Quadcopter unexpectedly ascends extremely fast

Hi,

I am fairly new to the ArduCopter and RC flying in general.  I have about 50 flights behind me and most have gone really well.  Today, I crashed twice as a result of my quad ascending very rapidly even though I was using moderate control inputs--it was definitely not behaving as it has in the past.  In both cases, I panicked, cut the throttle completely and was unable to recover before crashing into the ground.  The first flight was in loiter mode.  I then removed my brand new (now broken) gimbal and recalibrated the compass and accelerometers.  The second flight started in stabilize and I then switched to altitude hold mode as was trying to rule out a GPS issue.  The same basic thing happened.  Logs for both flights are attached.  Can someone help me make heads or tails of what is going on?

Frame: DJI F450

Controller: 3DR Pixhawk

Arducopter 3.2.1

Many thanks!

Colin

2015-04-03.log

2015-04-03b.log

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

Join diydrones

Email me when people reply –

Replies

  • Maybe this

    Common Problems

    1. High vibrations can lead to the copter rapidly climbing as soon as altitude hold is engaged.  Check theMeasuring Vibration and Vibration Dampening wiki pages for details on how to measure and reduce vibrations.
    • Hi Milo,

      I got the quad rebuilt (it goes a lot faster the second time when you know what you are doing) and did a bit of testing.  Vibration seems OK.  See the graph below.  The axis for each value is set to the allowable limits (+/-3 for x/y and -15 to -5 for z).  I also plotted the barometer's altitude since I was not just hovering the whole time.  You will see I was bringing it up and down fairly quickly in the second half.  It definitely was getting more throttle than when I saw problems in alt hold, though they are not directly comparable.  There are some excursions on the z-axis out of the range, but that does not appear to be vibration--its right at the point I am peaking and cut the throttle back quite a bit, so it was indeed accelerating quite aggressively.  Thoughts?3702557851?profile=original

      vibration.log

      Screen Shot 2015-04-03 at 7.02.06 PM.png

    • Hi _Milo_,

      Thanks.  I did this measurement back when I first built the quad, and while I do not recall the exact results, everything was well within the prescribed limits.  I have not changed anything with the quad that should have affected that.  I am still using the same stock DJI propellers and my controller is still mounted with the 3DR-supplied vibration dampening foam.  

      I should add that my recollection is not that it climbed right as I engaged alt hold.  I believe it was raising the throttle just a little bit from the level I had it at when hovering.  

      I will make new vibration measurements, but first I need to fix the damage.  Did the logs lead you to suspect that was the problem, or was it just the symptom?

      Thanks!

      • Sorry, no I haven't looked at the logs,

        I had a similar problem with Vector FC I upgraded motors on, and would get a huge uncommanded throttle, like WOT, it was from a vibration problem stemming from one of the motors. It would do this like in a alt hold mode, when I turned against the wind at speed, turning into the wind, would increase vibration.

        Even though you did a vibration test in hover, you may be getting more vibration at different throttle levels, that is enough to trigger this. Also even if you were fine before, if like a motor bearing is wearing out, or props nicked, etc...could be that you now have excessive vibration.

        • Good points.  Thank you.  Once I get it rebuilt, I will rerun the vibration test, though I am a little afraid to get airborne again without knowing the cause yet.  Is it safe to assume that if I stay in stabilize mode I am reasonably safe against this happening?  

          • Yes I'd say so, I'm still a noob on pixhawk, when it happened on my Vector switching out of alt hold mode would regain control, also the vector would kind of re-set itself. It would look pretty scary, but I kept enough altitude to recover.

            I have a video of what that behavior looked like if it might help, different controller but same issue.

            • Thanks again.  If you want to link to the video, I'd definitely have a look.  The behavior will long be seared in my memory ;-)

              • I thought it was some kind of tuning issue, because of the new motor/esc DJI secret sauce mix, but it turned out I had one bad motor (vibrations), tried balancing it, but I think it may just be like a magnet is slightly misplaced something is not equally balanced inside.

                https://www.youtube.com/watch?v=ffu5jgNL4Os

This reply was deleted.

Activity

Neville Rodrigues liked Neville Rodrigues's profile
Jun 30
Santiago Perez liked Santiago Perez's profile
Jun 21
More…