Strange loss of control in loiter

On my fifth flight today, something very strange happened while flying loiter mode. My copter started to drift from side to side as if trying to make a large circle, and I had to fight it hard to retain control. I have checked the logs carefully and I can't find an explanation. The large circles were captured nicely by the GPS, and apparently commanded by the flight controller (check Roll/DesRoll and Pitch/DesPitch) so I don't think this was a GPS, attitude estimation or other problem, though I could be mistaken. It also doesn't look like a mechanical issue either because none of the motor commands ever reached 100%. I'm at a loss to explain what happened - it's the first time I've seen loiter misbehaving like this, and I've done about 30 ArduCopter flights so far.

I'm attaching the dataflash log. I'm using the Pixhawk with the 3DR GPS. I'm running the ArduCopter master branch, updated on the 7th June at this commit: https://github.com/diydrones/ardupilot/commit/105382bf215043313de3e3a720970c926639909c

Brief summary of the flight:

  • Headed out flying east-south-east. I noticed a slight unexpected drift to the right (south).
  • Drift became severe, so I turned around and headed back west.
  • As I flew back, I noted the copter was tilting hard right. I fought the tilt with the stick and was able to continue to fly back home.
  • About 2/3 of the way back home, the tilts magically stopped without explanation.

Any insights much appreciated! Thanks :)

8.BIN

3691129091?profile=original

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

  • Is that a hydroelectric dam and if so I wonder if something there could have your compass throwing a fit?

    • It's a bridge. That could well have contributed to the TBE-like effect, though the underlying cause was a bad attitude estimate from the DCM algorithm, I think.

  • Is your board still attached to the frame/vibration dampening system?  I've had problems like this when a corner has come unattached.

    • Yes, all four corners are still attached to the vibration dampening mount.

      • The toilet bowl affect is usually cause by you compass. Did you run the compass motor test. How much noise do you have. What version board do you have. A 2.5 or 2.5.
        • The compass is fine.

          I eventually figured I had a classic case of "the leans". I've switched to EKF now, so far so good.

          https://groups.google.com/forum/#!topic/drones-discuss/_mrGyIGSkxo

          Google Groups
          Google Groups allows you to create and participate in online forums and email-based groups with a rich experience for community conversations.
This reply was deleted.

Activity

Neville Rodrigues liked Neville Rodrigues's profile
Jun 30
Santiago Perez liked Santiago Perez's profile
Jun 21
More…