Discussions

Altitude problem in AUTO mode and flipping over during descent: Crash

Posted by Dave Keck on August 2, 2011 at 8:15am

Hi all,

Yesterday all was going well with my second quad build, flying missions, loiter, ALT hold, etc.

 

But there is one issue that has been there for a while that finally led to a crash.  Ever since I’ve been using AUTO mode the copter does not hold its assigned altitude.  If I have the throttle up and the copter is gaining altitude, then when I switch to AUTO it will start the mission but will continue to gain altitude.  What I do to recover is switch to STABILIZE mode and lower it down slowly.  With my new quad it just won’t descend very fast so I lower the throttle a little more, then the copter will flip over (I don’t know the sweet spot yet and it gets too high to see the descent rate).  The previous two times  it has righted itself in one flip, yesterday it flipped and flipped and flipped and hit the ground from 50’ full power.  Only arms and motor mounts bent this time… and four broken props and divits in the neighbors back yard.

 

So I see two issues:

The first issue is why when switching to AUTO does it gain or lose altitude and not immediately descend or ascend to its assigned altitude?  This seems related to the throttle position when switching from STABILIZE to AUTO.

 

Second issue is why will the autopilot not give a stabile descent?  It seems that if I’m at altitude and cut the throttle the autopilot should know that it’s falling and at least provide enough power to keep it stabile and upright as it descends.

 

Thanks,

Dave

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

Join diydrones

Email me when people reply –
Follow

Replies

  • emile August 11, 2011 at 1:46pm
    My opinion is that it has something to do with the vertical accelartion fooling the DCM. Turbolence is there but it is not the main problem. What you see in Mikrokopter is what I would expect (would like) when descending.
    This is my experience.
    To get a stable quad on descent you need to lower P values a lot.
    This will give a stable quad but low authority on sticks.
    If you try to move forward while descending you will notice you have less authority on sticks. Sometimes even with full forward the copter on dsescent will stay on its place.
    Now imagine you are the APM while descending you will get some turbolence (naturally), but to try to correct you will give more power than needed because you think the quad is already tilting, this will cause the quad to start to oscillate and eventually flip.
    I'm trying to address this problem to the developers, but I think it is a hard work because you have to put your hands on the DCM...
    I wish I could do that.... :)
    CHeers,
    EMile
    experience.to - This website is for sale! - experience Resources and Information.
    This website is for sale! experience.to is your first and best source for information about experience . Here you will also find topics relating to…
  • Vincent Mees August 9, 2011 at 3:44pm

    Here's the log of the tests.

    10-08-11 12-31 4.log

  • Vincent Mees August 9, 2011 at 2:15pm

    Did some tests this evening concerning the uncontrolable behaviour when descending.

    First up was using EXPO 50% and then 100%. Conclusion: not a big difference than before.

    Then came changing the ESC timing from medium to high. Not a big difference than before.

    What I certainly notice is the fact that when descending, the quad at a certain point starts accelerating downwards and thats when the heavy oscillation starts. When I descend very slowly I can keep it stable. The problem is that I can't see the descend rate from the ground.

     

    Perhaps a proposal for a new mode ;-) Controlled descend rate: 1m/s for example 

     

    Looking forward to version 39 ;-)

  • Vincent Mees August 7, 2011 at 3:19pm

    Anyone else having problemsof instability with their diy frame whilst descending?

  • John C. August 2, 2011 at 8:23am
    There's obviously something else going on, but I had similar instability on descent when I first started. The fact of the matter is, you're dealing partially with regular rotorcraft dynamics.

    http://www.copters.com/aero/settling.html

    It's the reason the blackhawk crashed in Abbotobad...settling with power.

    The only way out of the instable silo of air created when descending rapidly is to have enough altitude to move forward, back or side to side into clean air.

    Hope this helps a bit...

    JC
    Helicopter Aviation
  • Dave Keck August 2, 2011 at 8:18am

    This is with 2.0.37 and also I've had this problem with previous builds and a different Qaud.

This reply was deleted.

Activity