11 Ways to Protect

Your Power Path

Design Tips and Tradeoffs Using TI’s Power Switches

TI.com I 2019
 Table of Contents
Introduction: Basics of Power Switches . . . . . . . . 3 Chapter 6: Overvoltage Protection . . . . . . . . . . . . 37

Chapter 1: Inrush Current Control . . . . . . . . . . . . 14 Chapter 7: Thermal Shutdown . . . . . . . . . . . . . . . 40

Chapter 2: Adjustable Current Limit . . . . . . . . . . . 18 Chapter 8: Current Monitoring . . . . . . . . . . . . . . . 42

Chapter 3: Short-Circuit Protection . . . . . . . . . . . 23 Chapter 9: Power-Good Signal . . . . . . . . . . . . . . . 48

Chapter 4: Reverse Current Blocking . . . . . . . . . . 28 Chapter 10: Safely Driving an Inductive Load . . . 52

Chapter 5: Reverse Polarity Protection . . . . . . . . 32 Chapter 11: Automotive Load Dump . . . . . . . . . . 58

11 Ways to Protect Your Power Path 2 Texas Instruments

 Introduction: Basics of Power Switches

Application Report help reduce system maintenance costs and maximize

equipment uptime.
SLVA927A–November 2017–Revised April 2019
Authors: Arthur Huang, Emily Roth Ideal diode, ORing controllers provide protection against
reverse-polarity conditions by monitoring an external
Abstract FET, significantly reducing power loss, and blocking
A Power Switch provides an electrical connection from reverse current. Whenever a transient event occurs,
a voltage source or ground to a load. It saves power the controller monitors and adjusts the external FET to
across multiple voltage rails and protects subsystems prevent damage to upstream components.
from damage. It also provides enhanced component
Smart high-side switches are for off-board load
protection, inrush current protection, and minimizes
protection. They provide additional diagnostic telemetry
printed-circuit board (PCB) size.
that monitors the output load current and detects short-
There are several power switch topologies with circuit and open-load events. Smart high-side switches
different functions that address different applications. have adjustable current limits, allowing more reliable
Load Switches establish the power switch foundation integration into applications with either large inrush
by providing safe and reliable distribution of power. current startup profiles or low peak currents. Adding a
Applications typically using load switches include power smart high-side switch to a design leads to a smarter
distribution, power sequencing, inrush current control, and more robust solution for driving capacitive, inductive,
and reduced current leakage. Integrated Power MUX and LED loads.
devices are similar to load switches but allow for multiple
Low-side switches connect the load to ground instead
input sources. This set of electronic switches is used
of providing a connection between a power supply and
to select and transition between two or more input
the load. By including an integrated flyback diode, low-
power paths to a single output while also providing input
side switches help eliminate inductive load transients
power protection.
by dissipating current in a circular loop. This allows
eFuses and Hot Swap controllers provide additional input them to drive inductive loads such as solenoids, relays,
power path protection functions such as current sense and motors.
monitoring, current limiting, undervoltage and overvoltage
This application report highlights the different topologies
protection, and thermal shutdown. This makes these
within the power switch portfolio, and provides
devices ideal for hot-plug and transient events that would
suggestions in choosing the correct solution for a faster
otherwise damage system components. These benefits
design time.

11 Ways to Protect Your Power Path 3 Texas Instruments

Introduction: Basics of Power Switches

www.ti.com

Table 1. Power Switch Topology Table

OUTPUT POWER
www.ti.com
POWER DISTRIBUTION INPUT POWER PROTECTION
PROTECTION
Load Switch Table 1. MUX
Power Power Switch
eFuse Topology Table Ideal Diode
Hot Swap Smart High- Low-Side
(2 input, 1 (Internal FET) (External FET) ORing Side Switch Switch
output) Controller OUTPUT POWER
POWER DISTRIBUTION INPUT POWER PROTECTION
Voltage Range 0 V to 18 V 2.8 V to 22 V 2.7 V to 60 V ±80 V ±75 V PROTECTION
6 V to 40 V 0 V to 100 V
Max Operating Current Load15Switch
A Power MUX
4.5 A eFuse
15 A HotN/A
Swap Ideal Diode
N/A Smart High-
12 A Low-Side
1A
(2 input, 1 (Internal FET) (External FET) ORing Side Switch Switch
Functions output) Controller
Inrush Current
Voltage Control
Range ✓ 18 V
0 V to 2.8 V ✓
to 22 V 2.7 V ✓
to 60 V ✓V
±80 ±75 V ✓ 40 V
6 V to 0 V to 100 V
Adjustable Current
Max Operating Limit
Current 15 A ✓A
4.5 ✓A
15 ✓
N/A N/A ✓A
12 1A
ReverseFunctions
Current Blocking ✓ ✓ ✓ ✓ ✓
Current
InrushSense
CurrentMonitoring
Control ✓ ✓ ✓ ✓ ✓
Short-Circuit
Adjustable Protection
Current Limit ✓ (1) ✓ ✓ ✓ ✓
Overvoltage
Reverse Protection
Current Blocking ✓ ✓ ✓ ✓ ✓
Reverse
Current Polarity Protection
Sense Monitoring ✓ ✓ ✓ ✓ ✓
Power GoodProtection
Short-Circuit Signal ✓✓(1) ✓ ✓ ✓ ✓
Inductive Load Compatibility
Overvoltage Protection ✓ ✓ ✓ ✓ ✓
Load-Dump
Reverse Compatibility
Polarity Protection ✓ ✓ ✓ ✓ ✓ ✓
Thermal
Power Shutdown
Good Signal ✓ ✓ ✓ ✓ ✓
Inductive Load Compatibility ✓ ✓
(1)
Self protected load switch
Load-Dump Compatibility ✓ ✓ ✓ ✓ ✓
Thermal Shutdown ✓ ✓ ✓ ✓ ✓

(1)
Input Protection Power Distribution
Self protected load switch

Ideal Diode Hot Swap

Controller
Input Protection Power Distribution
Load Switch Load
Ideal Diode Hot Swap
Controller
DC/DC Load
Hot Swap
Ideal Diode Controller Load Switch Load
eFuse
eFuse

DC/DC Load Switch Load

Hot Swap
Load
Ideal Diode Controller
eFuse
eFuse
Output Power
Power Load Switch
Mux VDS Clamp Protection
Load

High Side Switch Output Power Load

Power
Power
MuxDistribution VDS Clamp Protection

Figure 1. Power Switch Use

High Side Cases
Switch Load

Power Distribution

Figure 1. Typical
Figure Power
1. Power Switch
Switch UseUse Cases
Cases

2 to Protect
11 Ways BasicsYour
of Power
PowerSwitches
Path 4 SLVA927A – November 2017 – Revised AprilInstruments
Texas 2019
Submit Documentation Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

www.ti.com Load Switches

1 Load Switches

VIN VOUT

ON Control
Logic QOD

CT GND

Copyright © 2017, Texas Instruments Incorporated

Figure 2. Load Switch Block Diagram

Integrated load switches are electronic switches that turn power rails on and off. When the internal FET
turns on, current flows from the input to output and passes power to the downstream circuitry. When the
device is enabled, the rise time of the output voltage (VOUT) can be controlled by adjusting the capacitance
on an external pin (CT pin). When the device is disabled, the fall time of VOUT is controlled through the
quick output discharge (QOD). QOD pulls the output to ground whenever the device is turned off,
preventing the output from floating or entering an undetermined state.
Some common functions of load switches include power savings, power sequencing, and inrush current
control. Power savings is important in applications looking to minimize current dissipation and maximize
power efficiency. By disconnecting the supply from a load or subsystem, the switch minimizes power
drawn from inactive loads. Power sequencing is important in applications where individual voltage rails
need to be turned on and off in a specific order. By configuring the CT and QOD pins, the ramp-up and
power-down timing can be adjusted. Inrush current control protects systems that contain large bulk
capacitors near the load. When power is initially applied to the system, charging these capacitors can
result in a large inrush current that exceeds the nominal load current. If left unaddressed, this can cause
voltage rails to fall out of regulation due to the drop, resulting in the system entering an undesired state.
Load switches can mitigate the inrush current by using the CT pin to manage the rise time of the power
rail. This leads to a linear output slew rate with no voltage dips or external regulators required.

Table 2. Load Switch Examples (1)

VOLTAGE MAX
DESCRIPTION DEVICES TYPICAL Ron PACKAGE
RANGE CURRENT
TPS22918 1 V to 5.5 V 2A 52 mΩ SOT
Adjustable rise time, adjustable QOD
TPS22810 2.7 V to 18 V 2A 79 mΩ SOT
TPS22915 1.05 V to 5.5 V 2A 37 mΩ CSP
Space-constrained applications
TPS22916 1 V to 5.5 V 2A 60 mΩ CSP
Self protected with controlled rise time TPS22919 1.6 V to 5.5 V 1.5 A 90 mΩ SOT
Lowest ON-resistance, Power Good indication TPS22990 1 V to 5.5 V 10 A 3.9 mΩ SON
Fast turn-on time (≤ 65 µs), Power Good indication,
TPS22971 0.65 V to 3.6 V 3A 6.7 mΩ DSBGA
QOD, Thermal shutdown

(1)
For more information about load switches, visit TI.com/LoadSwitches.

SLVA927A
11 Ways – November
to Protect 2017
Your Power – Revised April 2019
Path 5 Basics of Power Switches 3
Texas Instruments
Submit Documentation Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

Power Multiplexing www.ti.com

2 Power Multiplexing

IN1 BFET1 HFET1

SS
Temp
SNS CSS

Control Logic
+ Gate Drivers

OUT

Protecion

Temp
IN2 SNS
BFET2 HFET2

Figure 3. Power MUX Block Diagram

Integrated Power MUX devices allow a system to transition between different power sources seamlessly.
If the main power supply fails, power multiplexing allows the system to switch to a backup power supply,
such as a battery, to preserve operating conditions. Power multiplexing can also provide switching
between two different voltage levels for subsystems that operate at two different voltages. In this scenario,
to prevent reverse current flow from VOUT into one of the VIN channels, reverse current protection (RCP)
blocks current from flowing back through the body diode. Power multiplexing also contains adjustable
current limits. If the current exceeds the threshold set by the switch, the switch clamps the channel and
prevent current from exceeding the limit. Furthermore, if the current limit forces the device to reach higher
temperatures, thermal shutdown will turn off the switch until it can operate at safe conditions again. Similar
to load switches, power MUX switches also contain inrush current control to prevent large transient current
events.
Power MUX devices can switch between different power rails in three general ways: manually,
automatically, or both. Manual switchovers occur with an external GPIO. Whenever the user wants to
switch between power rails, the enable pin is toggled and the output is powered by the other power rail.
Automatic switchover occurs whenever the primary power supply fails or is disconnected. When the
device detects the voltage drop, it automatically switches to the backup power rail. There are some Power
MUX solutions which offer the flexibility to be used in an automatic configuration and be controlled by a
manual control signal. This method can have a default (automatic) priority, but then be overridden by an
external microcontroller if needed.

Table 3. Integrated Power MUX Examples (1)

RECOMMENDED TYPICAL
DESCRIPTION DEVICE MAX CURRENT PACKAGE
VOLTAGE RANGE Ron
Automatic priority and manual override, adjustable
TPS2120 2.8 V to 22 V 3 A, each channel 62 mΩ CSP
current limit
Automatic priority and manual override, fast output
TPS2121 2.8 V to 22 V 4.5 A, each channel 56 mΩ QFN
switchover, adjustable current limit

(1)
For more information about power multiplexing, visit TI.com/PowerMux.

4 to Protect
11 Ways BasicsYour
of Power
PowerSwitches
Path 6 SLVA927A – November 2017 – Revised AprilInstruments
Texas 2019
Submit Documentation Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

www.ti.com eFuses

3 eFuses

VIN VOUT
OVP

EN/UVLO Control
Logic

dVdT

Current ILIM

GND Limit

Copyright © 2017, Texas Instruments Incorporated

Figure 4. eFuse Block Diagram

eFuses are integrated power protection switches that provide voltage and current protection during fault
events. These include short-circuit, overcurrent, overvoltage, undervoltage, and temperature events that
might otherwise damage downstream loads. During a short-circuit transient event, the current through the
eFuse increases very rapidly. The eFuse enables a fast-trip current threshold that terminates this rapid
increase in less than 200 ns, protecting the supply. If an overvoltage event occurs on the input (VIN), the
eFuse monitors the voltage across the internal FET and clamps the output voltage until the input falls
below the overvoltage threshold. eFuses also come with built-in overtemperature protection that shuts
down the FET if the junction temperature exceeds 150°C (typical). The eFuse either remains off (latch-off
version) or attempts to restart (auto retry version) the device after the junction temperature decreases.
eFuses offer many additional features similar to load switches including adjustable inrush current control
and reverse current protection.
Managing current flow from an active power bus to a subsidiary system can be a challenging task. As a
device is inserted or removed from a live supply, it is possible to see a very large spike in current during
the initial capacitor charging. An eFuse or Section 4 hot-swap controller ensures the safe insertion and
operation of these systems. Unlike hot swap controllers, eFuses contain an integrated FET which
minimizes total solution size. This allows eFuses to be used in applications such as power multiplexing.
By using two eFuses, each eFuse can control a power rail while providing reverse current protection for
its respective supply. eFuses are also UL 2367 certified, cutting down on system testing time.

Table 4. eFuse Examples (1)

RECOMMENDED MAX
TYPICA
DESCRIPTION DEVICE VOLTAGE CURRE PACKAGE
L Ron
RANGE NT
Overvoltage or undervoltage protection, QOD using FLT
TPS2595 2.7 V to 18 V 4A 34 mΩ SON
pin, adjustable current limit
Lowest Ron circuit-breaker device, accurate load monitoring,
TPS25982 2.5 V to 24 V 15 A 3 mΩ QFN
adjustable transient fault management
Back to back FETs, status monitoring, thermal shutdown, internal
TPS25942A 2.7 V to 18 V 5A 42 mΩ QFN
reverse current blocking
Reverse polarity protection, current sense output, adjustable SOP and
TPS2660 4.2 V to 55 V 2A 150 mΩ
current limit QFN
TPS1663 4.5 V to 60 V 6A 31 mΩ SOP and
Power limiting, overvoltage cut-off or voltage clamping functionality
QFN
SOP and
Power limiting, reverse current blocking, reverse polarity protection TPS2663 4.5 V to 60 V 6A 31 mΩ
QFN

(1)
For more information about eFuses, visit www.ti.com/efuses.

SLVA927A
11 Ways – November
to Protect 2017
Your Power – Revised April 2019
Path 7 Basics of Power Switches 5
Texas Instruments
Submit Documentation Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

Hot Swap www.ti.com

4 Hot Swap
Rsns
VIN VOUT

GATE

Rpg
OVP Telemetry PG
Isns Gate Logic
Control

Control Logic
UVLO

TIMER PWR GND

Copyright © 2017, Texas Instruments Incorporated

Figure 5. Hot Swap Block Diagram

Hot Swap controllers drive an external MOSFET that protects the system against hot swap events. Hot
Swap controllers do not integrate a MOSFET as eFuses do. The external MOSFET allows hot swap
controllers to operate at higher voltages and currents than eFuse devices. The controller monitors the gate
voltage of the external FET and adjusts the voltage depending on the situation. When the device is
inserted into a live power system, the controller measures the inrush current across RSNS. If the value
exceeds the programable current limit, the gate voltage is lowered and limits the current passing
downstream. If the power dissipated across the FET exceeds the programmable power limit, then the gate
voltage is reduced to lower the current flowing through RSNS. The overvoltage and undervoltage pins also
cut off the output voltage whenever the input voltage is not within specified thresholds.
To ensure that the external MOSFET remains within safe operating area (SOA), the hot swap controller
regulates the current limit at higher VDS voltages. The device also includes an assortment of telemetry that
monitors the operating conditions. The Power Good (PG) signal turns on whenever the power rail reaches
regulation, and some hot swap controllers contain PMBus monitoring that allows real-time feedback on the
device status.
Since hot swap controllers operate by controlling an external RSNS and MOSFET, they do not contain an
innate current limit. The external components allow the user to customize the solution size and power
requirements to fit their application.

Table 5. Hot Swap Examples (1)

DESCRIPTION DEVICE RECOMMENDED VOLTAGE RANGE PACKAGE
18-V analog devices, small footprint, easy-to-use TPS247xx 2.5 V to 18 V SOP or QFN
Meets 240-V requirements for high-end applications, similar to TPS247xx TPS2477x 2.5 V to 18 V QFN
LM25066A,
PMBus and I2C communication, balance between efficiency and accuracy 2.9 V to 17 V QFN
LM25066I
Higher voltage applications, PMBus and I2C communication, external FET
LM5066 10 V to 80 V PWP
temperature and failure sensing
Higher voltage applications, SOA protection, current limit LM5069 9 V to 80 V SOP
Negative voltage support, SOA protection, PMBus and I2C communication LM5064 –10 V to –80 V SOP
Negative voltage support with dual current limit, soft-start disconnect,
TPS2352x –10 V to –80 V SOP
ORing support, –200 V maximum rating
Circuit breaker function for severe current events, programmable fault
LM5067 –9 V to –80 V SOP
timer, PG output

(1)
For more information about Hot Swap controllers, visit www.ti.com/hotswap.

6 to Protect
11 Ways BasicsYour PowerSwitches
of Power Path 8 SLVA927A – November 2017 – Revised AprilInstruments
Texas 2019
Submit Documentation Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

www.ti.com Ideal Diode, ORing Controllers

5 Ideal Diode, ORing Controllers

VIN VOUT

Anode Cathode
GATE

Current
Pulldown

Control Logic

Copyright © 2017, Texas Instruments Incorporated

Figure 6. Ideal Diode Block Diagram

Ideal diode controllers control an external FET and, similar to a regular diode, can block reverse current
whenever a reverse voltage event occurs. Whenever one of these events occur, the controller shuts off
the FET and uses the body diode to prevent any transients from damaging upstream components. The
controller can also prevent against ground shorts at the input (VIN) by using the same method.
Ideal diode controllers can also protect against reverse polarity conditions, commonly caused by
connecting a battery incorrectly or mis-wiring a power supply. If the user accidentally switches the polarity
on VIN, an additional diode from the controller to GND can be included to prevent damage to the IC or the
power source. The controller also significantly lowers power dissipation normally found across diodes. By
driving the external FET instead of a diode, the voltage drop typically found across diode solutions can be
minimized.
Ideal diodes can also act as ORing controllers. Basic power redundancy architecture contains two or more
power supplies connected to a single load. ORing solutions allow the system to switch between power
sources if one were to fail, and can even connect power sources in parallel. This allows for uninterrupted
power and saves on redundant power supply costs.

Table 6. Ideal Diode Controller Examples (1)

RECOMMENDED TYPICAL FORWARD

DESCRIPTION DEVICE VOLTAGE QUIESCENT VOLTAGE PACKAGE
RANGE CURRENT THRESHOLD
Fast turnoff, voltage sensing TPS241x 3 V to 16.5 V N/A 10 mV SOP and SOIC
Automotive qualified, low Iq, high efficiency LM74700-Q1 3.2 V to 65 V 80 µA 20 mV SOT
Low side ORing controller, FET diagnostics LM5051 –6 V to –100 V 69 µA 45 mV SOIC
(2)
Low Iq, reverse current protection, integrated FET LM66100 1.5 V to 5.5 V 0.2 µA 79 mV SC-70

(1)
For more information about Ideal diode controllers, visit www.ti.com/idealdiode.
(2)
Typical forward voltage at Iout = 1 A.

SLVA927A
11 Ways – November
to Protect 2017
Your Power – Revised April 2019
Path 9 Basics of Power Switches 7
Texas Instruments
Submit Documentation Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

Smart High-Side Switches www.ti.com

6 Smart High-Side Switches

Drain (VS)
VIN VDS Clamp

Diagnostics
Diagnostics
and Gate Logic
Protection

Current Sense Current VOUT

Limit/Sense

CL

Copyright © 2017, Texas Instruments Incorporated

Figure 7. Smart High-side Switch Block Diagram

Smart high-side switches reliably drive off-board loads. These switches contain highly-adjustable and
selectable current limits that enable a system to be optimally designed for specific loads. By connecting an
external resistor to set the current-limit threshold, the switch protects the load and power supply from
overstressing during short-circuits to GND events or power-up conditions. This enables more reliable
designs by minimizing transient currents and supply droops. When the threshold is reached, a closed loop
activates and clamps the output current to the set value. A fault is then reported on the CS pin.
These switches also offer highly-accurate current sensing to provide real-time diagnostics to the system. A
current mirror sources current from VIN, reflecting this as voltage on the Current Sense (CS) pin. The CS
pin does not need to be calibrated, and can serve as a diagnostics report pin. Whenever an open load or
short happens, the voltage on the CS pin falls to 0 V. Whenever a current limit, thermal event, or an open
load or short in the off state occurs, the voltage is pulled up to its maximum threshold. High-accuracy
current monitoring and adjustable current limit are ideal for industrial applications like programmable logic
controllers, motor valves, servo drives and control units.
Another functionality of smart high-side switches is load-dump compatibility, which allows these devices to
connect directly to a 12-V battery without concerns about typical voltage and current transients. Additional
protection includes mitigation of large inrush current events that would otherwise damage downstream
components.
Smart high-side switches can be AEC-Q100 certified, allowing full integration into many automotive
applications that require a low on-resistance and high voltage tolerances to accommodate voltage spikes
and inrush current events. Some of these applications include front and rear lighting, seat heating,
infotainment, cluster, powertrain, and ADAS.

Table 7. Smart High-Side Examples (1) (2)

CURRENT
RECOMMENDED
SENSE CONTINUOUS LOAD TYPICA PACKA
DESCRIPTION DEVICE VOLTAGE
ACCURA CURRENT L Ron GE
RANGE
CY
Selectable current limit for design flexibility, low RON,
TPS1HA08-Q1 3 V to 40 V ±5% at 1 A 0 A to 12 A 8 mΩ SOP
small footprint, thermal sensing

Low standby current, highly accurate current sense, TPS1H100-Q1 3.5 V to 40 V ±3% at 1 A 0 A to 4 A 100 mΩ SOP
thermal shutdown TPS27S100 3.5 V to 40 V ±3% at 1 A 0 A to 4 A 80 mΩ SOP
Multi-channel support, fast hardware interrupts, low
TPSxH160-Q1 3.4 V to 40 V ±3% at 1 A 0 A to 1.8 A per channel 160 mΩ SOP
standby current, loss of GND diagnostics

(1)
Additional devices in various typical RON values are available.
(2)
For more information about smart high-side switches, see ti.com/smarthighsideswitch.

8 Basics of Power Switches SLVA927A – November 2017 – Revised April 2019

11 Ways to Protect Your Power Path 10 Texas
Submit Documentation Instruments
Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

www.ti.com Smart High-Side Switches

Table 7. Smart High-Side Examples (1) (2)

(continued)
CURRENT
RECOMMENDED
SENSE CONTINUOUS LOAD TYPICA PACKA
DESCRIPTION DEVICE VOLTAGE
ACCURA CURRENT L Ron GE
RANGE
CY
TPS2HB08-Q1 3 V to 28 V +3/-8% at 0 A to 8 A per channel 8 mΩ SOP
Dual-channel, programmable current limit
1A
Inductive load negative clamp with optimized slew rate, TPSxH000-Q1 3.4 V to 40 V - 0 A to 1A 1Ω SOP
global fault report
Highly accurate current limit, supports full diagnostics TPS1H200-Q1 3.4 V to 40 V - 0 A to 2.5 A 200 mΩ SOP
with the digital status output

SLVA927A – November 2017 – Revised April 2019 Basics of Power Switches 9

Submit
11 Ways Documentation
to Protect Feedback
Your Power Path 11 Texas Instruments
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches

Low-Side Switches www.ti.com

7 Low-Side Switches
Vcc Vcc

Load 1

IN1

Load 2

IN2

Load 7

IN7

Copyright © 2017, Texas Instruments Incorporated

Figure 8. Low-Side Switch Block Diagram

Low-side switches are used to connect and disconnect ground from a load, unlike the rest of the power
switch topologies. This configuration allows low-side switches to drive inductive loads; an internal flyback
diode prevents inductive transients from damaging the circuit and components. Whenever the switch is
opened, the inductive transients flow through the flyback diode and dissipate throughout the load. This
makes these devices ideal for motors, solenoids, and relays.
Low-side switches consists of two designs: Darlington pair arrays and low-side MOSFET solutions.
Darlington pair solutions can support higher voltage applications due to the higher voltage ratings of the
integrated BJTs, while the MOSFET solutions have lower on-resistances and lower leakage currents. Most
of the low-side switches contain seven channels, which can be tied in parallel to support higher current
operation.

Table 8. Low-side Switch Examples

RECOMMENDED MAX NUMBER OF
DESCRIPTION DEVICE PACKAGE
VOLTAGE RANGE CURRENT CHANNELS
SOIC, SOP,
500 mA per
Darlington pair BJTs, higher voltage support ULN2003A 0 V to 50 V 7 and DIP
channel
Packages
500 mA per
Darlington pair BJTs, 8-channel support ULN2803A 0 V to 50 V 8 SOIC
channel
Low-side MOSFET solutions, low on-resistance and current 600 mA per
TPL7407LA 0 V to 30 V 7 SOIC or SOP
leakage, power efficient channel

8 References
1. Texas Instruments, Basics of Load Switches Application Report
2. Texas Instruments, What is an eFuse? Application Report
3. Texas Instruments, Robust Hot Swap Design Application Report
4. Texas Instruments, Adjustable Current Limit of Smart High Side Switch Application Report

10 Basics of Power Switches SLVA927A – November 2017 – Revised April 2019

11 Ways to Protect Your Power Path 12 Submit DocumentationTexas Instruments
Feedback
Copyright © 2017–2019, Texas Instruments Incorporated
Introduction: Basics of Power Switches
www.ti.com Revision History

Revision History
NOTE: Page numbers for previous revisions may differ from page numbers in the current version.

Changes from Original (November 2018) to A Revision ................................................................................................ Page

• Changed content in abstract. ............................................................................................................ 1

• Changed headings, row 3, and row 7 in table 1. ...................................................................................... 2
• Added table notes to table 1. ............................................................................................................ 2
• Changed row 3 in table 2. ............................................................................................................... 3
• Changed content in section 2. ........................................................................................................... 4
• Changed row 1 and 2 in table 3. ........................................................................................................ 4
• Changed rows 2, 5, and 6 in table 4. ................................................................................................... 5
• Changed title and row 1, 2, and 4 in table 6. .......................................................................................... 7
• Changed content in section 6. ........................................................................................................... 8
• Changed rows 4, 5, and 6 in table 7. ................................................................................................... 8

SLVA927A – November 2017 – Revised April 2019 Revision History 11

Submit Documentation Feedback
11 Ways to Protect Your Power Path 13 Instruments Incorporated
Copyright © 2017–2019, Texas Texas Instruments
 Chapter 1: Inrush Current Control

Author: Alex Triano Some systems have restrictions on how much current
they can draw from the input. These restrictions may
Abstract
be set by safety standards, cable/connector ratings or
When a subsystem turns on, current initially flows to any the need to stay below the current rating of a fast-blow
capacitors or inductance at the load. This current is often fuse. A failure to manage inrush current can exceed
referred to as inrush current. If a switch turns on quickly restrictions and lead to compliance failures or damaged
(such as without slew-rate control), then the current cables, connectors or fuses.
may be large enough to interrupt other subsystems
A power switch that uses a metal-oxide semiconductor
connected to the same input rail. In some cases, inrush
field-effect transistor (MOSFET) to turn power on and
current during a short period of time can irreversibly
off can exceed the FET’s safe operating area (SOA) and
damage surrounding components or even the switch
damage the switch itself. Most power FET manufacturers
itself. It is possible to manage inrush current in several
include a graph listing how much current the switch can
ways, although all techniques result in a reduction of the
handle at a given drain-to-source voltage (VDS), ambient
speed at which the output voltage rises.
temperature and specified pulse duration. If a FET turns
The Dangers of Inrush Current on quickly, into a large load capacitance, the total energy
Inrush current may cause an upstream supply voltage passing through the FET is one-half capacitance-voltage-
to dip. If this voltage is supplying power to other squared (CV2) during the turn on-time (tON) duration. This
subsystems, those components may experience value can exceed the FET’s SOA curve and damage
misbehavior or a reset. the FET. If that same amount of energy is spread over a
In Figure 1, the switch is enabled, which causes a fast longer tON, then the FET has more time to dissipate that
transient voltage (dv/dt) and a large current spike on the heat and could be designed to survive.
input supply. This leads to a VSUPPLY voltage dip, which Common Techniques to Control Inrush Current
is visible while VSW_OUT continues to turn on. The voltage There are three traditional approaches and one newer
dip on VSUPPLY causes DC/DC converters to shut down; approach to controlling inrush current. The first is to use
VOUT decays until VSUPPLY recovers within the normal a passive element such as a series resistor or a negative
operating range. This scenario demonstrates how inrush temperature coefficient (NTC) thermistor to slowly charge
current on one load can cause other loads to experience the output. The second is to use a switch and manage
a reset.

Switch
VSW_EN
EN VSW_OUT VSUPPLY

Input VSUPPLY
supply
VSW_EN
VDCDC_OUT
DC / DC
VSW_OUT

ISUPPLY

VDCDC_OUT

Figure 1: Inrush current on VSW_OUT, causing a supply dip and reset on VDCDC_OUT.

11 Ways to Protect Your Power Path 14 Texas Instruments

Chapter 1: Inrush Current Control

the slew rate at which the output voltage is rising. The Slew-Rate Control
third is to use a current limiter or a device designed Controlling the rate at which the switch turns on directly
to provide a constant current until the switch is fully controls the rate at which the output voltage rises.
on. The newest approach is to use thermal regulation, Equation 1 calculates the inrush current as:
which provides maximum current while regulating the
temperature of the switch to prevent failure.
dVOUT (1)
IINRUSH = CLOAD ×
dt
Passive
Adding resistance in series with the supply limits current Slowing the switch speed (dVOUT/dt) reduces IINRUSH for a
to the load. A pure resistor would limit the peak inrush given, fixed CLOAD.
current to VIN/R, and then the current would begin
RC Time Constant
decreasing as the output becomes fully charged. This
For a discrete power switch, one common method to
technique will dissipate power through the resistor as
manage inrush current is to install a resistor-capacitor
well as limit the current that the load can draw once
(RC) that slows down the MOSFET switching speed, as
powered. But because of poor efficiency, this technique
shown in Figure 4.
is usually limited to low-current applications (often much
less than 1 A). Figure 2 shows a series resistor to limit
inrush current.
VIN
R
Load

EN1
VIN
Load

Figure 4: A discrete MOSFET with an RC rise time to limit

inrush current.
Figure 2: Series resistor to limit inrush current.

This is a simple method to control the switch speed

through an RC time constant. While the speed can be
If adding an NTC, the resistance will start off high when
reduced and the total turn-on time increased, note that
the switch is cool and power has not flowed through it.
the dv/dt speed is nonlinear; therefore, the inrush current
As power is applied, the high resistance will allow a small
is nonlinear. The switch will start to turn on slowly and will
amount of current through to the output. This will begin
exponentially increase as the switch becomes fully on.
to self-heat the device and cause the resistance to drop,
See Figure 5 for RC-based inrush current.
gradually allowing more and more current through to the
load until the NTC is fully on. Since NTC performance
relies heavily on ambient temperature, it may not be
suitable for applications needing a wide operating
temperature range. Figure 3 shows how to use a series
NTC thermistor to limit inrush current.

-to
VIN
Load

Figure 3: Using a series NTC thermistor to limit inrush current. Figure 5: RC-based inrush current.

11 Ways to Protect Your Power Path 15 Texas Instruments

Chapter 1: Inrush Current Control

Linear Soft Start or dV/dt IINRUSH_MAX = CLOAD ×

VOUT (2)
Many integrated power switches have a method of tON
controlling the output voltage rise time linearly. These
Constant Current/Current Limit Regulation
switches may come with a fixed or adjustable rise time.
Similar to a linear soft-start scheme, a constant current
In all cases, controlling the output voltage rise time
method of controlling inrush current would yield the
linearly means controlling a constant dVOUT/dt rate. In
same result if powering a purely capacitive load. If you
this case, if CLOAD is constant and dVOUT/dt is a constant
charge with a constant IINRUSH, then for a given CLOAD,
speed, then IINRUSH will also be a constant, as shown in
you will be charging at a constant dv/dt.
Figure 6.
The difference is the linear soft start was regulating
voltage, whereas a current limiter is regulating current.
The approaches will differ once you begin introducing
loads outside of a capacitor. If you have an integrated
circuit that activates before the output voltage is fully on,
then the linear soft-start method will have that current
superimposed. The constant current method will change
its slew rate, since some current is now going to the load
instead of to the output capacitor.
In Figure 7, linear soft start is maintaining a constant
slew rate for the output, whereas the constant current
Figure 6: Linear dv/dt to limit inrush current using the Texas limit will change its slew rate when the load turns on at
Instruments TPS22975 load switch. 2.5 V.
Constant current limit during startup can be helpful for
Linear soft start has a couple of advantages over an RC situations where loads may turn on while the switch
time constant approach, including simple and precise is still turning on. Constant-current-limit devices often
calculations for inrush current. There may be situations provide protection against events such as an output
where you have both a maximum inrush current short circuit. We will discuss current-limiting techniques
restriction and a maximum turn-on time requirement. In in more detail in the next chapter, but it is important to
these cases, if an RC time constant would either take too understand that it is one form of managing inrush current
long to turn on or have too high of an inrush current, then during startup.
a linear control scheme offers the best chance to satisfy
both requirements. If you can’t satisfy Equation 2, then Thermal Regulation
there is no solution other than changing the design In order to turn on as quick as possible but maintain
architecture to remove these restrictions, or moving to a protection for the switch, there is a final technique using
smaller CLOAD or a lower VOUT: thermal regulation. This can be useful in situations with

Inrush VIN VIN

limiter
VIN
Load
COUT VEN VEN

VOUT 2.5 V VOUT 2.5 V

Output load draws
2 A beginning at 2.5 V Current
2A limit
IC_LOAD IIN IIN 2A

ILOAD
Linear soft-start or dV/dt Constant current /
Current limit regulation

Figure 7: Linear dv/dt versus constant-current-limit regulation.

11 Ways to Protect Your Power Path 16 Texas Instruments

Chapter 1: Inrush Current Control

a large or unknown output capacitance (such as driving

an offboard load). Once the junction temperature rises to
a specified level, the device starts controlling the inrush
current profile in order to maintain a regulated junction
temperature. This will continue until the load capacitors
are fully charged, or until a timeout. To learn more
about a switch with a thermal regulation loop, see the
TPS2663 eFuse (or electronic fuse), also highlighted in
Figure 8.

Figure 8: Thermal regulation loop using the TPS2663 with 24-V VIN and
30 mF of output capacitance.

11 Ways to Protect Your Power Path 17 Texas Instruments

 Chapter 2 Adjustable Current Limit

Authors: Rakesh Panguloori and Lokesh Ghulyani affect the operation of system-critical loads connected to
the same power distribution bus, leading to an erroneous
Abstract
system function or system reset.
Current limiting is essential in many systems to restrict
To overcome such a scenario, you can use current-
transient overload currents to permissible levels.
limiting devices in each of the power paths to manage
Traditional current-limiting solutions such as discrete
overload and short-circuit events, and to guard sensitive
resettable fuse circuits are inferior in performance due
circuitry for reliable system operation.
to their inaccuracy, slower response, and a lack of
configurability and repeatability. This chapter discusses These parameters determine the performance of a
how integrated power switch solutions overcome the current-limiting device:
limitations of discrete current-limiting solutions. • Accuracy.
Integrated power switch solutions provide accurate and
• Adjustability.
adjustable current limit that helps limit overload currents
precisely, thereby reducing front-stage power budgets • Repeatability.
and the size of passive components, cables and printed
• Response time.
circuit board (PCB) traces. In addition, their accurate
current-limiting (or output power-limiting) function Traditional Ways to Implement Current Limiting
simplifies system design by complying with critical Fuses are traditionally considered protection devices,
safety standards. isolating overload or short-circuit faults from the main
The Need for Current Limiting system. While fuses are an inexpensive solution, they
can fall short of many of the protection requirements
The semiconductor components, passive filters, PCB
imposed on modern electronic equipment. The overload
traces and interfacing cables found inside electronic
current needs to be exceedingly higher (500%) than
equipment all have an inherent maximum current
the rated fuse current in order to produce a response
capability that when exceeded will cause them to
within milliseconds. This makes it extremely difficult to
overheat, possibly to an unrecoverable state. To
predict the precise overcurrent level at which the fuse will
overcome this, some kind of current-limiting device
open. A conservative fuse current rating selection may
restricts fault currents within the maximum current-rating
lead to a fuse blowup during inrush current events. In
limits and brings the system back to normal operation
addition, once the fuse blows during an overload event,
once the fault is cleared.
it has to be physically replaced, which increases system
In the example shown in Figure 1, a DC/DC converter downtime and maintenance costs. Figure 2 shows such
is supplying power to various internal subsystems, a scenario.
each requiring a specific voltage for proper operation.
An overload event in any one of these subsystems can

Step-down
µC
conversion

VIN
DC/DC
converter
Step-down Load Figure 2: A blown fuse needs to be physically replaced every time in
conversion = RL + CL order to resume normal operation.

Load A resettable fuse is a positive temperature coefficient

= RL + CL (PTC) device that increases its on-resistance with
temperature. During an overload event, the excessive
Figure 1: Power distribution in typical electronic equipment.
load current increases power loss, thereby increasing

11 Ways to Protect Your Power Path 18 Texas Instruments

Chapter 2: Adjustable Current Limit

its on-resistance. The higher on-resistance helps limit sense resistor Rs and P-channel N-channel P-channel
the overload current and protects the circuit. Unlike a transistor Q2 provide active current limiting.
physical fuse, PTCs allow current to flow after the fault is In steady-state operation, Q1 is fully turned on and load
cleared without replacing the device. current flows through the series sense resistor Rs and
Because PTCs are actuated by the heating effect of an discrete power switch Q1. The voltage drop across Rs
overcurrent load, their reaction time is limited to several generates bias voltage VBE for Q2, which is used to
milliseconds, as they have a natural dependency on the control Q1 in a closed-loop form. When the load current
ambient temperature, as shown in Figure 3. Another reaches high enough to bias Q2 on, Q2 reduces the
property of a resettable fuse is that its on-resistance gate-to-source voltage of Q1 to limit the load current to
increases after every reset, which raises concerns about VBE/Rs.
achieving repeatable performance over time. The response time of this discrete approach is much
better than a fuse or PTC, but still suffers from poor
accuracy. Current-limiting accuracy can be in the range
of 30% considering the variation of VBE voltage over
temperature. Another disadvantage is the power loss
in the sense resistor; for a 5-A application, dissipation
would be as high as 3.25 W (0.65 V × 5 A = 3.25 W).
It is possible to reduce the power loss by using either a
current-sense amplifier or a fast comparator to disable
pass switch Q1 in the event of overload, but this adds
complexity and cost. Discrete current-limiting circuits do
not incorporate thermal protection, so these solutions
need careful selection of the pass field-effect transistor
(FET) and prudent thermal design to keep the pass FET
Figure 3: Trip time vs. trip current temperature dependency. within safe operating area limits under extreme fault
conditions. All of these limitations often lead to a trade-
off between accuracy, cost, complexity, PCB size and
Discrete Current-Limiting Circuits power loss.
Figure 4 shows an example of a current-limit circuit eFuses
using discrete components. The resistor-capacitor An eFuse is an active current protection device with an
components around discrete power switch Q1 slow integrated FET used to limit currents to safe levels during
down the metal-oxide semiconductor field-effect fault conditions. The common elements of an eFuse are
transistor switching speed and provide inrush current a power switch to modulate the load current, a current-
control. In this implementation, the combination of series sensing element and control logic. As shown in Figure 5
on the folllowing page, the current-limit threshold can
Rs be set to a desired value through external resistor RLIM
at the ILIM pin of the eFuse, providing system flexibility in
VIN Q2 Q1
adjusting the current limit to a wide range.
In a steady-state condition, the integrated power
Load switch is fully turned on (just like a discrete switch) to
minimize voltage drop in the power path. If at any point
EN the load current reaches the set current-limit level, ILIM,
the gate control transitions the integrated power switch
into linear mode. This increases the on-resistance and
Figure 4: Active current-limiting circuit using discrete components. corresponding voltage drop across the eFuse to provide
a constant output current.

11 Ways to Protect Your Power Path 19 Texas Instruments

Chapter 2: Adjustable Current Limit

off the integrated power switch for junction temperatures

VIN
OVP VOUT above 150°C to ensure robust protection.

High-Side Switches
Smart power switches offer robust protection against
EN/UVLO Control overload and short-to-ground events through overcurrent
logic
protection and are available in Automotive Electronics
Council (AEC)-Q100 options. These devices provide a
dVdT
current-limiting function similar to eFuses but differ in
their ability to drive inductive offboard loads.
Current ILIM
limit As shown in Figure 7, smart power switches have
GND
options to use either an internal, fixed high current
limit (when the CL pin is ground) or an external lower
current-limit value (when the CL pin is tied to ground
through external resistor RCL). The adjustable low
Figure 5: Block diagram of an eFuse.
current limit greatly reduces the fault energy that the
internal power switch has to handle during a fault, in
Figure 6 shows the overload response of an eFuse with turn reducing stress on the power switch and improving
the current limit set at 4 A. As you can see, the eFuse system reliability.
responds quickly to an overload event (typically hundreds
of microseconds) and regulates the output current to ILIM.
Drain (VS)
Once the fault is removed, the eFuse recovers back to VIN V DS clamp

normal operation smoothly.

Diagnostics
An eFuse uses a current mirror circuit to measure the Diagnostics
and Gate logic
protection
path current without the need for an external sense
resistor (like Rs in Figure 4). This approach not only Current sense Current VOUT
saves board space but also eliminates the loss in sense limit/sense

resistors and provides better current-limiting accuracy

over discrete approaches—in the range of ±5% to ±8%.
CL
Another major advantage of eFuse devices is their
integrated overtemperature protection. The device shuts
Figure 7: Block diagram of a high-side switch.

VIN
The lower current limit also helps clamp the inrush
EN currents to a much lower value, as shown in Figure 8
on the following page. Lower fault currents save system-
VOUT level costs by minimizing PCB trace widths and cable
sizes, along with lower connector ratings and component
tolerances. For more information on the benefits of
adjustable current limiting with smart power switches,
see the application report, “Adjustable Current Limit of
I_IN
Smart High-Side Switches.”

Figure 6: Transient overload current limiting and recovery with the

TPS2595 eFuse.

11 Ways to Protect Your Power Path 20 Texas Instruments

Chapter 2: Adjustable Current Limit

Solution reference design provides an example of this

additional control loop.
There are safety and compliance standards such as
International Electrotechnical Commission (IEC) 61010-1
for industrial equipment power and energy limiting.
Complying with these standards means that precise
power limiting is also required.
Certain power supplies are inherently power-limited;
these power supplies come under National Electrical
Code (NEC) Class 2 or IEC 60950 standards. Achieving
Figure 8: Inrush current clamping with the TPS1H100-Q1 power limiting in these power supplies requires an
high-side switch. additional power-limiting circuit.

Power Limiting Implementing Power Limiting Using an eFuse

Power limiting is usually required to limit the power To precisely limit power, such as in industrial applications,
delivered to the load. For power supplies with a TI’s TPS2663 and TPS1663 eFuses offer an integrated
narrow voltage range, a current-limiting device would solution. These eFuses have an additional power-limiting
be sufficient to limit power. But for power supplies control loop, and the output power limit is configurable
with a wider voltage range, a power-limiting device is through the PLIM pin. The TPS2663 and TPS1663 can
required. Figure 9 illustrates a typical implementation limit output power with an accuracy of ±6% up to 150 W.
for limiting power. These devices also support load transients by enabling
current twice the programmed current limit (two times
Power or current the pulse current) to pass to the transient load.
limiting device
Figure 10 shows the power-limiting response of the
TPS2663. The output power limit is set to 100 W by the
+ Power
Load resistor on the PLIM pin; the device allows a pulse current
supply
– of 12 A and a pulse power of 282 W for a duration of
25 ms to support the load transients. For more details
Figure 9: Typical power limiting. on power limiting with TPS2663 devices, see the
data sheet.
The Need for Power Limiting
In most applications, it’s possible to achieve power
limiting with current limiting, but certain applications
need to precisely limit the energy or power delivered
into the load. Industrial systems such as programmable
logic controllers have a wide operating voltage range,
from 18 V to 36 V. In telecommunication systems, the
operating voltage range varies from 36 V to 72 V, thus
may require power limiting as well as current limiting.
Limiting current for a fixed output voltage or a smaller
variation in voltage can achieve power limiting. But
larger variations in voltage require an additional control
loop in order to limit power. This additional current
loop adjusts the current limit with voltage to maintain Figure 10: Power limiting with the TPS26633 (two times the pulse
constant power. TI’s Precision Power Limiting current support, ILIM = 6 A, VIN = 24 V).

11 Ways to Protect Your Power Path 21 Texas Instruments

Chapter 2: Adjustable Current Limit

U1

VIN OUT
IN OUT
R9
R1 R4 100k 91k
953k UVLO FLTb VIN
UVLO FLT_N
TCLOSE = 20 m

2
SHDNb ILIM U4
SHDN_N ILIM
U3 TLV170
V1 = 18 V2 OVP

VCC
C1 OVP RTN
V2 = 32 R7

1
1µ PAD D1 IN+
TD = 25 m 36.5k dV/dT R8 36.5k
DVDT
TR = 15 m GND OUT
TF = 15 m
PW = 25 m TPS26620_TRANS IN–

VEE
PER = 100 m C2
R3 {Cdvdt] PARAMETERS: R5 RL
34k Cdvdt = 22 n 12k C3 30
1µ R10
9.1k

Figure 11: Power limiting with the TPS2662, with an additional control loop.

Implementing Power Limiting with an Conclusion

External Control Loop Current limiting is a basic protection requirement in
It is possible to use current limiting to achieve power most modern electronic equipment. Although discrete
limiting for power supplies with a narrow input-voltage components work, they require more board space, are
range. Achieving power limiting in wider-voltage-range less efficient and can generally be less cost effective.
power supplies requires an additional control loop for TI integrated solutions, eFuses and high-side switches,
precise power limiting that adjusts the current limit based provide accurate current limiting, faster response times
on the output voltage. and can repair themselves without user intervention.
Figure 11 shows an implementation of the TPS2662 for
power limiting. The additional control loop uses a TLV170
amplifier to inject current proportional to the output
voltage into the ILIM pin for power limiting. Figure 12
provides the PSpice simulation results for the circuit
shown in Figure 11. For this simulation, the input voltage
varies from 18 V to 32 V. The TLV170 adjusts its current
limit and the output power is limited to less than 10 W.

Figure 12: Simulation results for power limiting with the TPS2662.

11 Ways to Protect Your Power Path 22 Texas Instruments

 Chapter 3 Short-Circuit Protection

Author: Alec Forbes The risk of a short circuit becomes higher in systems that
contain some or all of these characteristics:
Abstract
• Output cables. Over time, cabling can wear down
Short-circuit protection is critical to ensuring robust
and present exposed metal, which can make contact
reliability for output power protection, as short-circuit
with connectors.
failures can cause supply failures or even damage to
systems. The most common implementation of short- • Debris-filled environments. A common cause
circuit protection is through the use of a current-limited of short circuits is metallic debris shorting pins
output; however, it is important to understand the impact or traces.
of current limit on short-circuit reliability.
• Maintenance requirements. Any possibility
Short-circuit behavior and protection requirements vary of exposure to metal tools during maintenance
based on the output impedance, ranging from terminal increases the risk of an accidental short-circuit event.
shorts to a partial resistive short. Depending on the
short-circuit impedance and load profile, the current and • Strong voltage supplies. The more energy that a
energy profile will take different forms. power supply can provide, the higher the risk of a
short-circuit event causing a health hazard.
What is a Short Circuit?
Automotive systems, for example, regulate output short-
In personal electronics, industrial, automotive and
circuit protection through the Automotive Electronics
communication applications, short-circuit protection is
Council (AEC)-Q100-012 standard, which mandates
a safety requirement for all power distribution outputs.
that systems should be capable of surviving thousands
Short circuits occur whenever a charged conductor
of repetitive short-circuit events. The application report,
makes physical contact to a grounded element in
“AEC-Q100-012 Short-Circuit Reliability Test Results
the system. During a short circuit, the electrical short
for Smart High-Side Switches,” documents TI’s test
between power supply and ground can create a very low
process for high-side switches.
impedance path, causing a high, uncontrolled sustained
current flow. This type of flow is extremely dangerous for Protection Against Short Circuits
systems and can cause integrated circuit (IC) failure, as The primary danger during a short-circuit event is the
seen in Figure 1. Thus, any systems with a risk of short- sustained high current flow caused by low output
circuit events must implement robust protection. impedance. So in order to protect against short circuits,
the system must have output power protection that can
recognize a short-circuit event and then shut the system
off quickly and safely.
The best way to implement this type of protection is to
use a current-limiting switch on the output. This switch
will provide protection against a sustained uncontrolled
short-circuit current and shut off, protecting the output
circuitry. Because the output impedance is unknown
and often has an inductive element, the switch must be
able to safely demagnetize the inductance, and thus
needs a flyback mechanism. Figure 2 on the following
page shows the typical setup that has potential for
Figure 1: Example of IC failure. short circuit.
There are three primary short-circuit conditions that a
system must be able to withstand: terminal short circuits,
cable short circuits and resistive short circuits.

11 Ways to Protect Your Power Path 23 Texas Instruments

Chapter 3: Short-Circuit Protection

Cable To understand the short circuit, take a look at the total

13.5-V supply cumulative fault energy, which occurs in two portions:
• As the short-circuit current ramps up, the current
causes resistive power losses in the field-effect
Cable transistor (FET); however, they are low, as the FET is
EN
Gate VOUT in saturation mode and the resistance is very small.
driver

• After the switch hits its current limit and turns off,
CL Current
Load
small trace inductive elements in the short circuit
limit
create a negative voltage spike that causes the
voltage clamp in the switch to engage and the drain-
to-source voltage (VDS) to rise to 40 V. While the
Figure 2: Typical setup with potential for short circuit. inductance demagnetizes, this high VDS requires the
dissipation of more fault energy.
Terminal Short Circuit Because of the short length of the pulse, the total fault
The simplest type of short circuit is the terminal short energy is kept to a relatively safe 18 mJ, despite the
circuit, illustrated in Figure 3 with a TI high-side switch high current limit. This case is the easiest short circuit
providing an output protection stage. The output current to survive, as there are no series impedances that will
increases quickly until it hits a 90-A current limit, at which impact the output protection.
point the switch recognizes a short-circuit event and
Cable Short Circuit
turns the switch off, preventing dangerous current flow.
In the event of a short circuit through a cable, accounting
Because this is a terminal short circuit, there is minimal for the cable series impedance increases the difficulty of
series impedance, and the current-limiting output is protection. The waveform in Figure 4 on the following
easily able to protect the system from the short circuit. page shows a TI high-side switch short-circuited;
The rise and fall times of the short-circuit current are very however, this time the short circuit occurs at the load
quick, and the entire event is over in less than 50 µs. side of a 5-µH cable (a standard value for a 5-m cable).

Cable
13.5-V supply

Cable
EN
Gate VOUT
driver
Sh
ort
tog

CL
rou

Current
Load
n

limit
d

Figure 3: Terminal short on output.

11 Ways to Protect Your Power Path 24 Texas Instruments

Chapter 3: Short-Circuit Protection

Cable
13.5-V supply

Cable
EN
Gate VOUT
driver

nd
to grou
CL Current
Load
limit

Short
Figure 4: Cable short circuit across the load.

The extra series inductance significantly slows down the Most high-side switches list the inductive load energy
current rise and fall times, so the pulse takes 10 times peak, so make sure that the calculated energy from the
longer and the total fault energy is 75 mJ—four times maximum cable length and maximum current is below
higher than the terminal short circuit. Energy dissipation the rated energy. Figure 5 below shows these values for
occurs in two portions similar to the terminal short-circuit a TI high-side switch.
case—but now that the output inductance is 5 µH
Resistive Short Circuit
rather than just parasitics (typically less than 10 nH),
A resistive short circuit occurs when debris or corrosion
the inductor demagnetization contains a much higher
causes a partial short circuit or a load failure. In the short
percentage of the cumulative fault energy.
circuits described above, a low series resistance causes
The energy dissipated in the fault can be very large due the current to increase until it hits the current limit. When
to the high current, so the inductive clamp must be there is some series resistance with a high current limit,
rated high enough to dissipate this energy. Equation 1 there is a risk that the current will saturate before hitting
calculates the energy dissipated while discharging an the current limit. In the event of a series resistance, the
inductive element through a series clamp: output protection must have a low-enough current limit
1 VIN to register the short circuit as a fault event.
E=
2
× L × IPK² × (
VCLAMP–VIN ) (1)
Figure 6 on the following page shows a case where the
series resistance limits the short-circuit current to below
The high-side switch parameter, VCLAMP, is device- the output protection current limiter, so the device does
dependent, but it’s usually 40 V to 60 V. Therefore, with not turn off the output and current flows indefinitely.
even a small 5-µH cable inductance, a high current
During this time, there is a high, uncontrolled current flow
limit can create a large demagnetization energy that the
until the system hits thermal shutdown or a secondary
output protection stage must absorb. The input voltage
method of current monitoring turns the system off.
can also increase during demagnetization because of the
For reliability against resistive short circuits, you must
input inductance, increasing the total fault energy that
either design the overall system to manage the transient
must be dissipated.

Energy dissipation during turnoff, ETOFF Single pulse, LOUT = 5 mH, TJ,start = 125°C 200 mJ
Energy dissipation during turnoff, ETOFF Repetitive pulse, LOUT = 5 mH, TJ,start = 125°C 80 mJ
Figure 5: Inductive energy dissipation of a TI high-side switch.

11 Ways to Protect Your Power Path 25 Texas Instruments

Chapter 3: Short-Circuit Protection

Cable
13.5-V supply

Cable
EN
Gate VOUT
driver

d
un
gro
CL

to
Current
Load

ort
limit

Sh

Figure 6: Resistive short circuit.

current pulse or ensure that it will shut off by minimizing Protection Considerations
the output protection current limit. To protect against short circuits, it’s best practice to use
The “Improved Automotive Short Circuit Reliability an output protection stage with an integrated current
Through Adjustable Current Limiting” application note limit and inductive clamp. Typically, this protection
describes this event in greater detail. is implemented through a high-side switch with an
integrated output clamp, or through an eFuse or load
System Interruption
switch IC with an external flyback diode.
The previous sections covered how to ensure that the
When designing short-circuit protection, consider:
system will survive a short-circuit event. But even if the
system does survive without damage, a short-circuit • Minimizing the current-limiting value. Lowering the
event should cause as minimal of an interruption to the current limit minimizes the energy that must be
system as possible. dissipated in a short circuit. Because this energy is
a function of the square of current, minimizing the
During a short-circuit event, high current can overload
peak current significantly decreases the risk of failure.
the input supply and create an undervoltage lockout
Keep the current limit as close to the maximum
event for the entire system. To prevent this, it’s
working current as possible.
important that the short circuit is registered as quickly
as possible so to facilitate the implementation of short- • Confirming that the output protection can dissipate
circuit protection with solutions that have a very fast the required fault energy. Calculate the maximum
response time. short-circuit energy expected based on cable
To minimize the response time, many protection devices lengths and current limits, and check that the output
implement a fast-trip response that triggers within protection stage will be able to safely dissipate
nanoseconds of an overcurrent event. If the fast-trip that energy.
threshold is hit, the device instantly turns off to protect • Limiting input and output inductances. Inductances
the supply, and then slowly ramps up the output current in series with the short circuit significantly increase
safely to the chosen current limit. This keeps overshoot the demagnetization time and total fault energy that
during the short circuit from having an adverse effect on must be dissipated.
the supply.
• Guaranteeing the registration of a resistive short
circuit. In the event of a resistive short circuit, ensure
that either the current limit is low enough to recognize
a partial fault or that there is an alternative method

11 Ways to Protect Your Power Path 26 Texas Instruments

Chapter 3: Short-Circuit Protection

of overcurrent detection (like a microcontroller) overshoot. Understand how the response time will
continuously monitoring the load current. impact the maximum possible current.

• Understanding protection response times. In the By designing an output protection stage around these
event that a short circuit causes an overcurrent, principles, you can safely protect an output from a short
there is a response time before the ILIM circuitry is circuit and minimize the risk of damage.
able to limit the current, during which there is some

11 Ways to Protect Your Power Path 27 Texas Instruments

 Chapter 4 Reverse Current Blocking

Author: Alek Kaknevicius VREV

– +
Abstract
– +
Reverse current occurs when current travels from output
VREV VIN
to input rather than from input to output. This chapter IREVERSE System IREVERSE System

+ –
explains where reverse current comes from, why it can
be harmful to systems and how to modify a design to (a) (b)
protect against reverse current. Figure 2: Reverse polarity (a); reverse voltage (b).
What is Reverse Current?
Reverse current is when there is a higher voltage at the to input, reverse current spikes can also damage
output of a system than at the input, causing current cables and connectors. If you are using a MOSFET
to flow backwards through the system. There are two for load-switching applications, reverse current can
common sources of reverse voltage: the first is when travel backwards through its body diode if it becomes
power is disconnected from a system and the input forward-biased, because the FET output voltage is
voltage suddenly drops. During this time, a higher greater than the input voltage. This causes a linear rise in
voltage can be left at the output, which temporarily power dissipation across the body diode, expressed by
produces reverse current. See Figure 1. Equation 1:
P = IV: PD = IREV × VDROP (1)
V+ V+
VIN where PD is the power dissipated across the body diode,
0V IREV is reverse current through the device and VDROP is the
0V
IREVERSE CLOAD RLOAD voltage drop across the body diode.
If the heat generated by the power dissipation through
Figure 1: Reverse current caused by a sudden loss of power. the device exceeds the thermal rating of the device, then
combustion can occur. Therefore, it is essential to limit
reverse current flow, or reverse voltage.
The other cause of reverse current is when a metal-oxide
When Do You Need to Block Reverse Current?
semiconductor field-effect transistor (MOSFET) is used
There are several applications where blocking reverse
for a load-switching application and the body diode
current is necessary.
becomes forward-biased.
The output of a switch with a higher voltage than the Power Multiplexing
input causes a reverse voltage, which is what causes Power multiplexing is the practice of using a switching
reverse current. Don’t confuse this with negative voltage, circuit to choose one of multiple power supplies for
also known as reverse polarity. A negative voltage occurs a system, with the ability to switch between them.
in situations where the positive and negative terminals of Figure 3 on the following page shows this configuration.
a power supply are switched. In this case, what should If one of the power-supply voltages is higher than the
be connected to ground actually has a voltage from the other, then it’s possible for reverse current to occur
positive input to the system, as shown in Figure 2. This even when the other power rail has an “open” switch—
causes a different current phenomenon through the one example is when using FETs to switch the power
device than reverse polarity. For more information on supplies. The higher voltage at the output of the open
reverse polarity protection, see Chapter 5. FET causes reverse current to flow from the higher-
Why Do You Need to Block Reverse Current? voltage power supply, through the FET body diode and
into the lower-voltage power supply. Figure 4 on the
Reverse current can damage internal circuitry and
following page shows the application of 5 V to a system
power supplies. Depending on the path from output
while the switch for the 3.3-V rail is open.

11 Ways to Protect Your Power Path 28 Texas Instruments

Chapter 4: Reverse Current Blocking

System System
Power V1 V1 Power V1 V2
supply 1 supply 1

Power Power
supply 2 V2 supply 2 V2

Power supply 1 powers the system Power supply 2 powers the system

Figure 3: Power multiplexing configuration.

You can see in Figure 4 that with a simple FET solution reverse current will flow across the switch. To avoid this,
for the switch, reverse current is able to flow through the you should have a switch with reverse current blocking,
FET body diode, even when the switch is open. or a larger input than output capacitance.
In some systems, a supercapacitor holds up the output
System
5V when the input supply drops. This is sometimes referred
Power
supply 1 to as a “last gasp” circuit, which allows the system to
safely power down. It’s also a good idea to have reverse
current protection here in order to protect any upstream
IREVERSE supplies or components.
Power
supply 2 3.3 V How Do You Block Reverse Current?
There are several ways to block reverse current.
Current Diodes
Diodes are great for high-voltage, low-current
– + applications. However, diodes cause a forward-voltage
Voltage drop that increases the total power dissipation in the
system and drops the power supply by 0.6 V to 0.8 V.
Figure 4: Reverse current caused by power multiplexing.
This can cause decreased efficiency in the system and
a shortened battery life. A popular alternative is the use
ORing of a Schottky diode. They have lower forward-voltage
ORing is similar to power multiplexing, except that drops, but are more expensive and have higher reverse
rather than selecting one supply to power the system, current leakage, which could cause problems for the
the highest voltage always powers the system. Reverse system such as supply or battery damage.
current blocking is also necessary here, because each
Back-to-Back MOSFETs
ORing switch sees reverse current when the other switch
Using back-to-back MOSFETs is a powerful option,
is closed.
since it offers current blocking in both directions when
Sudden Loss of Input Power the MOSFETs are turned off. In comparison to the diode
When a closed switch suddenly loses power on its solution, there is a lower voltage drop from the power
input, there is the potential for reverse current. If the supply to the load. However, this implementation takes
capacitance on the output of the switch is larger than up a larger amount of space on the board, requiring
the input, then the voltage on the output will decay several components to build, as shown in Figure 5 on
more slowly. This means that while power is decaying, the following page.
the voltage on the output of the switch will fall more Aside from the large solution size and bill-of-materials
slowly than the input. During this period, the voltage on count, this solution also doubles the on-resistance
the output of the switch will be larger than the input, so of a single MOSFET solution and does not provide

11 Ways to Protect Your Power Path 29 Texas Instruments

Chapter 4: Reverse Current Blocking

P-channel or N-channel. This is usually not possible

Power
supply Load for a discrete MOSFET, since the body terminal is not
typically accessible.

Power Switches that Block Reverse Current

ON
Several types of power switches are capable of blocking
OFF reverse current.

Figure 5: Reverse current blocking with back-to-back Load Switches

MOSFETs disabled. Load switches integrate reverse current blocking by
switching the body terminal of the internal MOSFET. This
any reverse current blocking when the MOSFETs are allows the device to have a low on-resistance while still
turned on. providing protection to the supply. If the load switch is
off, then reverse current is always blocked.
Backwards MOSFET
Some devices offer always-on reverse current blocking,
If a MOSFET is positioned so that the body diode is
which will turn off the device if the output voltage is
facing from input to output, then there will be no reverse
greater than the input voltage by a certain voltage
current flow when the MOSFET is switched off (open)
threshold. The disadvantage with an always-on reverse
(see Figure 6).
current blocking scheme is that the output voltage needs
to be greater than the input voltage before the device
Power turns off, so some amount of reverse current will flow
Load
supply through the device before this occurs. Figure 7 shows
an example of the expected reverse current through a
load switch before reverse current blocking is enabled.
ON

OFF

Figure 6: Reverse current blocking using a single MOSFET disabled.

The disadvantage with this solution is that it is not

possible to switch the supply off, since there is always
a diode path between the source and the load. This is
a preferable status for diode ORing but not so much for
power multiplexing because you may need to prioritize a
lower voltage supply, such as when the main supply for
a system is a lower voltage but the backup battery is a Figure 7: Full-time reverse current protection (VIN = 3.0 V, VOUT ramp up
higher voltage. from 3.0 V to 3.3 V).

With an ORing circuit, the battery will power the system

until the voltage has been drained to below the main eFuses
supply voltage. A power multiplexing circuit will prevent With eFuses, a back-to-back MOSFET implementation
the battery from being used until the system decides to is most common due to the MOSFETs’ higher voltage
switch over to the backup battery. capability. In some devices, both MOSFETs are
Switching the MOSFET Body Terminal integrated into the device, but in others, there is a pin
To block reverse current with a single MOSFET, for driving an external second FET to create a back-to-
the body terminal must either be biased to the back configuration. Similar to load switches, there are
highest supply or ground, depending on if the FET is some eFuses that come with always-on reverse current

11 Ways to Protect Your Power Path 30 Texas Instruments

Chapter 4: Reverse Current Blocking

blocking by monitoring the difference between VOUT and Conclusion

VIN, while others only provide reverse current blocking Applications such as power multiplexing or a loss of
when turned off. power can lead to a reverse voltage event that will
Power Multiplexers induce reverse current. This can cause damage to both
power supply and system. A TI power switch is both a
Integrated power multiplexers safely handle power
size- and cost-effective solution helping prevent damage
multiplexing applications so that reverse current does
from reverse current. The TI power-switch portfolio has a
not flow from one supply into the other. Current power
variety of devices with different specifications for reverse
multiplexers adopt both the back-to-back and body
current protection, making them a good fit for a wide
terminal configurations, and all power multiplexers have
variety of applications.
reverse current blocking.

Ideal Diodes
Ideal diodes are controllers that drive an external
MOSFET with the body diode facing from supply to
load. These devices will automatically detect the input
and output voltage difference to always enable reverse
current blocking while maintaining a low on-resistance.

11 Ways to Protect Your Power Path 31 Texas Instruments

 Chapter 5 Reverse Polarity Protection

Author: Karikalan Selvaraj subsystems. A reverse-connected battery can damage

polarized components such as electrolytic capacitors, as
Abstract
shown in Figure 2.
Many front-end power system designs require protection
from damage caused by a reverse-connected input
power supply, a reversed battery connection or miswiring
of field power-supply lines. Electrolytic
capacitor at input Load
Schottky diodes are the traditional choice for protection of DC-DC converter
against reverse polarity conditions, but the associated Reverse current
power loss from the forward conduction requires careful
thermal management, leading to increased system cost Figure 2: Reverse-connected battery: polarized components.
and space. Plus, increased demand for higher power
density necessitates better and more efficient reverse
Common Reverse Polarity Protection Methods
polarity protection methods.
Popular methods to protect against reverse input polarity
Reverse Polarity Protection Overview employ either a Schottky diode or a P-channel metal-
Power system modules or subsystems powered from a oxide semiconductor field-effect transistor (MOSFET)
field power supply or battery power require protection with discrete components.
from reverse polarity connections caused by miswiring
Reverse Polarity Protection Using Schottky Diodes
during system maintenance or reinstallation. Most
The simplest method of reverse battery protection is to
front-end power systems need protection from dynamic
add a series diode at input of the system power path.
reverse polarity conditions that can occur during a surge
Figure 3 shows a reverse battery protection scheme
event or an inductive load disconnect from the battery.
using a Schottky diode. When the battery is installed
Reverse-connected supplies can cause damage to
correctly, load current flows in the forward direction of the
connected subsystems, circuits and components.
diode. If the battery is installed with the wrong polarity,
Many automotive battery-powered subsystems require the diode is reverse-biased and blocks reverse current,
protection from reverse battery connections during thereby protecting the load from a negative voltage.
a repair or jump-start from another vehicle, and from
transient reverse input conditions that can occur during a Blocks reverse current
vehicle’s operating lifetime. VBAT VOUT
Reverse Input Polarity Protection
Figure 1 shows battery input lines connected in reverse.
When this occurs, huge current flows through the
Load
electrostatic discharge (ESD) diode of microcontrollers
(MCUs), DC/DC converters or other integrated circuits,
and causes severe damage to battery-connected
Figure 3: Reverse input protection using a Schottky diode.

ESD diode of MCU Bulk capacitance placed at the output holds the output
Load
or DC-DC converter from falling immediately and can supply the load for a
Reverse current short time before the input supply recovers.

Figure 1: Reverse-connected battery: MCU or DC/DC converter.

11 Ways to Protect Your Power Path 32 Texas Instruments

Chapter 5: Reverse Polarity Protection

Drawbacks of using a Schottky diode for reverse normal operation, the body diode of the MOSFET is
battery protection include: forward-biased and conducts until the MOSFET turns
• Power dissipation: A forward conduction loss on. The MOSFET turns on quickly as the battery input
due to a forward voltage drop results in significant charges the gate through the current-limiting resistors.
efficiency loss at higher load currents. The MOSFET turns off during static reverse battery or
dynamic reverse battery conditions after the battery input
• Thermal management: Managing the power starts to swing negative, as the gate-source voltage
dissipation requires a heat sink, increasing total cost starts to go below MOSFET threshold voltage (Vth) and
and space. swings negative.
• Higher forward voltage drop: In systems with lower Reverse Polarity Protection vs.
voltage bus (3.3 V), a typical voltage drop of 0.4 V Reverse Current Blocking
in diode will result in lower voltage headroom for the
Reverse polarity protection, also known as reverse
downstream loads to operate.
hookup protection, prevents damage to a load from a
• Reverse leakage current: Reverse leakage negative voltage at the input during a reverse-connected
current of high-voltage Schottky diodes increases battery or an inductive load disconnect from a battery.
dramatically with junction temperatures, resulting in Reverse polarity protection does not necessarily block
higher power dissipation during reverse conduction. reverse current flowing into the battery from the load or
downstream DC/DC converters.
Reverse Polarity Protection Using
Discrete MOSFETs In many systems, large holdup capacitors provide
sufficient backup power during a short interruption
You can reduce the forward drop of the diode by
of the supply line or a shorted battery input so that
replacing the Schottky diode with a P-channel MOSFET
the subsystem can function uninterrupted or perform
and orienting its body diode in the same direction as
maintenance housekeeping tasks such as memory
the Schottky diode (shown in Figure 4). During normal
dump before turning off. Reverse current blocking
battery operation, the body diode of the MOSFET is
prevents reverse current from flowing back into the
forward-biased and conducts for a very short time until
battery from the load, and allows holdup capacitors
the MOSFET turns on when the gate voltage pulls below
to provide additional backup time for the subsystem
the source. When the battery polarity is reversed, the
to function during various dynamic reverse battery
gate-source voltage swings positive and the MOSFET
conditions or short interruptions.
turns off, protecting the downstream circuits from a
negative voltage. The reverse input protection shown in Figure 4 or
Figure 5 below does not block reverse current from
VBAT VOUT flowing back into the battery, since the MOSFETs are
on. In Figure 6 on the following page, the input power
goes from a positive to negative value. The P-channel
MOSFET remains on for a short duration of time, since
Load

VBAT VOUT

Figure 4: Reverse input protection using a P-channel MOSFET. Load

An alternate method of reverse battery protection is

using an N-channel MOSFET on the low side, such as
on the ground return path. The operating principle is
similar to the P-channel MOSFET in Figure 4. During Figure 5: Reverse input protection using an N-channel MOSFET.

11 Ways to Protect Your Power Path 33 Texas Instruments

Chapter 5: Reverse Polarity Protection

its VGS is still transitioning from a negative to positive This enables them to provide more efficient reverse
value. This results in a temporary negative voltage spike polarity protection than a Schottky diode or discrete
on the output. In contrast, the Schottky diode constantly MOSFET solution.
blocks reverse current from flowing into the supply and
Reverse Protection Using Ideal Diode Controllers
inherently provides both reverse polarity protection and
An ideal diode controller drives an external N-channel
reverse current blocking.
MOSFET to work as an ideal diode with a very
low forward voltage drop and no reverse current.
Figure 7 shows reverse polarity protection using TI’s
LM74700-Q1 controller.

V BATT V OUT

Voltage
Figure 6: Schottky diode vs. a P-channel MOSFET. TVS Regulator
GATE CATHODE
ANODE

VCAP LM74700
Drawbacks of using discrete MOSFETs include:
• Lack of reverse current blocking. MOSFETS are ON OFF
EN GND

not suitable for applications with a large holdup

capacitance. Reverse current blocking is possible Figure 7: Reverse polarity protection using the LM74700-Q1.
with an external comparator, but will increase cost
and space.

• Power dissipation at low input voltages. The The power MOSFET is connected in such way that its
P-channel MOSFET’s drain-to-source on-resistance body diode blocks reverse current when the MOSFET
(RDS(on)) increases drastically as the input voltage turns off. The forward voltage drop and power
reduces because the maximum gate-to-source dissipation are reduced significantly as the MOSFET
voltage (VGS) is limited by the input voltage range. turns on during forward conduction. Ideal diode
Power dissipation due to increased RDS(on) reduces controllers sense the reverse current flowing through the
efficiency in applications with a wide input voltage MOSFET and turn it off very quickly, allowing the body
range. For example, a 12-V automotive battery diode to block reverse current.
voltage can vary from 3.5 V during cold-crank An ideal diode controller offers a low regulated forward
conditions to 28 V during a jump-start scenario. voltage, true reverse current blocking, fast reverse
• P-channel MOSFET size and cost. For higher- current response, and very low shutdown current and
power loads, the size and cost of a P-channel operating quiescent current.
MOSFET can increase considerably. These features enable the ideal diode controller to
emulate an ideal diode in reverse polarity protection
• Low-side N-channel MOSFET. All systems
applications. Figure 8 on the following page shows the
may not be able to tolerate a jump in the system
response of the LM74700-Q1 to a reverse-connected
ground voltage during turn on and turn off or load
input supply. The LM74700-Q1 blocks reverse current
current transients.
completely by turning the MOSFET off when the input
Reverse Polarity Protection Using supply lines are connected in reverse polarity. The output
TI Power Switches does not dip negative, but instead remains at 0 V. The
TI’s ideal diode controllers and eFuse devices have load is isolated from the reverse input supply connection
integrated reverse polarity protection, reverse current and protected from damage caused by reverse current.
blocking capability and low forward conduction loss.

11 Ways to Protect Your Power Path 34 Texas Instruments

Chapter 5: Reverse Polarity Protection

output voltage negative. The LM74700-Q1 detects the

reverse current and turns the MOSFET off within 0.75 μs
to block reverse current and prevent the output from
going negative. It also prevents discharge of the bulk
Output remains protected holdup capacitors.
from negative voltage
Reverse Polarity Protection Using eFuses
The TPS2660 family of eFuse devices feature fully
integrated reverse input supply protection without the
need for additional components. These devices can
withstand a -60-V reverse voltage without damage.
Figure 10 compares the reverse input polarity protection
functionality of the TPS2660 eFuse against an external
Schottky diode. An integrated MOSFET replaces an
Figure 8: The LM74700-Q1’s response to reverse input polarity. external blocking diode, providing a fully integrated
reverse current blocking and reverse polarity solution.
Dynamic Reverse Polarity
Automotive subsystems powered through a car battery Input Output Input Output

are subjected to various electrical transients on their TPS266x

eFuse
power-supply lines and are expected to function without Hot-swap controller

interruption. One such transient condition is the dynamic GND

reverse polarity specified in International Organization for GND

Standardization (ISO) 7637-2 pulse 1, where a negative Figure 10: Integrated reverse polarity protection using the
TPS2660 eFuse.
transient voltage as low as –150 V is applied to the 12-V
battery supply line with 10-Ω generator impedance for
2 ms. Figure 11 shows the response of the TPS2660
Figure 9 shows the response of the LM74700-Q1 to eFuse to a reverse input supply (miswiring) of –57 V.
dynamic reverse polarity applied at its input. Before The TPS2660 blocks reverse current completely by
the test pulse is applied, the MOSFET is on and allows turning the MOSFET off in the event of a reverse input
the load current to pass through. When applying the supply and prevents damage to downstream circuits
ISO 7637-2 test pulse 1 at the battery input, the load and components.
current starts to reverse quickly and tries to pull the

VOUT

VGATE
Gate turns off quickly within 1 µs
VIN

TVS clamping at –42 V

IIN Figure 11: TPS2660 eFuse response to a reverse input supply.

Figure 9: The LM74700-Q1’s response to ISO 7637-2 pulse 1.

11 Ways to Protect Your Power Path 35 Texas Instruments

Chapter 5: Reverse Polarity Protection

Dynamic Reverse Polarity The benefits of using ideal diode controllers and eFuse
Industrial subsystems powered through an external devices for reverse polarity protection include:
field power supply are subjected to various electrical • Improved efficiency. A low forward conduction loss
transients on their power-supply lines and are expected leads to efficiency improvements.
to function without interruption. Positive and negative
• Integrated reverse polarity and reverse current
surges specified by International Electrotechnical
blocking.
Commission (IEC) 61000-4-5 are applied at the input,
where a negative transient voltage as low as –500 V is • Reduced system cost and space. Thermal
applied for 20 μs. management is simpler compared to a
Figure 12 shows the response of the TPS2660 eFuse Schottky diode.
to a –500-V negative surge on the input. The output Conclusion
remains protected from the negative surge and the
Using an N-Channel MOSFET-based ideal diode
output turns on after the surge event is over.
controller and fully integrated eFuse solutions provides
robust reverse polarity and reverse current blocking
with high power dissipation, thermal performance
and space savings against discrete diodes and
FET implementations.

Figure 12: TPS2660 eFuse response to –500-V surge at input.

11 Ways to Protect Your Power Path 36 Texas Instruments

 Chapter 6 Overvoltage Protection

Authors: Matthew Xiong and Praveen Durga Gorrela surge pulses. These systems require special protection
requirements, since both the intensity and duration of
Abstract
lightning transients are more severe.
Every electronic component is rated for a particular
Continuous overvoltages are overvoltage conditions
voltage level. The presence of any voltage higher than
that remain present for long periods and stress systems
the rated maximum voltage can cause damage to a
indefinitely. Continuous overvoltages can be caused
component or degrade its performance, which could
by the failure or miswiring of upstream power supplies,
lead to system failures. Therefore, it’s important to
voltage regulators, converters or the insertion of
prevent components from exposure to these higher
noncompliant adapters into a system.
voltages by adding overvoltage protection, which will
isolate any overvoltage condition from downstream Common Overvoltage Protection Methods
devices or clamp the output to a safe level. There are several different overvoltage protection
Common Sources of Overvoltages topologies for the two types of overvoltages.
Overvoltages applied to electronic components can Protection Against Transient Overvoltage Conditions
exceed avalanche breakdown voltage of internal device Since transient overvoltage conditions have short
structures, resulting in excessive current flow. This durations, the response time of any solution must be
excessive current flow can severely weaken, damage or fast enough to minimize the exposure of downstream
overheat the entire device. circuitry. ESD diodes, transient voltage suppressors
Increased demand for lower power consumption (TVSs) and Zener diodes can steer transient overcurrents
and cost drove technological improvements, such to the ground plane and clamp transient overvoltages
that many semiconductor components now operate so that downstream loads remain within the safe
at lower voltage ranges. Lowering the voltage of operating area.
individual components decreases overall system power Metal-oxide varistors (MOVs) can also clamp transient
consumption, which increases battery life, improves overvoltages, and absorb significantly more energy than
efficiency and indirectly lowers cost (since the system ESD diodes, TVSs and Zener diodes, but with a penalty
uses less energy). Cost is also directly lowered, since of slower response time. While TVSs can respond in
lower-voltage components are cheaper to manufacture nanoseconds, MOVs will respond in microseconds. As
than their higher-voltage counterparts. However, with a result, ESD diodes and TVSs are used for low-voltage
improved power efficiency and lower costs comes circuitry, while MOVs offer better protection for AC mains
a side effect—increased component sensitivity to or high-voltage DC stages.
overvoltage conditions.
For more information regarding ESD scenarios, causes
There are two types of overvoltages: transient and solutions, see the TI training, “ESD Essentials.”
overvoltages and continuous overvoltages. Transient
overvoltages are commonly caused by electrostatic Protection Against Continuous Overvoltage Conditions
discharge (ESD), electrical fast transients, voltage ringing While transient overvoltage protection solutions can
from hot-plug events, or inductive switching surges respond quickly to transient overvoltages, they are
from nearby power supplies or converters. Transient ill-suited for continuous overvoltage conditions, which
overvoltages can approach kilovolts but typically have characteristically last longer. After all, MOVs, ESD
short durations; for example, ESD events may last from diodes, TVSs and Zeners can only clamp and dissipate
60–100 ns and surge events can last about 20 µs. a certain amount of energy before failing. To protect
Telecommunications equipment may be exposed to against steady-state or continuous overvoltages, there
lightning surges—another form of transient overvoltage are two common methods: overvoltage lockout and
that can be orders of magnitude higher than regular overvoltage clamping.

11 Ways to Protect Your Power Path 37 Texas Instruments

Chapter 6: Overvoltage Protection

Overvoltage Lockout
Overvoltage lockout (also known as output voltage
cutoff) is the simpler and more common solution of the
two. TI power switches with an overvoltage lockout
feature generally have an OVLO pin that monitors the
input voltage rail through a configurable resistor divider.
Once the voltage at the OVLO pin increases beyond a
certain threshold, the internal comparator turns the pass
field-effect transistors off. Figure 1 is a typical block
diagram of an overvoltage lockout circuit implementation.

VIN
OUT
IN

Figure 2: TPS259573 overvoltage lockout response.

R1 Charge
pump
Driver
VIN(OV)
+
FLT
Overvoltage Clamp
OVLO
– Control logic
Overvoltage clamp is another way of responding when
R2 VOVLO GND
temporary overvoltage conditions occur in the system.
During temporary overvoltage events, it is best not to
interrupt the downstream load and clamp the system
Figure 1: Overvoltage lockout block diagram. voltage at a safe level, rather than cutting off the system
voltage. Temporary overvoltage may occur because of
For robust protection, faster response time is critical events such as:
in continuous overvoltage protection to ensure the • The load of an upstream, low-switching-frequency
downstream loads are not exposed to the overvoltage DC/DC converter drops suddenly.
condition for a long duration. For example, the TPS2595
eFuse responds to overvoltage condition within 3 µsec • An upstream DC/DC converter operates in pulse-
as shown in Table 1 below. frequency modulation at light loads.

Overvoltage lockout power switches will remain off as • The load is being powered by an unregulated,
long as the input voltage exceeds the set overvoltage secondary winding of a multioutput flyback converter.
lockout threshold. Once the input voltage falls below the
The duration of these events can range from few tens to
threshold, the device turns back on.
hundreds of microseconds.
Figure 2 shows the overvoltage lockout operation of the
When the input voltage exceeds a certain threshold, the
TPS259573. The overvoltage lockout set point is 17 V.
internal clamp activates and limits the output voltage
An overvoltage of 18 V is applied on a 12-V nominal
to the programmed level as shown in Figure 3 on the
operating voltage. Once the input voltage rises above
folowing page. This ensures that the load is not exposed
the set threshold, the internal field-effect transistor (FET)
to the high voltages present at the input supply while still
turns off and the output starts discharging. To report
powering the load.
the overvoltage event, the fault signal (FLT) is asserted
immediately after the pass FET turns off. As long as an overvoltage condition is present on the
input, the output voltage will clamp to the programmed

Parameter Test condition Typical value

tOVLO Overvoltage lockout response time VIN(OV) > VOVLO to FLT ¯ 3 µs
Table 1: TPS2595 overvoltage lockout response time.

11 Ways to Protect Your Power Path 38 Texas Instruments

Chapter 6: Overvoltage Protection

level. When the input drops below the output devices such as TVSs and Zener diodes. As an example,
clamp threshold, the clamp will release and normal the TPS259531’s output clamp response time is 5 µSec.
operation continues. Figure 4 shows the overvoltage clamp operation of the
During an input overvoltage condition, the gate of the TPS259531, which has an overvoltage clamp set to
FET is regulated to drop additional voltage across it and 5.7 V. An overvoltage of 8 V is applied on a 5-V nominal
clamp the output voltage. The power dissipation in the operating voltage. Once the input voltage rises above the
FET is the product of the voltage drop across it and the 5.7-V threshold, the gate of the internal FET is regulated
current flowing through, as shown in Equation 1: to clamp the output voltage at 5.7 V.
PD = (VIN – VOUT) × ILOAD (1)
The higher the input overvoltage and load current, the
higher the power dissipated in the FET to keep the
output voltage constant. As a result, there could be
significant heat dissipation in the internal FET and a rise
in the device’s junction temperature. If the overvoltage
condition persists, thermal shutdown can occur and
disconnect the load from the supply, at which point
the device will either stay latched off or start an auto-
retry cycle.
The benefit of a quick response is that it is possible to
lower or eliminate the dependence on external protection

Figure 4: TPS259531 overvoltage clamp response.

IN OUT

Charge Conclusion
pump Driver

+
To reliably protect power path circuits, overvoltage
OVCSEL OVC threshold
select
– Gate
control logic
protection is critical. Different system use cases demand
for different responses during overvoltage conditions
FLT such as overvoltage lock out or overvoltage clamping.
Over temperature
eFuse devices with integrated overvoltage protection
GND
protection provide flexible, fast and compact solutions to achieve
robust power path protection in your system.
Figure 3: Overvoltage clamp block diagram.

11 Ways to Protect Your Power Path 39 Texas Instruments

 Chapter 7 Thermal Shutdown

Author: Sreenath Unnikrishnan Thermal Shutdown in Power Switches

Thermal protection is an integral part of power switches,
Abstract
which themselves are used as input-protection devices
In power integrated circuits (ICs), thermal shutdown
or as high-side switches driving offboard loads. Monitors
protects the device from failure when its junction
are usually positioned in the power transistor area
temperature exceeds its safe limit. In power switches,
where most of the power dissipates. Some multichannel
fault conditions in load or thermal systems can cause
switches (TI’s TPS2HB08-Q1, for example) may include
high device junction temperatures.
more than one junction monitor.
When the junction temperature exceeds the threshold
In such multichannel power switch, the protection
for thermal shutdown, in order to self-protect, the
mechanism will only turn off the field-effect transistor
power switch turns itself off to reduce the power
(FET) with a high junction temperature; the switches
dissipation. This chapter describes the functionality of
pertaining to the other channels retain their functionality.
the thermal shutdown, the behavior of power switches
Figure 1 shows an example of the thermal
when reaching thermal shutdown conditions, and their
shutdown function.
recovery to normal operation when the fault condition is
no longer present.

Introduction FAULT (ST)

Many power ICs have a self-protection mechanism to

shut down the device in the event of a high junction VOUTx

temperature. Junction temperatures beyond normal

ENABLE
operation levels can occur if high power dissipation tRETRY

exists for a sufficient duration, or the heat dissipation TABS

THYS

from within the device to the rest of the system, or

ambient air is not effective. Without a built-in protection
TJ
t

mechanism, the junction temperature may exceed the

Switch is disabled. TJ Switch follows
TJ increases towards TABS due
decreases while switch is ENABLE. Normal
to an overload event
disabled operation
safe limit and permanently damage the device. In most
cases the protection mechanism turns off the function in Figure 1: Thermal shutdown function example.
the IC that caused the excessive power dissipation.
A thermal shutdown circuit typically detects that a power In Figure 1, the junction temperature rises due to
IC is overheating by measuring the absolute junction a temporary overload event. When the junction
temperature of the hottest areas on the chip. A thermal temperature reaches TABS, the switch function is
monitoring circuit translates the junction temperature disabled. As a result, excess power dissipation ends
to a voltage or current level that is then compared and the junction temperature starts to fall. When the
to a reference level (the absolute threshold, or TABS) temperature has decreased by THYS, the device can retry
corresponding to a temperature below the safe limit for turning back on. Normal operation will resume, since the
the IC. If the junction temperature is above TABS, then temporary overload is gone.
the device function will partially or fully shut down. The In cases where an overload condition persists, the device
implementation of a thermal hysteresis function requires will experience thermal shutdown again. Some power
that the junction temperature fall below a specified switches implement a minimum duration for tRETRY to
amount (set as THYS) before normal circuit function can reduce the average junction temperature.
resume. Thermal hysteresis ensures that the device
Higher-than-normal power dissipation can occur in
does not transition into and out of thermal shutdown at a
power ICs when the load current exceeds normal
high rate.
operational levels or when a high voltage drop appears

11 Ways to Protect Your Power Path 40 Texas Instruments

Chapter 7: Thermal Shutdown

across the FET with high inrush current. The heat Overload Protection
generated on the IC must transfer onto the printed circuit In power switches with current-limiting overload
board and then to ambient air. protection, the current is controlled to the current-limit
Equation 1 calculates the junction temperature (TJ) threshold level. While the current is limited to that safe
resulting from constant power dissipation (Pdiss) on an IC: level, the power dissipation can be quite high during
current limiting. Thermal shutdown occurs after a
TJ = TA + Pdiss × ΘJA (1)
duration of time under these conditions depending on
where ΘJA is the thermal resistance and TA is ambient the thermal impedance. The current from the supply to
temperature. the load will reduce.
Equation 2 calculates the power dissipation in the The thermal shutdown feature in power switches
FET as: supports an additional function—that of a fuse limiting
Pdiss = Vds × Ids (2) the duration of an overload fault event. In the case
of limiting the duration of an overload event, thermal
where Vds and Ids are the FET’s drain-to-source voltage
shutdown may also potentially help avoid damage
and current, respectively.
to upstream and downstream components and
The Vds and the Pdiss can be high when supporting high connectors. The combination of overload and built-in
inrush currents; for example, during initial turn-on into overtemperature protection ensures safe operating area
a capacitive load. When delivering a steady-state load protection for integrated power switches.
current (IL), Equation 3 expresses the power dissipated
in the FET with the on-resistance (Ron) as: Conclusion
Many TI power switches that are at risk for exposure
Pdiss = IL2 × Ron (3)
to fault conditions include a thermal shutdown function
Design Considerations as a protection mechanism. You must be aware of the
You will need to design the maximum device power absolute maximum temperature specification in the data
dissipation and thermal system parameters to keep sheet and design the thermal system and the worst-case
the junction temperature below the absolute maximum IC power dissipation to keep the junction temperature
temperature specification in the data sheet. The thermal below this level.
shutdown threshold is designed to protect within the
safe limit of the IC. If the power dissipation in the IC is
higher than the expected system design target—due to
a load fault, or if the thermal system is degraded due to
a defect—the actual junction temperature could exceed
the worst-case maximum design target and thermal
shutdown protection will kick in.

11 Ways to Protect Your Power Path 41 Texas Instruments

 Chapter 8: Current Monitoring

Authors: Abhinay Patil and Paul Kundmueller rising beyond a certain threshold and send a signal
to the central or graphics processing unit to quickly
Abstract
throttle back its performance, thus preventing the
Many applications require output load current monitoring. entire system from going into shutdown. Typically,
Actual needs vary from system to system, ranging this signal needs to be very fast (<10-µs latency).
from simple parameter data logging and reporting
to diagnostics and control. Multiple solutions are • Diagnostics. It’s possible to identify faults in a
available to monitor load current, each having their own system by looking at the current consumption of
advantages and challenges. Choosing the right solution different sections and comparing this consumption
requires an assessment of system requirements and to a threshold based on the expected value
trade-offs between factors that include performance, during normal operation. Any section consuming
cost, area and complexity. distinguishably high or low current could indicate a
fault. For example, in factory automation, wiring is
This chapter addresses the output load current-
a must for power and signal paths because of its
monitoring function available in integrated power
distribution across large factory environments. Digital
switches rather than stand-alone current-monitoring
output modules commonly specify the maximum
devices/solutions.
wire length as 100 m. With wires of this length,
Current-Monitoring System Applications the wire could accidentally break. Wire breaks are
The need for current monitoring in a whole system detectable by monitoring the current in the loop,
or a subsystem varies. Here are a few examples of either through an analog current monitor or an open-
applications that benefit from current monitoring: load detection feature.

• System power reporting. The most common Similarly, automotive driver assistance system
reason to monitor basic electrical parameters sensors and antennas are often connected via long
such as voltage and current is for reporting and cables due to their location in automobiles. Having
data-logging purposes. For example, in a data the ability to quickly alert and diagnose a wire break
center with thousands of servers, monitoring the is critical to ensure safe operation.
power consumption of individual server nodes
• Predictive maintenance. Some systems (servers,
helps redistribute the workload uniformly, thereby
line cards and programmable logic controller I/O
ensuring that the data center is running at its most
modules) monitor an electrical parameter’s profile
optimal efficiency.
and look for changes in the time/frequency signature
System current monitoring is applicable in factory to predict impending faults. This information helps
automation as well, especially in remote input/output system operators schedule maintenance before a
(I/O) modules that use M16 and M18 cables to fault occurs. Predictive maintenance can help avoid
distribute power. Having a way to monitor the current unplanned downtime, a loss of productivity and even
of the ports and the system is important in order to safety hazards. In these applications, the current
not exceed cable current ratings. Otherwise, damage monitor needs to have good resolution to capture
to the system could occur. subtle changes in the current, as well as sufficient
bandwidth to capture any transients or instantaneous
• Dynamic/real-time control. Some systems use
changes in the current.
load current measurements to implement real-time
control. In such applications, apart from accuracy, it’s Current-Monitoring Solutions
equally important to have high bandwidth and low There are different types of current monitoring solutions
latency. For example, in servers, the power monitor based on the sense element used (external/internal) and
can check if the overall system current is rapidly the type of output (analog/digital).

11 Ways to Protect Your Power Path 42 Texas Instruments

Chapter 8: Current Monitoring

External Sense Elements the temperature drift. Choose sense resistors with a low
Hot-swap controllers use an external resistor (RSENSE) in temperature coefficient, such as thin film or bulk metal.
the power path to sense the load current. The voltage The other sources of error stem mainly from an improper
drop across RSENSE is amplified by an internal high- printed circuit board (PCB) layout around the current-
side current-sense amplifier and then either driven as sense path. For best results, use a Kelvin sense
an analog current output to an external analog-to- connection, as shown in Figure 2.
digital converter (ADC), or fed into an internal ADC.
The main advantages of an external current-sense
Figure 1 shows current monitoring with an external
resistor-based solution are the ability to achieve high
sense element.
linearity and accuracy when sensing current over a wide
range. The disadvantages of an external current sense-
Source Load resistor are increased cost and board space, along with
+
the additional voltage drop, power dissipation and self-
–

Digital telemetry bus

heating in the sense resistor.
Analog IMON Internal Sense Elements
eFuses and smart high-side switches use internal
current-sense circuits and don’t need external
components in the power path for current sensing.
Figure 1: Block diagram of current monitoring with an external Current sensing is usually implemented using some
sense element.
form of current mirror circuit, which drives a fraction
of the main power field-effect transistor (FET) current
You’ll need to select the RSENSE value based on the on the current-monitor output pin. Figure 3 shows a
maximum system load current (IMAX) and the maximum block diagram of the TPS2663 eFuse, which uses an
sense voltage (VSNS,MAX) specified in the hot-swap integrated current-sense circuit to provide an analog
controller data sheet. Equation 1 calculates RSENSE current-monitor signal output on a dedicated pin (IMON).
based on these specifications:
RSENSE = VSNS,MAX/IMAX (1)

Current-sense accuracy is a function of the RSENSE

tolerance and the current-sense amplifier gain/
offset error, along with their associated temperature
coefficients. It is possible to minimize the error by
performing a system-level calibration with an accurate
load current reference, but it’s not possible to eliminate

Load current path

Figure 3: TPS2663 internal block diagram.

PCB PCB Some devices, like the TPS2596 eFuse, provide the
source load analog current-monitor output on a dual-purpose pin
trace trace
(ILIM), which is primarily used to set the adjustable current
Kelvin sense limit, as shown in Figure 4 on the following page. The
traces to voltage drop across the current-limiting resistor (RILM)
amplifier Sense resistor
provides a measure of the output load current.
VSENSE
The accuracy of an integrated current-sense solution
Figure 2: Kelvin sense layout for an external current-sense resistor.
is governed by the mismatches and other errors

11 Ways to Protect Your Power Path 43 Texas Instruments

Chapter 8: Current Monitoring

To ensure maximum use of the measurement’s dynamic

5 OUT
range, select RMON such that VMON at the full-scale load
current matches the full-scale voltage input of the ADC
x 660 µA/A or any other downstream circuit it feeds into.
2 µA
It is possible to improve sensitivity/resolution at low
2 dVdt
x21
currents by using a larger RMON, as it increases the
available signal voltage relative to the noise. However, a
Gate control larger RMON also increases the voltage swing at higher
Current limit amplifier 600 mV currents and might lead to clipping/saturation near
7 ILIM full scale due to headroom limitations on the IMON pin
driver circuit.
Figure 4: TPS2596 internal block diagram.
High-/Low-Current Digital Indication
Many solutions provide a digital output flag pin that is
associated with the internal circuits, which are factory- asserted/deasserted by comparing the load current with
trimmed to provide high accuracy across a range of load a certain threshold.
currents and operating temperatures. Some devices,
Smart high-side switches like the TPS27S100A provide
such as the TPS25982, also employ circuit techniques
an open-drain status output pin to flag the system that
like a copper-stabilizer amplifier to minimize offset error
an open load has occurred. One important aspect to
and drift over time and temperature. The use of factory
consider for open or light-load detection is the normal
trimming and other circuit techniques minimizes the need
operating range of the output current. To avoid false
for any additional calibration in the final system to achieve
fault reporting, the nominal operating range needs to be
highly accurate current measurements.
higher than the maximum open-load detection threshold
An internal current-sensing scheme offers significant listed in the data sheet. Table 1 on the following page
advantages over external sense resistor-based solutions is a reference to this specification in the TPS27S100A
by eliminating the added cost, size, voltage drop and data sheet.
heating associated with a sense resistor.
Additionally, the integrated analog current-monitor
Analog Output Types output (IMON) of the TPS25940, TPS25942, TPS25944,
The output is an analog current signal proportional to the TPS1663 and TPS2663 eFuses can be repurposed
load current. Equation 2 and Equation 3 express the together with the integrated comparator (PG, PGTH)
relationship between the output load current (ILOAD) and to provide an open or light-load detection function, as
the current-monitor signal (IMON): shown in Figure 5.
IMON = ILOAD × GMON (2)
Source IN OUT Load
where GMON is the current-monitoring gain.
PG
Alternately: PGTH
FLAG

IMON = ILOAD/KMON (3) IMON

where KMON is the current-sense ratio.

Connecting a resistor (RMON) from the current output pin

Figure 5: Open or light-load detection using the IMON, PGTH and
(IMON) to GND converts the sense current to a voltage, as PG pins.
expressed in Equation 4:
VMON = IMON × RMON (4)

11 Ways to Protect Your Power Path 44 Texas Instruments

Chapter 8: Current Monitoring

Electrical characteristics
5 V < VIN < 40 V; –40°C < TJ < 150°C unless otherwise specified
Parameter Test Conditions MIN TYP MAX Unit
Diagnostics
Open-load detection threshold in VEN = 0 V, when VIN – VOUT < V(ol,off), duration longer
V(ol,off) 1.4 1.8 2.6 V
off-state than td(ol,off). Open load detected.
Off-state output sink current with
I(ol,off) VEN = 0 V, VIN = VOUT = 24 V, TJ = 125°C. –150 µA
open load
Open-load detection-threshold VEN = 0 V, when VIN – VOUT < Vol, off, duration longer
td(ol,off) 600 µs
deglitch time in off state than tol,off. Open load detected.
Open-load detection threshold in VEN = 5 V, when IOUT < I(ol,on), duration longer than
I(ol,on) 2 6 10 mA
on state td(ol,on). Open load detected. Version A only.
Open-load detection-threshold VEN = 5 V, when IOUT < I(ol,on), duration longer than
td(ol,on) 700 µs
deglitch time in on-state td(ol,on). Open load detected.

Table 1: TPS27S100A electrical characteristics.

It’s also possible to achieve open or light-load detection

with other devices that have an analog current-monitor
output by adding an external comparator, as shown in
Figure 6.

Source IN OUT Load

FLAG
IMON

Vthreshold

Figure 6: Open or light-load detection using an external comparator.

Digital Telemetry Output Types Figure 7: LM5066I internal block diagram.

Another approach to digital current monitoring is

telemetry. This digital approach to current monitoring also defines the data format for storing power and
involves sensing an analog current and converting it to voltage data.
a digital signal through an integrated ADC, which is then TI offers many devices that conform to the PMBus
stored in internal digital registers. A digital bus such as standard. For more information on how PMBus is used
an I2C or Serial Peripheral Interface bus then reports the in a device, see the LM5066I data sheet. Sections 8.5.4
data back to a central processing unit. The highlighted through 8.5.8 in the data sheet thoroughly explain the
portions in Figure 7’s block diagram demonstrate data format and equations required to have the system
both the sensing and analog-to-digital conversion of digitally interpret analog current information through I2C.
the current.
Key Performance Specifications Relevant to
The data format of the stored analog current information Current Monitoring
can be defined by industry standards such as PMBus®
Here are some of the key specifications you should
or customized depending on the device. PMBus is a
consider while choosing a current-monitoring solution for
popular specification that helps standardize the reading
your application:
and writing of telemetry data through I2C. PMBus

11 Ways to Protect Your Power Path 45 Texas Instruments

Chapter 8: Current Monitoring

• Dynamic range refers to the range of load currents components such as gain error, offset error and
that the current-monitor circuit can sense and nonlinearity. While gain error dominates at higher
report accurately (see Figure 8). Signals outside the currents, offset error dominates at lower currents.
supported dynamic range would have a significantly For devices with an external current-sense resistor,
higher error in measurement, or may not be sensed the tolerance of the sense resistor also contributes
at all. The dynamic range is usually limited at the to the overall current-monitoring error. You will need
upper end by the headroom of the current-monitor to account for the temperature coefficient of the
amplifier circuit and on the lower end by noise and error components to determine the accuracy of the
offset errors. There’s usually a trade-off between measurement across temperatures.
accuracy at the lower or upper ends of the range.
• Resolution/sensitivity is a measure of the smallest
Using a higher gain or larger sense resistor in the
change in current that the monitoring circuit can
circuit improves the performance near zero scale
sense. There is usually a trade-off between resolution
by increasing the signal relative to the offset and
and dynamic range. For analog monitoring solutions,
noise, but leads to clipping and/or saturation of the
higher sensitivity means higher gain in the sense
output near full scale. Conversely, using a lower gain
circuit. For digital output solutions, higher sensitivity
or smaller sense resistor can ensure better linearity
also depends on the resolution of the ADC.
and accuracy near full scale, but leads to poor
measurement accuracy near zero scale because • Bandwidth is a measure of the fastest change
the measurement is dominated by offset and noise. in current that the monitoring circuit can sense.
Some current-monitoring devices use techniques Any signals within the circuit bandwidth will be
such as chopper-stabilized amplifiers, auto-zero reproduced faithfully, while signals outside the
amplifiers and multipoint trim to minimize offset bandwidth will be attenuated significantly, leading
errors and improve performance at the lower end of to erroneous measurements. The bandwidth also
the range. determines the circuit’s ability to detect instantaneous
changes in the current such as load transients, as
shown in Figure 9.
IMON

ILOAD
y
ty
vi
iti

x
ns
Se
cy
ra
cu
ac
DC

IMON
Insufficient
dynamic range

0
Insufficient Insufficient Time
sensitivity/ bandwidth
Insufficient resolution
0 dynamic range
Dynamic range ILOAD
Figure 8. Current-monitoring performance specifications. Figure 9. System impact of current-monitoring
performance specifications.

• DC accuracy is a measure of the absolute error/ • Latency is a measure of the delay between any
deviation of the current monitor’s output from change in the load current and the time at which
the expected value, expressed in terms of %FS that change is reflected in the measured output.
or %Reading. The error is a combination of Latency is usually a factor in digital current-

11 Ways to Protect Your Power Path 46 Texas Instruments

Chapter 8: Current Monitoring

monitoring solutions because of the analog-to-digital In addition to providing power-path control and
conversion time and propagation delay through the protection, power switches like hot-swap controllers,
digital circuits. eFuses and smart high-side switches can provide added
value by performing current monitoring as well. While
Conclusion
many power switches are available on the market, in
Table 2 lists the various power-switch solutions some cases it’s possible to narrow down the choice
offered by Texas Instruments with current- based on its integrated current-monitoring capabilities.
monitoring capabilities.

eFuse Hot-swap controller Smart high-side switches

Bus voltage 2.5 V–60 V ±80 V 3 V–65 V
Minimum sense current 20 mA Depends on external sense resistor 5 mA
Maximum sense current 15 A Depends on external sense resistor 80 A
Internal current-sense element Yes No Yes
Analog current monitor Yes Yes Yes
Digital current monitor No Yes No
Open/light-load detection Yes No Yes

Table 2: Power switches with current-monitoring capabilities.

11 Ways to Protect Your Power Path 47 Texas Instruments

 Chapter 9: Power-Good Signal

Author: Arthur Huang V OUT

IN
V CC

Abstract V IN
+
CIN
H
– ON MCU
Sensitive loads such as microprocessors or system- L
PG RST
CPU
FPGA
on-chips (SoCs) require stable input voltages to Upstream load GND Power switch
remain in regulation. If the voltage rail is experiencing
large amounts of inrush current or the application of
a high-capacitive load, the voltage could dip, causing Figure 1. Power-good reset configuration.
the device to enter an undervoltage lockout (UVLO)/
reset state or damage downstream circuitry. A power- logic integrated circuit (IC) ensures that the reset signal
good signal (PG) can help manage inrush current or will assert if any voltage rail falls outside of regulation.
control rail sequencing by connecting the signal to a Figure 2 shows this configuration by connecting PG
downstream load. from a DC-DC converter and a power switch together.
A power-good signal only asserts when the voltage V OUT1

rail reaches a value sufficient for normal operation. IN

Connecting a downstream enable pin or reset supervisor

CIN
+
V IN
– DC – DC
to the power-good signal ensures that the downstream converter
PG
V OUT1
load only operates with a stable, regulated voltage input. Upstream load
V CC1

Power-good functionality is also a good fit in applications V OUT2

V OUT2

such as power sequencing and power multiplexing. IN V CC2 MCU

CPU
CIN
+ FPGA
Common Power-Good Applications V IN H
– ON

L RST
PG
The following sections will include implementations of a GND
Upstream load Power switch
power-good signal.

Connecting to Downstream Loads Figure 2. Multiple power-good reset configuration.

Many systems require stable, regulated power inputs for
proper operation of downstream loads. If the power rail Systems that require robust input protection can
were to fall out of regulation, the load could malfunction use the power-good signal to reduce system stress
and cause the system to fail. Some downstream during startup. Some devices such as hot-swap
loads that depend on proper voltage rails include controllers monitor the stress across an external field-
microprocessors, field-programmable gate arrays effect transistor (FET) so that it remains within the safe
(FPGAs), SoCs and other sensitive components. operating area (SOA). Having the downstream load turn
As shown in Figure 1, the power-good signal can on during startup while the FET is not fully enhanced
provide feedback to a microcontroller (MCU) when the could add additional stress on the external FET and
voltage rail is on and stable. If the output of the power system. Connecting the power-good signal to the enable
switch contains a high-capacitive load, turning the MCU signal of the downstream load, as shown in Figure 3,
on when the output voltage is stable will manage the
inrush current. In other words, when the power-good Rsns
VOUT
VIN V CC
signal deasserts, the MCU will remain in an off state. DC/DC
GATE
Since the power-good output is usually open drain, Rpg

you will need a pullup resistor to bring the pin high Hot-swap Gate logic
PG
EN
controller
during regulation. GND

If multiple supply rails need regulating, connecting the

power-good signals together to a reset supervisor or Figure 3. Hot-swap power-good configuration.

11 Ways to Protect Your Power Path 48 Texas Instruments

Chapter 9: Power-Good Signal

ensures that the downstream load will only turn on when

the hot-swap FET is fully enhanced. CT
CT VOUT Load 1

NC VOUT

Power Sequencing Power R PU

supply 1 VIN VOUT

Many applications require controlled power-up and Power

VBIAS PG
CIN supply
power-down sequences to properly operate subsystems GPIO
ON
MCU GND
and downstream components. Sequencing is critical on TPS22990

voltage rails that need to turn on in a specific order to

ensure operational safety and reliability. Sequencing rails
also helps stagger the inrush current during power-up, CT VOUT Load 2

which reduces system stress and input voltage dip while CT

NC VOUT

also helping prevent potential reverse-bias conditions. Power

VIN VOUT R PU
supply 2

It is possible to use the power-good signal to sequence CIN

VBIAS PG

voltage rails. For example, by connecting the power- ON GND

good signal of a load switch to the enable pin of a TPS22990

second load switch, the second load switch will only Figure 4. Power sequencing.
turn on after the first load switch is stable. This condition
delays the turn on of the second load switch, effectively
prevent system reset. In order to prevent reverse current
staggering the voltage rails and reducing the inrush
conditions, you can use the power-good signal to toggle
current stress on the system. When the first switch is
the switchover between inputs. For more information on
disabled, the power-good signal pulls the EN/ON pin
reverse current blocking, see Chapter 4.
low and prevents the second switch from turning on. For
more information on inrush current, see Chapter 1. By connecting the power-good signal of a device like
a load switch to the gate of an external P-channel
Figure 4 demonstrates a power-sequencing application
FET (PFET), the power-good signal enables seamless
using power good. You can also add delay to the power-
switchover between two power rails and eliminates the
good signal to further delay the turn on/off sequencing of
need for a discrete timing circuit.
the rails, as shown in Figure 5.
As Figure 6 on the following page illustrates, when the
Power Multiplexing
TPS22990 power-good pin pulls high, the gate of the
Power multiplexing is a set of electronic switches used PFET also pulls high, and this turns off the FET. When
to select and transition multiple input power rails to the power-good pin asserts low, the gate of the PFET
a single output. Systems that include backup power pulls low and turns on the PFET, allowing current to flow
rails or backup batteries may require a fast switchover from IN2.
between supplies to minimize the output voltage dip and

VPGD VPGD VPGD

RPG1
RPG1 RPG1
PGD PGD PGD
Power Power Power
good good good
RPG2 RPG2
CPG CPG CPG

GND GND GND

A) Delay at rising edge only B) Long delay at rising edge, C) Short delay at rising edge and
Short delay at falling edge Long delay at falling edge or
Equal delays

Figure 5. Adding delay to the power-good output pin.

11 Ways to Protect Your Power Path 49 Texas Instruments

Chapter 9: Power-Good Signal

Power supply 1
TPS22953
System
One limitation of this circuitry is that the power-good
signal is only influenced by VOUT. When VOUT is present,
IN OUT
(4.5 V to 5.5 V) load
IN OUT

the power-good signal will assert even if the VOUT rail

R SNS1 = 10 k
BIAS SNS
R EN1 = 10 k CL
RSNS2 = 1.45 k

REN2 = 2.05 k
EN fluctuates. For example, if VOUT is 12-V nominal but a
PG
GND
PAD CT
high-capacitance load connects to the output, the input
voltage could dip down to 10 V. The power-good signal
R pullup

will remain asserted as long as the left FET is biased.

PG signal

Integrated Implementations
TPS22910A
Power supply 2
IN OUT Depending on the device, there are a few methods
(3.0 V to 3.6 V)
ON GND for implementing power good using an integrated IC.
The first method, VOUT/VIN power good is similar to the
Figure 6. Make-before-break power multiplexer schematic. discrete power-good solution. As Figure 8 illustrates,
the power-good signal asserts when the output voltage
Power-Good Signal Architecture reaches a certain percentage of VIN. For this example,
Power Good Signals can be implemented either the power-good signal will assert once VOUT reaches
discretely or with an integrated implementation. The 90% of VIN. This offers the simple functionality of power
following section will showcase both solutions good using a comparison between VIN and VOUT.

Discrete Implementation
Figure 7 shows a discrete implementation of a power- VIN VOUT
good signal. In this implementation, the assertion of
the power-good signal depends on the output voltage.
Power good asserts once the output voltage reaches Control PG
a certain threshold to bias the discrete FETs. In this logic
+
example, when VOUT is low, the voltage at the gate of -
FET Q2 pulls low, turning it off. This leaves the gate of
Q3 pulled up to V+, turning it on and pulling the power-
good signal to ground. When VOUT reaches a certain level
specified by the resistor divider, the gate on Q2 asserts Figure 8. Integrated VOUT/VIN power good.
high, which pulls the gate of Q3 low, turning Q3 off. This
pulls the power-good output high to VOUT, asserting the
However, VOUT/VIN power good comes with a limitation
power-good signal.
when turning on into a high-capacitive load. Since the
power-good signal only depends on the VIN-to-VOUT ratio,
VIN VOUT
power good will assert even if the voltage fluctuates. If
the downstream load is pulling a lot of current due to a
Q1 V+
high-capacitive load, there could potentially be a large
PG inrush current event during startup. As mentioned in
Chapter 1, charging the load capacitance could cause
Q3
the input voltage to dip. Since power good ultimately
Q2 depends on the VIN/VOUT ratio, the power-good signal will
ON still assert if VIN and VOUT dip.
VOUT/VIN power good also has another limitation at
lower voltage applications. At lower VIN, the power-
Figure 7. Discrete power-good feature. good signal could potentially not assert due to the on-
state resistance.

11 Ways to Protect Your Power Path 50 Texas Instruments

Chapter 9: Power-Good Signal

For example, looking at Equation 1, if the device is

VOUT
operating at 0.5 V at 1 A with an on-state resistance VIN

of 100 mΩ, the output voltage will be 0.4 V, which is

outside the range of 90% VIN. Therefore, the power-good
signal might not assert. Control PG
PG
logic
VOUT = VIN – ILOAD × RON (1)
An effective way to implement power good is by
measuring the on-resistance of the pass FET. Sometimes Figure 9. Integrated VGS power good.
called “VGS good,” this method asserts power good
when the VGS voltage rises to the point where the FET
on-resistance approaches its nominal value. A delay is causing an accidental assertion of the power-good pin.
also introduced into the circuitry to prevent any transient One limitation with VGS good involves VBIAS and VON
or inrush event from accidentally asserting VGS good. during startup. If the VBIAS and VON pins are high, and
After the voltage on the EN/ON pin exceeds the VIH level VIN starts to ramp, power good will assert since the VGS
of the device, the device will measure the on-resistance threshold remains high. Since VBIAS powers the voltage
and wait a certain time (tBLANK) before asserting high. If supplied to the gate, the VGS threshold will remain high.
the voltage on the EN/ON pin falls below VIL or the VGS of Conclusion
the FET falls below the threshold, power good deasserts. Sensitive applications that require stable input voltages
Figure 9 shows a common implementation of VGS can use a power good signal to remain within regulation.
power good. A power good signal can help manage inrush current or
VGS power good offers the advantage of working at lower control rail sequencing, which is useful in applications
VIN values since the on-state resistance fluctuations are such as power sequencing or power muxing.
small, and prevents transient or inrush conditions from

11 Ways to Protect Your Power Path 51 Texas Instruments

 Chapter 10: Safely Driving an Inductive Load

Author: Mahmoud Harmouch In the de-energizing state, the current decays from the
steady-state value to zero; a voltage spike proportional
Abstract
to the current slope appears across the switch. The
Inductive loads are relays, solenoids, electric motors voltage spike must be limited and safely dissipate the
and even loads connected through a long cable. Their stored energy, or it can damage the system.
impedance consists of both a resistance (R) and an
This chapter focuses on calculating inductive load
inductance (L) in series. The R value determines the
parameters for optimizing drive circuit capability. For both
steady-state current, and the L value determines the
high- and low-side switches, the energizing and de-
stored magnetic energy. This stored magnetic energy
energizing mechanism is the same.
in the inductor can cause system- or component-level
damage if not properly dissipated. Challenges of Inductive Loads
The opening or closure of a magnetic contact in a The voltage across an inductive load is time-dependent.
relay or solenoid requires the storage or dissipation of Equation 1 calculates the inductive load parameters
magnetic energy. In the case of an electric motor, this necessary for selecting a reliable drive circuit:
stored energy is necessary for mechanical rotation. dI(t)
V(t) = R I(t) + L (1)
Inductive loads are continuously energized and de- dt
energized for opening or closing contacts (relays Equation 1 shows that an overvoltage spike can
and solenoids) and rotation or idle (electric motors). occur when disconnecting an inductive load due to
Disconnecting an inductive load from an energized state a sudden change in current—the dI(t)/dt component.
creates a high-voltage spike that can lead to system The magnitude of this overvoltage spike is the element
damage. Safely de-energizing an inductive load requires that can cause component damage, and, if necessary,
the implementation of an appropriate clamp. should be clamped.
Introduction Energizing an Inductive Load
When encountering an inductive load, you must take An inductive load is energized when connected to a
into account the amount of energy stored, as this voltage source. The current ramps up exponentially to a
energy can damage components in the system if not steady-state value and magnetic energy is stored in the
properly managed. coil. Figure 1 illustrates this behavior, using a high-side
There are two states to consider when driving an switch to drive the load.
inductive load: I(t)
• Energizing or connecting the inductive load to a + VSW – I0 = VBAT /R

voltage source such as a battery. + I(t) + + 

t
VL
+ L VSW
• De-energizing or disconnecting the inductive load –
Inductive
VBAT – VOUT VBAT
from the voltage source. load +
t
R VR
You can reach either state by using switches such as a VOUT
– –
bipolar junction transistor, a power field-effect transistor, VBAT
or an integrated switch that connects one side of the t
SW initially open SW close

switch to the voltage source (the high-side switch) or Figure 1. Energizing an inductive load waveform.
to ground (the low-side switch). In some instances, the
integrated switch option may integrate a voltage clamp.
When energizing an inductive load in a resistor-inductor
In the energizing state, the switch drives the steady-
(RL) circuit, the voltage loop is a first-order differential
state load current and the inductance stores magnetic
equation (Equation 2 on the following page):
energy equal to half the L value and the square of the
load current.

11 Ways to Protect Your Power Path 52 Texas Instruments

Chapter 10: Safely Driving an Inductive Load

dI(t) Disconnecting an inductive load from a voltage source

VBAT = R I(t) + L (2)
dt generates a huge negative voltage spike across the
Solving Equation 2 gives you the instantaneous inductor, as shown in Figure 2. The voltage spike will
inductive load current. You can calculate the important appear across the switch and be much higher than the
circuit parameters using Equation 3: switch’s rated breakdown voltage. Adding an appropriate
voltage clamp will avoid having a voltage spike cause
VBAT –t

I(t) =

(1– e ) (3) severe damage.
R
The challenge at this stage is to select a switch that can Figure 2 shows the expected waveforms in the case
carry the steady-state load current and keep the junction of a disconnected energized inductive load through a
temperature below the thermal shutdown temperature. high-side switch, both with and without a clamp present
The best switch options include a low dropout voltage, in the system.
like in a low on-resistance switch.
I(t)
Table 1 shows the parameters introduced by VBAT /R

Equation 1, and their impact on the system. t

VOUT
SW is a FET and clamp is between drain and gate
VBAT
+ VSW – t
Stored + With Clamp
+ -(VCLAMP – VBAT)
Time Steady-state magnetic I(t) I(t) +
NO Clamp
Parameter constant current energy +
VCLAMP L VL
– -VSPIKE
2
VBAT – VOUT
1 L VBAT
Equation t = L/R I0 = VBAT/R
2
(R
) R
+
VBAT + VSPIKE
VSW

VR
Impact Predicts the Helps in RON Helps in – – –
energizing selection clamp circuit VCLAMP
time selection VBAT
t
Table 1: Calculated parameters of an energizing inductive load.

TSW,OFF

TDECAY
Figure 2. Waveform of an inductive load disconnected from a voltage
Disconnecting an Inductive Load from a source with a clamp (teal) and without a clamp (red).
Voltage Source
The voltage across an inductive load is expressed by Table 2 shows the effectiveness of implementing a
Equation 4: clamp voltage across the switch. The clamp slows down
dI(t) the current slope from 250 mA/10 µS to 250 mA/419 µS
V(t) = R I(t) + L (4)
dt and safely dissipates 3.125 mJ of stored energy. Without
A sudden change in the current (dI(t)/dt) induces a very the clamp, the 2512-V voltage spike can easily damage
high voltage as the current changes from VBAT/R to zero. the switch.

On-
state: Off- Stored Voltage
current state: magnetic spike or Voltage
DC switch current energy clamp across the
resistance Inductance on switch dI(t) switch
2

Clamp (R) (L) VBAT VBAT/R off 2

V
E = 1 L BAT
R
( ) TSW,OFF TDECAY
L
dt (VSW)
2512 V
Without
48 W 100 mH 12 V 250 mA 0 mA 3.125 mJ 10 µS 2500 V (high voltage
clamp
spike)
With
48 W 100 mH 12 V 250 mA 0 mA 3.125 mJ 419 µS 48 V 60 V
clamp

Table 2: Voltage levels caused by switching off an inductive load, with and without a clamp.

11 Ways to Protect Your Power Path 53 Texas Instruments

Chapter 10: Safely Driving an Inductive Load

VBAT VBAT VG
+ + +
I(t) + D
OFF State

VCLAMP
ON State
VL t
L G
VDS
– VOUT VDS
VCLAMP
Inductive + VBAT
load t
R VR - VOUT
S
VBAT I(t) + +
D – – VBAT
VCLAMP

+ L VL t
VBAT – VCLAMP
Inductive
G - VOUT
load + I(t)
V DS
R
VR I0 = VBAT/R
S
GND – – GND - - t
TDECAY

Figure 3. De-energizing an inductive load using a transient voltage suppression (TVS) clamp across the switch.

Safely De-Energizing an Inductive Load with Smart High-Side Switches,” offers more detailed
There are three ways to safely de-energize an inductive calculations for driving an inductive load.
load when disconnecting it from the voltage source: Let’s review the three methods.
• Use a clamp across the switch. Using a Clamp Across the Switch
• Use a clamp and blocking diode across the Using a clamp across the switch is the most effective
inductive load. method and covers most inductive loads. The clamp
is integrated in most smart high-side switch integrated
• Use a freewheeling diode across the inductive load.
circuits used in automotive and industrial systems.
In all three methods, the main objective is to keep the Figure 3 shows the de-energizing of an inductive load
voltage across the switch below the breakdown level and when using a clamp across the switch.
safely dissipate the stored energy in the inductive load.
Solving for Equation 5, where K represents the
Equation 5, for the clamping circuit, is based on a first- clamping method for the circuit shown in Figure 3,
order differential equation: results in Equation 6:
dI(t) (5)
RI(t) + L =K dI(t)
dt RI(t) + L = –(VCLAMP – VBAT) (6)
dt
where K is a constant depending on the clamp method
Table 3 lists the calculated parameters from Equation 6.
in each circuit. The application report, “How to Drive
Resistive, Inductive, Capacitive and Lighting Loads

Initial Stored
Inductance Resistance current energy Demagnetization time TVS dissipated energy

1 L VBAT
2
L R × I0
V
(
ED = L VCLAMP I0 – CLAMP
R
– VBAT
L R I0 = VBAT/R
2
( R
) tDEMAG = Ln(1 +
R VCLAMP – VBAT
) R
Ln(1 + V
R × I0
) )
CLAMP – VBAT

2
1 L VBAT VBAT
Simplified formula when VCLAMP > 3 × VBAT
2 R
( )V CLAMP– VBAT

Table 3: Parameters necessary for selecting a switch and clamp (clamp across switch).

11 Ways to Protect Your Power Path 54 Texas Instruments

Chapter 10: Safely Driving an Inductive Load

VG
VBAT VBAT
+ + + t
I(t) + D
VDS
VL V BAT + VCLAMP + VF
L G VDS VBAT
– V
VCLAMP + VF

+ OUT t
VOUT
R VR -
S
VBAT I(t) + + t
D – –
+ L VL - ( VCLAMP + VF)
I(t)
- VOUT

VCLAMP + VF
G
+
V DS I0 = VBAT/R
R
VR
S t
GND – – GND – – TDECAY

Figure 4. De-energizing an inductive load using a TVS clamp across the load.

Using a Clamp and Blocking Diode Across the Load Table 4 lists the calculated parameters from Equation 7.
Figure 4 shows the configuration in which the clamp The series diode is a must for preventing current from
and a blocking diode are placed across the load. flowing in the external TVS when in the on state.
Again, substituting the clamping voltage and forward
voltage for the blocking diode into Equation 5 generates
Equation 7:
dI(t)
RI(t) + L = –(VCLAMP + VF) (7)
dt

Initial Stored
Inductance Resistance current energy Demagnetization time TVS dissipated energy

1 L VBAT
2
R × I0 ( V
ED = L (VCLAMP + VF) I0 – CLAMP
+ VF
L R I0 = VBAT/R ( ) tDEMAG =
L
R
Ln(1 +
VCLAMP + VF
) R R
2 R
Ln(1 +
R × I0
VCLAMP + VF
) )
2
1 L VBAT
Simplified formula when VCLAMP > 2 × VBAT
2
(R
)
Table 4: Parameters necessary for selecting a switch and clamp (clamp and blocking diode across load).

11 Ways to Protect Your Power Path 55 Texas Instruments

Chapter 10: Safely Driving an Inductive Load

VBAT VBAT VG
+ + +
I(t) + D

VL t
VDS
L G
– V VDS
VF

+ OUT VBAT + VF
VBAT
t
R VR - VOUT
S
VBAT I(t) + +
D – – VBAT
+ L VL - VF t
– VOUT

VF
G
+ I(t)
V DS
R
VR I0 = VBAT/R
S
GND – – GND – – t
T DECAY

Figure 5. De-energizing an inductive load using a freewheeling diode across the load.

Using a Freewheeling Diode Across the Load Table 5 lists the calculated parameters from Equation 8.
Figure 5 shows the configuration using a freewheeling
diode across the load.
Using Equation 5 and substituting for K, Equation 8
generates the expression for this de-energizing method:
dI(t)
RI(t) + L = –VF (8)
dt

Initial Stored
Inductance Resistance current energy Demagnetization time Diode dissipated energy

)
2

L R I0 = VBAT/R
1 L VBAT
2
(
R
) tDEMAG =
L
R
Ln(1 +
R × I0
VF
)
R
(V
ED = L VF I0 – F Ln(1+
R
R × I0
VF
)

VF VBAT
Simplified formula 0.8 L
R R

Table 5: Parameters necessary for selecting a switch and freewheeling diode.

11 Ways to Protect Your Power Path 56 Texas Instruments

Chapter 10: Safely Driving an Inductive Load

Clamp across switch

Parameter (could be integrated) Clamp across load Freewheeling across load
VBAT + VF
Voltage stress across switch VCLAMP VBAT + VCLAMP + VF
Lowest stress
VBAT VBAT VBAT
Peak power VCLAMP × VCLAMP × VF ×
R R R

L V
L VBAT L VBAT Ln(1 + BAT )
Demagnetization time Ln(1 + ) R
Ln(1 +
VCLAMP + VF
) R VF
R VCLAMP – VBAT
Longest demagnetization time

2 VBAT VF
1 L VBAT VCLAMP 2

Dissipated energy
2
(R
)V CLAMP – VBAT
1 L VBAT
2
( )
R
0.8 L
R R
Lowest dissipated energy

Comment Most suitable for board space and Used if the switch is not Used if the demagnetization energy
cost because the clamp is small in capable of dissipating is relatively high in cases of high load
size and energy dissipates in the the energy; an additional currents and high inductance; long
switch clamp would increase demagnetization time TDECAY could
board space and cost be an issue; increases board space
and cost

Table 6: Pros and cons of the three clamping methods.

Selection Method Pros and Cons Conclusion

Table 6 summarizes all three ways to safely de-energize An inductive load can be driven by a high- or low-side
an inductive load. switch. The advantage of a high-side switch is the
capability of protection against short to ground and the
As previously stated, the clamp across the switch is
flexibility of load diagnostics. TI smart high-side switches
integrated in smart high-side switches. Smart high-
come in a variety of FET sizes and feature an integrated
side switches have the ability to dissipate the magnetic
clamp across the FET.
energy in most cases. In some applications, however,
the integrated clamp is not capable of dissipating the Using smart high-side switches, you must calculate the
magnetic energy, so an external clamp such as a TVS inductive load parameters and compare them against
diode is required. When using an external TVS diode, the device capability in terms of the on-resistance and
its clamp voltage should be lower than the integrated demagnetization energy capability. You will not need an
clamp’s lower limit minus the battery voltage, as external clamp if the switch’s integrated clamp is capable
expressed by Equation 9: of dissipating the demagnetization energy. If the clamp
(9) cannot dissipate that energy, you will need an external
VCLAMP(EXT) < VCLAMP(INT) – (VBAT + VF)
clamp, TVS diode or a freewheeling diode.
Table 7 summarizes the important parameters for smart
high- and low-side switches.

Low on- Short-to-ground Integrated Inductive energy Junction-to-ambient

Device family
resistance protection clamp capability thermal resistance
Smart high-side switches 4 4 4 4 4
Low-side switches 4 4 4

Table 7: Important parameters for driving an inductive load.

11 Ways to Protect Your Power Path 57 Texas Instruments

 Chapter 11: Automotive Load Dump

Author: Cameron Phillips

Abstract
In automotive applications, a load dump is one of the
largest transient pulses in modern electronics. This low-
impedance pulse, defined by International Organization A Load
for Standardization (ISO) 16750 pulse A and B, can be
devastating to integrated circuits (ICs), as voltages can
rise as high as 120 V for almost half a second.
These transient pulses occur when a generator delivering
Figure 2. Battery disconnected from car system.
current is abruptly disconnected from the system. Most
electronics qualified to withstand load dump will survive
the suppressed version of a pulse while other circuitry Since the alternator is always on when the engine is on,
clamps the unsuppressed waveform. This chapter will there is a high probability that a vehicle could experience
explore the details of load-dump transient pulses, their this battery disconnection scenario. Let’s look at the
causes and how devices are able to survive load dump. voltage profile of a load-dump transient and show the
danger that it poses to systems.
Causes of Load Dump
In modern automobiles, the alternator charges the ISO 16750-2
battery and powers the vehicle’s electronics when The ISO 16750-2 (2012) automotive standard highlights
the engine is running. An alternator mirrors the basic several different battery discontinuities, such as starting
properties of an inductor, which when spun by a belt profile (more commonly known as cold crank) and load
produces a magnetic field that converts to an electric dump. The standard also covers momentary drops in
current. Figure 1 illustrates the connection of a battery in supply voltage, ground reference and supply offset, as
parallel with an alternator and other loads of a vehicle. well as open- and short-circuit load conditions.

Pulse A (Unsuppressed)
Figure 3 shows the waveform of a load-dump pulse,
while defining the voltage levels that the alternator can
reach and how long it will take the voltage to dissipate.
Table 1 on the following page lists the values of Figure 3.
A Load
td
U

tr

US
0.9(US – UA) + UA
Figure 1. An alternator connected with a battery and load.

In Figure 1, the battery acts as a large capacitor,

keeping the voltage at a steady 13.5 V. However, if the car 0.1(US – UA) + UA
battery is disconnected from the line, then the inductive UA
0
nature of the alternator causes a very large voltage spike t

on the power line of the vehicle that could take almost Figure 3. Load-dump test A unsuppressed voltage waveform.
half a second to dissipate; see Figure 2. ISO 16750
defines this transient pulse as a load-dump pulse.

11 Ways to Protect Your Power Path 58 Texas Instruments

Chapter 11: Automotive Load Dump

Type of system Test

This clamp will dissipate the excess energy of
Parameter 12 V 24 V requirements the alternator and keep the voltage set at a level
Us 79 V–101 V 151 V–202 V defined in the pulse in Figure 5; Table 2 shows the
Ri 0.5 W–4 W 1 W–8 W 10 pulses subsequent values.
at 1-minute
td 40 ms–400 ms 100 ms–350 ms intervals U td
tr 10 ms 10 ms
tr
Table 1: Load-dump test A unsuppressed parameters.
US
0.9(US – UA) + UA
There is a large difference in the requirements between
12-V and 24-V automotive systems. A load-dump pulse
is a significant challenge for electronic designers because
most devices’ internal structure will break down before

US *
the voltage achieves the maximum level specified by
0.1(US – UA) + UA
ISO 16750-2. UA
0
Equation 1 calculates the maximum current that the IC t

can withstand from an alternator: Figure 5. Load-dump test B suppressed voltage waveform.
(US–Vbreakdown)/Ri = (79 V–40 V)/0.5 W = 78 A (1)

Type of system Test

There is a usually small clamp or electrostatic discharge
Parameter 12 V 24 V requirements
(ESD) structure within the IC that cannot withstand this Us* 35 V 58 V
level of current for 400 ms. With the pulse being such 5 pulses at
Ri 0.5 W–4 W 1 W–8 W
high energy, vehicle manufacturers usually implement 1-minute
td 40 ms–400 ms 100 ms–350 ms intervals
a centralized circuit that will clamp the pulse to a more
tr 10 ms 10 ms
manageable voltage.
Table 2: Load-dump test B suppressed parameters.
Pulse B (Suppressed)
Since the test A pulse is extremely high energy and
difficult to dissipate, most vehicles have a suppression Most vehicles require all battery-connected ICs to survive
circuit in parallel with the alternator and battery that a suppressed load dump.
clamps the voltage at a less-stressful level. This means Surviving Load Dump
that the internal circuit changes from Figure 2 to
The load-dump pulse is high energy and requires a
Figure 4.
high-power dissipation clamp to suppress the pulse.
To avoid this high dissipation, TI’s portfolio of power
switches includes a higher voltage tolerance instead of
trying to dissipate the entire pulse. This means that the
IC connected to the power line will simply allow the pulse
to occur without being damaged.
A Load
In TI smart high-side switches, the maximum accepted
voltage is 40 V to 45V depending on the switch. During a
suppressed load-dump event, the transient will not break
the switch, as it is lower than the maximum rating for
the device.
Figure 4. An alternator connected to a battery, load and central clamp.
Similarly, for TI’s ideal diode controller family, even though
the field-effect transistor (FET) is not integrated, it can

11 Ways to Protect Your Power Path 59 Texas Instruments

Chapter 11: Automotive Load Dump

still withstand a high transient during a suppressed load in ISO 16750-2 and is characterized for a suppressed
dump. As long as the main FET is rated with a high- pulse and unsuppressed pulse. The two pulses are
enough voltage to survive the suppressed load dump, nearly identical, with one having a central clamp to
the system should see no damage. dissipate a high portion of the energy and lowering
the voltage that the IC has to tolerate. TI’s portfolio of
Conclusion
smart high-side switches are designed to survive these
A car alternator helps charge the battery during normal
transients and can be used in automotive applications
operation, but due to its inductive nature, if the battery
connected directly to the battery.
becomes disconnected, the voltage on the power line
can spike up very high. This transient is represented

11 Ways to Protect Your Power Path 60 Texas Instruments

TI’s integrated power switch portfolio arms engineers with reliable solutions. From managing inrush
current to surviving stressful transients in an automotive environment, our diverse portfolio of
devices can distribute, protect or diagnose power path designs.

TI Worldwide Technical Support

TI Support11 Ways to Protect TI Training

Your Power Path

Thank you for your business. Find the answer to your support need or get
in touch with our support center at
www.ti.com/support
From technology fundamentals to advanced implementation, we offer
on-demand and live training to help bring your next-generation designs to
life. Get started now at

China: Design Tips and Tradeoffs Using TI’s Power Switches

training.ti.com
http://www.ti.com.cn/guidedsupport/cn/docs/supporthome.tsp China:
Japan: http://www.ti.com.cn/general/cn/docs/gencontent.tsp?contentId=71968
http://www.tij.co.jp/guidedsupport/jp/docs/supporthome.tsp Japan:
https://training.ti.com/jp

Technical support forums B011617

Search through millions of technical questions and answers at TI’s E2E™

Community (engineer-to-engineer) at
TI.com I 2019
e2e.ti.com
China:
http://www.deyisupport.com/
Japan:
http://e2e.ti.com/group/jp/

Important Notice: The products and services of Texas Instruments Incorporated and its subsidiaries described herein are sold subject to TI’s standard terms and conditions of sale.
Customers are advised to obtain the most current and complete information about TI products and services before placing orders. TI assumes no liability for applications assistance,
customer’s applications or product designs, software performance, or infringement of patents. The publication of information regarding any other company’s products or services
does not constitute TI’s approval, warranty or endorsement thereof.

The platform bar and E2E are trademarks of Texas Instruments. All other trademarks are the property of their respective owners.

© 2019 Texas Instruments Incorporated SLYY168

 IMPORTANT NOTICE AND DISCLAIMER

TI PROVIDES TECHNICAL AND RELIABILITY DATA (INCLUDING DATASHEETS), DESIGN RESOURCES (INCLUDING REFERENCE
DESIGNS), APPLICATION OR OTHER DESIGN ADVICE, WEB TOOLS, SAFETY INFORMATION, AND OTHER RESOURCES “AS IS”
AND WITH ALL FAULTS, AND DISCLAIMS ALL WARRANTIES, EXPRESS AND IMPLIED, INCLUDING WITHOUT LIMITATION ANY
IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT OF THIRD
PARTY INTELLECTUAL PROPERTY RIGHTS.
These resources are intended for skilled developers designing with TI products. You are solely responsible for (1) selecting the appropriate
TI products for your application, (2) designing, validating and testing your application, and (3) ensuring your application meets applicable
standards, and any other safety, security, or other requirements. These resources are subject to change without notice. TI grants you
permission to use these resources only for development of an application that uses the TI products described in the resource. Other
reproduction and display of these resources is prohibited. No license is granted to any other TI intellectual property right or to any third
party intellectual property right. TI disclaims responsibility for, and you will fully indemnify TI and its representatives against, any claims,
damages, costs, losses, and liabilities arising out of your use of these resources.
TI’s products are provided subject to TI’s Terms of Sale (www.ti.com/legal/termsofsale.html) or other applicable terms available either on
ti.com or provided in conjunction with such TI products. TI’s provision of these resources does not expand or otherwise alter TI’s applicable
warranties or warranty disclaimers for TI products.

Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265
Copyright © 2019, Texas Instruments Incorporated