0% found this document useful (0 votes)
15 views69 pages

Design Considerations For AM Transmitters

The document discusses considerations for designing AM radio transmitters, including regulatory requirements, reliability, redundancy, efficiency, and specific transmitter block diagrams and components. Key sections cover design criteria, modulation techniques, transmitter block diagrams, RF amplifier design principles and harmonic filtering.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
15 views69 pages

Design Considerations For AM Transmitters

The document discusses considerations for designing AM radio transmitters, including regulatory requirements, reliability, redundancy, efficiency, and specific transmitter block diagrams and components. Key sections cover design criteria, modulation techniques, transmitter block diagrams, RF amplifier design principles and harmonic filtering.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 69

DESIGN

CONSIDERATIONS
AM TRANSMITTERS
Overview
Design Criteria Power Supply
Modulation Circuit Board
TX Block Diagram Cooling
RF Amplifier Protection
Harmonic Filter
Combiner
Modulator
Exciter
Design Criteria
Early conceptual design starts with knowing a desired RF output power
and knowing the Regulatory Requirement governing the broadcast signal
such as:
– CE RED
– FCC CFR47 Part 73
– Innovation, Science and Economic Development Canada BETS5

These regulations govern items such as occupied bandwidth, spurious


and harmonic emissions, Audio quality, AM noise, FM noise, Electro-
magnetic compatibility, safety.
Design Criteria
Other design considerations include:
– Reliability
– Redundancy
– Dollars per watt ($/W)
– Power density (W/in3 )
– Power conversion efficiency (operating cost)
– Feature set

These considerations can quite often conflict with each other such as
redundancy and $/W.
Designers must understand and choose their guiding principals
Design Criteria
Nautel has always considered reliability one of the leading design
principal and as such has an internal engineering document guiding our
acceptable component stress.

Electrical stress, at its essence comes from two parameters:


– Voltage (Joules/coulomb) stress causing dielectric breakdown and
dielectric heating
– Current (Coulombs/second) stress causing excessive heating of
conductors (P= i2 R)
Design Criteria
Examples of Nautel design guidelines include:
– Resistors not to dissipate more than 50% of manufactures rating
– Capacitors operating voltage to be 66% of rating
– Voltage stress in air to be 5 V/mil (breakdown at sea level occurs at 75
V/mil in a uniform field)
– Transistor junction temperature not to exceed 75% of max rating under
normal operating conditions

These guidelines have helped establish a history of highly reliable


transmitters.
Design Criteria

Reliability and redundancy have also greatly influence Nautel designs.


Going back to our roots, it was a request for improved MTBF that drove
the first solid state designs as the Canadian Coast Guard needed ultra
reliable transmitters in northern Canada which tube designs of the day
could not meet.
Design Criteria

We continue to build the majority of our design redundancy in mind


including:
– multiple parallel amplifiers that can be hot serviced
– dual exciters
– dual low voltage power supplies
– Multiple parallel cooling fans
Design Criteria

Nautel also uses IPC


standards to guide the
construction of all
electronic assemblies.
Generic Transmitter Block Diagram
Antenna

AC Grid Antenna tuning or matching


unit

Transmitter
Control
Transformer Switch Power Supply Filter

Most
Amplifier
transmitters Exciter
fit the model Combiner
Amplifier
to the right
Modulation Design

• Modulation techniques
– Linear amplification
– Single Pulse Duration Modulation (PDM)
– Bi-Phase Pulse Duration Modulation (PDM)
– Multi Phase Pulse Duration Modulation (PDM)
– Pulse Step Modulation
– Pulse Step Modulation with step interpolation
Modulation Design

e = ( 1 + M sin ωm t) . sin(ωc t)

Where

M is modulation index
sin ωm t is modulating frequency
sin(ωc t) is the carrier frequency
Modulation Design
Consider a carrier wave (sine wave) of frequency fc and amplitude A given by:

Let m(t) represent the modulation waveform. For this example, we shall take the modulation to
be simply a sine wave of a frequency fm, a much lower frequency (such as an audio frequency)
than fc:

where M is the amplitude of the modulation. We shall insist that M<1 so that (1+m(t)) is
always positive. If M>1 then overmodulation occurs and reconstruction of message signal
from the transmitted signal would lead in loss of original signal. Amplitude modulation results
when the carrier c(t) is multiplied by the positive quantity (1+m(t)):
Modulation Design

Using trig identities, y(t) can be shown to be the sum


of three sine waves:

Therefore, the modulated signal has three components: the


carrier wave c(t) which is unchanged, and two pure sine waves
(known as sidebands) with frequencies slightly above and
below the carrier frequency fc.
Modulation Design
Envelope Elimination and Restoration
a.k.a. Kahn Technique

Envelope Term

RF Phase Term

X
Modulation Design

DSP Based Signal Generator Transmitter Power


Stage
Envelope Term
Envelope Input
i 2 (t ) + q 2 (t )

RF Phase Term
 q (t ) 
i (t ), q (t ) cos(ω c t + tan −1 
 i (t )
)

RF Input
X s (t )

Final Amplifier
AM Transmitter Block Diagram
AM Transmitter Block Diagram
RF Amplifier
The RF amplifier design is critical to the transmitter’s $/W, W/in3 ,
efficiency and reliability.
Usually, the amplifier will be the building block of a family of
models.
The power capability needs to be large enough to minimize
combiner and connectors costs but small enough not to drastically
affect power capability under failure. Shipping and servicing
weights are also very important.
RF Amplifier
Class D amplification has been the topography of choice since the
early 1980’s. Class D has excellent efficiency (theoretically 100%),
good $/W and W/in3
Choosing the switching device for the full bridge is critical as well.
Drain current, input capacitance, output capacitance, dv/dt rating,
thermal impedance of junction to case, case mechanics are all
considered.
The amplifier is designed always to be working into an inductive
load to minimize the switching loss and dv/dt.
RF Amplifier

Square wave produced by


Class D has a fundamental
and many odd harmonics
RF Amplifier
Amplifier efficiency is affected by 2 factors:
- FET on resistance (i2RDSon)
- Switching loss (1/2CV2F)

RDSon can be minimized by choosing a high current FET

Switching loss can be minimized by reducing the voltage across


the FET during turn on – this is accomplished by an inductive load
also known as zero voltage switching (ZVS)
RF Amplifier
RF Amplifier
Q1
B+

Class D Amplifier

Q2
FILTER

LOAD
RF DRIVE Q3
B+

Q4
RF Amplifier
Design sequence (iterative)

1) FET selection determines B+ (BVdss x derating)


2) Modulation depth selection determines PAV {PAV=B+/(mod+1)}
3) TX TPO and # of RF modules determine power per module
4) System impedance, # of RF modules and combiner turns ratio
determines filter input impedance and PA impedance
5) Filter input impedance determines filter component values

Power = Voltage2/resistance
PA testing and analysis to ensure PA power is reliable
RF Amplifier

Typical values for NX50


1) B+ = 400 VDC
2) PAV = 163 VDC
3) MOD Depth = 145%
4) # of RF modules = 20
5) Filter input Impedance = 4.17 Ω<25° or 4.60 Ω//+j9.87 Ω
6) Primary Turns = 6
7) PA impedance = 8.28 Ω//+j17.77 Ω
8) PA Power = 146 𝑉𝑉 𝑟𝑟𝑟𝑟𝑟𝑟 2 /8.28 Ω = 2574 watts*
9) 2574 watts x 20= 50 kW

*don’t forget square wave to sine conversion (4/π) and peak to


RMS (1/sqrt2)
RF Amplifier
Generalized Harmonic Filter Network - Shunt Notch
Engineering has

Output Impedance
Turns (Combiner)
designed a

@ 20 degrees

Turns (Cube)
Nominal FET
Phase Angle
PA(V) @ FP

Resistance

Secondary

Number of
Veffective

Efficiency
Peak Mod

Total Rac

Modules
(mOhm)

Primary
Rfet*2

Pload
Pout

Pfet

Rac
spreadsheet to

Xl
145.00% 163.27 25.00 133.22 2500.00 50 0.15 55.18 6.80 2500.00 0.98 3.17 7.50 6 1 20 50

Frequency Input Impedance per Input Inputs Power per Input Total Power (kW)

execute the required kHz


1710
rad/s
1.1E+07
R
3.78
+JX
1.76
mag
4.2
phase
25.0
n
1
RMS
75000
Peak
300125
RMS Peak
75 300.125
Carrier
50

calculations
Resistances Elementary 90 Degree T Section Impedances Adjusted Centre Resistance
Resistance Phase First T Second T Series Capacitor Shunt Notch 13.739
Input Centre L1a (uH) Z1 L (uH) C (nF) Z2 L (uH) C (nF) Cs (Ω) L3b (uH) C2n (nF) Ln (uH) Centre Resistance Adjustment Factor
3.78 13.7 0.164 7.20 0.67 12.92 26.2 2.44 3.55 30 2.79 3.16 0.305 1

efficiently
Note: decrease if measured capacitance is > 100% by 0.01 at a time until
Ideal Network (90 Degree T Sections) all inductors are calculated and then decrease by an additional 0.03.
Capacitors (nF) Inductors (uH) Confirm Voltage is less than 66% , Current is less than 70% and Power is
C1 C2 Cs L1 (ttl) L1 (per) L2 L3 Ln less than 66% of rated for all capacitors
12.92 3.16 3.102 0.834 0.83 3.110 5.23 0.3049

Stress Levels - Ideal Network Normalized Stress Levels


Power (kVAR) Current (rms Amps) Voltage (peak) Power (kVAR per nF) Current (rms Amps per nF)
C1 C2 Cs C1 C2 Cs L1 per L1 total L2 L3 Ln C1 C2 Cs C1 C2 Cs C1 C2 Cs
182 205 45 159 83 39 140.9 141 74 39 83 3242 6959 3287 14.1 65.0 14.5 12.3 26.4 12.5

Reduced Q Network Increased Q Network for First T Section


Measured Capacitance (nF) New Required Inductance (uH) Inductance (uH) C1 C1 Stress
C1 C2 Cs L1 (ttl) L2 L3 Ln L1 (ttl) L2 nF kVAR kVAR/nF Irms Vpk
12.00 3.00 3.00 0.75 2.36 3.93 0.32 1.03 3.38 12.00 256.4 21.365 181.81 3989.3
93% 95% 97% 89% 76% 75% 105% 123% 109% 93% 141% 151% 114% 123%

Stress Levels - Reduced Q Network Normalized Stress Levels - Reduced Q Network


Power (kVAR) Current (rms Amps) Voltage (peak) Power (kVAR per nF) Current (rms Amps per nF)
C1 C2 Cs C1 C2 Cs L1 per L1 total L2 L3 Ln C1 C2 Cs C1 C2 Cs C1 C2 Cs
136 161 47 133 72 39 141 141 74 39 72 2910 6316 3399 11.4 53.6 15.5 11.1 24.0 12.9
75% 78% 103% 83% 86% 100% 100% 100% 100% 100% 86% 90% 91% 103% 81% 82% 107% 90% 91% 103%

Loaded Q of L Sections Phase of L Sections +JX Phase


T1-in T1-out T2-in T2-out T1-in T1-out T2-in T2-out C1 C2 Cs L1 L2 L3 Ln Total Watts 427
1.65 0.16 1.68 0.22 59 9 59 13 25 136 161 47 -97 138 63 18 kVar Efficiency 99.43%
1200 1200 1200 1000 1000 1000 500 Q
Total Phase Delay (With Input Term): 165 114 134 39 -97 138 63 36 Watts
Total Calibration Phase Delay: 140
NG200, Zin: 13.4 + J3.7, 300 kW average, 1110 kW peak
Note: Requires series pairs for C2, Cannot use CS45 except for Cs
RF Amplifier

FET chosen for NX


RF Amplifier
RF Amplifier
Harmonic Filter
Harmonic filter needs to impedance match the antenna to the
amplifier to achieve designed power. (4.17 Ω<25° to 50 Ω)

The harmonic filter needs to attenuate harmonics produced in the


amplifiers.

The harmonic filter needs to provide isolation from external energy


which could damage the amplifiers.
Harmonic Filter
Harmonic Filter

L Networks are used to step up or step down impedance


ZS L ZS L

VAC
S ZL VAC
S
ZL
C C Low-pass

(low to high Z) (high to low Z)

ZS C ZS C

VAC L VAC
S
S ZL L ZL High-pass
Harmonic Filter
T networks are back RS
to back L networks X1 X3

AC
X2 RL

Three-element T Network

RS
XS1 XS2

AC
-XP1 R -XP2 RL
virtual
Harmonic Filter
Harmonic filter uses a double Tee with 3rd notch design to reduce
harmonics and impedance match the RF amp to the antenna.
Double Tee has the advantage of :
- keeping impedance matched as shunt C drifts with temperature
- Wide bandwidth
- Good attenuation

A series capacitor and shunt coil provide static and low freq
transient protection as well as transient suppressor device.
Combiner
The combiner needs to efficiently add the RF modules powers
while providing isolation between inputs.

Combiners can be categorized as those with balancing


resistors and those without.

Number of inputs needs to be greater than 4 for combiners


without balancing resistors in ordered for reasonable isolation
under failure.
Combiner
AM, for several generations, has used combiners without
balancing resistors.

AM Combiners have been both parallel (60°) and series.


Ampfet , NX = series
ND, XL, XR, NA= parallel

Series combining is broadband, less expensive, and uses a


smaller footprint
Combiner/Harmonic Filter
Voltage and Current Series Resonant High Pass Section
Measurement for Active
Protection Harmonic Filter
Output
PA_n

Forward and Reflected


Directional Coupler Ports for
RF Amplifiers
Spectrum Verification
PA_2

Adjustable Carbon Ball Gap


Combiner Transformers
PA_1 Calibrated Fast Spark Gap
(Pressurized Tritium)

Frequency Agile: Harmonic Filter Re-Tune in a Few Hours


Modulation
Modulation Design
Design
Pulse Width Modulator Switching
Duty Cycle
Input
Choke
B+ Filter Output to
Power
GND Freewheel Amplifier
Diode

• This circuit illustrates the simple switched modulator principle. It applies a fixed DC
supply voltage B+ through a series switch to a low pass filter.
• The switch is opened and closed at a fixed frequency of approximately 70 kHz
producing a square wave with a peak value of B+ at the filter input.
• The filter rejects the 70 kHz signal and produces a dc voltage at its output equal to
the average value of the square wave
– Example: one half of B+ for a 50/50 duty cycle.
• If the duty cycle is slowly varied, the DC output voltage may be adjusted from 0 to
B+.
Modulation Design
Astable
Origins of PWM Generator Multi-
Vibrator
∫ Comparator Output
to Pulse
Amplifiers
Integrator

Input Signal

• 70 kHz multi-vibrator can be passed to an integrator to produce the required saw-tooth


waveform.
• If an audio modulating signal is superimposed on the control voltage, then the width of the
pulses in the PWM control signal will vary accordingly.
• Waveforms depicting this process are shown on the next slide.
• It is important to understand that both the RF output level and the modulation information is
digitally encoded into the PWM control signal’s pulse width while its repetition frequency and
amplitude remain constant.
Modulation Design

PWM Generator Waveforms

• These waveforms illustrate how the width of the PWM pulses increase in response to a
rising control waveform input.
• Note the constant amplitude and repetition frequency of the PWM signal.
• The saw-tooth must have a very stable DC reference level and a very linear slope to
ensure high fidelity of the transmitter’s modulated output signal.
Modulation Design
AM Design

• PWM signals generated for


Carrier only – no audio applied

• PDM signals generated for


carrier with audio applied
Modulation Design
AM Design

• PWM signals generated for


Carrier only – no audio applied

• PDM signals generated for


carrier with audio applied
Modulation Design
AM Design
Digital Modulation Technology
Separate power processing stages for Envelope Modulation and RF Amplification are
Employed. This approach enjoys the following benefits:
•Optimal RF Transistor switching is maintained at all modulation levels. This cannot be
achieved when the RF transistors must both convert DC to RF and vary the RF envelope.
•Digitally controlled RF Transistor switching has nearly eliminated switching loss across
the AM band
•Allows for very high current capability transistors to reduce conduction loss
•Results in ultra high efficiency RF amplifier (up to 98% at 1710 kHz)
•RF Amplifier DC Supply can be shut down during transient events/VSWR to further
improve robustness
•All amplifiers see the same load at all power and modulation levels
Modulation Design
AM Design
The 9 Modulation phases are separately
synthesized digitally at 317 MSPS in the exciter 9 Phase Digital Modulation
FPGA. This results in very low quantization noise. Modulation Process
1

Each Modulator phase samples the desired 0.5

envelope at a rate of over 300 kHz. 0


0 1 2 3 4 5 6

The 9 Phase process samples the envelope at


-5
x 10
1 Phase Modulator
1

over 2.7 million samples per second. This rate 0.5

does not change with frequency. 0


0 1 2 3 4 5 6
-5
x 10

Distortion due to the modulation process is


9 Phase Modulation
1

essentially eliminated. 0.5

Reduced switching harmonic content allows for a


0
0 1 2 3 4 5 6
-5
x 10

very low Q modulation filter.

45
Modulation Design
AM Design
• Amplifier Design
Modulator Stage RF Amplifier Stage

SiC Rectifiers
Q1 Q3
(x3)
Q5

400 VDC
B+ Supply
Output+
Output-

Q2 Q4
Q6 Modulator
Filter

Digitally Controlled
Modulator Drivers Q7
(x3)
Digitally Controlled RF
Modulator FET Drivers (x4)
(x3)
Exciter
Exciter design has advanced greatly from early TTL and discretes
to a powerful DSP and FPGA.
The 2 signals being generated have not changed: RF drive (carrier
freq) and Mod Drive (PDM)
With the DSP, the exciter can now compensate for many of the
distortions in the amplifiers and modulators.
Modulator pinch-off, modulator filter roll-off and RF amplifier
incidental phase modulation are all corrected.
Exciter
“The NX Series of AM transmitters are the first high power AM
transmitters to be provided with Dynamic Pre-correction”

• Corrects primary forms of distortion


• Applies to analog + digital broadcasting
• More Linear – Clean Spectrum
• Adaptive envelope equalization
• Adaptive AM-AM correction
• Adaptive AM-PM correction
Exciter
Exciter
Pre-Correction Principle

An amplifier characteristic g(x)


may be corrected for with a
complementary characteristic h(x)
Gx
such that g(h(x)) = Gx
g(x) h(x) For this to be true, G h(x) = g-1(x)

x h(x) Gx
h g
Exciter
The FPGA has three correction sections in the forward path:
– Envelope equalization: Corrects for filtering effects in the modulator
(envelope magnitude and phase response versus frequency)
– AM/AM Correction: Corrects for amplitude error in the modulator due to
capacitive effects in the modulator FET. (Essentially AM distortion)
– AM/PM Correction: Corrects for phase error in the RF amplifier due to
capacitive effects in the RF FET. (IQM or IPM effects)

Additionally, it will be possible to correct for linear effects in the AM antenna


system using a filter in the DSP

51
Exciter
Magnitude Response
10
Modulator Filter
5 Equalizer
Composite
0

Magnitude (dB) -5

-10

-15

-20

-25

-30
0 20 40 60 80 100 120
Frequency (kHz)
NX Modulator Response
Exciter

AUI Screen
EQ Frequency
Response

EQ Frequency Response
Exciter
Group Delay (reference to carrier group delay)
2
Modulator Filter
1.5 Equalizer
Composite
1

Group Delay (us) 0.5

-0.5

-1

-1.5

-2
0 20 40 60 80 100 120
Frequency (kHz)
NX Modulator Response
Exciter

AUI Screen
EQ Filter
Delay

EQ Filter Delay
Exciter
No visible distortion in the trough with AM to AM correction

Performance Results - AM
Exciter

AUI Screen
AM-AM
Correction
Exciter

Nearly total elimination of


distortion at 25% modulation
(Distortion approx. 0.05%)
Exciter

AUI Screen
PM-AM
Correction
Power Supply
The AC power from the grid must be rectified so that it can be
switched at the desired carrier frequency. AC/DC power
conversions has two design possibilities:
• 50/60 Hz rectification ‘big iron’
• Switching power supplies
In a nutshell the ‘big iron’ is very reliable and has excellent $/W at
high power (>10 kW)
Switching power supplies offer higher power density (W/in3 ) , lower
weight and higher redundancy
Power Supply

Regulations regarding the AM noise on the output RF signal


would force the low pass filter following the rectifiers to be
excessive.
Nautel patented a ‘lines volt compensation’ technique early on
in its solid state designs that is continued to be used today.
Power Supply
Power supply design greatly affects the performance of the
transmitter within the user’s environment.

Big iron power supplies typically use SCR’s for soft start
and regulation but have power factor levels at approx. 0.93.
Power Supply
When running from generator, the transmitter’s regulation
circuits can make the load appear as a negative resistance
causing instability in the generator’s operation.

Also high power rectifiers can cause ‘line notching’ which


drastically reduces the available power from the generator.
Power Supply
Circuit Board Design
Circuit board creation represents a significant portion of the total
transmitter design effort.
Circuit boards typically use copper cladded epoxy resin (FR-4) but
sometimes employ high quality materials (ie. Teflon) depending on
the application.
FR-4 can operate at approx. 90 C before discolouration and
delamination.
The copper cladding is quantified in ‘ounces’ with each ounce
resulting in 0.0014” thk of copper cladding.
Circuit Board Design
Nautel has strict guidelines for the current that a PWB conductor
may carry versus the ounces of copper.

In high voltage applications, there are rules on creepage and


clearance distances that must be followed for maximum reliability.

Nautel has an internal 100 point checklist that every PWB must
pass before being manufactured.
Cooling
Transmitters employs multiple parallel DC cooling fans
which provide:
• Redundancy
• High efficiency
• Not sync’d to AC line frequency
Protection
Transmitter’s control function provides protection such as:

• SWR protection techniques (cutback, foldback,


shutback)
• Thermal protection – fan tach, heatsink monitor
• Over current protection
• Over modulation protection
Thank You

You might also like