so ive been messing around with this for a few weeks now and i think im overcomplicating it. basically i have an old yaesu g-450a rotator and i want to drive it from an arduino instead of the manual controller. the idea is to eventually tie it into my logging software so it just follows the beam heading automatically when i work a dx station.

i got the basic H-bridge motor driver wired up and i can turn the thing but the feedback pot is giving me grief. the voltage divider math isnt matching what im actually seeing on the analog pin, and i think its because the pot in the rotator isnt a clean linear taper. or maybe im just reading it wrong. ive got a 10-bit ADC so theoretically i should have plenty of resolution but the readings jump around like crazy especially when the motor is running.

has anyone done something similar with a raspberry pi instead? i was thinking maybe a pi would be easier since i could run hamlib directly on it and just write a python script. but then again the arduino response time is probably better for the actual motor control. maybe both? pi for the logic and arduino as a hardware interface? i dont know, starting to feel like this rabbit hole has no bottom.

yeah done almost exactly this, though i used a different rotator. the jumping ADC readings when the motor runs is almost certainly noise on your supply rail coupling into the analog reference. throw a 100uF cap across your Vcc and GND right at the arduino and a smaller ceramic like 100nF close to the AREF pin if you're using an external ref. also if your motor driver and arduino are sharing the same supply that's gonna cause you problems, run them off separate regulators even if they share a common ground.

the pi plus arduino combo is honestly what i ended up with too. pi runs hamlib and a little flask server that talks to my logging software over the network, arduino handles the PID loop for the motor and just accepts a target azimuth over serial. works really well once you iron out the noise stuff. the latency between the pi sending a command and the antenna actually moving is maybe a second or two which is totally fine for HF work.

not done the rotator thing specifically but i built a raspberry pi based rig controller for my ic-7300 and hamlib on the pi is pretty solid once you get the udev rules sorted out for the USB serial device, otherwise it renumbers on you every reboot and your scripts break. just something to watch for when you go down that road.

also on the pot linearity issue — some of those old rotator pots are genuinely worn out and you'll never get clean readings no matter what you do. might be worth measuring the resistance across the full sweep with a meter before you spend too much time on the software side of it.