Hello,
I am still struggling with this quadplane (3.5 m wingspan airframe, fly at around 28 kg with 4×22" props), especially since I changed the ESCs from Scorpion 80A PWM to Hobbywing Xrotor DroneCAN H130A. (Beofre that, flights were ay more
smooth)
I am experiencing oscillations in QRTL during the airbrake phase. This already caused an incident in the past, so before resuming flights I ran a QuickTune (via a Lua script). Unfortunately this did not improve the situation and I had another incident on this flight. Because of that, I don’t think the issue comes from the Q_A_RAT_* parameters tuned by QTUN, but rather from limits that are not restrictive enough.
On the logs, I can see divergent oscillations in pitch and pitch rate (and the same behavior in roll).
I am considering the following parameter changes to make the controller more constrained and less aggressive:
Q_A_ACCEL_P_MAX: 50000 → 30000
Q_A_ACCEL_R_MAX: 50000 → 30000
Q_A_ACCEL_Y_MAX: 50000 → 15000
Q_A_RAT_PIT_SMAX: 50 → 20
Q_A_RAT_RLL_SMAX: 50 → 20
Q_A_RAT_YAW_SMAX: 50 → 10
Q_A_RATE_P_MAX: 75 → 60
Q_A_RATE_R_MAX: 75 → 60
Q_A_RATE_Y_MAX: 75 → 30
Q_ANGLE_MAX: 3000 → 1500
Q_M_SLEW_DN_TIME: 0.3 → 0
Q_M_SLEW_UP_TIME: 0.3 → 0.1
I based these values on the Mugin EV350 and Foxtech Great Shark frame parameter files available here: https://github.com/Ardupilot/ardupilot/tree/master/Tools/Frame_params/QuadPlanes
And on realfight it does reduce the oscilations
Do you think these changes are coherent with the type of incident I’m seeing? Would you recommend other adjustments or checks in this situation?
Thanks in advance !
Logfile : Dropbox
It’s the last fight, but I had the same behavior on other flights since I changed the ESCs