0
\$\begingroup\$

I'm trying to design a low-noise amplifier circuit for the 8-20 MHz range. The circuit will be used to amplify received signals from an antenna, while blocking signals when the antenna is transmitting. I've managed to get an initial design with a single amplifier working, but I've been unsuccessful in my attempts to expand upon this design with two LNAs. Any help or guidance would be greatly appreciated.

The Design

The basic design I want is an RF path of Limiter -> Low-pass filter -> Amplifier -> Switch. The input signals to the board are +15V, GND, and a differential signal to control the switch (TR signal). My initial design that worked as expected is shown below:

v1 circuit schematic v1 circuit schematic

v1 circuit layout v1 circuit layout

To build off this, I modified the RF path for v2 to be Limiter -> Low-pass filter -> Amplifier -> Variable Attenuator -> Amplifier -> Switch. The goal is to have higher amplification, and use the attenuator to tune it to a value I like. There were several other auxiliary circuit design changes I made as well, but the basic RF circuitry remained very similar. This new design is shown below:

v2 circuit schematic v2 circuit schematic

v2 circuit layout v2 circuit layout

The RF components I chose to use are:

  • Limiter: RLM-521-2WL+
  • Filter: SCLF-21.4+
  • Amplifier: TSS-13LN+ - This is a slightly different amplifier than the one in v1. I will discuss this in my testing. The circuit controlling the amp is the manufacturers recommended circuit.
  • Attenuator: DAT-31R5A+
  • Switch: MSWA-2-20+

The Problem:

The circuit doesn't amplify as much as I'd need or expect. Here's all the testing I've done so far:

  • I checked all electrical connections and made sure all components on the board were mounted correctly.
  • I tested the circuit as a whole and found an amplification of only ~10 dB (I'd expect around 50 dB for two amps)
  • I verified that the TSS-13LN+ amplifiers did indeed work by plugging them into the v1 circuit board and using a bench-top power supply to give the required voltage. In this setup, the amplification was ~24 dB (as expected) and the current drawn by the amplifier was 150 mA.
  • I isolated the amplifier circuit from the rest of the board by:
    • Assembling only the amplifier control circuit on a bare board
    • Providing the 5V supply directly from a bench top power supply (with more than ample current supply)
    • Providing the TR logic signal directly from a bench top power supply
    • Connecting the VNA to either side of the amplifier using the RF test points
  • While isolated, I found the amplifier gain to be 10 dB (expected 24 dB) and current draw of 100 mA (expected 150 mA) while in the "on" state. This was the case for both amplifier subcircuits on the v2 board.

My Conclusion:

From my testing, I concluded that there is a flaw in the circuit layout for the amplifiers. The amplifiers aren't amplifying because they aren't drawing enough current, but I don't know why this is.

Does anybody have any clue as to what could be causing this problem, or what I can try/do to fix it? Any help or guidance at all would be greatly appreciated! Thanks!

\$\endgroup\$
2
  • \$\begingroup\$ A block diagram of both designs would help me understand the schematics. Also, where does the antenna connect? \$\endgroup\$
    – SteveSh
    Dec 16, 2023 at 2:03
  • \$\begingroup\$ Where was the attenuator during your "isolating the amplifiers" tests? And are you sure you're controlling the attenuator properly? \$\endgroup\$
    – SteveSh
    Dec 16, 2023 at 2:39

1 Answer 1

1
\$\begingroup\$

Wow, that's a pretty dense board for RF! I would have expected you making it a bit easier for yourself by spreading things out a little further. Bonus: keeping ~30 board thicknesses between switch mode power supplies and analog circuitry definitely helps make sure the fields from the fast-switching paths don't make it into signal lines. You of course still have the noise on the generated voltage, but as you demonstrate in your schematic, that can be cleaned up with enough capacitors and linear regulators. However, I'd recommend you use a less ancient regulator than the 7805 series; modern ICs have significantly higher psrr than that historic silicon!

That being said, design mistakes:

  • VG in your amplifier needs 0 or 5 V for control, even on 3.3 V supply (see data sheet) Your U10 only has a 3.3 V supply, but controls that line

  • your v1 actually had a linear regulator for the 5 V rail. I'd recommend you reestablish that, or you will feed switcher noise to the enable pin of your amplifiers. Behavior of amplifier with noisy VG is unspecified! This is my prime candidate for what causes the lack of gain. Try with an external clean 5 V to verify!

  • you don't need to invert TR using U10 at all! Just swap the in the differential input on u4, and you eliminated unwanted voltage reduction to 3 V

  • while I'd say separating the power supply, digital and analog sections spatially generally will improve emi and signal quality, it's a nice thing to have, and things would usually still work without. However, you're literally routing lines running straight to your amplifiers under the main inductor of your smps. That actually constitutes a design rule violation! Don't do that, you're essentially making that TR a secondary winding in an air-core transformer. Since this is TR, you might be pumping that such that it becomes unsanitary at the input of your inverter. Check with a scope!

  • but I think you already ruled out most of the above, so my next best guess is unwanted negative feedback from the output of the second to the input of the first. Insert a clean tone, and probe your board with a high-z probe to see at which control signal or supply you see what you shouldn't see!

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for the detailed response! I checked what you suggested, but I didn't find anything unexpected. However, I am unexperienced using oscilloscopes and may not be testing effectively. Do you have any tips or references I could use in testing for the unwanted negative feedback? I also used the oscilloscope to check the output waveforms, and found that the amplifiers are distorting the signal. The distortion appears in both v1 and v2 circuits, to differing degrees depending on the input signal strength. The input signals were within the amplifiers operating range. Thanks again! \$\endgroup\$
    – Theo
    Jan 8 at 17:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.