Discussions

ArduPlane crash - Log looks good - Can't find reason for crash

Posted by Brian Christal on September 4, 2015 at 6:24am

Please see attached log and screenshot.

Pixhawk
Flying wing
Arduplane 3.3

Flying mission with auto takeoff and waypoints. Plane perfectly executes take off and heads to first waypoint. Upon arrival at first waypoint plane pitches down until impact - even though desired pitch is up. No failsafes. No errors I can find. Z vibration low at time of failure. Any ideas?

Thanks!

-Brian

76_short8.zip

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

Join diydrones

Email me when people reply –

Replies

Andre K. September 5, 2015 at 7:16am

regarding your PM:

Yes, DesRoll & roll matches at line 48827 , but watch from ~49000 your roll does no longer follow and pitch starts dropping too.

It was late yesterday - it seems i were reading wrong lines, at time - must stop working too much on things :)
- Justin > Andre K. September 5, 2015 at 9:51pm
  
  So what would be the main cause of this?
  - Andre K. > Justin September 6, 2015 at 1:26am
    
    any problem that may cause contol surface(s) not move accordingly to PWM output. , wiring, servo power, servo, etc..
    - Brian Christal > Andre K. September 6, 2015 at 4:56am
      
      Can you see PWM commands in the log? The plane is still intact after the crash. It is a flying wing. Both elevon servos and control linkages are good. Seems like something was lost between Pixhawk Desired Pitch Up and the servo. Also, since it's a flying wing the plane will spiral down if you lose a servo. Can you tell from the log if the plane is spiraling? Seems like the log would show an oscillation. I will take a plane out later today and send it down in a spiral in stabilize mode and see.
Andre K. September 4, 2015 at 11:00am

look at the crazy uncommanded roll you have after log line 47600 , it banks over for no reason (RCOU.ch1 & ch2) does nothing to cause that - it tried to roll back later, but is unable,

most likely a control link fell apart or loose control surface.
- Brian Christal > Andre K. September 4, 2015 at 11:17am
  
  Thank you Andre for the excellent analysis. I suspected mechanical failure. Please help me understand how you see the uncommanded roll in the log?
  
  Thanks!
  
  Brian
  - Andre K. > Brian Christal September 4, 2015 at 11:47am
    
    I see a rapid roll (before you loose altitude) , your RCIN tels me it's not manually commanded, at the same time i see RCOUT.ch1 RCOUT.ch2 (your elevons servo signal) does not deviate from "normal"
    
    - then, as you got big roll angle, you can see autopilot try to correct the roll with very big servo output (compared to usual flight till then), but it does not help...
    - Brian Christal > Andre K. September 4, 2015 at 12:03pm
      
      Is this the right graph to show the roll? Thanks! Brian
      
      https://storage.ning.com/topology/rest/1.0/file/get/3702797567?profile=original
      - Andre K. > Brian Christal September 4, 2015 at 12:10pm
        
        I've been using ATT.Roll , but it'do.
        
        Brian Christal > Andre K. September 4, 2015 at 2:10pm
        
        Sorry to beat a dead horse, but I don't see the uncommanded roll. Roll and Desired Roll are somewhat aligned. Pitch and Desired Pitch clearly diverge.