System optimization and terrestrial RFI

It was nice seeing the Satnogs team again at Dayton this year! While there, I purchased a pair of M2 Eggbeater antennas (2M and 70cm) and put them up using a short run of high quality coax. The final run, from the outside entry box, enters a VHF/UHF lightning arrestor, then rides a short run of Times Microwave LMR600 coax into an airspy mini. I didn’t know what to expect, but I’ve been pleasantly surprised with the results. I am not yet ready for a tracking array, and this seems like the best I’ll have for some time. My location, on a ridge, allows me to see some satellites that get no more than 10 degrees over the horizon, which satisfies me in terms of performance when I can hear a satellite.

I still have noise and birdies and have been trying to track that down. Some of it may just be electronics in my home and near the antenna and/or the Raspberry Pi. I don’t think anything is ingressing via the coax, which has very good braid, crimped-on connectors, and is high quality. In an attempt to reduce the noise from the Rpi switching supply, I’ve recently put on a 3A Astron linear PS (variable voltage) set at 5.1V and removed the wall wart.

I know some of you have had great success in eliminating local terrestrial noise, but is there a point beyond which I’m unlikely to clean things up? I am slightly OCD when it comes to receiver performance, and I don’t want to endlessly pursue a receiving state that is impossible or very cost prohibitive.



nice to hear from a fellow OCD’er (;
Based on your description (would be nice to see some pics of the setup), I would look at grounding the antennas and RF choking the inside parts from the outside, thereby limiting the conducted noise that reaches the antenna via the coax.

The PI (and computers as well) can be pretty noisy, and conducted noise is quite hard to eliminate. First putting the PI in a metal case, then everything in a metal enclosure and mounting everything on a common plate. Choking on the USB and coax to try and eliminate what is possible in every step.
Have a look at my station image to se a implementation of this. Some things have changed since it was taken, but essentially the same.

Up at the antennas I would recommend a preamp box, either like the mast mounted preamps or a metal enclosure for LNA’s, with N connectors galvanically connected to the chassis. Preferrably grounded via the metal post/pipe down to a dedicated ground rod. Choking on the coax that goes to the receiver to isolate any noise coming from that direction.
Depending on the RF situation at your location, filters might improve on the reception, hooking a spectrum analyzer up to it and looking for strong signals that can be eliminated, like the ~100MHz can be pretty wild on a VHF antenna. Using preamps with BPF is the easy way to deal with this. Or building your own with commonly available modules like the ones from Mini-Circuits and the like. Filters can be placed in front of the receiver as it’s way easier to play with different configurations down in the shack (:

1 Like


Thanks for the reply. Sorry for the delay in replying.

I have done some of the things you mention. I have my Pi on an Astron linear variable voltage 3A power supply set at 5VDC, which is much cleaner than the Raspberry Pi “official” power supply, but it doesn’t seem to have helped much. I have toroids on the USB cable, and all of my coax is fed into a junction box on the outside of the house, where it enters lightning arrestors and has a direct 4 inch wide copper strap to a 12 foot ground rod. I think coupled noise is pretty well eliminated, even common mode. I don’t have a pre-amp at the antenna, and have considered dual 2M/440 preamps but with just a 50 foot run of high quality coax, I am not sure what that is going to buy me over a pre-am at the receiver.

Good advice on the spectrum analyzer, and I am going to do a local survey and see what the terrestrial noise sources are, and try to characterize them.

As far as filters, I have a GPIO Labs FM filter in line before the pre-amp, and it is very effective. It delivers a full 10dB more attenuantion than the Nooelect Flamingo I was previously using, and I’m quite pleased with it.

I’l also look into putting my Pi in a metal case, because right now I have it with several other Pi’s, naked, in a 3D printed Pi Rack (Printables)

Thanks again.