Crash after Autotune & "Prearm: ACRO_BAL_ROLL/PITCH" error

Hi everyone,

I’ve built a few drones before, but I ran into a really strange issue today that I haven’t seen on my previous builds. It resulted in a crash, and I’m trying to understand if this is a glitch or something I missed.

The Setup:

  • Frame: Holybro x650 - 15”

  • FC: CUAV Pixhawk v6x

  • Context: The drone was previously tuned and flying perfectly. I recently swapped to a lighter payload, so I reset to re-tune.

The Sequence of Events:

  1. Performed a flight to relearn hover thrust (AltHold).

  2. Initiated Autotune. It completed Roll and Pitch.

  3. I landed and disarmed after the Yaw tune started (to swap batteries).

  4. Upon rebooting to test the new tune (in AltHold), I got a blocking error: Prearm: ACRO_BAL_ROLL/PITCH.

The Troubleshooting & Crash: I don’t fly Acro, so this error was unexpected.

  • My values for ACRO_BAL_ROLL and PITCH were at 1.0.

  • I tried the documented defaults (2.0)—the error persisted.

  • The “Fix”: I randomly set them to 0.5, and the Prearm error cleared.

  • The Flight: I assumed since I wasn’t flying Acro, this setting wouldn’t matter. I armed and lifted off about 2 feet. The drone immediately had massive input lag (approx. 2 seconds) and felt like it was over-tilting (Acro behavior?).

  • It became uncontrollable due to the latency and crashed.

Current Status: Repaired the props. I reverted the Acro Balance params to default, and the Prearm error has returned.

My Theory:
I suspect the partial Autotune calculated ATC_ANG_RLL_P and PIT_P values that were mathematically incompatible with the default ACRO_BAL parameters (hence the prearm failure). By lowering ACRO_BAL to 0.5 to bypass the check, I likely masked a deeper issue - potentially critically low P-gains or a corrupted tune - which caused the sluggish response and “lag” in the control loop during the test flight.

Can anyone confirm if this relationship exists, or help me understand what happened?

Attachments:

Thanks for the help!