Discussions

APM 3.2 QUADCOPTER Err: EKF_CHECK-2 / Err: FAILSAFE_ EKF-1

Posted by Gonzalo Garat on December 15, 2014 at 4:26am

Hi all,

I'm attaching my log from our last flight as our little quad went completly bonkers before it crashed.

Short after takeoff it went completly out of control, no response at all from RC or Telemetry and then crashed near a pond (thank god).

But the mistery was about to begin when we started reviewing the log and found out that the trigger was the failsafe error  Err: EKF_CHECK-2 / Err: FAILSAFE_ EKF-1. 

In understand these two were recently added to APM, but this ain't the first time I see this sort of thing happening.

I'm attaching my log in hopes for some guidance.

Thank you all!

2014-12-14 11-10-14.log

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

Join diydrones

Tags: ekf, failsafe
Email me when people reply –
Follow

Replies

  • Mikko Korhonen September 20, 2016 at 3:17am

    i had same errors today. in gps mode, a sudden lost of altitude.

    I have had same thing before, so i was high enough to have time to change to STAB mode and make a save.

    Any info what is causing this? magnetic interference ?

    here´s log

  • Nicholas Bratton April 16, 2016 at 6:13pm

    Hi All,

    I've had a similar issue where my quad wouldn't respond and it ended up slamming into the side of a concrete building right before narrowly missing a parking lot of cars. Luckily nothing or nobody else was hurt but the quad. The 3 axis gimbal, Hero4 black, frame, etc are a total right off. The logs show similar Err: EKF_CHECK-2 / Err: FAILSAFE_ EKF-1 errors. I would appreciate any help root causing what went wrong. It was cold day in January and I'm not sure if this contributed to crash. 

     

    2016-01-18 16-51-45.log

    2016-01-18 16-51-45.bin

  • Robert Caraway March 9, 2016 at 4:35pm

    Paul

    I am new to the forum and need to find someone with expertise in interpreting error.  I seem to get servo fluctuation while on the ground in FBWA mode as well as other auto modes.  I think this is traced to my IMU2 being noisy.  Any D that I add to Roll seems to be amplified by the noise.  I am not sure if this is normal noise or not.  I can see the plane twitch in the air from the gyro noise being modulated though the servos.  Any idea how I can filter this out?  I have attached a file of my plane sitting on the ground, no wing, no wind, no motor.  I did lift the tail at one point in the file.

    Any help would be much appreciated.

    Scot

    2016-03-07 12-31-46 7.bin.jpg

  • Tuloski November 10, 2015 at 4:45am

    Yeah, perfect analysis...in fact close to the ground the magnetometer goes crazy...from 40 to 70 degree offset.

    Thanks

  • Tuloski November 6, 2015 at 5:18am

    Hello, I had the same problem and my EKF4.SM are high, up to 2.0. What does that mean? What EKF4.SM are? How GPS velocities are used? And what do you mean by GPS velocities? The SoG coming from GPS?

    It seems the vibrations were fine and no big error on yaw.

    I attached the log.

    2015-11-06 12-24-58.bin

  • Luca November 2, 2015 at 1:25pm

    Thanks for the feedback Paul! I know EKF is not being used as it only is applicable to the Pixhawk and I disabled it in the settings so it's an incorrect "labeling" as you say. Could these increasing errors be due to the gimbal that was glitching? I suspect it shouldn't happen anyhow as the DCM should keep it's bearing in time despite of the external mechanical influences (but not EM). I will try recalibrating the compass and see if the errors are reduced. Also I'm planning on purchasing a Neo GPS with magnetometer so I will be able to mount it further away from the DCM and fields. ATM the DCM is on the APM even if it is mounted on top of the drone. I do have a couple of HMC5883Ls around with 3.3V support (they use 2 transistors to adapt the 2 signal voltages instead of a dedicated TTL-3.3V IC but they work fine) and will try disabling the onboard one and mounting it externally if I don't see any improvements after the recalibration.

       FYI, I previously only had bad DCM errors and never say EKF errors. My GPS is mounted 10cms from the APM so it should be free. I have however encountere GPS glitches in the logs before, though not recently. BAD DCM errors come up when I rotate 360degrees and shortly after go away, but then again, these last 2 flights did not log any bad DCM errors (I did set the DCM/yaw divergence to "relaxed") apart from these incorrectly labeled EKF errors.

        I think I know why it didn't respond: falling behind the wet trees we lost telemetry and it fails to forced LAND when this happens apparently even if radio signal is good as I saw in the settings that this option was set to 1. Maybe I should disable it (I use a 433 MHz telemetry and it should reach over 1km in line of sight @ 20mW output which I set it to) so I can keep the drone in control at max without TX signal it will fail to LAND anyhow.

        I guess I have to run further tests to sort things out! Thanks again for the heads up!

  • Andy Piper June 9, 2015 at 8:40am

    Response from 3DR "Given you are using beta firmware that we do not provide support for, you would need to flash current stable firmware to demonstrate ekf issues." So that's a no then :(

    I tried setting EKF_GBIAS_PNOISE as you suggested and have not seen any TBE since although I probably haven't gone as high either. Overall copter control is good, so I'm happy with this. Thanks for your help.

  • Andy Piper May 26, 2015 at 1:27am

    Hi Paul,

    Thanks once again for the amazing analysis! This seems like quite a subtle issue, do you think 3DR would accept it as grounds for RMA? If the pixhawk is faulty I'd prefer to get it replaced.

    I'll have to check my logs from the same day for similar occurences. Should I expect to see the divergence disappear as the temperature increases again?

    In your graph above 3.5 deg/second doesn't seem like a lot - it's hard to see in the graph, but you are saying that even this much discrepancy can cause serious issues? Did DCM learn the offset as well or is that new to EKF?

    Have you seen this thread at all:

    http://diydrones.com/group/iris/forum/topics/iris-massive-yaw-issue

    ? He has even more severe gyro drift, but it settles down after a while. Could this also be temperature related?

    Thanks

    andy

  • Andy Piper May 25, 2015 at 7:29am

    Hi Paul,

    So I loaded 3.3rc5, straightened the mast, recalibrated the compass, set declination manually and redid compass mot. I ended up with lower offsets and a much higher level of agreement between the compasses. I took it for a flight test yesterday and all was good. Took it for another one today in some wind and while performance was good, I still ended up with some yaw errors and a small toilet bowel. The log is attached. You can see I go quite high and the errors and TBE occur at higher altitude. I'm guessing this is because the wind was stronger higher up, but it's still disconcerting and a little worrying. Note that I pretty much go straight up - there is almost no forward velocity. I've got this tuned about as accurately as I can now, so I'm not sure what else I can do.

    Thanks

    andy

    2015-05-25 12-39-54.bin

  • luke cooke May 13, 2015 at 12:16am

    I have also Had this error once before, After much reading I came to the conclusion that it must be a DCM error. Finally tracked it down to the a yaw error. But still have no idea what caused it???

    It did only appear once and I havent had it again.

    It would be good if the error codes where better documented, and EKF changed to DCM in the code for APM users.

This reply was deleted.

Activity