EN RU CN
Yaw Roll combination oscillation
  • Hello everyone,

    I seem to be having a problem where if I tilt my gimbal forward or back past a certain amount, a strange oscillation develops somewhere between the roll and yaw axis. If I turn off one or the other, it doesn't happen. This is on a heavier camera setup with 8108 motors. It seems like this could be a balance/inertia issue but I've tried my best to balance the whole rig. That and I find is strange that it's a lot less susceptible to this problem when rolling side to side. I've tried tuning it out, but can't seem to do that and get acceptable stabilization.

    Does anyone have any suggestions or a method through which I can try to sort this out?

    Thanks,

    Marc
  • Balance has to be spot on. Can you roll it with motors off and it stays where you let go? If not keep working on it. Next, look at redoing the ACC cal. May be that. Also, try it without Follow Mode on. If you can get it stable without follow mode, then, look to those settings as the culprit and keep trying to easy into those...Yaw only...hit the auto button, reset the board, increase dead band and so on.
  • Thanks for the reply Dave. You're blog posts on the other forum are actually what drove me to build yet another piece of camera hardware ;) To answer your question, balance is good on all axis. I'll definitely redo all the things on your list...though truthfully I've spent the last few days checking and re-checking the very same things. I do appreciate the confirmation that I'm refining things in the right way though. Definitely trying without follow mode for now to eliminate variables. Will take another stab at it today.

    Also tried various positions for the IMU (which I thought could be a contributing factor). Unfortunately no appreciable difference though between mounting centred on the Yaw axis (under the camera) or mounted directly beside the camera.
  • Getting closer. Sorry for the online diary...but maybe this triggers some thoughts from someone that has seen a similar process.

    Rebalanced: nothing changed. Recalibrated: nothing changed. Even disconnected the yaw expansion board to see if that helped and reconfigured for only Yaw and Roll on the main board: nothing changed. Removed remote: nothing changed.

    So far, it seems I have to use almost the full D values on Roll and Yaw (with Gyro High Sensitivity enabled) to get rid of this. Seems crazy as there are downside to that as well but it's getting me closer. Follow mode does not work too well with aggressive settings like this which makes it all the more confusing and hard to dial in.
  • mpk if you are using the 2.3b firmware go to follow mode and after checking the follow pitch roll yaw you will see two check boxes to start the roll and finish the roll. Put those to 90 degrees and you will stop that shake. What the gimbal is doing is that at zero degredes is trying to compensate what doesn´t need to...so starts shaking. Godd luck and let me know the result.
  • Thanks for explaining that kikojiu. That's one of the fields I have yet to play with in the GUI. Unfortunately this problem exists even before enabling Follow mode. That said, the byproduct of using the aggressive high D value tuning setup (that now "sort of" work in normal mode) with Follow mode is a high frequency oscillation when tilting up (more so than down). I think I will sort out follow mode after I get the normal operation dialled in.

    Maybe I'm just expecting too much from the gimbal but it seems like I'm so close to the performance I was expecting but just have to get over this specific problem.
  • There are many things which are not explained on the manual. This Alexmos guy is a genius so he takes for grande some things we need to be explained. Then we notice loads of small problems that together become a huge problem. Besides Mr. Alexmos does´t answer my e-mails or my messages here... :-(
    Another thing I don´t like is when people come here ask things when they have problems...they solve their problems and dom´t come here or help others...
  • I know what you mean. I'm actually enjoying the learning process and figuring out the way all the different things balance out. It makes for a better understanding how a device works...when there's time outside of production time that is. Luckily there are forums where people can discuss. Hopefully my learning process will help someone else down the road.
  • Back to square one as of today. for me. Crazy aggressive D values don't really work (as I expected) as sometimes it goes into high speed oscillation even under less strained circumstances. So I'm back to trying to see why the first issue of low frequency oscillation happens when I tilt the gimbal forward and back...
  • D as to be low. Ideally should be around 10. Try to identify which axis is causing the agressiveness. And switch off the gyro high sensitivity. It actually is like doubling the P values.
  • Indeed. I'd like to get D as low as possible too. Definitely understand the relationship/tradeoff there. It's all a balancing act that's tricky to sort out. Which is why I commented on being so hesitant with the aggressive settings. The strange thing is that there seems to be some relationship between axis that's causing this. Enabling and only one at a time doesn't cause the oscillation to surface.

    I will try again with "Gyro High Sensitivity" off, but as you said, it's essentially like doubling P and D values and I try to setup the settings as low as possible either way. Can't go too low since there needs to be torque/power/speed to move at least a 2kg load. When the settings are too low the gimbal doesn't perform enough stabilization.

    It's the interdependent relationship that I find confusing. Wish I could isolate it to only one axis. I can however say that Yaw requires the most D but not without substantial increase to D of roll. But as I said (and you mentioned as well), that unfortunately is not really a "solution" to the problem.
  • I posted this on my other thread and thought I'd continue here since the issues seem to have merged into one.

    Theory (that applies to my other underlying problem): The closer one gets to the extremes of travel the closer we get to a condition where two different axis line up. As a result we get to a sort of feedback situation where two motors achieve the same effect.

    For example: at 90 roll, the yaw and tilt motors effectively perform the same action.

    In between these extremes there is a transitionary state. You can see the yaw influenced for example when you move the tilt and roll motors (easier to see when those two motors are off). The system doesn't necessarily know which axis to correct and could go into a feedback/oscillation state.

    Is it possible that this is the cause of why the gimbal oscillates the further one pushes it from the centred balanced position?
  • Just wanted to mention that this behaviour has been concluded in this thead:
    http://forum.basecamelectronics.com/index.php?p=/discussion/564/imu-faulty#Item_9

    Essentially, you cannot tilt the frame too far in non-follow modes.