AQ6 - HEXA Has Small Fast Roll Oscillation

[rmsteele]

I am running AQ firmware 6.7-BETA r100 b1066 hw rev:2 - My HEXA fly's very well. Seems to be very stable even in very windy conditions. Looks very smooth when viewed from R/C pilot position at any distance including close up however, whenever it is viewed though FPV camera there is a small fast roll axis oscillation I have not been able to clear. It does not appear to be in the pitch or yaw axis. Only the roll. I have tried increasing the tilt rate D in steps of 100 as suggested in the troubleshooting but that does not seem to have any effect on this problem. If I were to guess I would say it is oscillating at most 3 degrees and at a rate of maybe 1.5 to 2 times per second. It will oscillate 5 or 6 times then stop oscillating for about 1 or 1.5 seconds then start oscillating again. All of these counts are estimates as it is difficult to count and time in flight. Any suggestions would be greatly appreciated.

[joebar.rc]

Does it also show up in the logs?, look at the motor output and aq_roll for example.
Also check out the z axis, maybe some vibration can be seen. Do you use rigid mounting of the flightcontroller or using soft dampeners?

And if the physics are spot on, well maybe some extra tuning can help

[rmsteele]

I am not seeing this in the logs or maybe I don't know what to look for. Below is a video from the onboard FPV camera point of view. You might at first think it is the camera that is moving and not the copter but I have checked that and it is absolutely not the camera but in fact the entire copter.

The AQ6 is mounted on rubber standoffs. This may still be a vibration issue however. When I look at the logs the vibration appears to be at 0.123 with an occasional spike at 0.125. I do not know what the scale is as it is not marked.

I am sure it is a still a PID setting that just needs to be fine-tuned. I just don't know which one or which direction. Any assistance would be greatly appreciated.

[Max]

Are you sure it's not the camera shaking? I see the arms moving in relation to the video frame, but not in relation to the trees in the distance, for example (especially watching the right front arm when it's close to the tree line, later in the vid). The hexa seems rock solid. How is the camera attached? Rubber standoffs for the FC can be hit/miss, really depending on the rest of the vibes in the frame. Yours seems to fly well. Post logs... an oscillation like that should be fairly obvious in the accel and motor graphs, I'd think.

-Max

[rmsteele]

Hi Max, no, I thought the same thing. The camera is hard fastened to the frame. It cannot move or shake separately.