Some random thing happened and I thought it might be good to write this down for the community.
I’ve been working on a 69mm 2S brushless prototype machine for a week now. FrSky R-XSR, TBS Unify Nano, 1.2g camera, 16x16 F4 stack, 1102 11500KV motors.
It had trouble from the beginning - motor 4 was randomly twitching. First we suspected a loose motor wire, so I redid all the wiring. It didn’t help it. I then swapped the 4in1 ESC and the machine started dipping to the right (turns out the replacement ESC had bad FETs). I then moved all 24 FETs from the one ESC to the other and made sure to move the FETs from ESC 4 to another one. After reassembly (new ESC, old FETs) the problem with motor 4 reappeared. Did I mention that I waterproofed everything before? Reflow was a massacre.
So, after reworking this damn thing for something like 3 days - I open the motors tab, spin up motor 4, it twitches - I turn off my transmitter (!) and it runs smoothly - turn the transmitter back on - motor starts twitching again - repeatable. Turns out the antennas of the R-XSR were pretty close to the BusyBee2 of ESC4 (nothing was touching, just too close).
The fix: A small piece of copper-foil covered with kapton-tape between the R-XSR and the 4in1. Seems to work for now.
Lesson learned: Micros/Nanos can be really painful to debug, expect the unexpected.
Edit: Another lesson learned - don’t conformal coat ESCs when not really necessary. Another of these 16x16 ESCs bit the dust, at least on FET burned out - took one trace on the PCB with it, possibly fixable, but only with a bridge-wire.