EN CN
Gimbal goes crazy with altitude?
  • webdr
    Im having a problem with my gimbal and I cant for the life of me figure it out.
    I have a Yun 1 gimbal and Alexmos board. Firmware is the latest one and so is software on the SimpleBGC_GUI. I get great video stability. Quad flies fast or slow, banks hard or hover, wind or no wind, its very stable. But as soon as I climb past, lets say 30 to 50 feet in height, the gimbal goes crazy and starts spinning on both axis, also pulsing. I have to land and kill the power to it, then restart and its fine until I climb up and start gaining some altitude. It doesn't matter if I climb fast or very slow... as soon as I get around 30 feet above ground it tilts and starts doing the "crazy chicken dance" as one member put it.

    I will say that its very cold, like 20 degrees F. I've initialized outside and let it acclimate to the cold before turning on. I also calibrated multiple times, checked balance many times and got it very precisely balanced. Ive also checked and unchecked the "initialize gyros on startup" with no difference.

    Any help would be appreciated.
  • webdr
    Anybody able to help?
  • Garug
    As there is nothing sensing the altitude, it can only be turbulence at that altitude. Are you flying near forest?
  • webdr
    It isn't a wind/turbulence problem because I can move the gimbal when its powered up and it always goes back to center. If the wind was moving it, it should go back to center or at least try to, not just start going haywire on both axis at the same time and not stop until I unplug it.

    Additionally, Ive flown in some pretty stiff winds down low and no issues, Ive also flown in dead calm where no trees are moving at all, started climbing and as soon as I near 30 feet or so, it goes crazy. That's not even above treetop level.

    Its got to be some other reason.
  • Zeb
    The only thing that is changing is the air pressure. Now the inertial measurement is solid state but is a physical IC with nano size sensors which are detecting movement and the resultant capacitive change provides the feedback to the system. If there was a fault with the IMU whereby the reduction in air pressure resulted in the above process being corrupted you would see the result you have described, this is a long shot and very obscure but I can't think of anything else that is physically changing in relation to height.
  • Garug
    If It would be Air-pressure, there would be days when it would vibrate at ground level and it would vibrate if changing to a flying site that is 50 ft higher than your current one. I think it is highly unlikely this small change of air-pressure would affect IMU like this even if the IMU would be broken.

    On this video after 0:50
    there is vibrations after passing certain level. (This is a Tarot gimbal, with some modification and the vibration dampening still needs work... Actually it will be totally rebuild with Basecam electronics )

    Sumer, winter, wind, calm, there is often a layer of air that is difficult even to penetrate with a multi copter it is not always there but easy to notice when it is. This is something I have experienced, but when now briefly looking in Google, this might be something that explains it: http://lukemweather.blogspot.fi/2011/06/nocturnal-boundary-layer-wind.html
  • webdr
    Heres a video showing how it happens:
  • Garug
    If you look the visible parts of frame, the frame appears to start vibrating before the gimbal loses control. I would try less stiff vibration dampening for the gimbal assembly and PID tuning. Now it seems to be just on the edge of loosing control.

    Note: cold makes many vibration dampers stiffer.
  • webdr
    If it was vibrations, why is it only when at 30+ feet in the air and not lower when it starts spinning erratically. Also, This rig's motors and props are perfectly balanced. I spent almost a week balancing all 4 props and motors dynamically. I can put my hand on the frame and cant feel any vibes. I think the shaking is the downwash from the rotors...

    What tuning aspect in the software could it be?
  • Garug
    But the vibrations come from the boundary layer and yes from the downwash from the rotors ;-)

    How does the gimbal act if you hold the multicopter in hand (motors do not need to run), tilt it 20 degrees and give it couple of sharp hits (not that powerful but just to get good vibes) Is there difference if you do this in cold?

    I would optimise the parameters so that you can tilt the multicopter some 40 degrees and give sharp hits and the gimbal is not disturbed. so not only for the minimal angular deviation. Using FW 2.4 and the FC functionality would hep too.

    I would start by lowering all PID values by 10%

    But I have to admit, it really looses the control strangely and consistently. Have you been flying in different weather conditions or at different time of the day, is it always the same?




  • AOPEN3434AOPEN3434
    hello,
    seems apparent that this is vibration issue hiy the throttle grater the vibration hiy the altitude.
    Balance your props?
  • webdr
    Ive balanced the props very well. Each one is balanced at the blades and the hub. Ive touched various parts of the quad while hovering and its one of the smoothest ones I have built.

    Im going to try Garug's advice of turning PID's down 10% and give the frame a few hit to see if it goes crazy
Add a Comment