Welcome to PhantomPilots.com

Sign up for a weekly email of the latest drone news & information

Compass problem with adv

Discussion in 'Pro/Adv Discussion' started by Rkcorp, Oct 13, 2016.

  1. Rkcorp

    Joined:
    Oct 13, 2016
    Messages:
    6
    Likes Received:
    0
    Hello, this morning i have made à flight and i have a problem during this.

    My drone no respond to remote à few moment and its crazy in air ! I have a yaw problem in log and speed problem too.

    The détails :

    Phantom Log Viewer - PhantomHelp.com

    Can you say me wath problem ?? My drone is warrantly, i must return my drone to Factory or not ?
     
  2. bigblue30us

    Joined:
    Sep 21, 2016
    Messages:
    8
    Likes Received:
    4
    Did you calibrate the compass before flight?


    Sent from my iPhone using Tapatalk
     
  3. Rkcorp

    Joined:
    Oct 13, 2016
    Messages:
    6
    Likes Received:
    0
    Yes i always calibrage compass before fly
     
  4. msinger

    Approved Vendor

    Joined:
    Oct 30, 2014
    Messages:
    18,936
    Likes Received:
    5,578
    Location:
    US
    It looks like you had a bad compass calibration. Instead of calibrating the compass before each flight, it would be better to follow this guide.

    Try calibrating your compass again in a location where you know there is no interference. And, then take it for a test flight to see if you run into the same issue.
     
    lonjsnyder and Oso like this.
  5. Rkcorp

    Joined:
    Oct 13, 2016
    Messages:
    6
    Likes Received:
    0
    The speed problem in report its not problem ?

    During flight he go in direction without control when compass error

    Very afraid
     
  6. BudWalker

    Joined:
    Jun 24, 2015
    Messages:
    774
    Likes Received:
    230
    Location:
    Santa Margarita, Ca
    What's the logic supporting the conclusion that it's a bad compass calibration? In the confirmed cases of bad compass calibration that I've seen the compass errors correlate with yaw.
     
  7. msinger

    Approved Vendor

    Joined:
    Oct 30, 2014
    Messages:
    18,936
    Likes Received:
    5,578
    Location:
    US
    Check out the fast switching between GPS mode and ATTI mode in the log. I don't know what else could cause something like that. Maybe a faulty compass?
     
  8. Rkcorp

    Joined:
    Oct 13, 2016
    Messages:
    6
    Likes Received:
    0
    The rapid change its make by me , the app say that when compass lost
     
  9. BudWalker

    Joined:
    Jun 24, 2015
    Messages:
    774
    Likes Received:
    230
    Location:
    Santa Margarita, Ca
    Isn't the switch to ATTI caused by the compass error?
     
  10. BudWalker

    Joined:
    Jun 24, 2015
    Messages:
    774
    Likes Received:
    230
    Location:
    Santa Margarita, Ca
    SPEED_ERROR_LARGE and YAW_ERROR_LARGE are both "compass errors".
     
  11. msinger

    Approved Vendor

    Joined:
    Oct 30, 2014
    Messages:
    18,936
    Likes Received:
    5,578
    Location:
    US
    That's what I'm thinking.

    What do you think caused this problem?
     
  12. Rkcorp

    Joined:
    Oct 13, 2016
    Messages:
    6
    Likes Received:
    0
    Its caused with me the app say make that during error compass
     
  13. BudWalker

    Joined:
    Jun 24, 2015
    Messages:
    774
    Likes Received:
    230
    Location:
    Santa Margarita, Ca
    Don't know. Been saying that a lot lately. I think in some cases the compass error detection algorithm yields a false positive. That's happened to me a couple of times. Seems like the lone unexplained compass error often occurs with a sharp pitch change. But, then all the other sharp pitch changes don't yield a compass error.

    @Rkcorp could you provide us with the .DAT file for this flight? It's on the P3, not the tablet. Go here to see how to retrieve it. The bottom half of that page will explain how to get the right .DAT. The .DAT will be large and can't be attached to a post here. You'll need to Dropbox it (or equivalent) and post a link to it.
     
  14. Oso

    Oso

    Joined:
    May 19, 2015
    Messages:
    2,533
    Likes Received:
    1,480
    Location:
    Sacramento CA
    I agree with you Mike. On the surface, this sounds like another classic example of calibrating in a bad (magnetically influenced) area. I'm looking forward to seeing the .DAT file info.
     
    #14 Oso, Oct 13, 2016
    Last edited: Oct 13, 2016
  15. Rkcorp

    Joined:
    Oct 13, 2016
    Messages:
    6
    Likes Received:
    0
  16. HWCM

    Joined:
    May 2, 2016
    Messages:
    657
    Likes Received:
    138
    Do not always calibrate before flight. This is a mistake. Calibrate it once and leave it. Unless the app tells you so.
     
  17. BudWalker

    Joined:
    Jun 24, 2015
    Messages:
    774
    Likes Received:
    230
    Location:
    Santa Margarita, Ca
    @Rkcorp I took a look at the .DAT and was unable to find the causes(s) of the "compass errors" in this flight. Although reported in the Go App as a compass error, the error is actually the nonGPSCause field being set to either YAW_ERROR_LARGE or SPEED_ERROR_LARGE. This error is not necessarily an indication of a magnetometer value error.

    I've seen several flights where these nonGPSCause errors were immediately preceded by an abrupt change in gyro or accelerometer value. That didn't happen here. The change in gyroX occurred after the error.
    upload_2016-10-14_5-58-11.png
    The motor data changed at this point but there was no control input. I.e., it was the Flight Controller, not the pilot or a burst of wind that caused this.

    The GyroZ drift rate was determined to be 0.1 deg/sec which is about average. An IMU calibration would likely lower it to .03 deg/sec.

    I looked for any magnetometer data that would support the possibility of a bad compass calibration. What I was hoping for was something what occurred in another flight I looked at. The pilot performed a test where the P3 was rotated CCW through 720 degrees. In the interval 60 thru -120 degrees the nonGPSCause field had the value COMPASS_ERROR_LARGE. The red line is the yaw (or heading) and comes mostly from the gyro data - magnetometer data has a very small impact on yaw. The green line is a heading value that is computed directly from the magnetometer data. It's shape and relation to the actual heading is exactly what would be expected if the compass wasn't calibrated. After this test the pilot did a more careful compass calibration and another test flight showed this problem was fixed.
    upload_2016-10-14_5-38-23.png

    The magMod data was normal throughout the flight.
    upload_2016-10-14_6-13-24.png

    I couldn't find any reason to conclude there was a compass calibration problem. Maybe I missed something. @Oso you mentioned that this looked like a classic compass calibration problem. What characteristics would indicate this?

    It's been suggested to re-calibrate and see if the error doesn't happen again. The problem is that if the error was random it probably won't happen on the next flight anyway. Maybe a better approach is to first try to reliably reproduce the error and then look at changing things.

    Mid-flight YAW_ERROR_LARGE or SPEED_ERROR_LARGE errors can be scary. But, in the ones that I've looked at navigation or flight control has never been compromised. I personally had a few of these and have ignored them. Actually, the first was close to some power lines close to my house. Subsequent to that flight I flew many time close to those power lines in an attempt to recreate the problem. Never happened again. BTW, replaying that flight in the Go App didn't give any indication that an error had occurred. (During the flight the message "Compass Error" was displayed). After a newer version of the Go App was installed the error was indicated in replay as a greyed out screen and something about a "GPS Error". Yesterday I loaded the .txt of this flight into my android and the Go App has no indication of the error on replay.

    I'm reluctant to make any suggestions because I don't really know what the cause is. But, as I stated above, I've had a few unexplained mid-flight nonGPSCause errors and have ignored them. If they start to occur more often then it's a trip to DJI repair.

    One thing I should mention is that if a compass error occurs immediately after launch get it on the ground ASAP. And, don't, under any circumstance initiate, an RTH.
     
  18. Oso

    Oso

    Joined:
    May 19, 2015
    Messages:
    2,533
    Likes Received:
    1,480
    Location:
    Sacramento CA

    Hi sir. My
    comment earlier was based only on what we have historically seen. It wasn't based on any actual .DAT info since we didn't have that at the time.

    For example, we have advised pilots countless times to not do a compass calibration in a magnetically influenced area. Even so, we routinely see new threads consistent with doing a compass calibration in a magnetically influenced area. Those threads often note that the pilot calibrated a compass before a flight, but then saw erratic flight and compass errors once launched and clear of the magnetic influence. Also, these threads are often started by relatively new forum members who have not searched the forum to see the countless similar threads.

    Prior to .DAT data being available, this thread seemed to fit the mold very well and looked like there was a high probability of this being a "good" calibration in a bad area. Your data analysis seems to show that this particular thread may be an exception.