Homebrew SDR Receiver

I bought a PCB carrying a TQP3M9009 LNA with a view to using this as the initial gain stage of my receiver. This cost me about £8 from eBay.

I figured the first thing to do with this was to make some basic measurements using my LiteVNA. To avoid having the output power of the amplifier damage the VNA I performed calibration and all measurements with a 20 dB attenuator on port 2. For posterity, I note that I powered the PCB directly from a 5V USB charger.

Here's a plot showing S21 and S11:

This seems about right to me — the measured gain is pretty similar to the gain reported in the part's datasheet. Here's a similar plot focused on the broadcast FM band:

This shows an obvious step in gain of around 1.7 dB at 100 MHz. I don't see a similar step in the S-parameters in the datasheet, so I don't think this is expected. (I wonder what causes it? Update: I think I've figured it out. See this log for details.)

That aside, it looks like this gives me 23-25 dB of gain across the whole band, with an input that's kinda matched to 50 ohms. So this might be good enough for some initial testing, though I do wonder if the step in gain in the middle of the broadcast FM band might cause problems later.

One outstanding design question is "how much gain does the receiver need?" In order to figure that out, it will be useful to know how much signal a broadcast FM signal might generate at the antenna. I figured I could estimate that by sticking an "antenna" (random wire) on my scope, setting the input to 50 ohms, and looking at the FFT of the broadcast FM band. So I did that, and here's what I saw:

The peaks here correspond to stations that the FM radio on my phone finds, so I'm confident I'm looking at signal.

The strongest signals have a magnitude of around -70 dBV RMS and weaker signals more like -90 dBV RMS.

I'll need to see something like 0 dBV RMS (about 2.8V peak to peak) at the ADC, meaning I'll need around 70-90 dB of gain.

While this is a crude estimate, it at least gives me a general sense of what I need to aim for, which is good enough for now.

In this project I'm attempting to build a homebrew SDR receiver. My initial goal is to be able to receive broadcast FM. I expect the result to be a janky prototype, rather than a polished product.

The high-level hardware design will follow the phasing direct conversion approach:

I have ideas about a few of the details (e.g. I plan to use a Si5351 for the oscillator), but much remains to be worked out. My general approach will be to build and test the individual stages in isolation to help refine the design.

I'm currently imagining that I'll stream the sampled I/Q to a PC and do the DSP there. I'd like to write the DSP code mostly from scratch, even if it's crude, but we'll see what happens...