MR60

PIxhawk Bug (Hw or SW ?) on compassmot procedure

Posted by Hugues on February 6, 2014 at 11:43am

Hi,

I did yesterday a compassmot on the newly received pixhawk with external GPS+Compass. I thought initially that everything went fine as the CLI screen gave me a 0% interference level as shown on this picture:

3691096087?profile=originalAfterwards, I launched Mission Planner and saw a very unstable Flight data HUD with the horizon dancing around as shown in this picture:

3691096130?profile=originalI did not understand immediately what was going wrong so I tried to reboot, to reconnect without success. Until I made the link with "compassmot".

I then checked my "Full parameters" list and found this BUG:

3691095983?profile=originalALL of the three COMPASS_MOT values were corrupted "non numerical". As soon as I manually wrote them as zeros again, it solved my compass issue.

Then I went back to the proud screen capture I took to show my succesful 0% interference and noticed in the CLI output something weird as can be seen on the above corresponding picture:

->above a certain throttle level the compassmot procedure has modified the compass mot values to "nan", whatever that means. But certainly non numerical values!

By the way I have repeated the compassmot procedure twice with the same effect each time.

Is this a known issue or a new discovered bug ? And is this a Pixhawk related bug and/or a firmware related bug ?

Thx for any help,

Hugues

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

Join diydrones

Email me when people reply –
Follow

Replies

  • Bill Isenberger April 30, 2014 at 6:13pm

    I'm also having some compassmot issues with firmware v3.1.3 and AMP 2.6 with external GPS/compass.

    http://diydrones.com/forum/topics/number-4-motor-on-quad-running-sl...

  • Bill Isenberger April 8, 2014 at 8:49am

    Wondering if this has anything to do with the on board compass being off.

    These are the offsets for the on board compass after calibration (external compass/GPS not plugged into the Pixhawk).
    compass_offsets.jpg


    After calibration with the Pixhawk pointing North (0 degrees), Mission Planner shows the Yaw (deg) as 309.77 (off by about 50 degrees).
    compass_wrong.jpg

  • Bill Isenberger March 29, 2014 at 8:20pm

    Any idea on when this will be fixed?

    Is there another firmware version that works with Pixhawk. I have a new Pixhawk and want to make sure it is working OK.

    • MR60
      Hugues > Bill Isenberger April 1, 2014 at 10:09am

      I understood it was on the list of issues the DEV will correct. Wa can only hope it shall at the latest come in release 3.2.

      But that's just a wish....

  • Bill Isenberger March 24, 2014 at 10:57am

    I changed motors on my quad and ran compassmot again. This time the results were 24%, even though the GPS is on a mast about 70mm above the props.

    pixh_compm.jpg

    • Bill Isenberger > Bill Isenberger March 24, 2014 at 1:03pm

      Man, that thing was all over the place in loiter.

      I replaced the Pixhawk with an APM 2.6 and ran compassmot. The result was 2% and is pretty stable in loiter. I think my Pixhawk is defective.

    • MR60
      Hugues > Bill Isenberger March 31, 2014 at 12:18pm
      Have you tried to measure your vibrations: zacc ? Are they low enough ?
    • Bill Isenberger > Hugues March 31, 2014 at 1:19pm

      Definitely some weird stuff with the Pixhawk and compassmot. Either NaN values or in the case of the different motors wrong values.

      Results of compassmot with Pixhawk  and APM on the same air frame after changing motors:

      Pixhawk 24%

      APM 2%

This reply was deleted.