Here's what I have, I took a Ardupilot(legacy) which is stabilized by an ArduIMU and then fed the stabilized aileron and elevator outputs and throttle to a KK v2.1(the gyro only one) setup for tricopter.

I think I'm seeing groundloops or some kind of interference in the outputs. Sometimes it seems like it's going to work and then I'll either see one motor act weak or when I hit stabilize it'll shut down the motors. Rudder is bypassing the Ardupilot and goes direct from Rx to the KK. I did this because the Ardupilot has problems with rudder And aileron unless I start cutting traces according to the manual.

If I arm the KK board(uses rudder to arm) before the IMU and Ardupilot finish initializing, all motors work. It seems to stabilize in the right direction.

Why would arming the KK board after everything else initializes mess up the KK? I figured the Ardupilot should just pass the signals as if it doesnt matter whether the signals go to a servo or another processor.

Is this from a groundloop or interference? I moved all the boards as far apart as I could, but is there something else basic I should do to prevent problems?

I was thinking it's ground or interference, but I'm wondering could it just be a power problem?

I see some difference in how it acts as battery voltage changes, so maybe the ESC just has a crummy BEC that cant handle all those little atmegas. ??????

Now I'm thinking it could be all 3 that makes it irratatic. The stabilize switch is on ch5 and sometimes when I hit that it shuts down the motors which get signal from the KK. Arming the KK is done on rudder, but it acts like it can be disarmed by ch5. It also seems to be sensitive to battery voltage and that's why I wonder if it's just crap BEC.

If it's interference would I see one channel "bleeding" into and changing another channel?

If it's power, would I see the same hit one channel and it could react on another?

I've triple checked where everything is wired.

Some sort of mix turned on, on TX ?? Like Helo mode instead of acro ??