Analyzing a fixed-wing UAV crash log with bad accelerometer health

Hello,
I’m looking for help with analyzing a crash log from a fixed-wing UAV.
Firmware version: ArduPlane 4.5.7.

My current suspicions are:

  • Tail-heavy CG

  • Poor / unreliable airspeed sensor data

  • Autopilot tuning issues

  • Bad accelerometer health

Any insights or guidance on what to check first would be greatly appreciated.

Welcome to the community.

No need to add “Need help” in the title of the post. All posts on this forum are from someone who “needs help”. That information is redundant, and just clutters the forum.

Bad accelerometer health is a serious issue. Fix that one first!

Here`s the log file. Thanks

You need to provide a .bin log file

Hi Stepan,

A big mistake is that you armed (in Mode Auto!) with ARMING_CHECK disabled. Could the files have been edited? It is noticeable that important GPS data was not logged at all.

Rolf

1 Like

CFIT - Controlled Flight into terrain. The plane was commanded to pitch down and it did.

The log file is missing a lot of data so it’s difficult to really put a lot of effort into the analysis.

1 Like

This is especially the case since no other information was provided. Maybe Stepan will catch up, he’s new to the forum after all.

I don’t think CFIT is entirely accurate. According to TECS, the target altitude should have remained stable.

Perhaps it was a “PCFIT” :wink: (due to poorly configured flight into terrain).

Rolf

2 Likes

Thanks everyone for taking part in the analysis.

Yes, I did notice that the autopilot commanded the aircraft into the ground.
The GPS data was intentionally removed on request.

The Arming_Check was performed using a modified version of Mission Planner, where the flag is temporarily set to 1 via a hotkey and automatically returns to 0 once the Arming_Check window is closed. Logging starts after ARM.

What is still unclear to me is why, after reaching the second mission waypoint, the aircraft suddenly pitched down aggressively, even though the mission altitude at that point was set to no less than 40–50 meters.

About PCFIT - plane wasn`t controlled by transmitter while flying

Thanks for confirming that. I am relieved that Arduplane is not the cause of such a corrupt log file. Could you please upload the original file?

Rolf

Sorry , i can`t upload full log but i have an update. The reason of crash was incorrect cruise and min/max speed(50m/s). Usually for this UAV it was 20m/s ± 4m/s , but after changing flight controller and updating to 4.5.7 it didn`t show speed differences in compare params from old params. So in order to achieve that speed FC forwarded plane down. Thanks for participating in review.