so ive been messing around with this for a few weeks now and i think im finally getting somewhere. basically i have an old prop pitch motor running my yagi and the controller that came with it died years ago and i was using a manual relay setup which was honestly terrible. so i picked up an arduino mega and a couple of current sensing boards and started writing some code to handle the position feedback from the pot on the rotator shaft.

the pot gives me a 0-5v sweep across the full 450 degrees of rotation and the arduino reads that fine on the analog pin, the problem im running into is the motor has a ton of inertia and by the time i tell it to stop it overshoots by like 8-10 degrees which is annoying when youre trying to nail down a specific azimuth. ive tried adding a braking delay in the code but it feels janky. wondering if anyone else has dealt with this and whether a PID loop is actually worth implementing or if theres a simpler fix. also considering tying a pi zero in to give it a web interface so i can control it from the logging software but thats a future problem

yeah PID is absolutely the right call for that inertia problem, i did basically the same thing with a green heron style clone i built and the overshoot was driving me crazy until i tuned a proper PID loop. the derivative term is what really helps with the braking — it kicks in as youre approaching the target and starts slowing the motor down before it gets there. took me a while to tune the Kd value right but once i got it dialed in the thing parks within 2 degrees consistently. theres a decent arduino PID library already written that works well, dont bother writing it from scratch.

one thing i also did was add a second slower speed relay so when its within like 15 degrees of target it switches to a slower creep speed and then the PID just handles final positioning. makes a huge difference with heavy beams.

the pi zero web interface part is really easy honestly, i run hamlib on a pi 3 and just pointed my logging software at it over the network. rotctld handles the whole thing. if your arduino can talk serial to the pi you could probably just write a simple python script that reads the position and accepts commands and then exposes that through rotctld. thats how mine works anyway. the hardest part was figuring out which rotator model to spoof in hamlib but there are generic options that work fine