Severe Pitch Oscillations in VTOL Reverse Transition

ArduPlane
, ,
1

Hi everyone,

I’m currently flight testing a 6m wingspan hybrid VTOL aircraft (quadplane configuration) with an all-up weight of ~110 kg, and I’m facing issues specifically during reverse transition.
Firmware - 4.6.2

Problem Description

During reverse transition, I observe significant pitch oscillations, especially while trying to bleed airspeed before switching modes.

Flight Profile

  • Flying in FBWA mode

  • Reducing throttle to zero to bleed airspeed

  • Preparing to switch to QLOITER

Observations

  • As the airspeed drops to ~27 m/s (current Q_ASSIST_SPEED for reverse transition), the aircraft:

    • Suddenly pitches up

    • Enters noticeable pitch oscillations

  • These oscillations are quite pronounced and repeatable

Parameter Setup

  • Q_ASSIST_SPEED (forward transition): 23 m/s

  • Q_ASSIST_SPEED (reverse transition): 27 m/s

We intentionally set a higher assist speed for reverse transition because:

  • Reverse transition appeared smoother at 27 m/s compared to 23 m/s

Log Analysis

  • From the logs, I noticed that:

    • The desired pitch demand drops to ~ 0

    • The controller does not seem to command pitch corrections

    • This happens when the aircraft is below Q_ASSIST_SPEED (27 m/s) during reverse transition

Key Questions

  1. Is it expected behavior for the desired pitch to go to zero and not command corrections below Q_ASSIST_SPEED in this phase?

  2. Could this indicate:

    • A limitation or behavior of FBWA + QuadPlane assist logic, or

    • A tuning issue (e.g., pitch PID, TECS interaction, or transition parameters)?

  3. Are there recommended practices for:

    • Managing airspeed bleed before reverse transition?

    • Avoiding pitch oscillations during this phase?

Additional Context

  • The issue is more pronounced in FBWA during deceleration

  • The oscillation begins right around the assist speed threshold

  • I came across a similar issue reported by another user where the root cause was the use of a fixed-wing ESC on quad motors

  • However, in my case, I am not using fixed-wing ESCs for the quad motors, so that particular cause does not apply here

Log Link - https://drive.google.com/file/d/1W-kP8qYd2xqjc5ZxZrUcQadmU6ZMNzUH/view?usp=sharing

Any insights, suggestions, or similar experiences would be really helpful

Thanks in advance!

2

How about switching to Q-hover mode during deceleration?
How can you set up two different Q_assist_speed for transition and reverse transition?

3

Currently, we decelerate in FBWA and then switch to QLOITER. The issue I foresee if we switch to QLOITER / QHOVER during deceleration is that the aircraft will pitch up abruptly to slow down which we want to avoid.
Regarding two different Q_Assist_Speed for transition and reverse transition, I change the speed parameter in flight before reverse transition.

4

We’ve been doing reverse transition in Q_althold mode for a long time. It doesn’t make the aircraft unstable and it won’t pitch up or down either. Aircraft will just slide without deceleration smoothly until you manually and slowly apply breaking. It has been always working and always best for us.

5

Do you mean Q-Hover?

But yes, I’ve used the same method and it can work very well.

6

Yes, I meant Q_hover. Thanks for correcting me. :slight_smile:

7

Since you only sent a tlog, there isn’t enough data to analyze. I’m just curious why your Q_A_ANG_RLL_P (3.3) and Q_A_ANG_PIT_P (3.4) are so low. This means your attitude control bandwidth is quite low; we need to look at the rate loop response to properly evaluate it