APM 2.6 DISARMING MOTORS IN MID AIR AND CRASH ( ERR:EKF FAILSAFE)

Hi guys,

Couple of days ago my quad crashed in a way i could never expect. I was flaying for couple of minutes when i saw some warnings on my laptop like "Bad compass health" but i continued flying.

When i was close to my launch point, quad automaticly swiched mode into "LAND", I changed it into AltHold, then RTL to see if GPS is functional and after 30sec or so quad just disarmed in mid air and crashed from 30 meters.

I know that i should avoid any warrnings and resolve them immediately but APM should NEVER disarm in mid air right?

So could any one help me understand what happend? I'm attaching my logs from APM and my GCS.

For what i can understand the "ERR:EKF_CHECK-2" and "ERR:EKF_CHECK-0" has something to do with it but could it disarm my quad?

Thanks for any help :)

2015-03-09 17-30-40.tlog

2015-03-09 18-09-49 1.bin

Replies

Markus Horeld March 15, 2015 at 3:17am

Sorry, no real answer, but I am really baffled, how often you read from EKF related crashes with APMs (which has no EKF at all) since release of Arducopter 3.2. Is it possible that the APMs have a problem with this code?
- Przemysław Kątny > Markus Horeld March 18, 2015 at 6:41am
  
  My gues is that perhaps is just becoming too much too handle for rather slow CPU of APM, I mean all of the calculations. I should also mention that durnig the flight I was recordin nearly all parameters to dataflash logs and I heard that it can drasticly slow down execution of some loops.
  
  I chcecked my logs with AMP Log Analyser which I found here (great tool by the way):
  
  http://www.rcgroups.com/forums/showthread.php?t=2151318
  
  and it clearly states that one of the problems was for example:
  
  "The number of long running main loops is 79 @ 129.909ms"
  
  So i undesrstand that it sits too long in some loops executing them, right?

This reply was deleted.

Discussions

APM 2.6 DISARMING MOTORS IN MID AIR AND CRASH ( ERR:EKF FAILSAFE)

Replies

Activity