I have my Raspberry Pi B mounted to a board with a ground plane antenna and am powering it with a USB power pack. Its portable and I can stick it outside and do some antenna experiments. I am working on a weather proof housing and long term power setup.
I was wondering how many people are powering their setup with ethernet POE. From what I have read it appears that I need a 48V POE injector and a POE splitter to 5V with a micro usb plug. The Raspberry Pi is the B version and not the B+. The network cable length will probably be about 50 ft from the power injector to the power splitter. The communications will be thru wifi and am only using the network cable to power the board.
I can safely say that many ground stations on the network are PoE powered (I own 3 of them like this). There are multiple options for PoE powering your RPi. 48V injector and PoE splitter that has an integrated 5V power supply is the easiest most universal way to go, but in case your board is a 3B+ and latest you can go for the PoE Hat directly.
That’s well within specs. Just make sure to crimp well and test the cable
Any recommendation on size of POE in watts. I am using a Raspberry Pi 3B and not a B+.
48V at 0.5A should be totally fine.
If you’re placing your RPi anywhere near your antenna and powering it via PoE, I would highly recommend doing testing to see if the PoE system is resulting in an increased noise floor. Same goes even if you are powering your RPi via some other means.
I have tested a few of the cheaper PoE injector/splitter units and found that they produce considerable noise on both 2m and 70cm. It is quite likely that many SatNOGS stations are not performing as well as they could be due to use of these kinds of splitters. Putting everything into a metal box may help, but there is still the possibility of noise coupling between PCBs within said box, and even through power rails.
Unfortunately, once installed, it’s not like you can just cut power to the PoE splitter to do an A/B test - as you are powering the RPi from this!
A ‘simple’ test would be to hook the in-situ antenna and SDR up to a (battery powered) laptop, and observe the band of interest. Turn the PoE injector and RPi on and off, and watch for changes in the observed noise floor. If it rises at all, then your station is equipment-limited.
From my experience tuning my three SatNOGS stations, physical separation is the best solution. If you’re running ethernet cable from your station to somewhere else, then I would suggest placing a preamp in a box near the antenna, and running coax instead. It doesn’t matter if the coax isn’t ultra-high-quality - as long as the preamp gain overcomes the cable loss, the system noise figure will not be degraded significantly, and system noise figure is what determines your stations sensitivity.