Maiden going perfectly then sudden flip over

Can someone take a look at this log from the maiden flight of my new quad. Here is the setup:

Holybro pixhawk 2.4.5 with V3.4.2 quad
Ublox NEO-M8N GPS
volo systems “the power” PDB
Mauch 100A current sensor
U7 v2 420kv motors, 17.5" CF props
rc tigermotor 80A esc’s
13,000 mah 6S battery

Finished build, calibrated esc’s, checked motors/ props/ direction, set and calibrated RC, Failsafes, calibrated current sensor, went outside and calibrated accel and compass. Good to fly, no messages, did a slow run-up, took off and hovered 2 feet above ground. Everything was perfect, solid control on pitch, roll and yaw. Hovered for about 1 min then suddenly copter flips over on its left side and crashes upside down about 20’ away. During crash I had zero control.

Thanks for any suggestions.
Luke

1.BIN (593.6 KB)

Sorry i don’t have time to root-cause this, but some observations:

  • the vehicle self-assessed itself as having a broken magnetometer
            {
               "duration" : 2.6308739185333252,
               "duration-units" : "seconds",
               "evilness" : 10,
               "evilness-is-deprecated" : "Use severity-score",
               "reason" : "The craft's assessment of its sensors indicate a problem",
               "sensor-name" : "MAG",
               "series" : [],
               "severity-score" : 10,
               "status" : "FAIL",
               "timestamp_start" : 190988994,
               "timestamp_stop" : 193619868
            }
  • the EKF self-assessed itself as in a borked state:
2016-11-18 11:02:12.36: MSG {TimeUS : 192392026, Message : EKF variance}

(it is vexing that the EKF status flags weren’t being logged; check your LOG_BITMASK)

Your magnetometers don’t agree with one-another:

You have VERY bad coupling between your second mag and your throttle:


So I would hazard a guess that you’ve got a high-current wire running right underneath your autopilot (if not wrapped several times around it :wink: Your vehicle’s estimate of its attitude has broken because your magnetic interference was so high.

The only reservation I have in this assesment are these lines from your log:

2016-11-18 11:02:11.38: MSG {TimeUS : 191410443, Message : EKF2 IMU0 is using optical flow}
2016-11-18 11:02:11.38: MSG {TimeUS : 191410686, Message : EKF2 IMU1 is using optical flow}
2016-11-18 11:02:12.36: MSG {TimeUS : 192392026, Message : EKF variance}
2016-11-18 11:02:13.59: MSG {TimeUS : 193619370, Message : Crash: Disarming}

That timing is awfully co-incidental. Why the OF messages just before the variance declaration?

Thanks for the info. I do have the current sensor mounted about 3" directly below the pixhawk. Perhaps that is the problem. I can try and mount it towards the side of the frame and test it. I will tie the copter down and run it up to normal current loads and see if I get any errors. Thanks again, Luke

Would it help to apply copper tape to the bottom of the mounting plate for the pixhawk?