Hi everyone,
Setup: X-UAV Talon (V-tail) · Pixhawk Cube Orange · QGroundControl · hand-launched · flying in Stabilized. Ailerons are on a single channel via a Y-cable (MAIN1). Two ruddervators on the V-tail.
What the aircraft does
From the first seconds after hand-launch it can roll left fine (aileron authority is there), but it won’t actually turn left — it banks but the heading barely changes. The aircraft turns right far too easily and left barely at all.
When switched from Stabilized to Loiter, it turns almost the opposite way from the target. The pilot then switches to Mission/Auto, but instead of coming back it just keeps flying roughly straight, in nearly the wrong direction.
Stabilized Mode Check: I always check the Stabilized mode before taking off. Right before these problematic flights, I physically rolled and pitched the plane in my hands to ensure the ailerons and elevator were reacting autonomously in the correct directions. There were no abnormalities in the control surface movements prior to launch.
Sensor Calibration: As for the gyroscopes and accelerometers, I performed a fresh calibration for both of them at the flying field right before these flights.
- The yaw rate integrator sits at its −0.2 limit ~83% of the flight, yet the aircraft still yaws +6…+8°/s to the right on average. The yaw controller looks fully saturated, fighting a constant right-yaw moment it can’t overcome.
- Sideslip is a constant ~−14° throughout (never near zero).
- The roll side also saturates (aileron output sits around −0.6, roll integrator railed ~97%), but I suspect that constant left-aileron is mostly the controller reacting to the right-yaw/right-roll tendency rather than the primary fault.
Possibly related issues from previous flights
I’m not sure if these earlier observations are connected to the same root cause:
-
Autotune completes fine with only the Roll and Pitch axes enabled, but it cannot complete when the Yaw axis is also enabled — it fails every time.
Is it to be expected that yaw autotune cannot be done on V-tail fuselages like the X-UAV Talon? -
On stick release in Stabilized, the aircraft drifts continuously to one side — but the direction of this drift changes from flight to flight, even with nothing changed intentionally between flights.
Given the yaw integrator is saturated at FW_YR_IMAX=0.2 basically the whole flight, does raising FW_YR_IMAX (and FW_RR_IMAX) to 0.4 make sense as a stopgap, or does that just mask the real problem?
I have all the flight logs (.ulg), CSVs, tlogs and the flight video, and can upload whichever are useful — just let me know which flights or topics you’d like to see. Flight Review link and Drive folder below.
[https://review.px4.io/plot_app?log=c416e6b1-9e60-4444-9eba-b08ab1ba1d7c]
[https://drive.google.com/drive/folders/1a_hS2JlEUrVSBbnE-0mKTSMW242gNf2E?usp=sharing]
Thanks a lot for reading — any pointers appreciated.
