HKPilot 32 (pixhawk clone) gps modes flyaway

For the past while I've been having problems with my compass (EKF and DCM errors resulting in change of heading) that have made it almost impossible to fly since I fly in simple mode. I was on arducopter 3.2.1 and after updating to 3.3 the errors involving the drifting heading have gone away but now I get a new error called ECM bad variance. Before the update when I was still getting EKF erros the quad was able to fly in all modes including loiter and pos hold without anything wrong, but today when I switched into loiter, the quad rolled to the right (about 40 degrees)  and flew in that direction until I switched back to alt hold or stabilize. This has never happened and I'm afraid I might have a flyaway if it switches to rtl because of a Fail Safe. Please help, I attached the log file below..

Btw I have a mounted Quanum LEA-6H GPS/Compass module and was wondering if there is a way to check if that compass is being used as the main compass and not the internal compass in the FC.

Thanks, Evan

136.BIN

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

Join diydrones

Email me when people reply –

Replies

  • Hey Mario, I've already tries flying with the compass connected to both the i2c and can bus ports and have seen no difference.
  • Hi Evan

    I did a closer look on the images you uploaded.
    According to my knowledge (and my own pixhawk with gps) the two-wire-cable of the GPS should be connected to the I2C-port of the pixhawk and not the CAN-BUS (except you have a real special GPS device)

    That would explain why your compass would't function properly.

    Best regards

    Mario

  • Hi Micheal, I just replaced the foam tape I was using to hold the FC on to the frame with a brick of squishy Styrofoam. I'll attach a photo of the flight controller and a log from when I flew it today. The same issue occurred today when I switched into loiter. The quad stayed in the same location and altitude i set loiter at but spun in a circle and rolled right.

    Michael Burmeister said:

    Looking at the logs it looks like you have bad vibrations in the flight controller.  In addition it looks like you were flying low enough that you were getting some GPS Doppler effect from the near by houses.  This caused sudden change in position that could have caused the copter to fly off in one direction.

    I would try and correct the vibration issues first and then do a test flight trying to stick to two flight modes and post the results.

    quad_front

    quad_top

  • Yes, I have many troubles too, you have support forum where developers follow, I visit too If I can help you, look your vers 3.3 and ask there, try always with a log attached by Dropbox or something similar http://discuss.ardupilot.org/

  • Looking at the logs it looks like you have bad vibrations in the flight controller.  In addition it looks like you were flying low enough that you were getting some GPS Doppler effect from the near by houses.  This caused sudden change in position that could have caused the copter to fly off in one direction.

    I would try and correct the vibration issues first and then do a test flight trying to stick to two flight modes and post the results.

  • Is there a way I can fix these issues?

    Is there another way we can talk because i'm having problems with diydrones that is not letting me reach the page unless i refresh the page multiple times.

    Thanks

    Cala said:

    It's a tipical movement that copter do when you are in loiter with compass issues, alt hold and stab don't uses compass; If you have int one disabled, you have to check your ext one, what is happening; in Mission planner you have flags to check compass issues during flight, I don't know if qcontrol have them, they are usefull.

    Log File C:\Users\ale\AppData\Local\Temp\tmpC9A4.tmp.log
    Size (kb) 4852.541015625
    No of lines 60849
    Duration 0:05:02
    Vehicletype ArduCopter
    Firmware Version V3.3.3
    Firmware Hash acf2e10c
    Hardware Type 
    Free Mem 0
    Skipped Lines 0

    Test: Autotune = NA - 
    Test: Balance/Twist = NA - 
    Test: Brownout = GOOD - 
    Test: Compass = FAIL - FAIL: Large compass offset params (X:216.00, Y:477.00, Z:61.00)
    FAIL: Large compass offset in MAG data (X:216.00, Y:477.00, Z:61.00)
    Large change in mag_field (84.29%)
    Max mag field length (1210.07) > recommended (550.00)

    Test: Dupe Log Data = GOOD - 
    Test: Empty = GOOD - 
    Test: Event/Failsafe = GOOD - 
    Test: GPS = GOOD - 
    Test: IMU Mismatch = FAIL - Check vibration or accelerometer calibration. (Mismatch: 1.52, WARN: 0.75, FAIL: 1.50)
    Test: Parameters = GOOD - 
    Test: PM = NA - 
    Test: Pitch/Roll = NA - 
    Test: Thrust = NA - 
    Test: VCC = GOOD -

  • Is there a way I can fix these issues?

    Is there another way we can talk because i'm having problems with diydrones that is not letting me reach the page unless i refresh the page multiple times.

    Thanks

    Cala said:

    It's a tipical movement that copter do when you are in loiter with compass issues, alt hold and stab don't uses compass; If you have int one disabled, you have to check your ext one, what is happening; in Mission planner you have flags to check compass issues during flight, I don't know if qcontrol have them, they are usefull.

    Log File C:\Users\ale\AppData\Local\Temp\tmpC9A4.tmp.log
    Size (kb) 4852.541015625
    No of lines 60849
    Duration 0:05:02
    Vehicletype ArduCopter
    Firmware Version V3.3.3
    Firmware Hash acf2e10c
    Hardware Type 
    Free Mem 0
    Skipped Lines 0

    Test: Autotune = NA - 
    Test: Balance/Twist = NA - 
    Test: Brownout = GOOD - 
    Test: Compass = FAIL - FAIL: Large compass offset params (X:216.00, Y:477.00, Z:61.00)
    FAIL: Large compass offset in MAG data (X:216.00, Y:477.00, Z:61.00)
    Large change in mag_field (84.29%)
    Max mag field length (1210.07) > recommended (550.00)

    Test: Dupe Log Data = GOOD - 
    Test: Empty = GOOD - 
    Test: Event/Failsafe = GOOD - 
    Test: GPS = GOOD - 
    Test: IMU Mismatch = FAIL - Check vibration or accelerometer calibration. (Mismatch: 1.52, WARN: 0.75, FAIL: 1.50)
    Test: Parameters = GOOD - 
    Test: PM = NA - 
    Test: Pitch/Roll = NA - 
    Test: Thrust = NA - 
    Test: VCC = GOOD -

  • It's a tipical movement that copter do when you are in loiter with compass issues, alt hold and stab don't uses compass; If you have int one disabled, you have to check your ext one, what is happening; in Mission planner you have flags to check compass issues during flight, I don't know if qcontrol have them, they are usefull.

    Log File C:\Users\ale\AppData\Local\Temp\tmpC9A4.tmp.log
    Size (kb) 4852.541015625
    No of lines 60849
    Duration 0:05:02
    Vehicletype ArduCopter
    Firmware Version V3.3.3
    Firmware Hash acf2e10c
    Hardware Type 
    Free Mem 0
    Skipped Lines 0

    Test: Autotune = NA - 
    Test: Balance/Twist = NA - 
    Test: Brownout = GOOD - 
    Test: Compass = FAIL - FAIL: Large compass offset params (X:216.00, Y:477.00, Z:61.00)
    FAIL: Large compass offset in MAG data (X:216.00, Y:477.00, Z:61.00)
    Large change in mag_field (84.29%)
    Max mag field length (1210.07) > recommended (550.00)

    Test: Dupe Log Data = GOOD - 
    Test: Empty = GOOD - 
    Test: Event/Failsafe = GOOD - 
    Test: GPS = GOOD - 
    Test: IMU Mismatch = FAIL - Check vibration or accelerometer calibration. (Mismatch: 1.52, WARN: 0.75, FAIL: 1.50)
    Test: Parameters = GOOD - 
    Test: PM = NA - 
    Test: Pitch/Roll = NA - 
    Test: Thrust = NA - 
    Test: VCC = GOOD -

  • btw I am using QgroundControl since APM planner doesn't seem to work with the current update and I am on a mac.


  • Hi Cala, I already have COMPASS_USE2 set to 0 (False), compass_external set to enabled and COMPASS_PRIMARY set to first compass. Right before you replied I went out to test it and see if it would still roll and flyaway in loiter mode, but this time i had no EKF errors while flying in alt hold. However when I switched to loiter, the quadcopter started going in circles getting bigger and bigger (toilet bowl) until I switched back to althold then it was fine. Right before flying I did a compass calibration and a motor compass interference calibration.

    Thanks for the response.


    Cala said:

    Hello Evan: I have the same controller and I disabled the int compass (compass use 2=0) thired of that err and never another problem.

    The Err denominations changed in new firms.

    To check compass, unplug ext one, move the copter and see if compass line moves in mission planner, then plug it and try again.

    Screen Shot 2016-07-23 at 9.20.17 PM.png

    Screen Shot 2016-07-23 at 9.20.30 PM.png

This reply was deleted.

Activity