so ive been messing around with a raspberry pi zero w for the past few weeks trying to get it to control my old Yaesu G-450A rotator and honestly its been equal parts fun and maddening. the rotator itself is fine, its the interface side that keeps biting me. i wrote a little python script that reads the azimuth voltage from the rotator controller through an ADS1115 ADC over i2c and its mostly working but i keep getting these weird jitter spikes where the reading will just randomly jump like 20-30 degrees for a sample or two and then settle back down. ive got some filtering in the code but its not totally cleaning it up.

the stepper driver side is actually working pretty solid, using an L298N module to drive the rotator motor direction and its responding to the az commands from hamlib rotctld which is running on the same pi. wsjtx on my main machine can talk to it over the network and it does actually point the antenna where i tell it to, most of the time.

i think the jitter might be electrical noise getting into the ADC lines since the motor and the signal wires are kind of all bundled together in the cable run i did. gonna try some shielding or maybe just separate the runs. anyone dealt with this kind of noise pickup on i2c ADC stuff in a shack environment? the RF alone in there is probably not helping either.

yeah the i2c noise thing is a real pain, especially if you have any length on those lines at all. i2c really doesnt like long cable runs, like even a meter or two can start causing issues depending on the capacitance. but if its specifically correlated with when the motor is running id look at the power supply first before anything else. are you powering the pi and the L298N from the same supply? because the motor switching transients can absolutely get back onto the power rail and from there into your ADC reference. try throwing a big electrolytic cap across the motor driver supply and a smaller ceramic right at the ADS1115 VCC pin and see if that helps at all. also make sure your grounds are all tied together at one point, star grounding basically, instead of daisy chained which causes all kinds of weird loops.

the RF environment thing is worth thinking about too. i had a similar setup a while back and adding ferrite beads on the i2c lines and on the power leads to the pi made a noticeable difference. just small clip on ones, nothing fancy. also might want to add some pull up resistors on SDA and SCL if you havent already, the internal ones on the pi are a bit weak for anything beyond real short traces.

not sure if this helps but ive been using a nano instead of a pi for rotator stuff and just letting it handle the low level ADC reading and motor control, then it talks to the pi over serial for the hamlib side. kinda separates the concerns and the nano is way more tolerant of electrical noise in my experience, the atmega ADC isnt perfect but its less fussy about the i2c stuff. might be overcomplicating things for your setup though depends how much you want to tinker with it lol. the wsjtx integration part sounds like it's working better than mine ever did honestly