Well, it’s settled. I have a major antenna switch RFI ingress problem. As much as 30 dB on some frequencies. Sigh.

SDR Community provides a really nice Spectrum Analyzer application for use with its RSP hardware. I thought this would be a great tool to use in place to check the output signals from homebrew antenna switch.

Shown above is my worst case noise effecting HF reception. My white trace is the switch output with the relays OFF, i.e. no output. My yellow trace is with the output relay switched ON. In both cases, there is no antenna connected to the switch input. Just a 50 ohm terminator.

With the spectrum analyzer sweeping a 20 kHz chunk at 9.5 MHz, you can see the OFF trace at around -130 dBm, which is roughly my noise floor in the receiver. On the other hand, my yellow trace is around -102 dBm with the relay turned ON.

This means my antenna switch RFI ingress is around 30 dB of added noise. Not good. In fact, very bad. My situation is better on other frequencies, but in all cases, at least 15 dB of added noise.

What’s really interesting is that the added noise completely changes the phasing characteristics of my dual channel receivers, and makes deep nulls of local noise impossible.

Antenna Switch RFI Ingress Source

Now, I have to find the antenna switch RFI ingress source, and see if I can mitigate the problem. My first guess is the Arduino microcontroller. My second option is local power supplies in the shack. Or both.

Time to take the board back to the lab. I will try bypassing the relay modules. I will try controlling the relay modules from a separate power supply with the Arduino disconnected. If the Arduino is the problem, I guess I can use some rotary and toggle switches to control the relays. Stay tuned.