Very Strange Attitude Oscillation (it's really bizarre)

Hello All,

Flight Logs
Video (Please take a look at the video starting from 10 seconds

After some attitude PID control, I have run out of options why copter is behaving this way. Normally, you’d expect if a controller is tuned faulty (e.g MC_PITCH_P, MC_PITCHRATE_P, I,D) you would see an oscillation or a twitchy or a sluggish behaviour on PITCH axis.

The problem we have is that axis between Motor 1 and Motor 2 y=x line in ENU body frame behaves differently than the axis between Motor 3 and Motor 4 y=-x in ENU body frame. On the y=-x axis, the multicopter holds the attitude without oscillations or too much overshoot, but on the y=x axis, the multicopter overshoots therefore oscillates crazily, until it levels itself. I don’t think its related to software or control parameters, but rather a hardware issue.

We suspect one of the ESC’s (motor 2) might be faulty in some way. Do you think that might be the case?