Bi-Copter YAW oscillation

Copter 4.1.5 - set to Frame class 10 (Bi-copter)
Can I beg some advice please?
I’m building a bi-copter based on the Matek F765 WSE FC and am having trouble:with significant oscillation in YAW that I can’t seem to dial out.

To improve matters, I actually just reduced the YAW PIDs to their minimum settings. It’s still wagging its tail! I can’t believe it.
I then almost got a fly-away, the rotors started spooling up to high speed, luckily i was able to grab it and pull the battery. I was in Stabilise mode, so would a GPS glitch still trigger a RTH and cause the props to spin up?
Any help would be most appreciated.
2022-03-08 17-45-30.bin (869.8 KB)

  1. Regarding GPS glitches and runaways: in your log, I saw a few RC failsafes, which caused disarms, and no attempts to go RTH. The biggest current peak I saw is related, most likely, to a significant momentary pitch (and roll), which, due to ATC_ANGLE_BOOST=1, increased the throttle such that the altitude is roughly the same.

  2. Bicopter is quite a beast with regards to PIDs. I have a growing experience with CoaxCopters, and they are very counter-intuitive for those with experience in normal copters. Bicopter might be similar in some axis combinations. In particular, I believe the PIDs for yaw shall be increased in your case compared to the defaults, since the oscillations you see are, quite likely, not the usual control overshoot oscillations, but a result of slow response of servos caused by insufficient output.

I do have a bicopter, in fact, but for various reasons I did not yet try to fly or tune it. The preliminary mini-tests suggest me, however, that the default settings are way too small for my vehicle.

Thank you for taking the time to reply. You are very right in that setting up a bicopter is very tricky… especially so as there’s very little information out there. I’ve tried INAV too - with little success.
Your view on the responsiveness of the servos is very interesting!
I’ve had recent success by significantly reducing the amount of travel available to the servos.
I’m in the middle of a redesign at the moment to get round a hardware issue, but I’ll certainly try your suggestion regarding increasing the rates above standard.:+1: