 26:25: Type
00: IN/OUT
01: Zero-length packet
11: Disable channel command
Bit 24: Terminate (last entry for the selected channel/endpoint)
Bits 23:16 PTXQSAV[7:0]: Periodic transmit request queue space available
Indicates the number of free locations available to be written in the periodic transmit request
queue. This queue holds both IN and OUT requests.
00: Periodic transmit request queue is full
01: 1 location available
10: 2 locations available
bxn: n locations available (0 ≤ n ≤ 8)
Others: Reserved
Bits 15:0 PTXFSAVL[15:0]: Periodic transmit data FIFO space available
Indicates the number of free locations available to be written to in the periodic Tx FIFO.
Values are in terms of 32-bit words
0000: Periodic Tx FIFO is full
0001: 1 word available
0010: 2 words available
bxn: n words available (where 0 ≤ n ≤ PTXFD)
Others: Reserved

29.15.25 OTG host all channels interrupt register (OTG_HAINT)

Address offset: 0x414
Reset value: 0x0000 0000
When a significant event occurs on a channel, the host all channels interrupt register
interrupts the application using the host channels interrupt bit of the core interrupt register
(HCINT bit in OTG_GINTSTS). This is shown in Figure 342. There is one interrupt bit per
channel, up to a maximum of 16 bits. Bits in this register are set and cleared when the
application sets and clears bits in the corresponding host channel-x interrupt register.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

HAINT[15:0]

r r r r r r r r r r r r r r r r

RM0402 Rev 6 1029/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 HAINT[15:0]: Channel interrupts
One bit per channel: Bit 0 for Channel 0, bit 15 for Channel 15

29.15.26 OTG host all channels interrupt mask register

(OTG_HAINTMSK)
Address offset: 0x418
Reset value: 0x0000 0000
The host all channel interrupt mask register works with the host all channel interrupt register
to interrupt the application when an event occurs on a channel. There is one interrupt mask
bit per channel, up to a maximum of 16 bits.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

HAINTM[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 HAINTM[15:0]: Channel interrupt mask
0: Masked interrupt
1: Unmasked interrupt
One bit per channel: Bit 0 for channel 0, bit 15 for channel 15

1030/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.27 OTG host port control and status register (OTG_HPRT)

Address offset: 0x440
Reset value: 0x0000 0000
This register is available only in host mode. Currently, the OTG host supports only one port.
A single register holds USB port-related information such as USB reset, enable, suspend,
resume, connect status, and test mode for each port. It is shown in Figure 342. The rc_w1
bits in this register can trigger an interrupt to the application through the host port interrupt
bit of the core interrupt register (HPRTINT bit in OTG_GINTSTS). On a port interrupt, the
application must read this register and clear the bit that caused the interrupt. For the rc_w1
bits, the application must write a 1 to the bit to clear the interrupt.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
PTCTL
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PSPD[1:0]
[3]

r r rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
POC PEN
PTCTL[2:0] PPWR PLSTS[1:0] Res. PRST PSUSP PRES POCA PENA PCDET PCSTS
CHNG CHNG

rw rw rw rw r r rw rs rw rc_w1 r rc_w1 rc_w1 rc_w1 r

Bits 31:19 Reserved, must be kept at reset value.

Bits 18:17 PSPD[1:0]: Port speed
Indicates the speed of the device attached to this port.
01: Full speed
10: Low speed
11: Reserved

Bits 16:13 PTCTL[3:0]: Port test control

The application writes a nonzero value to this field to put the port into a Test mode, and the
corresponding pattern is signaled on the port.
0000: Test mode disabled
0001: Test_J mode
0010: Test_K mode
0011: Test_SE0_NAK mode
0100: Test_Packet mode
0101: Test_Force_Enable
Others: Reserved
Bit 12 PPWR: Port power
The application uses this field to control power to this port, and the core clears this bit on an
overcurrent condition.
0: Power off
1: Power on
Bits 11:10 PLSTS[1:0]: Port line status
Indicates the current logic level USB data lines
Bit 10: Logic level of OTG_DP
Bit 11: Logic level of OTG_DM
Bit 9 Reserved, must be kept at reset value.

RM0402 Rev 6 1031/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 8 PRST: Port reset

When the application sets this bit, a reset sequence is started on this port. The application
must time the reset period and clear this bit after the reset sequence is complete.
0: Port not in reset
1: Port in reset
The application must leave this bit set for a minimum duration of at least 10 ms to start a
reset on the port. The application can leave it set for another 10 ms in addition to the
required minimum duration, before clearing the bit, even though there is no maximum limit
set by the USB standard.
High speed: 50 ms
Full speed/Low speed: 10 ms
Bit 7 PSUSP: Port suspend
The application sets this bit to put this port in suspend mode. The core only stops sending
SOFs when this is set. To stop the PHY clock, the application must set the port clock stop
bit, which asserts the suspend input pin of the PHY.
The read value of this bit reflects the current suspend status of the port. This bit is cleared
by the core after a remote wakeup signal is detected or the application sets the port reset bit
or port resume bit in this register or the resume/remote wakeup detected interrupt bit or
disconnect detected interrupt bit in the core interrupt register (WKUPINT or DISCINT in
OTG_GINTSTS, respectively).
0: Port not in suspend mode
1: Port in suspend mode
Bit 6 PRES: Port resume
The application sets this bit to drive resume signaling on the port. The core continues to
drive the resume signal until the application clears this bit.
If the core detects a USB remote wakeup sequence, as indicated by the port resume/remote
wakeup detected interrupt bit of the core interrupt register (WKUPINT bit in
OTG_GINTSTS), the core starts driving resume signaling without application intervention
and clears this bit when it detects a disconnect condition. The read value of this bit indicates
whether the core is currently driving resume signaling.
0: No resume driven
1: Resume driven
When LPM is enabled and the core is in L1 state, the behavior of this bit is as follow:
1. The application sets this bit to drive resume signaling on the port.
2. The core continues to drive the resume signal until a predetermined time specified in
BESLTHRS[3:0] field of OTG_GLPMCFG register.
3. If the core detects a USB remote wakeup sequence, as indicated by the port
L1Resume/Remote L1Wakeup detected interrupt bit of the core interrupt register
(WKUPINT in OTG_GINTSTS), the core starts driving resume signaling without application
intervention and clears this bit at the end of resume.This bit can be set or cleared by both
the core and the application. This bit is cleared by the core even if there is no device
connected to the host.
Bit 5 POCCHNG: Port overcurrent change
The core sets this bit when the status of the port overcurrent active bit (bit 4) in this register
changes.
Bit 4 POCA: Port overcurrent active
Indicates the overcurrent condition of the port.
0: No overcurrent condition
1: Overcurrent condition
Bit 3 PENCHNG: Port enable/disable change
The core sets this bit when the status of the port enable bit 2 in this register changes.

1032/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 2 PENA: Port enable

A port is enabled only by the core after a reset sequence, and is disabled by an overcurrent
condition, a disconnect condition, or by the application clearing this bit. The application
cannot set this bit by a register write. It can only clear it to disable the port. This bit does not
trigger any interrupt to the application.
0: Port disabled
1: Port enabled
Bit 1 PCDET: Port connect detected
The core sets this bit when a device connection is detected to trigger an interrupt to the
application using the host port interrupt bit in the core interrupt register (HPRTINT bit in
OTG_GINTSTS). The application must write a 1 to this bit to clear the interrupt.
Bit 0 PCSTS: Port connect status
0: No device is attached to the port
1: A device is attached to the port

29.15.28 OTG host channel x characteristics register (OTG_HCCHARx)

Address offset: 0x500 + 0x20 * x, (x = 0 to 11)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
ODD
CHENA CHDIS DAD[6:0] MCNT[1:0] EPTYP[1:0] LSDEV Res.
FRM

rs rs rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

EPDIR EPNUM[3:0] MPSIZ[10:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 CHENA: Channel enable

This field is set by the application and cleared by the OTG host.
0: Channel disabled
1: Channel enabled
Bit 30 CHDIS: Channel disable
The application sets this bit to stop transmitting/receiving data on a channel, even before
the transfer for that channel is complete. The application must wait for the Channel disabled
interrupt before treating the channel as disabled.
Bit 29 ODDFRM: Odd frame
This field is set (reset) by the application to indicate that the OTG host must perform a
transfer in an odd frame. This field is applicable for only periodic (isochronous and interrupt)
transactions.
0: Even frame
1: Odd frame
Bits 28:22 DAD[6:0]: Device address
This field selects the specific device serving as the data source or sink.

RM0402 Rev 6 1033/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 21:20 MCNT[1:0]: Multicount

This field indicates to the host the number of transactions that must be executed per frame
for this periodic endpoint. For non-periodic transfers, this field is not used
00: Reserved. This field yields undefined results
01: 1 transaction
10: 2 transactions per frame to be issued for this endpoint
11: 3 transactions per frame to be issued for this endpoint
Note: This field must be set to at least 01.
Bits 19:18 EPTYP[1:0]: Endpoint type
Indicates the transfer type selected.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 LSDEV: Low-speed device
This field is set by the application to indicate that this channel is communicating to a low-
speed device.
Bit 16 Reserved, must be kept at reset value.
Bit 15 EPDIR: Endpoint direction
Indicates whether the transaction is IN or OUT.
0: OUT
1: IN
Bits 14:11 EPNUM[3:0]: Endpoint number
Indicates the endpoint number on the device serving as the data source or sink.
Bits 10:0 MPSIZ[10:0]: Maximum packet size
Indicates the maximum packet size of the associated endpoint.

29.15.29 OTG host channel x interrupt register (OTG_HCINTx)

Address offset: 0x508 + 0x20 * x, (x = 0 to 11)
Reset value: 0x0000 0000
This register indicates the status of a channel with respect to USB- and AHB-related events.
It is shown in Figure 342. The application must read this register when the host channels
interrupt bit in the core interrupt register (HCINT bit in OTG_GINTSTS) is set. Before the
application can read this register, it must first read the host all channels interrupt
(OTG_HAINT) register to get the exact channel number for the host channel-x interrupt
register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the OTG_HAINT and OTG_GINTSTS registers.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
FRM
Res. Res. Res. Res. Res. DTERR BBERR TXERR Res. ACK NAK STALL Res. CHH XFRC
OR

rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

1034/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 31:11 Reserved, must be kept at reset value.

Bit 10 DTERR: Data toggle error.
Bit 9 FRMOR: Frame overrun.
Bit 8 BBERR: Babble error.
Bit 7 TXERR: Transaction error.
Indicates one of the following errors occurred on the USB.
CRC check failure
Timeout
Bit stuff error
False EOP
Bit 6 Reserved, must be kept at reset value.
Bit 5 ACK: ACK response received/transmitted interrupt.
Bit 4 NAK: NAK response received interrupt.
Bit 3 STALL: STALL response received interrupt.
Bit 2 Reserved, must be kept at reset value.
Bit 1 CHH: Channel halted.
Indicates the transfer completed abnormally either because of any USB transaction error or
in response to disable request by the application.

Bit 0 XFRC: Transfer completed.

Transfer completed normally without any errors.

29.15.30 OTG host channel x interrupt mask register (OTG_HCINTMSKx)

Address offset: 0x50C + 0x20 * x, (x = 0 to 11)
Reset value: 0x0000 0000
This register reflects the mask for each channel status described in the previous section.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DTERR FRM BBERR TXERR STALL XFRC
Res. Res. Res. Res. Res. Res. ACKM NAKM Res. CHHM
M ORM M M M M

rw rw rw rw rw rw rw rw rw

Bits 31:11 Reserved, must be kept at reset value.

Bit 10 DTERRM: Data toggle error mask.
0: Masked interrupt
1: Unmasked interrupt
Bit 9 FRMORM: Frame overrun mask.
0: Masked interrupt
1: Unmasked interrupt

RM0402 Rev 6 1035/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 8 BBERRM: Babble error mask.

0: Masked interrupt
1: Unmasked interrupt
Bit 7 TXERRM: Transaction error mask.
0: Masked interrupt
1: Unmasked interrupt
Bit 6 Reserved, must be kept at reset value.
Bit 5 ACKM: ACK response received/transmitted interrupt mask.
0: Masked interrupt
1: Unmasked interrupt
Bit 4 NAKM: NAK response received interrupt mask.
0: Masked interrupt
1: Unmasked interrupt
Bit 3 STALLM: STALL response received interrupt mask.
0: Masked interrupt
1: Unmasked interrupt
Bit 2 Reserved, must be kept at reset value.
Bit 1 CHHM: Channel halted mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed mask
0: Masked interrupt
1: Unmasked interrupt

29.15.31 OTG host channel x transfer size register (OTG_HCTSIZx)

Address offset: 0x510 + 0x20 * x, (x = 0 to 11)
Reset value: 0x0000 0000

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DO
DPID[1:0] PKTCNT[9:0] XFRSIZ[18:16]
PNG

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

XFRSIZ[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

1036/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 31 DOPNG: Do Ping

This bit is used only for OUT transfers. Setting this field to 1 directs the host to do PING
protocol.
Note: Do not set this bit for IN transfers. If this bit is set for IN transfers, it disables the
channel.
Bits 30:29 DPID[1:0]: Data PID
The application programs this field with the type of PID to use for the initial transaction. The
host maintains this field for the rest of the transfer.
00: DATA0
10: DATA1
11: SETUP (control) / reserved (non-control)
Bits 28:19 PKTCNT[9:0]: Packet count
This field is programmed by the application with the expected number of packets to be
transmitted (OUT) or received (IN).
The host decrements this count on every successful transmission or reception of an OUT/IN
packet. Once this count reaches zero, the application is interrupted to indicate normal
completion.
Bits 18:0 XFRSIZ[18:0]: Transfer size
For an OUT, this field is the number of data bytes the host sends during the transfer.
For an IN, this field is the buffer size that the application has reserved for the transfer. The
application is expected to program this field as an integer multiple of the maximum packet
size for IN transactions (periodic and non-periodic).

29.15.32 Device-mode registers

These registers must be programmed every time the core changes to device mode

29.15.33 OTG device configuration register (OTG_DCFG)

Address offset: 0x800
Reset value: 0x0220 0000
This register configures the core in device mode after power-on or after certain control
commands or enumeration. Do not make changes to this register after initial programming.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
ERRAT NZLSO
Res. Res. PFIVL[1:0] DAD[6:0] Res. DSPD[1:0]
IM HSK
rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 1037/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 31:16 Reserved, must be kept at reset value.

Bit 15 ERRATIM: Erratic error interrupt mask
1: Mask early suspend interrupt on erratic error
0: Early suspend interrupt is generated on erratic error
Bit 13 Reserved, must be kept at reset value.
Bits 12:11 PFIVL[1:0]: Periodic frame interval
Indicates the time within a frame at which the application must be notified using the end of
periodic frame interrupt. This can be used to determine if all the isochronous traffic for that
frame is complete.
00: 80% of the frame interval
01: 85% of the frame interval
10: 90% of the frame interval
11: 95% of the frame interval
Bits 10:4 DAD[6:0]: Device address
The application must program this field after every SetAddress control command.
Bit 3 Reserved, must be kept at reset value.
Bit 2 NZLSOHSK: Non-zero-length status OUT handshake
The application can use this field to select the handshake the core sends on receiving a
nonzero-length data packet during the OUT transaction of a control transfer’s status stage.
1:Send a STALL handshake on a nonzero-length status OUT transaction and do not send
the received OUT packet to the application.
0:Send the received OUT packet to the application (zero-length or nonzero-length) and send
a handshake based on the NAK and STALL bits for the endpoint in the device endpoint
control register.
Bits 1:0 DSPD[1:0]: Device speed
Indicates the speed at which the application requires the core to enumerate, or the
maximum speed the application can support. However, the actual bus speed is determined
only after the chirp sequence is completed, and is based on the speed of the USB host to
which the core is connected.
00: Reserved
01: Reserved
10: Reserved
11: Full speed (USB 1.1 transceiver clock is 48 MHz)

29.15.34 OTG device control register (OTG_DCTL)

Address offset: 0x804
Reset value: 0x0000 0002

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
DS
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. BESL Res. Res.
RJCT
rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PO
CGO SGO CGI SGI GON GIN RWU
Res. Res. Res. Res. PRG TCTL[2:0] SDIS
NAK NAK NAK NAK STS STS SIG
DNE
rw w w w w rw rw rw r r rw rw

1038/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 31:19 Reserved, must be kept at reset value.

Bit 18 DSBESLRJCT: Deep sleep BESL reject
Core rejects LPM request with BESL value greater than BESL threshold programmed.
NYET response is sent for LPM tokens with BESL value greater than BESL threshold. By
default, the deep sleep BESL reject feature is disabled.
Bits 17:12 Reserved, must be kept at reset value.
Bit 11 POPRGDNE: Power-on programming done
The application uses this bit to indicate that register programming is completed after a
wakeup from power down mode.
Bit 10 CGONAK: Clear global OUT NAK
Writing 1 to this field clears the Global OUT NAK.
Bit 9 SGONAK: Set global OUT NAK
Writing 1 to this field sets the Global OUT NAK.
The application uses this bit to send a NAK handshake on all OUT endpoints.
The application must set the this bit only after making sure that the Global OUT NAK
effective bit in the core interrupt register (GONAKEFF bit in OTG_GINTSTS) is cleared.
Bit 8 CGINAK: Clear global IN NAK
Writing 1 to this field clears the Global IN NAK.
Bit 7 SGINAK: Set global IN NAK
Writing 1 to this field sets the Global non-periodic IN NAK.The application uses this bit to
send a NAK handshake on all non-periodic IN endpoints.
The application must set this bit only after making sure that the Global IN NAK effective bit
in the core interrupt register (GINAKEFF bit in OTG_GINTSTS) is cleared.
Bits 6:4 TCTL[2:0]: Test control
000: Test mode disabled
001: Test_J mode
010: Test_K mode
011: Test_SE0_NAK mode
100: Test_Packet mode
101: Test_Force_Enable
Others: Reserved
Bit 3 GONSTS: Global OUT NAK status
0:A handshake is sent based on the FIFO status and the NAK and STALL bit settings.
1:No data is written to the Rx FIFO, irrespective of space availability. Sends a NAK
handshake on all packets, except on SETUP transactions. All isochronous OUT packets are
dropped.

RM0402 Rev 6 1039/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 2 GINSTS: Global IN NAK status

0:A handshake is sent out based on the data availability in the transmit FIFO.
1:A NAK handshake is sent out on all non-periodic IN endpoints, irrespective of the data
availability in the transmit FIFO.
Bit 1 SDIS: Soft disconnect
The application uses this bit to signal the USB OTG core to perform a soft disconnect. As
long as this bit is set, the host does not see that the device is connected, and the device
does not receive signals on the USB. The core stays in the disconnected state until the
application clears this bit.
0:Normal operation. When this bit is cleared after a soft disconnect, the core generates a
device connect event to the USB host. When the device is reconnected, the USB host
restarts device enumeration.
1:The core generates a device disconnect event to the USB host.
Bit 0 RWUSIG: Remote wakeup signaling
When the application sets this bit, the core initiates remote signaling to wake up the USB
host. The application must set this bit to instruct the core to exit the suspend state. As
specified in the USB 2.0 specification, the application must clear this bit 1 ms to 15 ms after
setting it.
If LPM is enabled and the core is in the L1 (sleep) state, when the application sets this bit,
the core initiates L1 remote signaling to wake up the USB host. The application must set
this bit to instruct the core to exit the sleep state. As specified in the LPM specification, the
hardware automatically clears this bit 50 µs (TL1DevDrvResume) after being set by the
application. The application must not set this bit when bRemoteWake from the previous
LPM transaction is zero (refer to REMWAKE bit in GLPMCFG register).

Table 207 contains the minimum duration (according to device state) for which the Soft
disconnect (SDIS) bit must be set for the USB host to detect a device disconnect. To
accommodate clock jitter, it is recommended that the application add some extra delay to
the specified minimum duration.

Table 207. Minimum duration for soft disconnect

Operating speed Device state Minimum duration

Full speed Suspended 1 ms + 2.5 µs

Full speed Idle 2.5 µs
Full speed Not Idle or suspended (Performing transactions) 2.5 µs

1040/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.35 OTG device status register (OTG_DSTS)

Address offset: 0x808
Reset value: 0x0000 0010
This register indicates the status of the core with respect to USB-related events. It must be
read on interrupts from the device all interrupts (OTG_DAINT) register.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. DEVLNSTS[1:0] FNSOF[13:8]

r r r r r r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
SUSP
FNSOF[7:0] Res. Res. Res. Res. EERR ENUMSPD[1:0]
STS

r r r r r r r r r r r r

Bits 31:24 Reserved, must be kept at reset value.

Bits 23:22 DEVLNSTS[1:0]: Device line status
Indicates the current logic level USB data lines.
Bit [23]: Logic level of D+
Bit [22]: Logic level of D-
Bits 21:8 FNSOF[13:0]: Frame number of the received SOF
Bits 7:4 Reserved, must be kept at reset value.
Bit 3 EERR: Erratic error
The core sets this bit to report any erratic errors.
Due to erratic errors, the OTG_FS controller goes into suspended state and an interrupt is
generated to the application with Early suspend bit of the OTG_GINTSTS register (ESUSP
bit in OTG_GINTSTS). If the early suspend is asserted due to an erratic error, the application
can only perform a soft disconnect recover.
Bits 2:1 ENUMSPD[1:0]: Enumerated speed
Indicates the speed at which the OTG_FS controller has come up after speed detection
through a chirp sequence.
11: Full speed using embedded FS PHY
Others: reserved
Bit 0 SUSPSTS: Suspend status
In device mode, this bit is set as long as a suspend condition is detected on the USB. The
core enters the suspended state when there is no activity on the USB data lines for a period
of 3 ms. The core comes out of the suspend:
– When there is an activity on the USB data lines
– When the application writes to the remote wakeup signaling bit in the OTG_DCTL register
(RWUSIG bit in OTG_DCTL).

RM0402 Rev 6 1041/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.36 OTG device IN endpoint common interrupt mask register

(OTG_DIEPMSK)
Address offset: 0x810
Reset value: 0x0000 0000
This register works with each of the OTG_DIEPINTx registers for all endpoints to generate
an interrupt per IN endpoint. The IN endpoint interrupt for a specific status in the
OTG_DIEPINTx register can be masked by writing to the corresponding bit in this register.
Status bits are masked by default.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
INEPN INEPN ITTXFE XFRC
Res. Res. NAKM Res. Res. Res. Res. Res. Res. TOM Res. EPDM
EM MM MSK M

rw rw rw rw rw rw rw

Bits 31:14 Reserved, must be kept at reset value.

Bit 13 NAKM: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bits 12:10 Reserved, must be kept at reset value.
Bit 9 Reserved, must be kept at reset value.
Bit 8 Reserved, must be kept at reset value.
Bit 7 Reserved, must be kept at reset value.
Bit 6 INEPNEM: IN endpoint NAK effective mask
0: Masked interrupt
1: Unmasked interrupt
Bit 5 INEPNMM: IN token received with EP mismatch mask
0: Masked interrupt
1: Unmasked interrupt
Bit 4 ITTXFEMSK: IN token received when Tx FIFO empty mask
0: Masked interrupt
1: Unmasked interrupt
Bit 3 TOM: Timeout condition mask (Non-isochronous endpoints)
0: Masked interrupt
1: Unmasked interrupt

1042/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 2 Reserved, must be kept at reset value.

Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

29.15.37 OTG device OUT endpoint common interrupt mask register

(OTG_DOEPMSK)
Address offset: 0x814
Reset value: 0x0000 0000
This register works with each of the OTG_DOEPINTx registers for all endpoints to generate
an interrupt per OUT endpoint. The OUT endpoint interrupt for a specific status in the
OTG_DOEPINTx register can be masked by writing into the corresponding bit in this
register. Status bits are masked by default.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
OUT STS
NAK BERR OTEPD XFRC
Res. Res. Res. Res. Res. PKT Res. Res. PHSR STUPM Res. EPDM
MSK M M M
ERRM XM
rw rw rw rw rw rw rw rw

Bits 31:14 Reserved, must be kept at reset value.

Bit 13 NAKMSK: NAK interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 12 BERRM: Babble error interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bits 11:10 Reserved, must be kept at reset value.
Bit 9 Reserved, must be kept at reset value.
Bit 8 OUTPKTERRM: Out packet error mask
0: Masked interrupt
1: Unmasked interrupt
Bit 7 Reserved, must be kept at reset value.
Bit 6 Reserved, must be kept at reset value.
Bit 5 STSPHSRXM: Status phase received for control write mask
0: Masked interrupt
1: Unmasked interrupt

RM0402 Rev 6 1043/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 4 OTEPDM: OUT token received when endpoint disabled mask. Applies to control OUT
endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 3 STUPM: STUPM: SETUP phase done mask. Applies to control endpoints only.
0: Masked interrupt
1: Unmasked interrupt
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDM: Endpoint disabled interrupt mask
0: Masked interrupt
1: Unmasked interrupt
Bit 0 XFRCM: Transfer completed interrupt mask
0: Masked interrupt
1: Unmasked interrupt

29.15.38 OTG device all endpoints interrupt register (OTG_DAINT)

Address offset: 0x818
Reset value: 0x0000 0000
When a significant event occurs on an endpoint, a OTG_DAINT register interrupts the
application using the device OUT endpoints interrupt bit or device IN endpoints interrupt bit
of the OTG_GINTSTS register (OEPINT or IEPINT in OTG_GINTSTS, respectively). There
is one interrupt bit per endpoint, up to a maximum of 16 bits for OUT endpoints and 16 bits
for IN endpoints. For a bidirectional endpoint, the corresponding IN and OUT interrupt bits
are used. Bits in this register are set and cleared when the application sets and clears bits in
the corresponding device endpoint-x interrupt register (OTG_DIEPINTx/OTG_DOEPINTx).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

OEPINT[15:0]

r r r r r r r r r r r r r r r r
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

IEPINT[15:0]

r r r r r r r r r r r r r r r r

Bits 31:16 OEPINT[15:0]: OUT endpoint interrupt bits

One bit per OUT endpoint:
Bit 16 for OUT endpoint 0, bit 19 for OUT endpoint 3.
Bits 15:0 IEPINT[15:0]: IN endpoint interrupt bits
One bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 3 for endpoint 3.

1044/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.39 OTG all endpoints interrupt mask register

(OTG_DAINTMSK)
Address offset: 0x81C
Reset value: 0x0000 0000
The OTG_DAINTMSK register works with the device endpoint interrupt register to interrupt
the application when an event occurs on a device endpoint. However, the OTG_DAINT
register bit corresponding to that interrupt is still set.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

OEPM[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

IEPM[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 OEPM[15:0]: OUT EP interrupt mask bits

One per OUT endpoint:
Bit 16 for OUT EP 0, bit 19 for OUT EP 3
0: Masked interrupt
1: Unmasked interrupt
Bits 15:0 IEPM[15:0]: IN EP interrupt mask bits
One bit per IN endpoint:
Bit 0 for IN EP 0, bit 3 for IN EP 3
0: Masked interrupt
1: Unmasked interrupt

29.15.40 OTG device VBUS discharge time register

(OTG_DVBUSDIS)
Address offset: 0x0828
Reset value: 0x0000 17D7
This register specifies the VBUS discharge time after VBUS pulsing during SRP.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

VBUSDT[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

RM0402 Rev 6 1045/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 VBUSDT[15:0]: Device VBUS discharge time
Specifies the VBUS discharge time after VBUS pulsing during SRP. This value equals:
VBUS discharge time in PHY clocks / 1 024
Depending on your VBUS load, this value may need adjusting.

29.15.41 OTG device VBUS pulsing time register

(OTG_DVBUSPULSE)
Address offset: 0x082C
Reset value: 0x0000 05B8
This register specifies the VBUS pulsing time during SRP.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

DVBUSP[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 DVBUSP[15:0]: Device VBUS pulsing time. This feature is only relevant to OTG1.3.
Specifies the VBUS pulsing time during SRP. This value equals:
VBUS pulsing time in PHY clocks / 1 024

29.15.42 OTG device IN endpoint FIFO empty interrupt mask register

(OTG_DIEPEMPMSK)
Address offset: 0x834
Reset value: 0x0000 0000
This register is used to control the IN endpoint FIFO empty interrupt generation
(TXFE_OTG_DIEPINTx).

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

INEPTXFEM[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

1046/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 INEPTXFEM[15:0]: IN EP Tx FIFO empty interrupt mask bits
These bits act as mask bits for OTG_DIEPINTx.
TXFE interrupt one bit per IN endpoint:
Bit 0 for IN endpoint 0, bit 3 for IN endpoint 3
0: Masked interrupt
1: Unmasked interrupt

29.15.43 OTG device control IN endpoint 0 control register

(OTG_DIEPCTL0)
Address offset: 0x900
Reset value: 0x0000 0000
This section describes the OTG_DIEPCTL0 register for USB_OTG FS. Nonzero control
endpoints use registers for endpoints 1–3.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
NAK
EPENA EPDIS Res. Res. SNAK CNAK TXFNUM[3:0] STALL Res. EPTYP Res.
STS

rs rs w w rw rw rw rw rs r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USBA
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. MPSIZ[1:0]
EP

r rw rw

Bit 31 EPENA: Endpoint enable

The application sets this bit to start transmitting data on the endpoint 0.
The core clears this bit before setting any of the following interrupts on this endpoint:
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting data on an endpoint, even before the
transfer for that endpoint is complete. The application must wait for the endpoint disabled
interrupt before treating the endpoint as disabled. The core clears this bit before setting the
endpoint disabled interrupt. The application must set this bit only if endpoint enable is
already set for this endpoint.
Bits 29:28 Reserved, must be kept at reset value.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for an endpoint after a SETUP packet is received on
that endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 TXFNUM[3:0]: Tx FIFO number
This value is set to the FIFO number that is assigned to IN endpoint 0.

RM0402 Rev 6 1047/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 21 STALL: STALL handshake

The application can only set this bit, and the core clears it when a SETUP token is received
for this endpoint. If a NAK bit, a Global IN NAK or Global OUT NAK is set along with this bit,
the STALL bit takes priority.
Bit 20 Reserved, must be kept at reset value.
Bits 19:18 EPTYP: Endpoint type
Hardcoded to ‘00’ for control.
Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status
1: The core is transmitting NAK handshakes on this endpoint.
When this bit is set, either by the application or core, the core stops transmitting data, even
if there are data available in the Tx FIFO. Irrespective of this bit’s setting, the core always
responds to SETUP data packets with an ACK handshake.
Bit 16 Reserved, must be kept at reset value.
Bit 15 USBAEP: USB active endpoint
This bit is always set to 1, indicating that control endpoint 0 is always active in all
configurations and interfaces.
Bits 14:2 Reserved, must be kept at reset value.
Bits 1:0 MPSIZ[1:0]: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes

29.15.44 OTG device IN endpoint x control register (OTG_DIEPCTLx)

Address offset: 0x900 + 0x20 * x, (x = 1 to 5)
Reset value: 0x0000 0000
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SD0
EO
SODD PID/ NAK
EPENA EPDIS SNAK CNAK TXFNUM[3:0] STALL Res. EPTYP[1:0] NUM/
FRM SEVN STS
DPID
FRM

rs rs w w w w rw rw rw rw rw rw rw r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USBA
Res. Res. Res. Res. MPSIZ[10:0]
EP

rw rw rw rw rw rw rw rw rw rw rw rw

1048/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 31 EPENA: Endpoint enable

The application sets this bit to start transmitting data on an endpoint.
The core clears this bit before setting any of the following interrupts on this endpoint:
– SETUP phase done
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the endpoint disabled interrupt. The application must set this bit only if endpoint
enable is already set for this endpoint.
Bit 29 SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to odd frame.
Bit 28 SD0PID: Set DATA0 PID
Applies to interrupt/bulk IN endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous IN endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for OUT endpoints on a transfer completed interrupt,
or after a SETUP is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 TXFNUM[3:0]: Tx FIFO number
These bits specify the FIFO number associated with this endpoint. Each active IN endpoint
must be programmed to a separate FIFO number.
This field is valid only for IN endpoints.
Bit 21 STALL: STALL handshake
Applies to non-control, non-isochronous IN endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes priority.
Only the application can clear this bit, never the core.
Bit 20 Reserved, must be kept at reset value.
Bits 19:18 EPTYP[1:0]: Endpoint type
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt

RM0402 Rev 6 1049/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 17 NAKSTS: NAK status

It indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit:
For non-isochronous IN endpoints: The core stops transmitting any data on an IN endpoint,
even if there are data available in the Tx FIFO.
For isochronous IN endpoints: The core sends out a zero-length data packet, even if there
are data available in the Tx FIFO.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.
Bit 16 EONUM: Even/odd frame
Applies to isochronous IN endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk IN endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1
Bit 15 USBAEP: USB active endpoint
Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
Bits 14:11
Reserved, must be kept at reset value.

Bits 10:0 MPSIZ[10:0]: Maximum packet size

The application must program this field with the maximum packet size for the current logical
endpoint. This value is in bytes.

1050/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.45 OTG device IN endpoint x interrupt register (OTG_DIEPINTx)

Address offset: 0x908 + 0x20 * x, (x = 0 to 5)
Reset value: 0x0000 0080
This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 342. The application must read this register when the IN
endpoints interrupt bit of the core interrupt register (IEPINT in OTG_GINTSTS) is set.
Before the application can read this register, it must first read the device all endpoints
interrupt (OTG_DAINT) register to get the exact endpoint number for the device endpoint-x
interrupt register. The application must clear the appropriate bit in this register to clear the
corresponding bits in the OTG_DAINT and OTG_GINTSTS registers.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PKTD IN IN EP
Res. Res. NAK Res. Res. Res. Res. TXFE ITTXFE TOC Res. XFRC
RPSTS EPNE EPNM DISD

rc_w1 rc_w1 r rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bits 31:14 Reserved, must be kept at reset value.

Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 Reserved, must be kept at reset value.
Bit 11 PKTDRPSTS: Packet dropped status
This bit indicates to the application that an ISOC OUT packet has been dropped. This bit
does not have an associated mask bit and does not generate an interrupt.
Bit 10 Reserved, must be kept at reset value.
Bit 9 Reserved, must be kept at reset value.
Bit 8 Reserved, must be kept at reset value.
Bit 7 TXFE: Transmit FIFO empty
This interrupt is asserted when the Tx FIFO for this endpoint is either half or completely
empty. The half or completely empty status is determined by the Tx FIFO Empty Level bit in
the OTG_GAHBCFG register (TXFELVL bit in OTG_GAHBCFG).
Bit 6 INEPNE: IN endpoint NAK effective
This bit can be cleared when the application clears the IN endpoint NAK by writing to the
CNAK bit in OTG_DIEPCTLx.
This interrupt indicates that the core has sampled the NAK bit set (either by the application
or by the core). The interrupt indicates that the IN endpoint NAK bit set by the application
has taken effect in the core.
This interrupt does not guarantee that a NAK handshake is sent on the USB. A STALL bit
takes priority over a NAK bit.

RM0402 Rev 6 1051/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 5 INEPNM: IN token received with EP mismatch

Indicates that the data in the top of the non-periodic TxFIFO belongs to an endpoint other
than the one for which the IN token was received. This interrupt is asserted on the endpoint
for which the IN token was received.
Bit 4 ITTXFE: IN token received when Tx FIFO is empty
Indicates that an IN token was received when the associated Tx FIFO (periodic/non-
periodic) was empty. This interrupt is asserted on the endpoint for which the IN token was
received.
Bit 3 TOC: Timeout condition
Indicates that the core has detected a timeout condition on the USB for the last IN token on
this endpoint.
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.

29.15.46 OTG device IN endpoint 0 transfer size register

(OTG_DIEPTSIZ0)
Address offset: 0x910
Reset value: 0x0000 0000
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the endpoint enable bit in the device control endpoint 0 control registers
(EPENA in OTG_DIEPCTL0), the core modifies this register. The application can only read
this register once the core has cleared the endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–3.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. PKTCNT[1:0] Res. Res. Res.

rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. XFRSIZ[6:0]

rw rw rw rw rw rw rw

Bits 31:21 Reserved, must be kept at reset value.

1052/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 20:19 PKTCNT[1:0]: Packet count

Indicates the total number of USB packets that constitute the transfer size amount of data for
endpoint 0.
This field is decremented every time a packet (maximum size or short packet) is read from
the Tx FIFO.
Bits 18:7 Reserved, must be kept at reset value.
Bits 6:0 XFRSIZ[6:0]: Transfer size
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet from the external memory is written to
the Tx FIFO.

29.15.47 OTG device IN endpoint transmit FIFO status register

(OTG_DTXFSTSx)
Address offset: 0x918 + 0x20 * x, (x = 0 to 5)
Reset value: 0x0000 0200
This read-only register contains the free space information for the device IN endpoint Tx
FIFO.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

INEPTFSAV[15:0]

r r r r r r r r r r r r r r r r

Bits 31:16 Reserved, must be kept at reset value.

Bits 15:0 INEPTFSAV[15:0]: IN endpoint Tx FIFO space available
Indicates the amount of free space available in the endpoint Tx FIFO.
Values are in terms of 32-bit words:
0x0: Endpoint Tx FIFO is full
0x1: 1 word available
0x2: 2 words available
0xn: n words available
Others: Reserved

RM0402 Rev 6 1053/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.48 OTG device IN endpoint x transfer size register (OTG_DIEPTSIZx)

Address offset: 0x910 + 0x20 * x, (x = 1 to 5)
Reset value: 0x0000 0000
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using the endpoint enable bit in the OTG_DIEPCTLx registers (EPENA bit in
OTG_DIEPCTLx), the core modifies this register. The application can only read this register
once the core has cleared the endpoint enable bit.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. MCNT[1:0] PKTCNT[9:0] XFRSIZ[18:16]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

XFRSIZ[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.

Bits 30:29 MCNT[1:0]: Multi count
For periodic IN endpoints, this field indicates the number of packets that must be transmitted
per frame on the USB. The core uses this field to calculate the data PID for isochronous IN
endpoints.
01: 1 packet
10: 2 packets
11: 3 packets
Bits 28:19 PKTCNT[9:0]: Packet count
Indicates the total number of USB packets that constitute the transfer size amount of data for
this endpoint.
This field is decremented every time a packet (maximum size or short packet) is read from
the Tx FIFO.
Bits 18:0 XFRSIZ[18:0]: Transfer size
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet from the external memory is written to the
Tx FIFO.

1054/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.49 OTG device control OUT endpoint 0 control register

(OTG_DOEPCTL0)
Address offset: 0xB00
Reset value: 0x0000 8000
This section describes the OTG_DOEPCTL0 register. Nonzero control endpoints use
registers for endpoints 1–3.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
NAK
EPENA EPDIS Res. Res. SNAK CNAK Res. Res. Res. Res. STALL SNPM EPTYP[1:0] Res.
STS

w r w w rs rw r r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USBA
Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. MPSIZ[1:0]
EP

r r r

Bit 31 EPENA: Endpoint enable

The application sets this bit to start transmitting data on endpoint 0.
The core clears this bit before setting any of the following interrupts on this endpoint:
– SETUP phase done
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application cannot disable control OUT endpoint 0.
Bits 29:28 Reserved, must be kept at reset value.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit on a transfer completed interrupt, or after a SETUP
is received on the endpoint.
Bit 26 CNAK: Clear NAK
A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 Reserved, must be kept at reset value.
Bit 21 STALL: STALL handshake
The application can only set this bit, and the core clears it, when a SETUP token is received
for this endpoint. If a NAK bit or Global OUT NAK is set along with this bit, the STALL bit
takes priority. Irrespective of this bit’s setting, the core always responds to SETUP data
packets with an ACK handshake.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Bits 19:18 EPTYP[1:0]: Endpoint type
Hardcoded to 2’b00 for control.

RM0402 Rev 6 1055/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 17 NAKSTS: NAK status

Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit, the core stops receiving data, even if
there is space in the Rx FIFO to accommodate the incoming packet. Irrespective of this bit’s
setting, the core always responds to SETUP data packets with an ACK handshake.
Bit 16 Reserved, must be kept at reset value.
Bit 15 USBAEP: USB active endpoint
This bit is always set to 1, indicating that a control endpoint 0 is always active in all
configurations and interfaces.
Bits 14:2 Reserved, must be kept at reset value.
Bits 1:0 MPSIZ[1:0]: Maximum packet size
The maximum packet size for control OUT endpoint 0 is the same as what is programmed in
control IN endpoint 0.
00: 64 bytes
01: 32 bytes
10: 16 bytes
11: 8 bytes

29.15.50 OTG device OUT endpoint x interrupt register (OTG_DOEPINTx)

Address offset: 0xB08 + 0x20 * x, (x = 0 to 5)
Reset value: 0x0000 0080
This register indicates the status of an endpoint with respect to USB- and AHB-related
events. It is shown in Figure 342. The application must read this register when the OUT
endpoints interrupt bit of the OTG_GINTSTS register (OEPINT bit in OTG_GINTSTS) is set.
Before the application can read this register, it must first read the OTG_DAINT register to
get the exact endpoint number for the OTG_DOEPINTx register. The application must clear
the appropriate bit in this register to clear the corresponding bits in the OTG_DAINT and
OTG_GINTSTS registers.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
STSPH OTEP EP
Res. Res. NAK BERR Res. Res. Res. Res. Res. Res. STUP Res. XFRC
SRX DIS DISD

rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1 rc_w1

Bits 31:14 Reserved, must be kept at reset value.

Bit 13 NAK: NAK input
The core generates this interrupt when a NAK is transmitted or received by the device. In
case of isochronous IN endpoints the interrupt gets generated when a zero length packet is
transmitted due to unavailability of data in the Tx FIFO.
Bit 12 BERR: Babble error interrupt
The core generates this interrupt when babble is received for the endpoint.

1056/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bits 11:10 Reserved, must be kept at reset value.

Bit 9 Reserved, must be kept at reset value.
Bit 8 Reserved, must be kept at reset value.
Bit 7 Reserved, must be kept at reset value.
Bit 6 Reserved, must be kept at reset value.
Bit 5 STSPHSRX: Status phase received for control write
This interrupt is valid only for control OUT endpoints. This interrupt is generated only after
OTG_FS has transferred all the data that the host has sent during the data phase of a
control write transfer, to the system memory buffer. The interrupt indicates to the application
that the host has switched from data phase to the status phase of a control write transfer.
The application can use this interrupt to ACK or STALL the status phase, after it has
decoded the data phase.
Bit 4 OTEPDIS: OUT token received when endpoint disabled
Applies only to control OUT endpoints.
Indicates that an OUT token was received when the endpoint was not yet enabled. This
interrupt is asserted on the endpoint for which the OUT token was received.
Bit 3 STUP: SETUP phase done
Applies to control OUT endpoint only.Indicates that the SETUP phase for the control
endpoint is complete and no more back-to-back SETUP packets were received for the
current control transfer. On this interrupt, the application can decode the received SETUP
data packet.
Bit 2 Reserved, must be kept at reset value.
Bit 1 EPDISD: Endpoint disabled interrupt
This bit indicates that the endpoint is disabled per the application’s request.
Bit 0 XFRC: Transfer completed interrupt
This field indicates that the programmed transfer is complete on the AHB as well as on the
USB, for this endpoint.

RM0402 Rev 6 1057/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.15.51 OTG device OUT endpoint 0 transfer size register

(OTG_DOEPTSIZ0)
Address offset: 0xB10
Reset value: 0x0000 0000
The application must modify this register before enabling endpoint 0. Once endpoint 0 is
enabled using the endpoint enable bit in the OTG_DOEPCTL0 registers (EPENA bit in
OTG_DOEPCTL0), the core modifies this register. The application can only read this
register once the core has cleared the endpoint enable bit.
Nonzero endpoints use the registers for endpoints 1–5.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

PKTCNT
Res. STUPCNT[1:0] Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

Res. Res. Res. Res. Res. Res. Res. Res. Res. XFRSIZ[6:0]

rw rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.

Bits 30:29 STUPCNT[1:0]: SETUP packet count
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bits 28:20 Reserved, must be kept at reset value.
Bit 19 PKTCNT: Packet count
This field is decremented to zero after a packet is written into the Rx FIFO.
Bits 18:7 Reserved, must be kept at reset value.
Bits 6:0 XFRSIZ[6:0]: Transfer size
Indicates the transfer size in bytes for endpoint 0. The core interrupts the application only
after it has exhausted the transfer size amount of data. The transfer size can be set to the
maximum packet size of the endpoint, to be interrupted at the end of each packet.
The core decrements this field every time a packet is read from the Rx FIFO and written to
the external memory.

1058/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.15.52 OTG device OUT endpoint x control register

(OTG_DOEPCTLx)
Address offset: 0xB00 + 0x20 * x, (x = 1 to 5)
Reset value: 0x0000 0000
The application uses this register to control the behavior of each logical endpoint other than
endpoint 0.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
SD1 SD0
EO
PID/ PID/ NAK
EPENA EPDIS SNAK CNAK Res. Res. Res. Res. STALL SNPM EPTYP[1:0] NUM/
SODD SEVN STS
DPID
FRM FRM

rs rs w w w w rw/rs rw rw rw r r

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
USBA
Res. Res. Res. Res. MPSIZ[10:0]
EP

rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 EPENA: Endpoint enable

Applies to IN and OUT endpoints.
The application sets this bit to start transmitting data on an endpoint.
The core clears this bit before setting any of the following interrupts on this endpoint:
– SETUP phase done
– Endpoint disabled
– Transfer completed
Bit 30 EPDIS: Endpoint disable
The application sets this bit to stop transmitting/receiving data on an endpoint, even before
the transfer for that endpoint is complete. The application must wait for the endpoint
disabled interrupt before treating the endpoint as disabled. The core clears this bit before
setting the endpoint disabled interrupt. The application must set this bit only if endpoint
enable is already set for this endpoint.
Bit 29 SD1PID: Set DATA1 PID
Applies to interrupt/bulk IN and OUT endpoints only. Writing to this field sets the endpoint
data PID (DPID) field in this register to DATA1.
SODDFRM: Set odd frame
Applies to isochronous IN and OUT endpoints only. Writing to this field sets the Even/Odd
frame (EONUM) field to odd frame.
Bit 28 SD0PID: Set DATA0 PID
Applies to interrupt/bulk OUT endpoints only.
Writing to this field sets the endpoint data PID (DPID) field in this register to DATA0.
SEVNFRM: Set even frame
Applies to isochronous OUT endpoints only.
Writing to this field sets the Even/Odd frame (EONUM) field to even frame.
Bit 27 SNAK: Set NAK
A write to this bit sets the NAK bit for the endpoint.
Using this bit, the application can control the transmission of NAK handshakes on an
endpoint. The core can also set this bit for OUT endpoints on a transfer completed interrupt,
or after a SETUP is received on the endpoint.

RM0402 Rev 6 1059/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bit 26 CNAK: Clear NAK

A write to this bit clears the NAK bit for the endpoint.
Bits 25:22 Reserved, must be kept at reset value.
Bit 21 STALL: STALL handshake
Applies to non-control, non-isochronous OUT endpoints only (access type is rw).
The application sets this bit to stall all tokens from the USB host to this endpoint. If a NAK
bit, Global IN NAK, or Global OUT NAK is set along with this bit, the STALL bit takes
priority. Only the application can clear this bit, never the core.
Applies to control endpoints only (access type is rs).
The application can only set this bit, and the core clears it, when a SETUP token is received
for this endpoint. If a NAK bit, Global IN NAK, or Global OUT NAK is set along with this bit,
the STALL bit takes priority. Irrespective of this bit’s setting, the core always responds to
SETUP data packets with an ACK handshake.
Bit 20 SNPM: Snoop mode
This bit configures the endpoint to Snoop mode. In Snoop mode, the core does not check
the correctness of OUT packets before transferring them to application memory.
Bits 19:18 EPTYP[1:0]: Endpoint type
This is the transfer type supported by this logical endpoint.
00: Control
01: Isochronous
10: Bulk
11: Interrupt
Bit 17 NAKSTS: NAK status
Indicates the following:
0: The core is transmitting non-NAK handshakes based on the FIFO status.
1: The core is transmitting NAK handshakes on this endpoint.
When either the application or the core sets this bit:
The core stops receiving any data on an OUT endpoint, even if there is space in the Rx
FIFO to accommodate the incoming packet.
Irrespective of this bit’s setting, the core always responds to SETUP data packets with an
ACK handshake.
Bit 16 EONUM: Even/odd frame
Applies to isochronous IN and OUT endpoints only.
Indicates the frame number in which the core transmits/receives isochronous data for this
endpoint. The application must program the even/odd frame number in which it intends to
transmit/receive isochronous data for this endpoint using the SEVNFRM and SODDFRM
fields in this register.
0: Even frame
1: Odd frame
DPID: Endpoint data PID
Applies to interrupt/bulk OUT endpoints only.
Contains the PID of the packet to be received or transmitted on this endpoint. The
application must program the PID of the first packet to be received or transmitted on this
endpoint, after the endpoint is activated. The application uses the SD0PID register field to
program either DATA0 or DATA1 PID.
0: DATA0
1: DATA1

1060/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 15 USBAEP: USB active endpoint

Indicates whether this endpoint is active in the current configuration and interface. The core
clears this bit for all endpoints (other than EP 0) after detecting a USB reset. After receiving
the SetConfiguration and SetInterface commands, the application must program endpoint
registers accordingly and set this bit.
Bits 14:11 Reserved, must be kept at reset value.
Bits 10:0 MPSIZ[10:0]: Maximum packet size
The application must program this field with the maximum packet size for the current logical
endpoint. This value is in bytes.

29.15.53 OTG device OUT endpoint x transfer size register

(OTG_DOEPTSIZx)
Address offset: 0xB10 + 0x20 * x, (x = 1 to 5)
Reset value: 0x0000 0000
The application must modify this register before enabling the endpoint. Once the endpoint is
enabled using endpoint enable bit of the OTG_DOEPCTLx registers (EPENA bit in
OTG_DOEPCTLx), the core modifies this register. The application can only read this
register once the core has cleared the endpoint enable bit.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
RXDPID/
Res. PKTCNT[9:0] XFRSIZ[18:16]
STUPCNT[1:0]

r/rw r/rw rw rw rw rw rw rw rw rw rw rw rw rw rw

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

XFRSIZ[15:0]

rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw

Bit 31 Reserved, must be kept at reset value.

RM0402 Rev 6 1061/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Bits 30:29 RXDPID[1:0]: Received data PID

Applies to isochronous OUT endpoints only.
This is the data PID received in the last packet for this endpoint.
00: DATA0
10: DATA1

STUPCNT[1:0]: SETUP packet count

Applies to control OUT endpoints only.
This field specifies the number of back-to-back SETUP data packets the endpoint can
receive.
01: 1 packet
10: 2 packets
11: 3 packets
Bits 28:19 PKTCNT[9:0]: Packet count
Indicates the total number of USB packets that constitute the transfer size amount of data for
this endpoint.
This field is decremented every time a packet (maximum size or short packet) is written to
the Rx FIFO.
Bits 18:0 XFRSIZ[18:0]: Transfer size
This field contains the transfer size in bytes for the current endpoint. The core only interrupts
the application after it has exhausted the transfer size amount of data. The transfer size can
be set to the maximum packet size of the endpoint, to be interrupted at the end of each
packet.
The core decrements this field every time a packet is read from the Rx FIFO and written to
the external memory.

29.15.54 OTG power and clock gating control register (OTG_PCGCCTL)

Address offset: 0xE00
Reset value: 0x200B 8000
This register is available in host and device modes.

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16

Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res. Res.

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
PHY ENL1 PHY GATE STPP
Res. Res. Res. Res. Res. Res. Res. Res. SUSP Res. Res.
SLEEP GTG SUSP HCLK CLK

r r rw r rw rw

Bits 31:8 Reserved, must be kept at reset value.

Bit 7 SUSP: Deep Sleep
This bit indicates that the PHY is in Deep Sleep when in L1 state.
Bit 6 PHYSLEEP: PHY in Sleep
This bit indicates that the PHY is in the Sleep state.

1062/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Bit 5 ENL1GTG: Enable sleep clock gating

When this bit is set, core internal clock gating is enabled in Sleep state if the core cannot
assert utmi_l1_suspend_n. When this bit is not set, the PHY clock is not gated in Sleep
state.
Bit 4 PHYSUSP: PHY suspended
Indicates that the PHY has been suspended. This bit is updated once the PHY is suspended
after the application has set the STPPCLK bit.
Bits 3:2 Reserved, must be kept at reset value.
Bit 1 GATEHCLK: Gate HCLK
The application sets this bit to gate HCLK to modules other than the AHB Slave and Master
and wakeup logic when the USB is suspended or the session is not valid. The application
clears this bit when the USB is resumed or a new session starts.
Bit 0 STPPCLK: Stop PHY clock
The application sets this bit to stop the PHY clock when the USB is suspended, the session
is not valid, or the device is disconnected. The application clears this bit when the USB is
resumed or a new session starts.

29.15.55 OTG_FS register map

The table below gives the USB OTG register map and reset values.

Table 208. OTG_FS register map and reset values

Register
Offset
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name

VBVALOEN
VBVALOVA
BVALOVAL

AVALOVAL
HSHNPEN

BVALOEN

AVALOEN
CURMOD

HNGSCS
OTGVER

DHNPEN

SRQSCS
CIDSTS

HNPRQ
BSVLD
ASVLD

EHEN
DBCT

OTG_

SRQ
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.

GOTGCTL
0x000

Reset value 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0
HNSSCHG
ADTOCHG

SRSSCHG
DBCDNE

HNGDET

SEDET
OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.

Res.
Res.
GOTGINT
0x004

Reset value 0 0 0 - 0 0 0
PTXFELVL

GINTMSK
TXFELVL

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.

GAHBCFG
0x008

Reset value 0 0 0
HNPCAP
SRPCAP

PHYSEL
FDMOD
FHMOD

OTG_
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

Res.
Res.
Res.

TRDT TOCAL
GUSBCFG
0x00C

Reset value 0 0 0 1 0 1 0 0 1 0 0 0
RXFFLSH
TXFFLSH
AHBIDL

CSRST
FCRST
PSRST

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

TXFNUM
GRSTCTL
0x010

Reset value 1 0 0 0 0 0 0 0 0 0 0

RM0402 Rev 6 1063/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Table 208. OTG_FS register map and reset values (continued)

Register
Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name

IPXFR/INCOMPISOOUT

GONAKEFF
ENUMDNE

GINAKEFF
USBSUSP
WKUPINT

CIDSCHG

ISOODRP
HPRTINT

IISOIXFR

USBRST
DISCINT

RSTDET

NPTXFE
SRQINT

OTGINT
OEPINT

RXFLVL
LPMINT

ESUSP
IEPINT
PTXFE
HCINT

CMOD
EOPF
OTG_

MMIS
Res.

Res.
Res.

Res.
Res.

SOF
GINTSTS
0x014

Reset value 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0

IPXFRM/IISOOXFRM

GONAKEFFM
ENUMDNEM

GINAKEFFM
USBSUSPM
ISOODRPM
CIDSCHGM

IISOIXFRM
RSTDETM

NPTXFEM
RXFLVLM
LPMINTM

ESUSPM
PTXFEM

USBRST
DISCINT

OTGINT
OEPINT

EOPFM

MMISM
SRQIM

IEPINT
PRTIM

SOFM
WUIM

OTG_
HCIM

Res.

Res.
Res.

Res.
Res.

Res.
GINTMSK
0x018

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
STSPHST

OTG_
Res.
Res.
Res.
Res.

Res.
Res.

GRXSTSR FRMNUM PKTSTS DPID BCNT EPNUM

(Device mode)

0x01C Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

GRXSTSR PKTSTS DPID BCNT CHNUM

(Host mode)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
STSPHST

OTG_
Res.
Res.
Res.
Res.

Res.
Res.

GRXSTSP FRMNUM PKTSTS DPID BCNT EPNUM

(Device mode)

0x020 Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

GRXSTSP PKTSTS DPID BCNT CHNUM

(Host mode)

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

RXFD
GRXFSIZ
0x024
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0

OTG_
HNPTXFSIZ/
NPTXFD/TX0FD NPTXFSA/TX0FSA
0x028 OTG_
DIEPTXF0
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
OTG_
Res.

NPTXQTOP NPTQXSAV NPTXFSAV

HNPTXSTS
0x02C
Reset value 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
PWRDWN

PS2DET
DCDEN
BCDEN
VBDEN

DCDET
SDEN
PDEN

SDET
PDET

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

GCCFG
0x038

Reset value 0 0 0 0 0 0 X X X X

1064/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Table 208. OTG_FS register map and reset values (continued)

Register
Offset

31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12

10
11

9
8
7
6
5
4
3
2
1
0
name
OTG_CID PRODUCT_ID
0x03C
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0

REMWAKE
L1RSMOK
SNDLPM

LPMACK
ENBESL

L1DSEN
SLPSTS

L1SSEN

LPMEN
OTG_ LPMR LPM LPM
Res.
Res.
Res.
LPMCHIDX BESLTHRS BESL
GLPMCFG CNTSTS RCNT RSP
0x054

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

OTG_
PTXFSIZ PTXSA
0x100 HPTXFSIZ
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
OTG_
INEPTXFD INEPTXSA
0x104 DIEPTXF1
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
OTG_
INEPTXFD INEPTXSA
0x108 DIEPTXF2
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0
. . .
. . .
. . .
. . .
OTG_
INEPTXFD INEPTXSA
0x114 DIEPTXF5
Reset value 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0
FSL

FSLSS
OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
S
HCFG
0x400 PCS

Reset value 0 0 0
RLDCTRL

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

FRIVL
HFIR
0x404

Reset value 0 1 1 1 0 1 0 1 0 0 1 1 0 0 0 0 0

OTG_
FTREM FRNUM
0x408 HFNUM
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1
OTG_
PTXQTOP PTXQSAV PTXFSAVL
0x410 HPTXSTS
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

HAINT
HAINT
0x414
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

OTG_
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

HAINTM
HAINTMSK
0x418
Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
POCCHNG

PENCHNG

PCDET
PSUSP

PCSTS
PLSTS
PPWR

POCA
PRES

PENA
PRST

OTG_ PSP
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.
Res.

Res.

PTCTL
HPRT D
0x440

Reset value 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RM0402 Rev 6 1065/1163

1122
 .
.
.
.
0x530
0x528
0x528
0x520
0x510
0x508
0x508
0x500

0x52C
0x50C
Offset

1066/1163
.
.
.
.
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
name

HCINT1
HCINT1
HCINT0
HCINT0

HCTSIZ1
HCTSIZ0

HCCHAR1
HCCHAR0
Register

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value

HCINTMSK1
HCINTMSK0

0
0
0
0
DOPNG Res. Res. Res. CHENA DOPNG Res. Res. Res. CHENA 31

0
0
0
0
Res. Res. Res. CHDIS Res. Res. Res. CHDIS 30

0
0
0
0

DPID
DPID
Res. Res. Res. ODDFRM Res. Res. Res. ODDFRM 29

0
0
0
0

Res. Res. Res. Res. Res. Res. 28

0
0
0
0

Res. Res. Res. Res. Res. Res. 27

0
0
0
0

Res. Res. Res. Res. Res. Res. 26

0
0
0
0

Res. Res. Res. Res. Res. Res. 25

USB on-the-go full-speed (OTG_FS)

DAD
DAD

0
0
0
0

Res. Res. Res. Res. Res. Res. 24

0
0
0
0

Res. Res. Res. Res. Res. Res. 23

PKTCNT
PKTCNT

0
0
0
0

Res. Res. Res. Res. Res. Res. 22

0
0
0
0

Res. Res. Res. Res. Res. Res. 21

MCNT MCNT

0
0
0
0

Res. Res. Res. Res. Res. Res. 20

0
0
0
0

Res. Res. Res. Res. Res. Res. 19

EPTYP EPTYP

0
0
0
0

Res. Res. Res. Res. Res. Res. 18

0
0
0
0

Res. Res. Res. LSDEV Res. Res. Res. LSDEV

RM0402 Rev 6
17

0
0
Res. Res. Res. Res. Res. Res. Res. Res. 16

.
.
.
.
0
0
0
0

Res. Res. Res. EPDIR Res. Res. Res. EPDIR 15

0
0
0
0

Res. Res. Res. Res. Res. Res. 14

0
0
0
0

Res. Res. Res. Res. Res. Res. 13

0
0
0
0

Res. Res. Res. Res. Res. Res. 12

EPNUM
EPNUM

0
0
0
0

Res. Res. Res. Res. Res. Res. 11

0
0
0
0

0
0
0
0
0
0

DTERRM DTERR DTERR DTERRM DTERR DTERR 10

0
0
0
0

0
0
0
0
0
0

FRMORM FRMOR FRMOR FRMORM FRMOR FRMOR 9

XFRSIZ
XFRSIZ

0
0
0
0

0
0
0
0
0
0

BBERRM BBERR BBERR BBERRM BBERR BBERR 8

0
0
0
0

0
0
0
0
0
0
Table 208. OTG_FS register map and reset values (continued)

TXERRM TXERR TXERR TXERRM TXERR TXERR 7

0
0
0
0

0
0

NYET Res. NYET Res. 6

0
0
0
0

0
0
0
0
0
0

ACKM ACK ACK ACKM ACK ACK 5

MPSIZ
MPSIZ

0
0
0
0

0
0
0
0
0
0

NAKM NAK NAK NAKM NAK NAK 4

0
0
0
0

0
0
0
0
0
0

STALLM STALL STALL STALLM STALL STALL 3

0
0
0
0

0
0

Res. AHBERR Res. Res. AHBERR Res. 2

0
0
0
0

0
0
0
0
0
0

CHHM CHH CHH CHHM CHH CHH 1

0
0
0
0

0
0
0
0
0
0

XFRCM XFRC XFRC XFRCM XFRC XFRC 0

RM0402
 0x818
0x814
0x810
0x808
0x804
0x800
0x670
0x668
0x660

0x66C
Offset
RM0402

DSTS
DCTL

OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_

DCFG

DAINT
name

HCINT11

DIEPMSK
HCTSIZ11

DOEPMSK
Register

Reset value
Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
HCCHAR11

HCINTMSK11

0
0
0
Res. Res. Res. Res. Res. DOPNG Res. Res. CHENA 31

0
0
0
Res. Res. Res. Res. Res. Res. Res. CHDIS 30

0
0
0

DPID
Res. Res. Res. Res. Res. Res. Res. ODDFRM 29

0
0
0

Res. Res. Res. Res. Res. Res. Res. 28

0
0
0

Res. Res. Res. Res. Res. Res. Res. 27

0
0
0

Res. Res. Res. Res. Res. Res. Res. 26

0
0
0

Res. Res. Res. Res. Res. Res. Res. 25

DAD

0
0
0

Res. Res. Res. Res. Res. Res. Res. 24

0
0
0
0

Res. Res. Res. Res. Res. Res. 23

OEPINT
PKTCNT

LN
STS
DEV

0
0
0
0

Res. Res. Res. Res. Res. Res. 22

0
0
0
0

Res. Res. Res. Res. Res. Res. 21

MCNT

0
0
0
0

Res. Res. Res. Res. Res. Res. 20

0
0
0
0

Res. Res. Res. Res. Res. Res. 19

EPTYP

0
0
0
0

0
Res. Res. DSBESLRJCT Res. Res. Res. 18

0
0
0
0

Res. Res. Res. Res. Res. Res. LSDEV

RM0402 Rev 6
17

0
0
0
Res. Res. Res. Res. Res. Res. Res. 16

0
0
0
0

0
Res. Res. Res. ERRATIM Res. Res. EPDIR 15

0
0
0
0

0
Res. Res. Res. XCVRDLY Res. Res. 14

FNSOF

0
0
0
0
0
0

NAKMSK NAKM Res. Res. Res. Res. 13

0
0
0
0
0

0
BERRM Res. Res. Res. Res. 12
EPNUM

PFIVL

0
0
0
0

0
Res. Res. POPRGDNE 0 Res. Res. 11

0
0
0
0

0
0
0
0

Res. Res. CGONAK DTERRM DTERR 10

0
0
0
0

0
0
0
0

Res. Res. SGONAK FRMORM FRMOR 9

XFRSIZ

0
0
0
0
0

0
0
0
0

OUTPKTERRM Res. CGINAK BBERRM BBERR 8

0
0
0

0
0
0
0

DAD
Table 208. OTG_FS register map and reset values (continued)

Res. Res. Res. SGINAK TXERRM TXERR 7

IEPINT

0
0
0
0

0
0

Res. INEPNEM Res. Res. 6

0
0
0
0

0
0
0
0

Res. INEPNMM Res. TCTL ACKM ACK 5

MPSIZ

0
0
0
0
0

0
0
0
0

OTEPDM ITTXFEMSK Res. NAKM NAK 4

0
0
0
0
0

0
0
0
0

STUPM TOM EERR GONSTS Res. STALLM STALL 3

0
0
0

0
0
0

Res. Res. GINSTS NZLSOHSK Res. Res. 2

ENUMSPD

0
0
0
0
0

0
1
0
0
0

EPDM EPDM SDIS CHHM CHH 1

DSPD

0
0
0
0
0

0
0
0
0
0

XFRCM XFRCM SUSPSTS RWUSIG XFRCM XFRC 0

USB on-the-go full-speed (OTG_FS)

1067/1163
1122
 0x938
0x930
0x928
0x920
0x918
0x910
0x908
0x900
0x834
0x828

0x82C
0x81C
Offset

1068/1163
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
name

OTG_DIE

DIEPINT1
DIEPINT0
USPULSE

DIEPCTL1
DIEPCTL0
OTG_DVB

DIEPTSIZ1
DIEPTSIZ0
PEMPMSK

DTXFSTS1
DTXFSTS0
Register

DVBUSDIS

Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
Reset value
DAINTMSK

0
0
Res. Res. Res. EPENA Res. Res. Res. EPENA Res. Res. Res. 0
31

0
0
0
0

Res. Res. EPDIS Res. Res. Res. EPDIS Res. Res. Res. 30

0
0
0

MCN
Res. Res. SODDFRM/SD1PID Res. Res. Res. Res. Res. Res. Res. 29

0
0
0

Res. Res. SD0PID/SEVNFRM Res. Res. Res. Res. Res. Res. Res. 28

0
0
0
0

Res. Res. SNAK Res. Res. Res. SNAK Res. Res. Res. 27

0
0
0
0

Res. Res. CNAK Res. Res. Res. CNAK Res. Res. Res. 26

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 25

USB on-the-go full-speed (OTG_FS)

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 24

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res.

OEPM

23

PKTCNT
TXFNUM
TXFNUM

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 22

0
0
0
0

Res. Res. STALL Res. Res. Res. STALL Res. Res. Res. 21

0
0

0
Res. Res. Res. Res. Res. Res. Res. Res. Res. 20

PKT
CNT

0
0
0

0
0
Res. Res. Res. Res. Res. Res. Res. 19
EPTYP EPTYP

0
0
0
0

Res. Res. Res. Res. Res. Res. Res. Res. 18

0
0
0
0

Res. Res. NAKSTS Res. Res. Res. NAKSTS Res. Res. Res.

RM0402 Rev 6
17

0
0
0

Res. Res. EONUM/DPID Res. Res. Res. Res. Res. Res. Res. 16

0
0
1
0

0
0
0
0
0

Res. USBAEP Res. Res. USBAEP 15

0
0

0
0
0
0
0

Res. Res. Res. Res. Res. 14

0
0
0
0

0
0
0
0
0

NAK Res. Res. NAK Res. 13

0
0

0
0
0
0
1

Res. Res. Res. Res. Res. 12

0
0
0
0

0
0
0
0
0

PKTDRPSTS Res. Res. PKTDRPSTS Res. 11

0
0
0

0
0
0
1
1

Res. Res. Res. Res. 10

0
0
0

1
1
0
0
1

Res. Res. Res. Res. 9

XFRSIZ

0
0
0

0
0
0
1
1

Res. Res. Res. Res. 8

0
1
0
1
0

0
0
0
1
1
IEPM
Table 208. OTG_FS register map and reset values (continued)

TXFE Res. TXFE Res. 7

VBUSDT

DVBUSP

INEPTFSAV
INEPTFSAV

0
0
0
0
0

0
0
0
0
0
1

INEPTXFEM

INEPNE INEPNE Res. 6

0
0
0
0
0

0
0
0
0
1
0

INEPNM INEPNM Res. 5

MPSIZ

0
0
0
0
0

0
0
0
0
1
1

ITTXFE ITTXFE Res. 4

0
0
0
0
0

0
0
0
0
1
0

TOC TOC Res. 3

XFRSIZ

0
0
0

0
0
0
0
0
1

Res. Res. Res. 2

0
0
0
0
0

0
0
0
0
0
0
1

EPDISD EPDISD 1
MPSIZ

0
0
0
0
0

0
0
0
0
0
0
1

XFRC XFRC 0
RM0402
 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0x940

0xB08
0xB00
0x9B0
0x9B8
0x9A8
0x9A0
Offset
RM0402

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
name

DIEPINT5
DIEPCTL5
DIEPCTL2

DOEPINT0
DIEPTSIZ5
DTXFSTS5
Register

Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value

DOEPCTL0

0
0
0
Res. EPENA Res. Res. Res. EPENA EPENA 31

0
0
0
0
Res. EPDIS Res. Res. EPDIS EPDIS 30
MCNT

0
0
0
Res. Res. Res. Res. SODDFRM SODDFRM 29

0
0
0
Res. Res. Res. Res. SD0PID/SEVNFRM SD0PID/SEVNFRM 28

0
0
0
Res. SNAK Res. Res. SNAK 0 SNAK 27

0
0
0
0
Res. CNAK Res. Res. CNAK CNAK 26

0
0
0

Res. Res. Res. Res. 25

0
0
0

Res. Res. Res. Res. 24

0
0
0

Res. Res. Res. Res. 23

PKTCNT
TXFNUM
TXFNUM

0
0
0

Res. Res. Res. Res. 22

0
0
0
0

Res. STALL Res. Res. STALL STALL 21

0
0
Res. SNPM Res. Res. Res. Res. 20

0
0
0
0

Res. Res. Res. 19

EPTYP EPTYP EPTYP

0
0
0
0

Res. Res. Res. 18

0
0
0
0

Res. NAKSTS Res. Res. NAKSTS NAKSTS

RM0402 Rev 6
17

0
0
0

Res. Res. Res. Res. EONUM/DPID EONUM/DPID 16

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

1
0
0
0

0
Res. USBAEP Res. USBAEP USBAEP 15

0
0
Res. Res. Res. Res. Res. 14

0
0
0

0
NAK Res. NAK Res. Res. 13

0
0
0
BERR Res. Res. Res. Res. 12

0
0

0
Res. Res. PKTDRPSTS Res. Res. 11

0
0
0

0
Res. Res. Res. 10

0
0
0

1
Res. Res. Res. 9

XFRSIZ

0
0
0

0
Res. Res. Res. 8

0
1
0
0

0
Table 208. OTG_FS register map and reset values (continued)

Res. Res. TXFE 7

INEPTFSAV

0
0
0
0

0
Res. Res. INEPNE 6

0
0
0
0
0

0
STSPHSRX Res. INEPNM 5
MPSIZ
MPSIZ

0
0
0
0
0

0
OTEPDIS Res. ITTXFE 4

0
0
0
0
0

0
STUP Res. TOC 3

0
0
0

0
Res. Res. Res. 2

0
0
0
0
0
0

0
EPDISD EPDISD 1
MPSIZ

0
0
0
0
0
0

0
XFRC XFRC 0
USB on-the-go full-speed (OTG_FS)

1069/1163
1122
 .
.
.
.

0xE00
0xB30
0xB28
0xB20
0xB10

0xBB0
0xBA8
0xBA0
Offset

1070/1163
.
.
.
.

OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
OTG_
name

PCGCCTL
DOEPINT5
DOEPINT1
Register

Reset value
Reset value
Reset value

Reset value
Reset value
Reset value
Reset value
Reset value

DOEPCTL5
DOEPCTL1

DOEPTSIZ5
DOEPTSIZ1
DOEPTSIZ0

0
0
Res. Res. Res. EPENA Res. Res. EPENA Res. 31

0
0
0
0
Res. RXDPID/ Res. EPDIS RXDPID/ Res. EPDIS 0 30
P

STUPCNT STUPCNT
STU

CNT

0
0
0
0
0
Res. Res. SODDFRM Res. SODDFRM 29

0
0
0
0
Res. Res. SD0PID/SEVNFRM Res. SD0PID/SEVNFRM Res. 28

0
0
0
0
Res. Res. SNAK Res. SNAK Res. 27

0
0
0
0
Res. Res. CNAK Res. CNAK Res. 26

0
0
Res. Res. Res. Res. Res. Res. 25
USB on-the-go full-speed (OTG_FS)

0
0
Res. Res. Res. Res. Res. Res. 24

0
0
Res. Res. Res. Res. Res. Res. 23

PKTCNT
PKTCNT

0
0
Res. Res. Res. Res. Res. Res. 22

0
0
0
0

Res. Res. STALL Res. STALL Res. 21

0
0
0
0

Res. Res. SNPM Res. SNPM Res. 20

0
0
0
0
0

Res. Res. Res. PKTCNT 19

EPTYP
EP
TYP

0
0
0
0

Res. Res. Res. Res. 18

0
0
0
0

Res. Res. NAKSTS Res. NAKSTS Res.

RM0402 Rev 6
17

0
0
0
0

Res. Res. EONUM/DPID Res. EONUM/DPID Res. 16

.
.
.
.

0
0
0
0

Res. Res. USBAEP Res. USBAEP Res. 15

0
0
Res. Res. Res. Res. Res. Res. 14

0
0
0
0

Res. NAK Res. NAK Res. Res. 13

0
0
0
0

Res. BERR Res. BERR Res. Res. 12

0
0

Res. Res. Res. Res. Res. Res. 11

0
0
0
0

Res. Res. Res. Res. 10

0
0
0
0

Res. Res. Res. Res. 9

Refer to Section 2.2 on page 49 for the register boundary addresses.

XFRSIZ
XFRSIZ

0
0
0
0

Res. Res. Res. Res. 8

0
0
0
0

0
Table 208. OTG_FS register map and reset values (continued)

SUSP Res. Res. Res. 7

0
0
0
0

0
0

PHYSLEEP Res. Res. 6

0
0
0
0
0
0

0
0

ENL1GTG STSPHSRX STSPHSRX 5

MPSIZ
MPSIZ

0
0
0
0
0
0

0
0

PHYSUSP OTEPDIS OTEPDIS 4

0
0
0
0
0
0
0

Res. STUP STUP 3

XFRSIZ

0
0
0
0
0

Res. Res. Res. 2

0
0
0
0
0
0

0
0

GATEHCLK EPDISD EPDISD 1

0
0
0
0
0
0

0
0

STPPCLK XFRC XFRC 0

RM0402
RM0402 USB on-the-go full-speed (OTG_FS)

29.16 OTG_FS programming model

29.16.1 Core initialization

The application must perform the core initialization sequence. If the cable is connected
during power-up, the current mode of operation bit in the OTG_GINTSTS (CMOD bit in
OTG_GINTSTS) reflects the mode. The OTG_FS controller enters host mode when an “A”
plug is connected or device mode when a “B” plug is connected.
This section explains the initialization of the OTG_FS controller after power-on. The
application must follow the initialization sequence irrespective of host or device mode
operation. All core global registers are initialized according to the core’s configuration:
1. Program the following fields in the OTG_GAHBCFG register:
– Global interrupt mask bit GINTMSK = 1
– Rx FIFO non-empty (RXFLVL bit in OTG_GINTSTS)
– Periodic Tx FIFO empty level
2. Program the following fields in the OTG_GUSBCFG register:
– HNP capable bit
– SRP capable bit
– OTG_FS timeout calibration field
– USB turnaround time field
3. The software must unmask the following bits in the OTG_GINTMSK register:
OTG interrupt mask
Mode mismatch interrupt mask
4. The software can read the CMOD bit in OTG_GINTSTS to determine whether the
OTG_FS controller is operating in host or device mode.

RM0402 Rev 6 1071/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

29.16.2 Host initialization

To initialize the core as host, the application must perform the following steps:
1. Program the HPRTINT in the OTG_GINTMSK register to unmask
2. Program the OTG_HCFG register to select full-speed host
3. Program the PPWR bit in OTG_HPRT to 1. This drives VBUS on the USB.
4. Wait for the PCDET interrupt in OTG_HPRT0. This indicates that a device is
connecting to the port.
5. Program the PRST bit in OTG_HPRT to 1. This starts the reset process.
6. Wait at least 10 ms for the reset process to complete.
7. Program the PRST bit in OTG_HPRT to 0.
8. Wait for the PENCHNG interrupt in OTG_HPRT.
9. Read the PSPD bit in OTG_HPRT to get the enumerated speed.
10. Program the HFIR register with a value corresponding to the selected PHY clock 1
11. Program the FSLSPCS field in the OTG_HCFG register following the speed of the
device detected in step 9. If FSLSPCS has been changed a port reset must be
performed.
12. Program the OTG_GRXFSIZ register to select the size of the receive FIFO.
13. Program the OTG_HNPTXFSIZ register to select the size and the start address of the
Non-periodic transmit FIFO for non-periodic transactions.
14. Program the OTG_HPTXFSIZ register to select the size and start address of the
periodic transmit FIFO for periodic transactions.
To communicate with devices, the system software must initialize and enable at least one
channel.

29.16.3 Device initialization

The application must perform the following steps to initialize the core as a device on power-
up or after a mode change from host to device.
1. Program the following fields in the OTG_DCFG register:
– Device speed
– Non-zero-length status OUT handshake
– Periodic Frame Interval
2. Clear the DCTL.SDIS bit. The core issues a connect after this bit is cleared.
3. Program the OTG_GINTMSK register to unmask the following interrupts:
– USB reset
– Enumeration done
– Early suspend
– USB suspend
– SOF
4. Wait for the USBRST interrupt in OTG_GINTSTS. It indicates that a reset has been
detected on the USB that lasts for about 10 ms on receiving this interrupt.
5. Wait for the ENUMDNE interrupt in OTG_GINTSTS. This interrupt indicates the end of
reset on the USB. On receiving this interrupt, the application must read the OTG_DSTS

1072/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

register to determine the enumeration speed and perform the steps listed in Endpoint
initialization on enumeration completion on page 1095.
At this point, the device is ready to accept SOF packets and perform control transfers on
control endpoint 0.

29.16.4 Host programming model

Channel initialization
The application must initialize one or more channels before it can communicate with
connected devices. To initialize and enable a channel, the application must perform the
following steps:
1. Program the OTG_GINTMSK register to unmask the following:
2. Channel interrupt
– Non-periodic transmit FIFO empty for OUT transactions (applicable when
operating in pipelined transaction-level with the packet count field programmed
with more than one).
– Non-periodic transmit FIFO half-empty for OUT transactions (applicable when
operating in pipelined transaction-level with the packet count field programmed
with more than one).
3. Program the OTG_HAINTMSK register to unmask the selected channels’ interrupts.
4. Program the OTG_HCINTMSK register to unmask the transaction-related interrupts of
interest given in the host channel interrupt register.
5. Program the selected channel’s OTG_HCTSIZx register with the total transfer size, in
bytes, and the expected number of packets, including short packets. The application
must program the PID field with the initial data PID (to be used on the first OUT
transaction or to be expected from the first IN transaction).
6. Program the OTG_HCCHARx register of the selected channel with the device’s
endpoint characteristics, such as type, speed, direction, and so forth. (The channel can
be enabled by setting the channel enable bit to 1 only when the application is ready to
transmit or receive any packet).

Halting a channel
The application can disable any channel by programming the OTG_HCCHARx register with
the CHDIS and CHENA bits set to 1. This enables the OTG_FS host to flush the posted
requests (if any) and generates a channel halted interrupt. The application must wait for the
CHH interrupt in OTG_HCINTx before reallocating the channel for other transactions. The
OTG_FS host does not interrupt the transaction that has already been started on the USB.
Before disabling a channel, the application must ensure that there is at least one free space
available in the non-periodic request queue (when disabling a non-periodic channel) or the
periodic request queue (when disabling a periodic channel). The application can simply
flush the posted requests when the request queue is full (before disabling the channel), by
programming the OTG_HCCHARx register with the CHDIS bit set to 1 which automatically
clears the CHENA bit to 0.
The application is expected to disable a channel on any of the following conditions:

RM0402 Rev 6 1073/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

1. When an STALL, TXERR, BBERR or DTERR interrupt in OTG_HCINTx is received for

an IN or OUT channel. The application must be able to receive other interrupts
(DTERR, Nak, data, TXERR) for the same channel before receiving the halt.
2. When a DISCINT (disconnect device) interrupt in OTG_GINTSTS is received. (The
application is expected to disable all enabled channels).
3. When the application aborts a transfer before normal completion.

Operational model
The application must initialize a channel before communicating to the connected device.
This section explains the sequence of operation to be performed for different types of USB
transactions.
• Writing the transmit FIFO
The OTG_FS host automatically writes an entry (OUT request) to the periodic/non-
periodic request queue, along with the last 32-bit word write of a packet. The
application must ensure that at least one free space is available in the periodic/non-
periodic request queue before starting to write to the transmit FIFO. The application
must always write to the transmit FIFO in 32-bit words. If the packet size is non-32-bit
word aligned, the application must use padding. The OTG_FS host determines the
actual packet size based on the programmed maximum packet size and transfer size.

Figure 343. Transmit FIFO write task

Start

Read OTG_HPTXSTS /OTG_HNPTXSTS

registers for available FIFO and queue
spaces

Wait for NPTXFE/PTXFE interrupt in 1 MPS

No or LPS FIFO space
OTG_GINTSTS available?

Yes Yes

Write 1 packet
data to transmit
FIFO

More
packets to
send?

No

MPS: Maximum packet size Done

LPS: Last packet size ai15673c

• Reading the receive FIFO

The application must ignore all packet statuses other than IN data packet (bx0010).

1074/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 344. Receive FIFO read task

• Bulk and control OUT/SETUP transactions

A typical bulk or control OUT/SETUP pipelined transaction-level operation is shown in
Figure 345. See channel 1 (ch_1). Two bulk OUT packets are transmitted. A control
SETUP transaction operates in the same way but has only one packet. The
assumptions are:
– The application is attempting to send two maximum-packet-size packets (transfer
size = 1, 024 bytes).
– The non-periodic transmit FIFO can hold two packets (128 bytes for FS).
– The non-periodic request queue depth = 4.
• Normal bulk and control OUT/SETUP operations
The sequence of operations in (channel 1) is as follows:
1. Initialize channel 1
2. Write the first packet for channel 1
3. Along with the last word write, the core writes an entry to the non-periodic request
queue
4. As soon as the non-periodic queue becomes non-empty, the core attempts to send an
OUT token in the current frame
5. Write the second (last) packet for channel 1
6. The core generates the XFRC interrupt as soon as the last transaction is completed
successfully
7. In response to the XFRC interrupt, de-allocate the channel for other transfers
8. Handling non-ACK responses

RM0402 Rev 6 1075/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Figure 345. Normal bulk/control OUT/SETUP

Application AHB Host USB Device

1 init_reg(ch_1)
init_reg(ch_2) Non-Periodic Request
1 write_tx_fifo 4
Queue
Assume that this queue
(ch_1) 1 3 can hold 4 entries.
MPS
set_ch_en 2
(ch_2)
ch_1
2 write_tx_fifo
(ch_1) 1 ch_2
_
MPS OU T
set_ch_en
(ch_2)
5 ch_1

ch_2 D AT A0
MPS

3
ACK
set_ch_en
(ch_2) IN

4 D AT A0
5
RxFLvl interruptt

ACK
1 ch_1
read_rx_sts
MPS
read_rx_fifo O UT
ch_2
set_ch_en
ch_2 D AT A1
(ch_2)
MPS
ch_2

7 XferCompl interrupt
ACK
6
De-allocate IN
(ch_1)

D AT A1
RxFLvl interruptt

1 6 ACK
read_rx_stsre MPS
ad_rx_fifo

RxFLvl interruptt 8 ch_2

read_rx_sts
7
XferCompl interruptt

Disable
(ch_2) 9
RxFLvl interruptt

read_rx_sts 11 10
ChHltd interruptt

De-allocate 12
(ch_2) 13
MSv36018V1

1. The grayed elements are not relevant in the context of this figure.

1076/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

The channel-specific interrupt service routine for bulk and control OUT/SETUP
transactions is shown in the following code samples.
• Interrupt service routine for bulk/control OUT/SETUP and bulk/control IN
transactions
a) Bulk/control OUT/SETUP
Unmask (NAK/TXERR/STALL/XFRC)
if (XFRC)
{
Reset Error Count
Mask ACK
De-allocate Channel
}
else if (STALL)
{
Transfer Done = 1
Unmask CHH
Disable Channel
}
else if (NAK or TXERR )
{
Rewind Buffer Pointers
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
else
{
Reset Error Count
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}

RM0402 Rev 6 1077/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

else if (ACK)
{
Reset Error Count
Mask ACK
}
The application is expected to write the data packets into the transmit FIFO when the
space is available in the transmit FIFO and the request queue. The application can
make use of the NPTXFE interrupt in OTG_GINTSTS to find the transmit FIFO space.
b) Bulk/control IN
Unmask (TXERR/XFRC/BBERR/STALL/DTERR)
if (XFRC)
{
Reset Error Count
Unmask CHH
Disable Channel
Reset Error Count
Mask ACK
}
else if (TXERR or BBERR or STALL)
{
Unmask CHH
Disable Channel
if (TXERR)
{
Increment Error Count
Unmask ACK
}
}
else if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel
}
}
else if (ACK)
{
Reset Error Count
Mask ACK
}

1078/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

else if (DTERR)
{
Reset Error Count
}
The application is expected to write the requests as and when the request queue space is
available and until the XFRC interrupt is received.
• Bulk and control IN transactions
A typical bulk or control IN pipelined transaction-level operation is shown in Figure 346.
See channel 2 (ch_2). The assumptions are:
– The application is attempting to receive two maximum-packet-size packets
(transfer size = 1 024 bytes).
– The receive FIFO can contain at least one maximum-packet-size packet and two
status words per packet (72 bytes for FS).
– The non-periodic request queue depth = 4.

RM0402 Rev 6 1079/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Figure 346. Bulk/control IN transactions

Application AHB Host USB Device

1 init_reg(ch_1)
init_reg(ch_2) Non-Periodic Request
1 write_tx_fifo 4
Queue
Assume that this queue
(ch_1) 1 3 can hold 4 entries.
MPS
set_ch_en 2
(ch_2)
ch_1
ch
c 1
2 write_tx_fifo
(ch_1) 1 ch_2
MPS OU T
set_ch_en
(ch_2)
5 ch_1

ch_2 D AT A0
MPS

3
ACK
set_ch_en
(ch_2) IN

4 D AT A0
5
RxFLvl interrupt

ACK
1 ch_1
read_rx_sts
MPS
read_rx_fifo O UT
ch_2
set_ch_en
ch_2 D AT A1
(ch_2)
MPS
ch_2

7 r ompl interru
XferC r pt
ACK
K
6
De-allocate IN
(ch_1)

D AT A1
RxFLvl interrupt

1 6 ACK
read_rx_stsre MPS
ad_rx_fifo

RxFLvl interrupt 8 ch_2

read_rx_sts
7
XferCompl interrupt

Disable
(ch_2) 9
RxFLvl interrupt

read_rx_sts 11 10
ChHltd interrupt

De-allocate 12
(ch_2) 13
ai15675b

1. The grayed elements are not relevant in the context of this figure.

1080/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

The sequence of operations is as follows:

1. Initialize channel 2.
2. Set the CHENA bit in OTG_HCCHAR2 to write an IN request to the non-periodic
request queue.
3. The core attempts to send an IN token after completing the current OUT transaction.
4. The core generates an RXFLVL interrupt as soon as the received packet is written to
the receive FIFO.
5. In response to the RXFLVL interrupt, mask the RXFLVL interrupt and read the received
packet status to determine the number of bytes received, then read the receive FIFO
accordingly. Following this, unmask the RXFLVL interrupt.
6. The core generates the RXFLVL interrupt for the transfer completion status entry in the
receive FIFO.
7. The application must read and ignore the receive packet status when the receive
packet status is not an IN data packet (PKTSTS in OTG_GRXSTSR ≠ 0b0010).
8. The core generates the XFRC interrupt as soon as the receive packet status is read.
9. In response to the XFRC interrupt, disable the channel and stop writing the
OTG_HCCHAR2 register for further requests. The core writes a channel disable
request to the non-periodic request queue as soon as the OTG_HCCHAR2 register is
written.
10. The core generates the RXFLVL interrupt as soon as the halt status is written to the
receive FIFO.
11. Read and ignore the receive packet status.
12. The core generates a CHH interrupt as soon as the halt status is popped from the
receive FIFO.
13. In response to the CHH interrupt, de-allocate the channel for other transfers.
14. Handling non-ACK responses
• Control transactions
Setup, data, and status stages of a control transfer must be performed as three
separate transfers. setup-, data- or status-stage OUT transactions are performed
similarly to the bulk OUT transactions explained previously. Data- or status-stage IN
transactions are performed similarly to the bulk IN transactions explained previously.
For all three stages, the application is expected to set the EPTYP field in

RM0402 Rev 6 1081/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

OTG_HCCHAR1 to control. During the setup stage, the application is expected to set
the PID field in OTG_HCTSIZ1 to SETUP.
• Interrupt OUT transactions
A typical interrupt OUT operation is shown in Figure 347. The assumptions are:
– The application is attempting to send one packet in every frame (up to 1 maximum
packet size), starting with the odd frame (transfer size = 1 024 bytes)
– The periodic transmit FIFO can hold one packet (1 KB)
– Periodic request queue depth = 4
The sequence of operations is as follows:
1. Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_HCCHAR1.
2. Write the first packet for channel 1.
3. Along with the last word write of each packet, the OTG_FS host writes an entry to the
periodic request queue.
4. The OTG_FS host attempts to send an OUT token in the next (odd) frame.
5. The OTG_FS host generates an XFRC interrupt as soon as the last packet is
transmitted successfully.
6. In response to the XFRC interrupt, reinitialize the channel for the next transfer.

1082/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 347. Normal interrupt OUT

Application AHB Host USB Device

1 init_reg(ch_1)
init_reg(ch_2) Periodic Request Queue
1 write_tx_fifo 3
Assume that this queue
1 can hold 4 entries.
(ch_1)
MPS
set_ch_en 2
(ch_2) ch_1 4
2 ch_2
_

3
Odd
OU T
(micro)
frame
D ATA0
M PS

5
6 XferCompl interrupt AC K

init_reg(ch_1) 4 IN

write_tx_fifo
(ch_1) 1 5
MPS D ATA0

RxFLvl interrupt
ACK
read_rx_sts
1
read_rx_fifo
6 MPS

RxFLvl interrupt
read_rx_sts ch_1
7 8
XferCompl interrupt
ch 2
ch_2
init_reg(ch_2)
9
set_ch_en
(ch_2)

Even
OU T (micro)
XferCompl interrupt frame
D ATA1
init_reg(ch_1)
MPS

write_tx_fifo 1
MPS ACK
(ch_1)
IN

D ATA1

MSv36020V1

1. The grayed elements are not relevant in the context of this figure.
• Interrupt service routine for interrupt OUT/IN transactions
a) Interrupt OUT
Unmask (NAK/TXERR/STALL/XFRC/FRMOR)

RM0402 Rev 6 1083/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

if (XFRC)
{
Reset Error Count
Mask ACK
De-allocate Channel
}
else
if (STALL or FRMOR)
{
Mask ACK
Unmask CHH
Disable Channel
if (STALL)
{
Transfer Done = 1
}
}
else
if (NAK or TXERR)
{
Rewind Buffer Pointers
Reset Error Count
Mask ACK
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
{
Re-initialize Channel (in next b_interval - 1 Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK
}

1084/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

The application uses the NPTXFE interrupt in OTG_GINTSTS to find the

transmit FIFO space.
Interrupt IN
Unmask (NAK/TXERR/XFRC/BBERR/STALL/FRMOR/DTERR)
if (XFRC)
{
Reset Error Count
Mask ACK
if (OTG_HCTSIZx.PKTCNT == 0)
{
De-allocate Channel
}
else
{
Transfer Done = 1
Unmask CHH
Disable Channel
}
}
else
if (STALL or FRMOR or NAK or DTERR or BBERR)
{
Mask ACK
Unmask CHH
Disable Channel
if (STALL or BBERR)
{
Reset Error Count
Transfer Done = 1
}
else
if (!FRMOR)
{
Reset Error Count
}
}
else
if (TXERR)
{
Increment Error Count
Unmask ACK
Unmask CHH
Disable Channel
}
else

RM0402 Rev 6 1085/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}
else
Re-initialize Channel (in next b_interval - 1 /Frame)
}
}
else
if (ACK)
{
Reset Error Count
Mask ACK
}

• Interrupt IN transactions
The assumptions are:
– The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame, starting with odd (transfer size = 1 024 bytes).
– The receive FIFO can hold at least one maximum-packet-size packet and two
status words per packet (1 031 bytes).
– Periodic request queue depth = 4.
• Normal interrupt IN operation
The sequence of operations is as follows:
1. Initialize channel 2. The application must set the ODDFRM bit in OTG_HCCHAR2.
2. Set the CHENA bit in OTG_HCCHAR2 to write an IN request to the periodic request
queue.
3. The OTG_FS host writes an IN request to the periodic request queue for each
OTG_HCCHAR2 register write with the CHENA bit set.
4. The OTG_FS host attempts to send an IN token in the next (odd) frame.
5. As soon as the IN packet is received and written to the receive FIFO, the OTG_FS host
generates an RXFLVL interrupt.
6. In response to the RXFLVL interrupt, read the received packet status to determine the
number of bytes received, then read the receive FIFO accordingly. The application
must mask the RXFLVL interrupt before reading the receive FIFO, and unmask after
reading the entire packet.
7. The core generates the RXFLVL interrupt for the transfer completion status entry in the
receive FIFO. The application must read and ignore the receive packet status when the
receive packet status is not an IN data packet (PKTSTS in GRXSTSR ≠ 0b0010).
8. The core generates an XFRC interrupt as soon as the receive packet status is read.
9. In response to the XFRC interrupt, read the PKTCNT field in OTG_HCTSIZ2. If the
PKTCNT bit in OTG_HCTSIZ2 is not equal to 0, disable the channel before re-

1086/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

initializing the channel for the next transfer, if any). If PKTCNT bit in OTG_HCTSIZ2 =
0, reinitialize the channel for the next transfer. This time, the application must reset the
ODDFRM bit in OTG_HCCHAR2.

RM0402 Rev 6 1087/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Figure 348. Normal interrupt IN

Application AHB Host USB Device

1 init_reg(ch_1)
init_reg(ch_2) Periodic Request Queue
1 write_tx_fifo 3
Assume that this queue
1 can hold 4 entries.
(ch_1)
MPS
set_ch_en 2
(ch_2) ch 1
ch_1 4
2 ch_2

3
Odd
OU T
(micro)
frame
D ATA0
M PS

5
6 Xfer Com pl i nterr upt AC K

init_reg(ch_1) 4 IN

write_tx_fifo
(ch_1) 1 5
MPS D ATA0

RxFLvl interrupt
ACK
read_rx_sts
1
read_rx_fifo
6 MPS

RxFLvl interrupt
read_rx_sts ch 1
ch_1
7 8
XferCompl interrupt
ch_2
init_reg(ch_2)
9
set_ch_en
(ch_2)

Even
OU T (micro)
Xfferr Compl interrr upt frame
D ATA1
init_reg(ch_1)
MPS

write_tx_fifo 1
MPS ACK
(ch_1)
IN

D ATA1

ai15676b

1. The grayed elements are not relevant in the context of this figure.
• Isochronous OUT transactions
A typical isochronous OUT operation is shown in Figure 348. The assumptions are:
– The application is attempting to send one packet every frame (up to 1 maximum

1088/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

packet size), starting with an odd frame. (transfer size = 1 024 bytes).
– The periodic transmit FIFO can hold one packet (1 KB).
– Periodic request queue depth = 4.
The sequence of operations is as follows:
1. Initialize and enable channel 1. The application must set the ODDFRM bit in
OTG_HCCHAR1.
2. Write the first packet for channel 1.
3. Along with the last word write of each packet, the OTG_FS host writes an entry to the
periodic request queue.
4. The OTG_FS host attempts to send the OUT token in the next frame (odd).
5. The OTG_FS host generates the XFRC interrupt as soon as the last packet is
transmitted successfully.
6. In response to the XFRC interrupt, reinitialize the channel for the next transfer.
7. Handling non-ACK responses

RM0402 Rev 6 1089/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Figure 349. Isochronous OUT transactions

Application AHB Host USB Device

1 init_reg(ch_1)
init_reg(ch_2) Periodic Request
1 write_tx_fifo 3
Queue
Assume that this queue
(ch_1) 1 can hold 4 entries.
MPS
set_ch_en 2
(ch_2) ch_1 4
2 ch_2

3
Odd
OUT
(micro)
frame
D A T A0
5 MPS
6
XferCompl interrupt

init_reg(ch_1)
4 IN

write_tx_fifo
(ch_1) 1
MPS
5 D A T A0

RxFLvl interrupt

1
read_rx_sts 6 MPS
read_rx_fifo

RxFLvl interrupt

read_rx_sts
7 ch_1
8
XferCompl interrupt
ch 2
ch_2
init_reg(ch_2)
9
set_ch_en
(ch_2)

Even
OUT (micro)
frame
D A T A0
MPS
XferCompl interrupt
init_reg(ch_1)
IN
write_tx_fifo 1
(ch_1)
MPS
DA TA
0

MSv36022V1

1. The grayed elements are not relevant in the context of this figure.
• Interrupt service routine for isochronous OUT/IN transactions
Code sample: isochronous OUT
Unmask (FRMOR/XFRC)
if (XFRC)

1090/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

{
De-allocate Channel
}
else
if (FRMOR)
{
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
De-allocate Channel
}
Code sample: Isochronous IN
Unmask (TXERR/XFRC/FRMOR/BBERR)
if (XFRC or FRMOR)
{
if (XFRC and (OTG_HCTSIZx.PKTCNT == 0))
{
Reset Error Count
De-allocate Channel
}
else
{
Unmask CHH
Disable Channel
}
}
else
if (TXERR or BBERR)
{
Increment Error Count
Unmask CHH
Disable Channel
}
else
if (CHH)
{
Mask CHH
if (Transfer Done or (Error_count == 3))
{
De-allocate Channel
}

RM0402 Rev 6 1091/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

else
{
Re-initialize Channel
}
}
• Isochronous IN transactions
The assumptions are:
– The application is attempting to receive one packet (up to 1 maximum packet size)
in every frame starting with the next odd frame (transfer size = 1 024 bytes).
– The receive FIFO can hold at least one maximum-packet-size packet and two
status word per packet (1 031 bytes).
– Periodic request queue depth = 4.
The sequence of operations is as follows:
1. Initialize channel 2. The application must set the ODDFRM bit in OTG_HCCHAR2.
2. Set the CHENA bit in OTG_HCCHAR2 to write an IN request to the periodic request
queue.
3. The OTG_FS host writes an IN request to the periodic request queue for each
OTG_HCCHAR2 register write with the CHENA bit set.
4. The OTG_FS host attempts to send an IN token in the next odd frame.
5. As soon as the IN packet is received and written to the receive FIFO, the OTG_FS host
generates an RXFLVL interrupt.
6. In response to the RXFLVL interrupt, read the received packet status to determine the
number of bytes received, then read the receive FIFO accordingly. The application
must mask the RXFLVL interrupt before reading the receive FIFO, and unmask it after
reading the entire packet.
7. The core generates an RXFLVL interrupt for the transfer completion status entry in the
receive FIFO. This time, the application must read and ignore the receive packet status
when the receive packet status is not an IN data packet (PKTSTS bit in
OTG_GRXSTSR ≠ 0b0010).
8. The core generates an XFRC interrupt as soon as the receive packet status is read.
9. In response to the XFRC interrupt, read the PKTCNT field in OTG_HCTSIZ2. If
PKTCNT ≠ 0 in OTG_HCTSIZ2, disable the channel before re-initializing the channel
for the next transfer, if any. If PKTCNT = 0 in OTG_HCTSIZ2, reinitialize the channel
for the next transfer. This time, the application must reset the ODDFRM bit in
OTG_HCCHAR2.

1092/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 350. Isochronous IN transactions

Application AHB Host USB Device

1 init_reg(ch_1)
init_reg(ch_2) Periodic Request
1 write_tx_fifo 3
Queue
Assume that this queue
(ch_1) 1 can hold 4 entries.
MPS
set_ch_en 2
(ch_2) ch 1
ch_1 4
2 ch_2

3
Odd
OUT
(micro)
frame
DATA0
5 MPS
6
XferCompl interrupt

init_reg(ch_1)
4 IN

write_tx_fifo
(ch_1) 1
MPS
5 DATA0

RxFLvl interrupt

1
read_rx_sts 6 MPS
read_rx_fifo

RxFLvl interrupt

read_rx_sts
7 ch 1
ch_1
8
XferCompl interrupt
ch_2
init_reg(ch_2)
9
set_ch_en
(ch_2)

Even
OUT (micro)
frame

DATA0
MPS
XferCompl interrupt
init_reg(ch_1)
IN
write_tx_fifo 1
(ch_1)
MPS

DATA0

MSv36021V1

1. The grayed elements are not relevant in the context of this figure.
• Selecting the queue depth
Choose the periodic and non-periodic request queue depths carefully to match the
number of periodic/non-periodic endpoints accessed.
The non-periodic request queue depth affects the performance of non-periodic

RM0402 Rev 6 1093/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

transfers. The deeper the queue (along with sufficient FIFO size), the more often the
core is able to pipeline non-periodic transfers. If the queue size is small, the core is
able to put in new requests only when the queue space is freed up.
The core’s periodic request queue depth is critical to perform periodic transfers as
scheduled. Select the periodic queue depth, based on the number of periodic transfers
scheduled in a microframe. If the periodic request queue depth is smaller than the
periodic transfers scheduled in a microframe, a frame overrun condition occurs.
• Handling babble conditions
OTG_FS controller handles two cases of babble: packet babble and port babble.
Packet babble occurs if the device sends more data than the maximum packet size for
the channel. Port babble occurs if the core continues to receive data from the device at
EOF2 (the end of frame 2, which is very close to SOF).
When OTG_FS controller detects a packet babble, it stops writing data into the Rx
buffer and waits for the end of packet (EOP). When it detects an EOP, it flushes already
written data in the Rx buffer and generates a Babble interrupt to the application.
When OTG_FS controller detects a port babble, it flushes the Rx FIFO and disables the
port. The core then generates a port disabled interrupt (HPRTINT in OTG_GINTSTS,
PENCHNG in OTG_HPRT). On receiving this interrupt, the application must determine
that this is not due to an overcurrent condition (another cause of the port disabled
interrupt) by checking POCA in OTG_HPRT, then perform a soft reset. The core does
not send any more tokens after it has detected a port babble condition.

29.16.5 Device programming model

Endpoint initialization on USB reset
1. Set the NAK bit for all OUT endpoints
– SNAK = 1 in OTG_DOEPCTLx (for all OUT endpoints)
2. Unmask the following interrupt bits
– INEP0 = 1 in OTG_DAINTMSK (control 0 IN endpoint)
– OUTEP0 = 1 in OTG_DAINTMSK (control 0 OUT endpoint)
– STUPM = 1 in OTG_DOEPMSK
– XFRCM = 1 in OTG_DOEPMSK
– XFRCM = 1 in OTG_DIEPMSK
– TOM = 1 in OTG_DIEPMSK
3. Set up the data FIFO RAM for each of the FIFOs
– Program the OTG_GRXFSIZ register, to be able to receive control OUT data and
setup data. If thresholding is not enabled, at a minimum, this must be equal to 1
max packet size of control endpoint 0 + 2 words (for the status of the control OUT
data packet) + 10 words (for setup packets).
– Program the OTG_DIEPTXF0 register (depending on the FIFO number chosen) to
be able to transmit control IN data. At a minimum, this must be equal to 1 max
packet size of control endpoint 0.
4. Program the following fields in the endpoint-specific registers for control OUT endpoint
0 to receive a SETUP packet
– STUPCNT = 3 in OTG_DOEPTSIZ0 (to receive up to 3 back-to-back SETUP
packets)

1094/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

At this point, all initialization required to receive SETUP packets is done.

Endpoint initialization on enumeration completion

1. On the Enumeration Done interrupt (ENUMDNE in OTG_GINTSTS), read the
OTG_DSTS register to determine the enumeration speed.
2. Program the MPSIZ field in OTG_DIEPCTL0 to set the maximum packet size. This
step configures control endpoint 0. The maximum packet size for a control endpoint
depends on the enumeration speed.
At this point, the device is ready to receive SOF packets and is configured to perform control
transfers on control endpoint 0.

Endpoint initialization on SetAddress command

This section describes what the application must do when it receives a SetAddress
command in a SETUP packet.
1. Program the OTG_DCFG register with the device address received in the SetAddress
command
2. Program the core to send out a status IN packet

Endpoint initialization on SetConfiguration/SetInterface command

This section describes what the application must do when it receives a SetConfiguration or
SetInterface command in a SETUP packet.
1. When a SetConfiguration command is received, the application must program the
endpoint registers to configure them with the characteristics of the valid endpoints in
the new configuration.
2. When a SetInterface command is received, the application must program the endpoint
registers of the endpoints affected by this command.
3. Some endpoints that were active in the prior configuration or alternate setting are not
valid in the new configuration or alternate setting. These invalid endpoints must be
deactivated.
4. Unmask the interrupt for each active endpoint and mask the interrupts for all inactive
endpoints in the OTG_DAINTMSK register.
5. Set up the data FIFO RAM for each FIFO.
6. After all required endpoints are configured; the application must program the core to
send a status IN packet.
At this point, the device core is configured to receive and transmit any type of data packet.

Endpoint activation
This section describes the steps required to activate a device endpoint or to configure an
existing device endpoint to a new type.

RM0402 Rev 6 1095/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

1. Program the characteristics of the required endpoint into the following fields of the
OTG_DIEPCTLx register (for IN or bidirectional endpoints) or the OTG_DOEPCTLx
register (for OUT or bidirectional endpoints).
– Maximum packet size
– USB active endpoint = 1
– Endpoint start data toggle (for interrupt and bulk endpoints)
– Endpoint type
– Tx FIFO number
2. Once the endpoint is activated, the core starts decoding the tokens addressed to that
endpoint and sends out a valid handshake for each valid token received for the
endpoint.

Endpoint deactivation
This section describes the steps required to deactivate an existing endpoint.
1. In the endpoint to be deactivated, clear the USB active endpoint bit in the
OTG_DIEPCTLx register (for IN or bidirectional endpoints) or the OTG_DOEPCTLx
register (for OUT or bidirectional endpoints).
2. Once the endpoint is deactivated, the core ignores tokens addressed to that endpoint,
which results in a timeout on the USB.
Note: The application must meet the following conditions to set up the device core to handle
traffic:
NPTXFEM and RXFLVLM in the OTG_GINTMSK register must be cleared.

Operational model
SETUP and OUT data transfers:
This section describes the internal data flow and application-level operations during data
OUT transfers and SETUP transactions.
• Packet read
This section describes how to read packets (OUT data and SETUP packets) from the
receive FIFO.
1. On catching an RXFLVL interrupt (OTG_GINTSTS register), the application must read
the receive status pop register (OTG_GRXSTSP).
2. The application can mask the RXFLVL interrupt (in OTG_GINTSTS) by writing to
RXFLVLM = 0 (in OTG_GINTMSK), until it has read the packet from the receive FIFO.
3. If the received packet’s byte count is not 0, the byte count amount of data is popped
from the receive data FIFO and stored in memory. If the received packet byte count is
0, no data is popped from the receive data FIFO.
4. The receive status readout of the packet of FIFO indicates one of the following:
a) Global OUT NAK pattern:
PKTSTS = Global OUT NAK, BCNT = 0x000, EPNUM = (0x0),
DPID = (0b00).
These data indicate that the global OUT NAK bit has taken effect.
b) SETUP packet pattern:
PKTSTS = SETUP, BCNT = 0x008, EPNUM = Control EP Num,

1096/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

DPID = DATA0. These data indicate that a SETUP packet for the specified
endpoint is now available for reading from the receive FIFO.
c) Setup stage done pattern:
PKTSTS = Setup Stage Done, BCNT = 0x0, EPNUM = Control EP Num,
DPID = (0b00).
These data indicate that the setup stage for the specified endpoint has completed
and the data stage has started. After this entry is popped from the receive FIFO,
the core asserts a setup interrupt on the specified control OUT endpoint.
d) Data OUT packet pattern:
PKTSTS = DataOUT, BCNT = size of the received data OUT packet (0 ≤ BCNT
≤ 1 024), EPNUM = EPNUM on which the packet was received, DPID = Actual
Data PID.
e) Data transfer completed pattern:
PKTSTS = Data OUT transfer done, BCNT = 0x0, EPNUM = OUT EP Num on
which the data transfer is complete, DPID = (0b00).
These data indicate that an OUT data transfer for the specified OUT endpoint has
completed. After this entry is popped from the receive FIFO, the core asserts a
transfer completed interrupt on the specified OUT endpoint.
5. After the data payload is popped from the receive FIFO, the RXFLVL interrupt
(OTG_GINTSTS) must be unmasked.
6. Steps 1–5 are repeated every time the application detects assertion of the interrupt line
due to RXFLVL in OTG_GINTSTS. Reading an empty receive FIFO can result in
undefined core behavior.
Figure 351 provides a flowchart of the above procedure.

Figure 351. Receive FIFO packet read

SETUP transactions

RM0402 Rev 6 1097/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

This section describes how the core handles SETUP packets and the application’s
sequence for handling SETUP transactions.
• Application requirements
1. To receive a SETUP packet, the STUPCNT field (OTG_DOEPTSIZx) in a control OUT
endpoint must be programmed to a non-zero value. When the application programs the
STUPCNT field to a non-zero value, the core receives SETUP packets and writes them
to the receive FIFO, irrespective of the NAK status and EPENA bit setting in
OTG_DOEPCTLx. The STUPCNT field is decremented every time the control endpoint
receives a SETUP packet. If the STUPCNT field is not programmed to a proper value
before receiving a SETUP packet, the core still receives the SETUP packet and
decrements the STUPCNT field, but the application may not be able to determine the
correct number of SETUP packets received in the setup stage of a control transfer.
– STUPCNT = 3 in OTG_DOEPTSIZx
2. The application must always allocate some extra space in the receive data FIFO, to be
able to receive up to three SETUP packets on a control endpoint.
– The space to be reserved is 10 words. Three words are required for the first
SETUP packet, 1 word is required for the setup stage done word and 6 words are
required to store two extra SETUP packets among all control endpoints.
– 3 words per SETUP packet are required to store 8 bytes of SETUP data and 4
bytes of SETUP status (setup packet pattern). The core reserves this space in the
receive data FIFO to write SETUP data only, and never uses this space for data
packets.
3. The application must read the 2 words of the SETUP packet from the receive FIFO.
4. The application must read and discard the setup stage done word from the receive
FIFO.
• Internal data flow
1. When a SETUP packet is received, the core writes the received data to the receive
FIFO, without checking for available space in the receive FIFO and irrespective of the
endpoint’s NAK and STALL bit settings.
– The core internally sets the IN NAK and OUT NAK bits for the control IN/OUT
endpoints on which the SETUP packet was received.
2. For every SETUP packet received on the USB, 3 words of data are written to the
receive FIFO, and the STUPCNT field is decremented by 1.
– The first word contains control information used internally by the core
– The second word contains the first 4 bytes of the SETUP command
– The third word contains the last 4 bytes of the SETUP command
3. When the setup stage changes to a data IN/OUT stage, the core writes an entry (setup
stage done word) to the receive FIFO, indicating the completion of the setup stage.
4. On the AHB side, SETUP packets are emptied by the application.
5. When the application pops the setup stage done word from the receive FIFO, the core
interrupts the application with an STUP interrupt (OTG_DOEPINTx), indicating it can
process the received SETUP packet.
6. The core clears the endpoint enable bit for control OUT endpoints.

• Application programming sequence

1098/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

1. Program the OTG_DOEPTSIZx register.

– STUPCNT = 3
2. Wait for the RXFLVL interrupt (OTG_GINTSTS) and empty the data packets from the
receive FIFO.
3. Assertion of the STUP interrupt (OTG_DOEPINTx) marks a successful completion of
the SETUP data transfer.
– On this interrupt, the application must read the OTG_DOEPTSIZx register to
determine the number of SETUP packets received and process the last received
SETUP packet.

Figure 352. Processing a SETUP packet

Wait for STP in OTG_DOEPINTx

rem_supcnt=
rd_reg(OTG_DOEPTSIZx)

setup_cmd[31:0 = mem[4 – 2 * rem_supcnt]

setup_cmd[63:32] = mem[5 – 2 * rem_supcnt]

Find setup cmd type

Read ctrl_rd/wr/2 stage Write

2-stage

setup_np_in_pkt setup_np_in_pkt rcv_out_pkt

Data IN phase Status IN phase Data OUT phase

MSv37035V1

• Handling more than three back-to-back SETUP packets

Per the USB 2.0 specification, normally, during a SETUP packet error, a host does not send
more than three back-to-back SETUP packets to the same endpoint. However, the USB 2.0
specification does not limit the number of back-to-back SETUP packets a host can send to
the same endpoint. When this condition occurs, the OTG_FS controller generates an
interrupt (B2BSTUP in OTG_DOEPINTx).

• Setting the global OUT NAK

Internal data flow:
1. When the application sets the Global OUT NAK (SGONAK bit in OTG_DCTL), the core
stops writing data, except SETUP packets, to the receive FIFO. Irrespective of the

RM0402 Rev 6 1099/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

space availability in the receive FIFO, non-isochronous OUT tokens receive a NAK
handshake response, and the core ignores isochronous OUT data packets
2. The core writes the Global OUT NAK pattern to the receive FIFO. The application must
reserve enough receive FIFO space to write this data pattern.
3. When the application pops the Global OUT NAK pattern word from the receive FIFO,
the core sets the GONAKEFF interrupt (OTG_GINTSTS).
4. Once the application detects this interrupt, it can assume that the core is in Global OUT
NAK mode. The application can clear this interrupt by clearing the SGONAK bit in
OTG_DCTL.
Application programming sequence:
1. To stop receiving any kind of data in the receive FIFO, the application must set the
Global OUT NAK bit by programming the following field:
– SGONAK = 1 in OTG_DCTL
2. Wait for the assertion of the GONAKEFF interrupt in OTG_GINTSTS. When asserted,
this interrupt indicates that the core has stopped receiving any type of data except
SETUP packets.
3. The application can receive valid OUT packets after it has set SGONAK in OTG_DCTL
and before the core asserts the GONAKEFF interrupt (OTG_GINTSTS).
4. The application can temporarily mask this interrupt by writing to the GONAKEFFM bit in
the OTG_GINTMSK register.
– GONAKEFFM = 0 in the OTG_GINTMSK register
5. Whenever the application is ready to exit the Global OUT NAK mode, it must clear the
SGONAK bit in OTG_DCTL. This also clears the GONAKEFF interrupt
(OTG_GINTSTS).
– CGONAK = 1 in OTG_DCTL
6. If the application has masked this interrupt earlier, it must be unmasked as follows:
– GONAKEFFM = 1 in OTG_GINTMSK

• Disabling an OUT endpoint

The application must use this sequence to disable an OUT endpoint that it has enabled.
Application programming sequence:

1100/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

1. Before disabling any OUT endpoint, the application must enable Global OUT NAK
mode in the core.
– SGONAK = 1 in OTG_DCTL
2. Wait for the GONAKEFF interrupt (OTG_GINTSTS)
3. Disable the required OUT endpoint by programming the following fields:
– EPDIS = 1 in OTG_DOEPCTLx
– SNAK = 1 in OTG_DOEPCTLx
4. Wait for the EPDISD interrupt (OTG_DOEPINTx), which indicates that the OUT
endpoint is completely disabled. When the EPDISD interrupt is asserted, the core also
clears the following bits:
– EPDIS = 0 in OTG_DOEPCTLx
– EPENA = 0 in OTG_DOEPCTLx
5. The application must clear the Global OUT NAK bit to start receiving data from other
non-disabled OUT endpoints.
– SGONAK = 0 in OTG_DCTL

• Transfer Stop Programming for OUT endpoints

The application must use the following programing sequence to stop any transfers (because
of an interrupt from the host, typically a reset).
Sequence of operations:
1. Enable all OUT endpoints by setting
– EPENA = 1 in all N/A_DOEPCTLx registers.
2. Flush the RxFIFO as follows
– Poll N/A_GRSTCTL.AHBIDL until it is 1. This indicates that AHB master is idle.
– Perform read modify write operation on N/A_GRSTCTL.RXFFLSH =1
– Poll N/A_GRSTCTL.RXFFLSH until it is 0, but also using a timeout of less than 10
milli-seconds (corresponds to minimum reset signaling duration). If 0 is seen
before the timeout, then the RxFIFO flush is successful. If at the moment the
timeout occurs, there is still a 1, (this may be due to a packet on EP0 coming from
the host) then go back (once only) to the previous step (“Perform read modify write
operation”).
3. Before disabling any OUT endpoint, the application must enable Global OUT NAK
mode in the core, according to the instructions in “Setting the global OUT NAK on
page 1099”. This ensures that data in the RxFIFO is sent to the application
successfully. Set SGONAK = 1 in N/A_DCTL
4. Wait for the GONAKEFF interrupt (N/A_GINTSTS)
5. Disable all active OUT endpoints by programming the following register bits:
– EPDIS = 1 in registers N/A_DOEPCTLx
– SNAK = 1 in registers N/A_DOEPCTLx
6. Wait for the EPDIS interrupt in N/A_DOEPINTx for each OUT endpoint programmed in
the previous step. The EPDIS interrupt in N/A_DOEPINTx indicates that the

RM0402 Rev 6 1101/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

corresponding OUT endpoint is completely disabled. When the EPDIS interrupt is

asserted, the following bits are cleared:
– EPENA = 0 in registers N/A_DOEPCTLx
– EPDIS = 0 in registers N/A_DOEPCTLx
– SNAK = 0 in registers N/A_DOEPCTLx

• Generic non-isochronous OUT data transfers

This section describes a regular non-isochronous OUT data transfer (control, bulk, or
interrupt).
Application requirements:
1. Before setting up an OUT transfer, the application must allocate a buffer in the memory
to accommodate all data to be received as part of the OUT transfer.
2. For OUT transfers, the transfer size field in the endpoint’s transfer size register must be
a multiple of the maximum packet size of the endpoint, adjusted to the word boundary.
– transfer size[EPNUM] = n × (MPSIZ[EPNUM] + 4 – (MPSIZ[EPNUM] mod 4))
– packet count[EPNUM] = n
– n>0
3. On any OUT endpoint interrupt, the application must read the endpoint’s transfer size
register to calculate the size of the payload in the memory. The received payload size
can be less than the programmed transfer size.
– Payload size in memory = application programmed initial transfer size – core
updated final transfer size
– Number of USB packets in which this payload was received = application
programmed initial packet count – core updated final packet count
Internal data flow:
1. The application must set the transfer size and packet count fields in the endpoint-
specific registers, clear the NAK bit, and enable the endpoint to receive the data.
2. Once the NAK bit is cleared, the core starts receiving data and writes it to the receive
FIFO, as long as there is space in the receive FIFO. For every data packet received on
the USB, the data packet and its status are written to the receive FIFO. Every packet
(maximum packet size or short packet) written to the receive FIFO decrements the
packet count field for that endpoint by 1.
– OUT data packets received with bad data CRC are flushed from the receive FIFO
automatically.
– After sending an ACK for the packet on the USB, the core discards non-
isochronous OUT data packets that the host, which cannot detect the ACK, re-
sends. The application does not detect multiple back-to-back data OUT packets
on the same endpoint with the same data PID. In this case the packet count is not
decremented.
– If there is no space in the receive FIFO, isochronous or non-isochronous data
packets are ignored and not written to the receive FIFO. Additionally, non-
isochronous OUT tokens receive a NAK handshake reply.
– In all the above three cases, the packet count is not decremented because no data
are written to the receive FIFO.

1102/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

3. When the packet count becomes 0 or when a short packet is received on the endpoint,
the NAK bit for that endpoint is set. Once the NAK bit is set, the isochronous or non-
isochronous data packets are ignored and not written to the receive FIFO, and non-
isochronous OUT tokens receive a NAK handshake reply.
4. After the data are written to the receive FIFO, the application reads the data from the
receive FIFO and writes it to external memory, one packet at a time per endpoint.
5. At the end of every packet write on the AHB to external memory, the transfer size for
the endpoint is decremented by the size of the written packet.
6. The OUT data transfer completed pattern for an OUT endpoint is written to the receive
FIFO on one of the following conditions:
– The transfer size is 0 and the packet count is 0
– The last OUT data packet written to the receive FIFO is a short packet
(0 ≤ packet size < maximum packet size)
7. When either the application pops this entry (OUT data transfer completed), a transfer
completed interrupt is generated for the endpoint and the endpoint enable is cleared.
Application programming sequence:
1. Program the OTG_DOEPTSIZx register for the transfer size and the corresponding
packet count.
2. Program the OTG_DOEPCTLx register with the endpoint characteristics, and set the
EPENA and CNAK bits.
– EPENA = 1 in OTG_DOEPCTLx
– CNAK = 1 in OTG_DOEPCTLx
3. Wait for the RXFLVL interrupt (in OTG_GINTSTS) and empty the data packets from the
receive FIFO.
– This step can be repeated many times, depending on the transfer size.
4. Asserting the XFRC interrupt (OTG_DOEPINTx) marks a successful completion of the
non-isochronous OUT data transfer.
5. Read the OTG_DOEPTSIZx register to determine the size of the received data
payload.

• Generic isochronous OUT data transfer

This section describes a regular isochronous OUT data transfer.
Application requirements:
1. All the application requirements for non-isochronous OUT data transfers also apply to
isochronous OUT data transfers.
2. For isochronous OUT data transfers, the transfer size and packet count fields must
always be set to the number of maximum-packet-size packets that can be received in a
single frame and no more. Isochronous OUT data transfers cannot span more than 1
frame.
3. The application must read all isochronous OUT data packets from the receive FIFO
(data and status) before the end of the periodic frame (EOPF interrupt in
OTG_GINTSTS).
4. To receive data in the following frame, an isochronous OUT endpoint must be enabled
after the EOPF (OTG_GINTSTS) and before the SOF (OTG_GINTSTS).
Internal data flow:

RM0402 Rev 6 1103/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

1. The internal data flow for isochronous OUT endpoints is the same as that for non-
isochronous OUT endpoints, but for a few differences.
2. When an isochronous OUT endpoint is enabled by setting the endpoint enable and
clearing the NAK bits, the Even/Odd frame bit must also be set appropriately. The core
receives data on an isochronous OUT endpoint in a particular frame only if the
following condition is met:
– EONUM (in OTG_DOEPCTLx) = FNSOF[0] (in OTG_DSTS)
3. When the application completely reads an isochronous OUT data packet (data and
status) from the receive FIFO, the core updates the RXDPID field in OTG_DOEPTSIZx
with the data PID of the last isochronous OUT data packet read from the receive FIFO.
Application programming sequence:
1. Program the OTG_DOEPTSIZx register for the transfer size and the corresponding
packet count
2. Program the OTG_DOEPCTLx register with the endpoint characteristics and set the
endpoint enable, ClearNAK, and Even/Odd frame bits.
– EPENA = 1
– CNAK = 1
– EONUM = (0: Even/1: Odd)
3. Wait for the RXFLVL interrupt (in OTG_GINTSTS) and empty the data packets from the
receive FIFO
– This step can be repeated many times, depending on the transfer size.
4. The assertion of the XFRC interrupt (in OTG_DOEPINTx) marks the completion of the
isochronous OUT data transfer. This interrupt does not necessarily mean that the data
in memory are good.
5. This interrupt cannot always be detected for isochronous OUT transfers. Instead, the
application can detect the INCOMPISOOUT interrupt in OTG_GINTSTS.
6. Read the OTG_DOEPTSIZx register to determine the size of the received transfer and
to determine the validity of the data received in the frame. The application must treat
the data received in memory as valid only if one of the following conditions is met:
– RXDPID = DATA0 (in OTG_DOEPTSIZx) and the number of USB packets in
which this payload was received = 1
– RXDPID = DATA1 (in OTG_DOEPTSIZx) and the number of USB packets in
which this payload was received = 2
The number of USB packets in which this payload was received =
Application programmed initial packet count – core updated final packet count
The application can discard invalid data packets.

• Incomplete isochronous OUT data transfers

This section describes the application programming sequence when isochronous OUT data
packets are dropped inside the core.
Internal data flow:
1. For isochronous OUT endpoints, the XFRC interrupt (in OTG_DOEPINTx) may not
always be asserted. If the core drops isochronous OUT data packets, the application

1104/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

could fail to detect the XFRC interrupt (OTG_DOEPINTx) under the following
circumstances:
– When the receive FIFO cannot accommodate the complete ISO OUT data packet,
the core drops the received ISO OUT data
– When the isochronous OUT data packet is received with CRC errors
– When the isochronous OUT token received by the core is corrupted
– When the application is very slow in reading the data from the receive FIFO
2. When the core detects an end of periodic frame before transfer completion to all
isochronous OUT endpoints, it asserts the incomplete isochronous OUT data interrupt
(INCOMPISOOUT in OTG_GINTSTS), indicating that an XFRC interrupt (in
OTG_DOEPINTx) is not asserted on at least one of the isochronous OUT endpoints. At
this point, the endpoint with the incomplete transfer remains enabled, but no active
transfers remain in progress on this endpoint on the USB.
Application programming sequence:
1. Asserting the INCOMPISOOUT interrupt (OTG_GINTSTS) indicates that in the current
frame, at least one isochronous OUT endpoint has an incomplete transfer.
2. If this occurs because isochronous OUT data is not completely emptied from the
endpoint, the application must ensure that the application empties all isochronous OUT
data (data and status) from the receive FIFO before proceeding.
– When all data are emptied from the receive FIFO, the application can detect the
XFRC interrupt (OTG_DOEPINTx). In this case, the application must re-enable
the endpoint to receive isochronous OUT data in the next frame.
3. When it receives an INCOMPISOOUT interrupt (in OTG_GINTSTS), the application
must read the control registers of all isochronous OUT endpoints (OTG_DOEPCTLx) to
determine which endpoints had an incomplete transfer in the current microframe. An
endpoint transfer is incomplete if both the following conditions are met:
– EONUM bit (in OTG_DOEPCTLx) = FNSOF[0] (in OTG_DSTS)
– EPENA = 1 (in OTG_DOEPCTLx)
4. The previous step must be performed before the SOF interrupt (in OTG_GINTSTS) is
detected, to ensure that the current frame number is not changed.
5. For isochronous OUT endpoints with incomplete transfers, the application must discard
the data in the memory and disable the endpoint by setting the EPDIS bit in
OTG_DOEPCTLx.
6. Wait for the EPDISD interrupt (in OTG_DOEPINTx) and enable the endpoint to receive
new data in the next frame.
– Because the core can take some time to disable the endpoint, the application may
not be able to receive the data in the next frame after receiving bad isochronous
data.

• Stalling a non-isochronous OUT endpoint

This section describes how the application can stall a non-isochronous endpoint.

RM0402 Rev 6 1105/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

1. Put the core in the Global OUT NAK mode.

2. Disable the required endpoint
– When disabling the endpoint, instead of setting the SNAK bit in OTG_DOEPCTL,
set STALL = 1 (in OTG_DOEPCTL).
The STALL bit always takes precedence over the NAK bit.
3. When the application is ready to end the STALL handshake for the endpoint, the
STALL bit (in OTG_DOEPCTLx) must be cleared.
4. If the application is setting or clearing a STALL for an endpoint due to a
SetFeature.Endpoint Halt or ClearFeature.Endpoint Halt command, the STALL bit must
be set or cleared before the application sets up the status stage transfer on the control
endpoint.

Examples
This section describes and depicts some fundamental transfer types and scenarios.
• Bulk OUT transaction
Figure 353 depicts the reception of a single Bulk OUT data packet from the USB to the AHB
and describes the events involved in the process.

Figure 353. Bulk OUT transaction

Host USB Device Application

init _out_ep
XFRSIZ = 64 bytes
1 PKTCNT = 1
2 Wr_reg(OTG_DOEPTSIZx)

O UT EPENA = 1
CN AK = 1
3 Wr_reg(OTG_DOEPCTLx)
64 bytes
4 6
xact_1
AC K RXFLVL iintr
OTG_DO
EPCTLx
, NAK=1
idle until intr
5 PKTCN
T0

XFRSIZ
r =0 rcv_out _pkt()
On new xfer
OU T XF or RxFIFO
int r RC not em pty
7
NA K
idle until intr
8

MS36931V1

After a SetConfiguration/SetInterface command, the application initializes all OUT endpoints

by setting CNAK = 1 and EPENA = 1 (in OTG_DOEPCTLx), and setting a suitable
XFRSIZ and PKTCNT in the OTG_DOEPTSIZx register.

1106/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

1. host attempts to send data (OUT token) to an endpoint.

2. When the core receives the OUT token on the USB, it stores the packet in the Rx FIFO
because space is available there.
3. After writing the complete packet in the Rx FIFO, the core then asserts the RXFLVL
interrupt (in OTG_GINTSTS).
4. On receiving the PKTCNT number of USB packets, the core internally sets the NAK bit
for this endpoint to prevent it from receiving any more packets.
5. The application processes the interrupt and reads the data from the Rx FIFO.
6. When the application has read all the data (equivalent to XFRSIZ), the core generates
an XFRC interrupt (in OTG_DOEPINTx).
7. The application processes the interrupt and uses the setting of the XFRC interrupt bit
(in OTG_DOEPINTx) to determine that the intended transfer is complete.

IN data transfers
• Packet write
This section describes how the application writes data packets to the endpoint FIFO when
dedicated transmit FIFOs are enabled.
1. The application can either choose the polling or the interrupt mode.
– In polling mode, the application monitors the status of the endpoint transmit data
FIFO by reading the OTG_DTXFSTSx register, to determine if there is enough
space in the data FIFO.
– In interrupt mode, the application waits for the TXFE interrupt (in OTG_DIEPINTx)
and then reads the OTG_DTXFSTSx register, to determine if there is enough
space in the data FIFO.
– To write a single non-zero length data packet, there must be space to write the
entire packet in the data FIFO.
– To write zero length packet, the application must not look at the FIFO space.
2. Using one of the above mentioned methods, when the application determines that
there is enough space to write a transmit packet, the application must first write into the
endpoint control register, before writing the data into the data FIFO. Typically, the
application, must do a read modify write on the OTG_DIEPCTLx register to avoid
modifying the contents of the register, except for setting the endpoint enable bit.
The application can write multiple packets for the same endpoint into the transmit FIFO, if
space is available. For periodic IN endpoints, the application must write packets only for one
microframe. It can write packets for the next periodic transaction only after getting transfer
complete for the previous transaction.

• Setting IN endpoint NAK

Internal data flow:

RM0402 Rev 6 1107/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

1. When the application sets the IN NAK for a particular endpoint, the core stops
transmitting data on the endpoint, irrespective of data availability in the endpoint’s
transmit FIFO.
2. Non-isochronous IN tokens receive a NAK handshake reply
– Isochronous IN tokens receive a zero-data-length packet reply
3. The core asserts the INEPNE (IN endpoint NAK effective) interrupt in OTG_DIEPINTx
in response to the SNAK bit in OTG_DIEPCTLx.
4. Once this interrupt is seen by the application, the application can assume that the
endpoint is in IN NAK mode. This interrupt can be cleared by the application by setting
the CNAK bit in OTG_DIEPCTLx.
Application programming sequence:
1. To stop transmitting any data on a particular IN endpoint, the application must set the
IN NAK bit. To set this bit, the following field must be programmed.
– SNAK = 1 in OTG_DIEPCTLx
2. Wait for assertion of the INEPNE interrupt in OTG_DIEPINTx. This interrupt indicates
that the core has stopped transmitting data on the endpoint.
3. The core can transmit valid IN data on the endpoint after the application has set the
NAK bit, but before the assertion of the NAK Effective interrupt.
4. The application can mask this interrupt temporarily by writing to the INEPNEM bit in
OTG_DIEPMSK.
– INEPNEM = 0 in OTG_DIEPMSK
5. To exit endpoint NAK mode, the application must clear the NAK status bit (NAKSTS) in
OTG_DIEPCTLx. This also clears the INEPNE interrupt (in OTG_DIEPINTx).
– CNAK = 1 in OTG_DIEPCTLx
6. If the application masked this interrupt earlier, it must be unmasked as follows:
– INEPNEM = 1 in OTG_DIEPMSK

• IN endpoint disable
Use the following sequence to disable a specific IN endpoint that has been previously
enabled.
Application programming sequence:

1108/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

1. The application must stop writing data on the AHB for the IN endpoint to be disabled.
2. The application must set the endpoint in NAK mode.
– SNAK = 1 in OTG_DIEPCTLx
3. Wait for the INEPNE interrupt in OTG_DIEPINTx.
4. Set the following bits in the OTG_DIEPCTLx register for the endpoint that must be
disabled.
– EPDIS = 1 in OTG_DIEPCTLx
– SNAK = 1 in OTG_DIEPCTLx
5. Assertion of the EPDISD interrupt in OTG_DIEPINTx indicates that the core has
completely disabled the specified endpoint. Along with the assertion of the interrupt, the
core also clears the following bits:
– EPENA = 0 in OTG_DIEPCTLx
– EPDIS = 0 in OTG_DIEPCTLx
6. The application must read the OTG_DIEPTSIZx register for the periodic IN EP, to
calculate how much data on the endpoint were transmitted on the USB.
7. The application must flush the data in the endpoint transmit FIFO, by setting the
following fields in the OTG_GRSTCTL register:
– TXFNUM (in OTG_GRSTCTL) = Endpoint transmit FIFO number
– TXFFLSH in (OTG_GRSTCTL) = 1
The application must poll the OTG_GRSTCTL register, until the TXFFLSH bit is cleared by
the core, which indicates the end of flush operation. To transmit new data on this endpoint,
the application can re-enable the endpoint at a later point.

• Transfer Stop Programming for IN endpoints

The application must use the following programing sequence to stop any transfers (because
of an interrupt from the host, typically a reset).
Sequence of operations:
1. Disable the IN endpoint by setting:
– EPDIS = 1 in all N/A_DIEPCTLx registers
2. Wait for the EPDIS interrupt in N/A_DIEPINTx, which indicates that the IN endpoint is
completely disabled. When the EPDIS interrupt is asserted the following bits are
cleared:
– EPDIS = 0 in N/A_DIEPCTLx
– EPENA = 0 in N/A_DIEPCTLx
3. Flush the TxFIFO by programming the following bits:
– TXFFLSH = 1 in N/A_GRSTCTL
– TXFNUM = “FIFO number specific to endpoint” in N/A_GRSTCTL
4. The application can start polling till TXFFLSH in N/A_GRSTCTL is cleared. When this
bit is cleared, it ensures that there is no data left in the Tx FIFO.

• Generic non-periodic IN data transfers

Application requirements:

RM0402 Rev 6 1109/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

1. Before setting up an IN transfer, the application must ensure that all data to be
transmitted as part of the IN transfer are part of a single buffer.
2. For IN transfers, the transfer size field in the endpoint transfer size register denotes a
payload that constitutes multiple maximum-packet-size packets and a single short
packet. This short packet is transmitted at the end of the transfer.
– To transmit a few maximum-packet-size packets and a short packet at the end of
the transfer:
Transfer size[EPNUM] = x × MPSIZ[EPNUM] + sp
If (sp > 0), then packet count[EPNUM] = x + 1.
Otherwise, packet count[EPNUM] = x
– To transmit a single zero-length data packet:
Transfer size[EPNUM] = 0
Packet count[EPNUM] = 1
– To transmit a few maximum-packet-size packets and a zero-length data packet at
the end of the transfer, the application must split the transfer into two parts. The
first sends maximum-packet-size data packets and the second sends the zero-
length data packet alone.
First transfer: transfer size[EPNUM] = x × MPSIZ[epnum]; packet count = n;
Second transfer: transfer size[EPNUM] = 0; packet count = 1;
3. Once an endpoint is enabled for data transfers, the core updates the transfer size
register. At the end of the IN transfer, the application must read the transfer size
register to determine how much data posted in the transmit FIFO have already been
sent on the USB.
4. Data fetched into transmit FIFO = Application-programmed initial transfer size – core-
updated final transfer size
– Data transmitted on USB = (application-programmed initial packet count – core
updated final packet count) × MPSIZ[EPNUM]
– Data yet to be transmitted on USB = (Application-programmed initial transfer size
– data transmitted on USB)
Internal data flow:
1. The application must set the transfer size and packet count fields in the endpoint-
specific registers and enable the endpoint to transmit the data.
2. The application must also write the required data to the transmit FIFO for the endpoint.
3. Every time a packet is written into the transmit FIFO by the application, the transfer size
for that endpoint is decremented by the packet size. The data is fetched from the
memory by the application, until the transfer size for the endpoint becomes 0. After
writing the data into the FIFO, the “number of packets in FIFO” count is incremented
(this is a 3-bit count, internally maintained by the core for each IN endpoint transmit
FIFO. The maximum number of packets maintained by the core at any time in an IN
endpoint FIFO is eight). For zero-length packets, a separate flag is set for each FIFO,
without any data in the FIFO.
4. Once the data are written to the transmit FIFO, the core reads them out upon receiving
an IN token. For every non-isochronous IN data packet transmitted with an ACK

1110/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

handshake, the packet count for the endpoint is decremented by one, until the packet
count is zero. The packet count is not decremented on a timeout.
5. For zero length packets (indicated by an internal zero length flag), the core sends out a
zero-length packet for the IN token and decrements the packet count field.
6. If there are no data in the FIFO for a received IN token and the packet count field for
that endpoint is zero, the core generates an “IN token received when Tx FIFO is empty”
(ITTXFE) interrupt for the endpoint, provided that the endpoint NAK bit is not set. The
core responds with a NAK handshake for non-isochronous endpoints on the USB.
7. The core internally rewinds the FIFO pointers and no timeout interrupt is generated.
8. When the transfer size is 0 and the packet count is 0, the transfer complete (XFRC)
interrupt for the endpoint is generated and the endpoint enable is cleared.
Application programming sequence:
1. Program the OTG_DIEPTSIZx register with the transfer size and corresponding packet
count.
2. Program the OTG_DIEPCTLx register with the endpoint characteristics and set the
CNAK and EPENA (endpoint enable) bits.
3. When transmitting non-zero length data packet, the application must poll the
OTG_DTXFSTSx register (where x is the FIFO number associated with that endpoint)
to determine whether there is enough space in the data FIFO. The application can
optionally use TXFE (in OTG_DIEPINTx) before writing the data.

• Generic periodic IN data transfers

This section describes a typical periodic IN data transfer.
Application requirements:
1. Application requirements 1, 2, 3, and 4 of Generic non-periodic IN data transfers on
page 1109 also apply to periodic IN data transfers, except for a slight modification of
requirement 2.
– The application can only transmit multiples of maximum-packet-size data packets
or multiples of maximum-packet-size packets, plus a short packet at the end. To

RM0402 Rev 6 1111/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

transmit a few maximum-packet-size packets and a short packet at the end of the
transfer, the following conditions must be met:
transfer size[EPNUM] = x × MPSIZ[EPNUM] + sp
(where x is an integer ≥ 0, and 0 ≤ sp < MPSIZ[EPNUM])
If (sp > 0), packet count[EPNUM] = x + 1
Otherwise, packet count[EPNUM] = x;
MCNT[EPNUM] = packet count[EPNUM]
– The application cannot transmit a zero-length data packet at the end of a transfer.
It can transmit a single zero-length data packet by itself. To transmit a single zero-
length data packet:
– transfer size[EPNUM] = 0
packet count[EPNUM] = 1
MCNT[EPNUM] = packet count[EPNUM]
2. The application can only schedule data transfers one frame at a time.
– (MCNT – 1) × MPSIZ ≤ XFERSIZ ≤ MCNT × MPSIZ
– PKTCNT = MCNT (in OTG_DIEPTSIZx)
– If XFERSIZ < MCNT × MPSIZ, the last data packet of the transfer is a short
packet.
– Note that: MCNT is in OTG_DIEPTSIZx, MPSIZ is in OTG_DIEPCTLx, PKTCNT
is in OTG_DIEPTSIZx and XFERSIZ is in OTG_DIEPTSIZx
3. The complete data to be transmitted in the frame must be written into the transmit FIFO
by the application, before the IN token is received. Even when 1 word of the data to be
transmitted per frame is missing in the transmit FIFO when the IN token is received, the
core behaves as when the FIFO is empty. When the transmit FIFO is empty:
– A zero data length packet would be transmitted on the USB for isochronous IN
endpoints
– A NAK handshake would be transmitted on the USB for interrupt IN endpoints
Internal data flow:
1. The application must set the transfer size and packet count fields in the endpoint-
specific registers and enable the endpoint to transmit the data.
2. The application must also write the required data to the associated transmit FIFO for
the endpoint.
3. Every time the application writes a packet to the transmit FIFO, the transfer size for that
endpoint is decremented by the packet size. The data are fetched from application
memory until the transfer size for the endpoint becomes 0.
4. When an IN token is received for a periodic endpoint, the core transmits the data in the
FIFO, if available. If the complete data payload (complete packet, in dedicated FIFO

1112/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

mode) for the frame is not present in the FIFO, then the core generates an IN token
received when Tx FIFO empty interrupt for the endpoint.
– A zero-length data packet is transmitted on the USB for isochronous IN endpoints
– A NAK handshake is transmitted on the USB for interrupt IN endpoints
5. The packet count for the endpoint is decremented by 1 under the following conditions:
– For isochronous endpoints, when a zero- or non-zero-length data packet is
transmitted
– For interrupt endpoints, when an ACK handshake is transmitted
– When the transfer size and packet count are both 0, the transfer completed
interrupt for the endpoint is generated and the endpoint enable is cleared.
6. At the “Periodic frame Interval” (controlled by PFIVL in OTG_DCFG), when the core
finds non-empty any of the isochronous IN endpoint FIFOs scheduled for the current
frame non-empty, the core generates an IISOIXFR interrupt in OTG_GINTSTS.
Application programming sequence:
1. Program the OTG_DIEPCTLx register with the endpoint characteristics and set the
CNAK and EPENA bits.
2. Write the data to be transmitted in the next frame to the transmit FIFO.
3. Asserting the ITTXFE interrupt (in OTG_DIEPINTx) indicates that the application has
not yet written all data to be transmitted to the transmit FIFO.
4. If the interrupt endpoint is already enabled when this interrupt is detected, ignore the
interrupt. If it is not enabled, enable the endpoint so that the data can be transmitted on
the next IN token attempt.
5. Asserting the XFRC interrupt (in OTG_DIEPINTx) with no ITTXFE interrupt in
OTG_DIEPINTx indicates the successful completion of an isochronous IN transfer. A
read to the OTG_DIEPTSIZx register must give transfer size = 0 and packet count = 0,
indicating all data were transmitted on the USB.
6. Asserting the XFRC interrupt (in OTG_DIEPINTx), with or without the ITTXFE interrupt
(in OTG_DIEPINTx), indicates the successful completion of an interrupt IN transfer. A
read to the OTG_DIEPTSIZx register must give transfer size = 0 and packet count = 0,
indicating all data were transmitted on the USB.
7. Asserting the incomplete isochronous IN transfer (IISOIXFR) interrupt in
OTG_GINTSTS with none of the aforementioned interrupts indicates the core did not
receive at least 1 periodic IN token in the current frame.

• Incomplete isochronous IN data transfers

This section describes what the application must do on an incomplete isochronous IN data
transfer.
Internal data flow:
1. An isochronous IN transfer is treated as incomplete in one of the following conditions:
a) The core receives a corrupted isochronous IN token on at least one isochronous
IN endpoint. In this case, the application detects an incomplete isochronous IN
transfer interrupt (IISOIXFR in OTG_GINTSTS).
b) The application is slow to write the complete data payload to the transmit FIFO
and an IN token is received before the complete data payload is written to the
FIFO. In this case, the application detects an IN token received when Tx FIFO
empty interrupt in OTG_DIEPINTx. The application can ignore this interrupt, as it

RM0402 Rev 6 1113/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

eventually results in an incomplete isochronous IN transfer interrupt (IISOIXFR in

OTG_GINTSTS) at the end of periodic frame.
The core transmits a zero-length data packet on the USB in response to the
received IN token.
2. The application must stop writing the data payload to the transmit FIFO as soon as
possible.
3. The application must set the NAK bit and the disable bit for the endpoint.
4. The core disables the endpoint, clears the disable bit, and asserts the endpoint disable
interrupt for the endpoint.
Application programming sequence:
1. The application can ignore the IN token received when Tx FIFO empty interrupt in
OTG_DIEPINTx on any isochronous IN endpoint, as it eventually results in an
incomplete isochronous IN transfer interrupt (in OTG_GINTSTS).
2. Assertion of the incomplete isochronous IN transfer interrupt (in OTG_GINTSTS)
indicates an incomplete isochronous IN transfer on at least one of the isochronous IN
endpoints.
3. The application must read the endpoint control register for all isochronous IN endpoints
to detect endpoints with incomplete IN data transfers.
4. The application must stop writing data to the Periodic Transmit FIFOs associated with
these endpoints on the AHB.
5. Program the following fields in the OTG_DIEPCTLx register to disable the endpoint:
– SNAK = 1 in OTG_DIEPCTLx
– EPDIS = 1 in OTG_DIEPCTLx
6. The assertion of the endpoint disabled interrupt in OTG_DIEPINTx indicates that the
core has disabled the endpoint.
– At this point, the application must flush the data in the associated transmit FIFO or
overwrite the existing data in the FIFO by enabling the endpoint for a new transfer
in the next microframe. To flush the data, the application must use the
OTG_GRSTCTL register.

• Stalling non-isochronous IN endpoints

This section describes how the application can stall a non-isochronous endpoint.
Application programming sequence:

1114/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

1. Disable the IN endpoint to be stalled. Set the STALL bit as well.

2. EPDIS = 1 in OTG_DIEPCTLx, when the endpoint is already enabled
– STALL = 1 in OTG_DIEPCTLx
– The STALL bit always takes precedence over the NAK bit
3. Assertion of the endpoint disabled interrupt (in OTG_DIEPINTx) indicates to the
application that the core has disabled the specified endpoint.
4. The application must flush the non-periodic or periodic transmit FIFO, depending on
the endpoint type. In case of a non-periodic endpoint, the application must re-enable
the other non-periodic endpoints that do not need to be stalled, to transmit data.
5. Whenever the application is ready to end the STALL handshake for the endpoint, the
STALL bit must be cleared in OTG_DIEPCTLx.
6. If the application sets or clears a STALL bit for an endpoint due to a
SetFeature.Endpoint Halt command or ClearFeature.Endpoint Halt command, the
STALL bit must be set or cleared before the application sets up the status stage
transfer on the control endpoint.
Special case: stalling the control OUT endpoint
The core must stall IN/OUT tokens if, during the data stage of a control transfer, the host
sends more IN/OUT tokens than are specified in the SETUP packet. In this case, the
application must enable the ITTXFE interrupt in OTG_DIEPINTx and the OTEPDIS interrupt
in OTG_DOEPINTx during the data stage of the control transfer, after the core has
transferred the amount of data specified in the SETUP packet. Then, when the application
receives this interrupt, it must set the STALL bit in the corresponding endpoint control
register, and clear this interrupt.

29.16.6 Worst case response time

When the OTG_FS controller acts as a device, there is a worst case response time for any
tokens that follow an isochronous OUT. This worst case response time depends on the AHB
clock frequency.
The core registers are in the AHB domain, and the core does not accept another token
before updating these register values. The worst case is for any token following an
isochronous OUT, because for an isochronous transaction, there is no handshake and the
next token could come sooner. This worst case value is 7 PHY clocks when the AHB clock
is the same as the PHY clock. When the AHB clock is faster, this value is smaller.
If this worst case condition occurs, the core responds to bulk/interrupt tokens with a NAK
and drops isochronous and SETUP tokens. The host interprets this as a timeout condition
for SETUP and retries the SETUP packet. For isochronous transfers, the Incomplete
isochronous IN transfer interrupt (IISOIXFR) and Incomplete isochronous OUT transfer
interrupt (IISOOXFR) inform the application that isochronous IN/OUT packets were
dropped.

Choosing the value of TRDT in OTG_GUSBCFG

The value in TRDT (OTG_GUSBCFG) is the time it takes for the MAC, in terms of PHY
clocks after it has received an IN token, to get the FIFO status, and thus the first data from
the PFC block. This time involves the synchronization delay between the PHY and AHB
clocks. The worst case delay for this is when the AHB clock is the same as the PHY clock.
In this case, the delay is 5 clocks.

RM0402 Rev 6 1115/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

Once the MAC receives an IN token, this information (token received) is synchronized to the
AHB clock by the PFC (the PFC runs on the AHB clock). The PFC then reads the data from
the SPRAM and writes them into the dual clock source buffer. The MAC then reads the data
out of the source buffer (4 deep).
If the AHB is running at a higher frequency than the PHY, the application can use a smaller
value for TRDT (in OTG_GUSBCFG).
Figure 354 has the following signals:
• tkn_rcvd: Token received information from MAC to PFC
• dynced_tkn_rcvd: Doubled sync tkn_rcvd, from PCLK to HCLK domain
• spr_read: Read to SPRAM
• spr_addr: Address to SPRAM
• spr_rdata: Read data from SPRAM
• srcbuf_push: Push to the source buffer
• srcbuf_rdata: Read data from the source buffer. Data seen by MAC
To calculate the value of TRDT, refer to Table 206: TRDT values (FS) .

Figure 354. TRDT max timing case

1116/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

29.16.7 OTG programming model

The OTG_FS controller is an OTG device supporting HNP and SRP. When the core is
connected to an “A” plug, it is referred to as an A-device. When the core is connected to a
“B” plug it is referred to as a B-device. In host mode, the OTG_FS controller turns off VBUS
to conserve power. SRP is a method by which the B-device signals the A-device to turn on
VBUS power. A device must perform both data-line pulsing and VBUS pulsing, but a host can
detect either data-line pulsing or VBUS pulsing for SRP. HNP is a method by which the B-
device negotiates and switches to host role. In Negotiated mode after HNP, the B-device
suspends the bus and reverts to the device role.

A-device session request protocol

The application must set the SRP-capable bit in the core USB configuration register. This
enables the OTG_FS controller to detect SRP as an A-device.

Figure 355. A-device SRP

Suspend 6

DRV_VBUS 1

2 5

VBUS_VALID

VBUS pulsing
A_VALID

4 7
3 Data line pulsing Connect
D+

Low
D-

ai15681c

1. DRV_VBUS = VBUS drive signal to the PHY

VBUS_VALID = VBUS valid signal from PHY
A_VALID = A-peripheral VBUS level signal to PHY
D+ = Data plus line
D- = Data minus line
The following points refer and describe the signal numeration shown in the Figure 355:
1. To save power, the application suspends and turns off port power when the bus is idle
by writing the port suspend and port power bits in the host port control and status
register.
2. PHY indicates port power off by deasserting the VBUS_VALID signal.
3. The device must detect SE0 for at least 2 ms to start SRP when VBUS power is off.
4. To initiate SRP, the device turns on its data line pull-up resistor for 5 to 10 ms. The
OTG_FS controller detects data-line pulsing.
5. The device drives VBUS above the A-device session valid (2.0 V minimum) for VBUS
pulsing.
The OTG_FS controller interrupts the application on detecting SRP. The session

RM0402 Rev 6 1117/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

request detected bit is set in Global interrupt status register (SRQINT set in
OTG_GINTSTS).
6. The application must service the session request detected interrupt and turn on the
port power bit by writing the port power bit in the host port control and status register.
The PHY indicates port power-on by asserting the VBUS_VALID signal.
7. When the USB is powered, the device connects, completing the SRP process.

B-device session request protocol

The application must set the SRP-capable bit in the core USB configuration register. This
enables the OTG_FS controller to initiate SRP as a B-device. SRP is a means by which the
OTG_FS controller can request a new session from the host.

Figure 356. B-device SRP

Suspend 6
1
VBUS_VALID

2
B_VALID
3

DISCHRG_VBUS

4
SESS_END

5 8
Data line pulsing Connect
DP

DM Low

7
CHRG_VBUS VBUS pulsing

ai15682c

1. VBUS_VALID = VBUS valid signal from PHY

B_VALID = B-peripheral valid session to PHY
DISCHRG_VBUS = discharge signal to PHY
SESS_END = session end signal to PHY
CHRG_VBUS = charge VBUS signal to PHY
DP = Data plus line
DM = Data minus line
The following points refer and describe the signal numeration shown in the Figure 356:
1. To save power, the host suspends and turns off port power when the bus is idle.
The OTG_FS controller sets the early suspend bit in the core interrupt register after 3
ms of bus idleness. Following this, the OTG_FS controller sets the USB suspend bit in
the core interrupt register.
The OTG_FS controller informs the PHY to discharge VBUS.
2. The PHY indicates the session’s end to the device. This is the initial condition for SRP.
The OTG_FS controller requires 2 ms of SE0 before initiating SRP.
For a USB 1.1 full-speed serial transceiver, the application must wait until VBUS
discharges to 0.2 V after BSVLD (in OTG_GOTGCTL) is deasserted. This discharge

1118/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

time can be obtained from the transceiver vendor and varies from one transceiver to
another.
3. The OTG_FS core informs the PHY to speed up VBUS discharge.
4. The application initiates SRP by writing the session request bit in the OTG control and
status register. The OTG_FS controller perform data-line pulsing followed by VBUS
pulsing.
5. The host detects SRP from either the data-line or VBUS pulsing, and turns on VBUS.
The PHY indicates VBUS power-on to the device.
6. The OTG_FS controller performs VBUS pulsing.
The host starts a new session by turning on VBUS, indicating SRP success. The
OTG_FS controller interrupts the application by setting the session request success
status change bit in the OTG interrupt status register. The application reads the session
request success bit in the OTG control and status register.
7. When the USB is powered, the OTG_FS controller connects, completing the SRP
process.

A-device host negotiation protocol

HNP switches the USB host role from the A-device to the B-device. The application must set
the HNP-capable bit in the core USB configuration register to enable the OTG_FS controller
to perform HNP as an A-device.

Figure 357. A-device HNP

1. DPPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DP line inside the PHY.
DMPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DM line inside the PHY.
The following points refer and describe the signal numeration shown in the Figure 357:
1. The OTG_FS controller sends the B-device a SetFeature b_hnp_enable descriptor to
enable HNP support. The B-device’s ACK response indicates that the B-device
supports HNP. The application must set host Set HNP enable bit in the OTG control

RM0402 Rev 6 1119/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

and status register to indicate to the OTG_FS controller that the B-device supports
HNP.
2. When it has finished using the bus, the application suspends by writing the port
suspend bit in the host port control and status register.
3. When the B-device observes a USB suspend, it disconnects, indicating the initial
condition for HNP. The B-device initiates HNP only when it must switch to the host role;
otherwise, the bus continues to be suspended.
The OTG_FS controller sets the host negotiation detected interrupt in the OTG
interrupt status register, indicating the start of HNP.
The OTG_FS controller deasserts the DM pull down and DM pull down in the PHY to
indicate a device role. The PHY enables the OTG_DP pull-up resistor to indicate a
connect for B-device.
The application must read the current mode bit in the OTG control and status register
to determine device mode operation.
4. The B-device detects the connection, issues a USB reset, and enumerates the
OTG_FS controller for data traffic.
5. The B-device continues the host role, initiating traffic, and suspends the bus when
done.
The OTG_FS controller sets the early suspend bit in the core interrupt register after 3
ms of bus idleness. Following this, the OTG_FS controller sets the USB suspend bit in
the core interrupt register.
6. In Negotiated mode, the OTG_FS controller detects the suspend, disconnects, and
switches back to the host role. The OTG_FS controller asserts the DM pull down and
DM pull down in the PHY to indicate its assumption of the host role.
7. The OTG_FS controller sets the connector ID status change interrupt in the OTG
interrupt status register. The application must read the connector ID status in the OTG
control and status register to determine the OTG_FS controller operation as an A-
device. This indicates the completion of HNP to the application. The application must
read the Current mode bit in the OTG control and status register to determine host
mode operation.
8. The B-device connects, completing the HNP process.

B-device host negotiation protocol

HNP switches the USB host role from B-device to A-device. The application must set the
HNP-capable bit in the core USB configuration register to enable the OTG_FS controller to
perform HNP as a B-device.

1120/1163 RM0402 Rev 6

RM0402 USB on-the-go full-speed (OTG_FS)

Figure 358. B-device HNP

1. DPPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DP line inside the PHY.
DMPULLDOWN = signal from core to PHY to enable/disable the pull-down on the DM line inside the PHY.
The following points refer and describe the signal numeration shown in the Figure 358:
1. The A-device sends the SetFeature b_hnp_enable descriptor to enable HNP support.
The OTG_FS controller’s ACK response indicates that it supports HNP. The application
must set the device HNP enable bit in the OTG control and status register to indicate
HNP support.
The application sets the HNP request bit in the OTG control and status register to
indicate to the OTG_FS controller to initiate HNP.
2. When it has finished using the bus, the A-device suspends by writing the port suspend
bit in the host port control and status register.
The OTG_FS controller sets the Early suspend bit in the core interrupt register after 3
ms of bus idleness. Following this, the OTG_FS controller sets the USB suspend bit in
the core interrupt register.
The OTG_FS controller disconnects and the A-device detects SE0 on the bus,
indicating HNP. The OTG_FS controller asserts the DP pull down and DM pull down in
the PHY to indicate its assumption of the host role.
The A-device responds by activating its OTG_DP pull-up resistor within 3 ms of
detecting SE0. The OTG_FS controller detects this as a connect.
The OTG_FS controller sets the host negotiation success status change interrupt in the
OTG interrupt status register, indicating the HNP status. The application must read the
host negotiation success bit in the OTG control and status register to determine host

RM0402 Rev 6 1121/1163

1122
USB on-the-go full-speed (OTG_FS) RM0402

negotiation success. The application must read the current Mode bit in the core
interrupt register (OTG_GINTSTS) to determine host mode operation.
3. The application sets the reset bit (PRST in OTG_HPRT) and the OTG_FS controller
issues a USB reset and enumerates the A-device for data traffic.
4. The OTG_FS controller continues the host role of initiating traffic, and when done,
suspends the bus by writing the port suspend bit in the host port control and status
register.
5. In Negotiated mode, when the A-device detects a suspend, it disconnects and switches
back to the host role. The OTG_FS controller deasserts the DP pull down and DM pull
down in the PHY to indicate the assumption of the device role.
6. The application must read the current mode bit in the core interrupt (OTG_GINTSTS)
register to determine the host mode operation.
7. The OTG_FS controller connects, completing the HNP process.

1122/1163 RM0402 Rev 6

RM0402 Debug support (DBG)

30 Debug support (DBG)

30.1 Overview
The STM32F412xx is built around a Cortex®-M4 with FPU core which contains hardware
extensions for advanced debugging features. The debug extensions allow the core to be
stopped either on a given instruction fetch (breakpoint) or data access (watchpoint). When
stopped, the core’s internal state and the system’s external state may be examined. Once
examination is complete, the core and the system may be restored and program execution
resumed.
The debug features are used by the debugger host when connecting to and debugging the
STM32F412xx MCUs.
Two interfaces for debug are available:
• Serial wire
• JTAG debug port

Figure 359. Block diagram of STM32 MCU and Cortex®-M4 with FPU-level
debug support

STM32F4xx debug support

Cortex-M4 debug s uppo rt
Bus matrix
DCode
interface
i
D

Cortex-M4 Data
System
core interface
d

JTMS/
te

SWDIO