Plane enters spiral/toilet-bowl loop during roll in AutoTune (Factory PIDs)

Archive Plane 4.3
1

Hi everyone,

I am experiencing an issue where my plane enters a continuous, uncommanded circling/toilet-bowl behavior, but it happens specifically when attempting to roll while flying in AutoTune mode.

Vehicle Details:

  • Airframe: fixed wing with span of 158cm, with vtail
  • Flight Controller: Matek H743
  • Firmware Version: ArduPlane V4.7.0
  • PID Settings: Factory defaults / stock parameters

Symptom Description:
I am taking off directly in AutoTune mode as recommended by the documentation for a maiden flight. The takeoff and climb are perfectly normal, straight, and stable. However, the moment I input a roll command to start the tuning process, the plane fails to complete a crisp roll and instead locks into a widening, continuous circling pattern (similar to a toilet-bowl effect).

Because this is happening entirely within AutoTune mode from takeoff to touchdown, I understand that it shouldn’t be a GPS/Compass position-hold conflict.

What I have checked on the ground:

  • Manual Mode: Ailerons move in the correct direction with transmitter stick inputs.
  • FBWA Mode: Tilting the plane by hand causes the ailerons to deflect correctly to counter the roll (defending the horizon).
  • Trims: Transmitter trims are centered at zero. Mechanical surfaces look flush on the bench.

I am unsure if this is a severe aerodynamic stall/adverse yaw issue due to the weak or its too strong factory PID gains, a subtle control reversal I am missing under load, or something else entirely.

I have attached my Dataflash log (.bin file) below. Could someone please review my EKF status, servo outputs, and rate responses during the roll attempts to help me pinpoint the cause?

Thank you for your help!

2

Roll seems to be totally unresponsive to inputs. If the control inputs are working properly on the bench then the controls might not be enough, or strong enough.

image
image1252Ă—676 71.5 KB

3

so the problem is mechanically? im sure that the control surface is big enough and i confirm it in manual mode, I am actually thinking that the attitude roll direction is reversed and does not follow desRoll. am i wrong?, and i commanded on rolling to right but the plane is rolling indefinitely to the left

4

this is another maiden flight using another plane but the problem is the same toilet bowl

5

When it comes to the mechanical side of the plane I can only go by what you’ve said. If manual and FBWA checks are correct there needs to be something else that’s preventing the plane from doing what it’s told. The desired roll is to the right, and the plane is rolling left. I can see that in the graph. So the controller is reporting the same thing you’re saying. That tells me the controller orientation is correct, and the response is correct. What I can’t validate is the physical control movement.

Also look at center of gravity, or sources of drag. Did a cover come off? Did the battery shift? Is the servo intermittent?

6

thank you so much, as you said the problem was the roll authority not so strong, i increase the range of aileron servo to the maximum mechanical range, now it flys great and autotune is working as intended.