I have been trying to eliminate EKF_YAW reset event as soon as it takes off in QStabilized, QLoiter, QHover modes (AKA Toilet bowl). I moved the GPS from AFT to FWD, but no change. Did multiple calibrations, no change applied Magfit, no change.
The NYT QUAD X-Motor Tailsitter is using MicoAir F405 V2. It starts to hover at 1/4 of the throttle and after each flight motors are cold, and the PID loop looks good. I am conducting the test without wings so no elevons are attached. So I am solely relying on motor torque to control the Yaw.
As I understand, after takeoff a Yaw reset causes the tailsitter to chase the new direction but it cannot catch in time thus goes into constant hunting.
RCIN_4 and RCOUT 1,2,3,4. Frequency domain, and PID settings image below. Let me know what else you would like to see.
