RF-Amp

From Land Boards Wiki
Jump to navigation Jump to search

RF-Amp P1943-720px.jpg

RF Amplifier Features

  • Class A (Common Emitter) Amplifier
    • Emitter resistor bypassed with capacitor for higher AC gain
    • Transformer coupled input/output for impedance matching
    • Capacitively coupled input
  • Single 2N3904 NPN transistor
  • Applications
    • IF Amplifier
    • Antenna Amplifier
    • Microphone amplifier (with minor modifications)
  • As RF Amplifier
    • Measured Gain Bandwidth (GBW) of 150
    • Measured Gain @7 MHz, +25.2 dB at 12V
    • Measured Gain @9 MHz, +24.7 dB at 12V
    • Measured Gain @30 MHz, +13.8 dB at 12V
  • Input connectors: SMA, BNC, or direct solder coax
  • +12V nominal power
    • 12-14 VDC
    • 5 mm terminal block for power
  • 49x49mm card
  • 4x 4-40 mounting holes

Source

RF Amplifier Design

Schematic

RF Amp Schematic-4.PNG

LT Spice Simulation

As Built - Rev 1

RF-AMP-LTSPICE XFMRS.PNG

Additional Resistor

  • Insert 4.7 ohm resistor to emitter bypass capacitor
    • Reduces maximum gain
    • Increases gain over 1-30 MHz bandwidth
    • Possible better for Antenna Amplifier application at lower frequencies
  • +22.4 dB at 9 MHz

RF-AMP-Rev2 LTSPICE.PNG

Charlie Morris Design Calculations

Beta DC

  • Geometric mean min/max beta at operating current
    • =sqrt(100*300) = 173

Beta AC

  • Gain bandwidth product divided by operating frequency
    • Assume operating frequency of 9 MHz (IF frequency)
    • = 300/9 = 33.3

DC Operating Point

  • Max HFE RF gain at CE current of 10 mA
    • If Vce = 6V, this is 60 mW power dissipation
  • Assume Ve (voltage across emitter resistor) = 1/10 Vcc = 12V/10 = 1.2V
    • R3 is Re (emitter resistor) = 1.2V/0.01A = 120 ohms
  • VCE = 0.7V (typical from data sheet)
  • V(emitter) at 10% of Vcc rule of thumb = 1.2V
  • V(base) = V(emitter) + VCE = 1.9V
  • Base current is collector current divided by Beta DC
    • Biasing resistors = 10x current needed by base current
    • 10 mA in C-E, beta DC less = 10 mA/173 = 58 uA
    • 10x the current in the biasing resistors = 580 uA (calculated)
  • R2 is 1.9V at 580 uA = 3.29K use 3.3K
  • R1 sources current to R2 and transistor base
    • Voltage = Vcc (12V) - 1.9V = 10.1V
    • Current = 577 uA + 58 uA = 635 uA
    • R1 = 10.1 / .635 mA = 15.9K, use 15K

Measured DC operating point

  • Measured with no input
  • Vcc = 11.96V
  • Current draw = 12 mA
    • Quick test for wiring and more or less correct parts
    • Expected 11 mA - close enough
  • +BUFF = 11.84V
    • 0.12V which is 12 mA through R4 10 ohms - expected
  • V emitter = 1.41V
    • 1.41V/12 Ohms = 11.75 mA close to 12 mA total measured current
  • V on input divider = 2.06V
    • Vbase + 0.7V - close
    • Measured Vbe = 2.06-1.41 = 0.65 - close

Input resistance

  • Xc for 0.1uF cap from emitter to ground
    • C=0.1uF
    • F=10MHz
    • 1/2*pi*F*C = 0.16 ohms
  • Parallel resistors R1, R2 paralleled with transistor input impedance
    • R1=15K, R2=3.3K
    • Transistor resistance = Beta AC (33.3) times re
      • re = 26 / Ie (10 mA in mA) = 26/10 = 2.6
      • SSDRA uses 25 as constant - close enough
        • 26 comes from Ebers-Moll approximation
      • Beta AC * re = 33.3*2.6 = 83.3 ohms - predominates
    • All in parallel are 80.8 ohms

Gain calculation

  • Approximation
  • Ic = 0.01A
  • Rc = 200
  • Vrc = 2V
  • Gain = Vrc / vt
    • vt = 26 mV at room temperature
    • Gain = 2V / .026V = 79.2 V/V
    • Gain = +37 dB

Input/Output Transformers

FT37-43 10 Turns.PNG

Tracks

RF-Amp-tracks.PNG

Input Transformer

  • Input Transformer (T1 on Charlie's - T2 on this board)
  • Need to calculate turns ratio
  • 50:80.8 Ohms
  • n = sqrt(Zout/Zin)sqrt(80.8/50) = 1.27 turns ratio
  • Turns choices
  • Minimum number of turns
  • Rule of thumb - want Xl (coil impedance smallest value) to be least 4-5X the load
    • Load = 80.8 ohms
    • 5 * 80.8 ohms = 404.2 ohms minimum
      • More turns = larger capacitance and drops bandwidth
    • Toroid is FT37-43
    • From Toroid page
      • Xl = 404.4 at 9 MHz is 4.5 turns, round up to 5
    • Try nearest integer numbers turns ratios
      • 5:6 = 6% error
      • 6:8 = -4.6%
      • 7:9 = -1.1% << good choice
      • 8:10 = +1.7%
      • 9:11 = +4.0%
      • 10:13 = -2.19%
  • Use 7:9 turns ratio for optimal input transformer

RF-Amp-T2.PNG

Output Transformer

  • Output transformer (T2 on Charlie's - T1 on this board)
  • T2 - different than Charlie's design since my Crystal filters are all 50 ohms in/out
  • SSDRA suggest presenting 200 ohm load to the collector
    • Can't find reference in SSDRA
    • Reflecting back 50 ohms load to 200 ohm collector...
  • 200:50 ohms
  • n = sqrt(200/50) = 2.0:1 turns ratio
  • 10:5 turns
    • 10 turns primary (on transistor collector)
      • 10 turns = 35 uH
    • 5 turns secondary (towards output)
      • 5 turns = 8.75 uH
    • 15 turns = 9.5 in

RF-Amp-T1.PNG

Charlie's Notes

IF Amp 0046A.jpg

IF Amp 0046B.jpg

IF Amp 0046C.jpg

IF Amp 0047A.jpg

IF Amp 0047B.jpg

IF Amp 0047C.jpg

NanoVNA Measurements

  • Goal: Measure RF-Amp performance using a NanoVNA running NanoSaver software on PC
  • S21 (gain) needs to be measured with a 40 dB attenuator on input to RF-Amp to avoid compression on the output
  • S11 (reflection) input impedance can't be measured with input 40 dB attenuator because S11 just ends up measuring the attenuator
    • Output should be terminated to 50 ohms for S11 measurement
  • DC current = 12 mA

Measure S21

  • Put 40 dB attenuator on RF-Amp input, measure S21 at output
    • NanoVNA provides 50 ohm load to RF-Amp to properly terminate output
  • Measure S21 with 9:11 input transformer
    • S21 @ 100 KHz = -8 dB dB
    • S21 @ 1.45 MHz = 35.4 dB (peak gain)
    • S21 @ 9.1 MHz = 24.3 dB
    • S21 @ 16 MHz = 20.1 dB
    • S21 @ 30 MHz = 12.7 dB
  • Peak gain justifies use of 40 dB attenuator to protect NanoVNA

RF-Amp S21 40dBAttenInput 1-30MHz.png

LTspice vs NanoVNA

  • LTspice simulation was pretty similar to NanoVNA results
    • -10 dB at 100 KHz
    • +32 dB at peak
    • Lower output at higher frequencies

RF-Amp S21 LTspice-vs-NanoVNA 1-30MHz.png

Measure Input Compression

  • Is there compression if the NanoVNA drives the input directly?
    • Test by driving directly from NanoVNA set to CW = 9 MHz
    • Measured output with scope - not clipped at 9 MHz
      • Approx. 1Vpp input = +22.1 dBm gain which matches the S21 with the attenuator on the input
      • Vpp = 12.4V with 50 Ohm load resistor
    • Starts clipping at 7 Mhz and down
  • Therefore, can measure input impedance at 9 MHz
  • Other evidence of compression
    • Compare S21 gain with no input attenuator, put external 40 dB RF Attenuators on output of RF-Amp to protect NanoVNA input
    • S21 shows lower gain in lower frequencies so clipping/compression is happening
    • Was: 35 dB at 1.4 MHz
    • Is: 23.1 dB at 1.5 MHz
  • Due to compression can't accurately measure lower frequencies with attenuator at output
  • Compression below 7 MHz matches what was on scope

RF-Amp S21 40dBAttenOutput 1-30MHz.png

W2AEW S11 Measurement Method

  • Can't drive the RF Amp directly from the NanoVNA
    • High output level from the NanoVNA overdrives the RF Amp
    • W2AEW provides a way of driving the RF Amp card without overdriving and still measure S11

  • Calibrate NanoVNA using External 30 dB Attenuator
  • Scan 1-30 Mhz
    • Overdriven at 1 MHz which "swamps" the RF Amp
  • Re-calibrated at 1.5-31.5 MHz
    • Peak gain at 1 MHz = 32 dB
      • Does not overdrive the Amp or NanoVNA
      • Downsize is a lot of noise in the return loss
  • Tested two units
    • Unit 1 has a 7:10 input transformer (T2) ratio
    • Unit 2 has a 7:9 input transformer (T2) ratio

Unit 1

  • 9 MHz measurements
    • VSWR = 1.172
    • S11 (Return Loss) = -22.014 dB
    • S21 (Gain) = +23.624 dB

RF-Amp W2AEW S21 1-30MHz.png

RF-Amp W2AEW S11 1-30MHz.png

Unit 2

  • 9 MHz measurement
    • VSWR = 1.182
    • S11 (Return Loss) = -21.565 dB
    • S21 (Gain) = +24.656 dB
  • 20 dB gain at 15 MHz
    • Gain Bandwidth (GBW) = ~150
    • GBW is a good predictor of gain at particular frequencies
    • Calculated Gain of 14 dB at 30 MHz - measured at +12.8 dB
    • Measured at +26 dB at 7 MHz

RF-Amp U2 W2AEW S21 1-30MHz.png

RF-Amp U2 W2AEW S11 1-30MHz.png

Compare RF Amp vs Kits and Parts Amp

Modified to use as Microphone Amp

MicAmp P1946-720px.jpg

  • Charlie's video

Charlie's Schematic/Calculations

Mic Amp 1A.jpg

Mic Amp 1B.jpg

Mic Amp 2.jpg

Test with Electret Microphone

MicAmp P1947-720px.jpg

MicAmp P1949-720px.jpg

  • Charlie assume voltage/current - didn't measure
    • Chose to determine Electret operating point through measurement
  • DC powered
  • AC coupled output
  • 13.8VDC (max) power
  • Attach decade resistor box
  • Adjust resistance to get 4V across mic at 13.8 VDC supply
  • Selected value = 33K pullup to 13.8V gets 4V across mic
  • 2.5V out with 12V supply

Wiring up Mic to Amp

KY-037.jpg

  • Install Electret Condenser Microphone on small perf board
  • Cable using 18" RG-174 coax to input of RF Amp card

MicAmp P952-720px.jpg

Schematic Mods

RF Amp Schematic-MODS.png

Part Value Changes

  • No transformers
    • Transformers replaced by passives/jumpers
  • R1 - 15K = OK
  • R2 - 3K (small difference vs 3.3K on RF Amp)
  • R3 - 120 = OK
  • R5 - 50 ohms
    • Install R5 to simulate balanced modulator 50 Ohm load
    • Install R5 on long leads to easily remove
  • 33K pullup to bias Electret mic
  • T1 primary winding - 560 ohm
  • C1 - 0.1 uF
  • C2 - 10 uF
  • C3 - 47 uF
    • Add capacitor from Vc point (transistor collector and 560 ohm resistor) to T1 output side
  • Install output SMA connector

RF-Amp MICAMP.PNG

Tested

  • In application output goes to Balanced Modulator
    • Output level should be +7dBm for ADE-1 Mixers
  • Tested into AudioAmp386 - works

Mic Amp LTspice Simulation

  • Low frequency response can be improved by increasing the value of the emitter bypass capacitor

Mic Amp LTSpice Sim.PNG

Video

Assembly Sheet