I’m replying here because I think I ran into essentially the same issue on a front-2-motor tilt QuadPlane in X configuration, and I’m hoping to get updated guidance on the best way to handle it in 2026.
I attached:
Airframe / setup
-
QuadPlane tiltrotor with front two motors tilting together
-
Not vectored yaw
-
Q_FRAME_CLASS = 1
-
Q_FRAME_TYPE = 1 (X frame)
-
Q_TILT_TYPE = 0 (continuous tilt)
-
Q_TILT_MASK = 5 (front motors 1 and 3 tilt)
Motors follow the standard Quad-X layout:
-
Motor 1 = front right = CCW
-
Motor 2 = rear left = CCW
-
Motor 3 = front left = CW
-
Motor 4 = rear right = CW
So yaw is being generated through differential motor RPM, not differential tilt angle.
Important background
What makes this especially confusing is that we had a smaller test plane with this same general front-2-tilt concept, and it transitioned smoothly. We did not see this yaw feedback issue there.
The logs I’m posting here is from a tilt rotor test for a ~40lb aircraft that failed to transition properly as seen in the attached video
What I’m seeing
I’ve now seen a very similar failure mode as to what other people described with yaw control issues on multiple transition attempts with tilt angles of 20, 45, and 60 degrees.
In the logs:
-
when C3 > C1, we see a CW / nose-right yaw
-
when C1 > C3, we see a CCW / nose-left yaw
On previous tilt rotor attempts on a 35lb aircraft (not in the video or logs) one flight at around 45 degrees, we attempted transition twice and got a very strong CW/right yaw.
On another flight at 60 degrees, the aircraft first had a smaller yaw transient and then later developed a much more violent yaw event.
The pattern that stands out is that the problem appears when the front motor outputs stop matching and begin separating.
Why I think this is happening
My current theory is that this is exactly the issue described earlier in this thread for front-2-tilt aircraft in X frame.
In pure Quad-X multicopter logic, if the controller wants a yaw correction, it changes the RPM of the CW pair vs CCW pair. That makes sense in vertical flight.
But once the front two motors are partially tilted forward, the same RPM difference is no longer acting only through normal multicopter yaw torque. It also creates a horizontal thrust asymmetry between the front-left and front-right motors.
For example:
-
if Motor 3 (front left) is increased
-
and Motor 1 (front right) is decreased
-
while both front motors are partially tilted forward
then the front-left motor is producing more forward thrust than the front-right motor, which creates a nose-right / CW yaw moment.
So it looks like the controller may be trying to correct yaw in the normal VTOL sense, but in the intermediate tilt region the actual aircraft response may become the opposite sign, or at least a much stronger yaw moment than intended.
This would also explain why the problem seems worse at 60 degrees than at 45 degrees.
What I still do not understand
What I’m trying to pin down is:
-
Is the front motor differential mainly an attempted yaw correction from the VTOL controller?
- I really think it’s just yaw, but roll can technically also cause differential thrust
-
Or is some of that split actually being driven by roll/pitch corrections in Quad-X mixing, which then turns into yaw because the front motors are tilted?
-
Has anyone gotten a front-two-tilt QuadPlane in X frame to transition reliably without this issue, and if so, what was the real fix? For now assume I cannot use vectored yaw in this arrangement. We’re right now looking into running a lua script that sets motor based yaw PID values to 0 in FBWA and re enabling them to the previous default values in any other flight mode.
What I’m asking
Has there been any real solution for this in the last few years?
What is the best path forward here?
-
Switch to H frame instead of X?
-
Keep X, but significantly reduce yaw P / yaw aggressiveness?
-
Reduce Q_TILT_MAX and keep the intermediate tilt region shorter?
-
Increase Q_TILT_RATE_DN so the aircraft spends less time in the dangerous partially tilted regime?
-
Is this still fundamentally a limitation of using normal Quad-X control logic on a front-2-tilt aircraft in the intermediate transition region?
I’d especially appreciate input from anyone who has actually solved this on a larger front-2-tilt aircraft, since our smaller test platform worked fine but the larger ~40 lb class aircraft is where this became a serious issue.
If useful, I can also post:
