so ive been messing around with this for a few weeks now and i think im overcomplicating it honestly. the idea is pretty simple — i want the arduino to read azimuth from my yaesu g-800 rotator controller (tapping the voltage on the pot output like everyone does) and then compare it against a target bearing i set, then drive the rotator relay until it gets close enough. like within 3 or 4 degrees is fine for HF.

i got the analog read part working no problem, voltage divider is set up and im getting sensible numbers out of it. the part thats killing me is the dead band logic. like when do i tell it to stop? i had it overshooting constantly last night and i couldnt figure out if it was the loop speed or just my relay debounce being garbage. tried adding a delay but that felt hacky.

also wondering if i should just scrap the arduino and throw a raspberry pi at this instead. pi has more headroom for eventually adding like a web interface or whatever but feels like overkill for what is basically just a comparator with relays. anyone gone either route and have thoughts?

yeah i did almost exactly this about two years ago with a nano and a g-450. the overshooting thing drove me insane for a while. what fixed it for me was not running the relay continuously — i basically implemented a really crude PWM-ish thing where if the error is large i run the relay solid, but inside like 10 degrees i start pulsing it in short bursts and letting the rotator coast. rotators have a ton of mechanical momentum and if youre just slamming the relay on until you hit the target youre always gonna overshoot. the dead band alone isnt enough.

as for pi vs arduino, i ended up sticking with the arduino for the actual control loop because its deterministic and i didnt want linux scheduling messing with timing. but i did add an esp8266 alongside it for the wifi and web interface piece. best of both worlds kinda. the pi is great but it feels wrong using it just to toggle some relays, you know?

the momentum thing is real, especially on the bigger rotators. one thing worth looking at is whether your controller has a brake — some of them engage a mechanical brake when you release the control voltage and that changes the coasting behavior a lot. might be worth checking the g-800 schematic to see whats actually happening when your relay opens.

i havent done a rotator project specifically but ive done a bunch of arduino stuff for shack automation, band decoder for my amp, that kind of thing. for anything where timing is sloppy and mechanics are involved i usually just tune the dead band empirically — like start at plus or minus 5 degrees and keep widening it until the hunting stops. its not elegant but it works. you can always tighten it back up if you need better accuracy later.