Strange Vibration Issue

I first posted this in Unassigned, but it wasn't getting any traction.

Hi All,

I have been chasing jello in my video for about seven months now. Somewhere along the line I've introduced a new vibration problem that I don't understand and haven't been able to fix.

In addition to the typical jello that is caused by prop/motor vibration, I am now seeing a kind of jitter in the pitch axis. It is not oscillation, as far as I can tell. It is more random than that. It happens whenever I give the aircraft horizontal stick. It also doesn't seem to be prop/motor vibration, but rather something else.

When I first noticed the issue I was in the middle of a series of gimbal mounting changes. At first I thought that there was some interaction between the airframe, flight controller, and gimbal vibration dampers. However, when I went back to an earlier version of the gimbal mount, that has almost no damping, the problem persisted. It also persists through autotune, and repeated motor/prop balancing efforts. I can't take the aircraft back to a state where it doesn't have this jitter.

The linked video, recorded on my goggle DVR, shows the problem. My FPV cam is hard-mounted to the airframe, so it is actual airframe movement that shows up in the video.

Has anyone ever encountered anything like this?

I've also included a screen shot of the Pixhawk's vibration log. It shows quite a bit of movement around the z-axis.

There is also a pic of the aircraft for reference.

https://www.dropbox.com/s/40s6hfy3pl17elk/Jitter_DVR.avi?dl=0

https://www.dropbox.com/s/i5xk6fxjkk8da7w/27m_s%20z-axis.jpg?dl=0

https://www.dropbox.com/s/3zx3nrp5anb276j/View%20of%20Aircraft.jpg?...

Aircraft info:

ebay TBS Disco knock-off frame

HobbyKing folding arms

Pixhawk, APM3.3

Foxeer HS1177 on homemade mount

Feiyu Tech 2-axis gimbal on homemade mount

KDE 2315 motors

APC 10x4.5 MR props

Dragonlink Micro Receiver

ImmersionRC 600mW VTx

Cheap MinimOSD clone, The reason for the glitchy video in cold weather

Maytech PDB with integrated regulators

Multistar 5200mAh, 4S battery

Some other junk

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

Join diydrones

Email me when people reply –

Replies

  • Hi Graham,

    Thanks for the reply and your suggestions.

    There is one more moment arm that I could eliminate. That is the external static pressure port. It has very little mass, but it's worth a try at this point. I guess I could also try flying with the GPS mast folded down, but this hasn't changed from the original aircraft configuration.

    I haven't had the sack to approach the aircraft in flight, but I have spent many hours with the thing resting on or strapped to my workbench (with vibration damping). I have been struggling with prop/motor vibration for a long time. It's been such a strange mystery that I even had all of the motors reworked under warranty. Three of the four improved considerably. The fourth (which was actually replaced altogether) is worse, and I'm still working on it as I have time. Vib on three motors is between 0.75-1.25m/s^2. Vib on the fourth is around 5m/s^2. I'm with you in thinking that there is still the possibility that my video issue could be due to this last motor/prop, as I've now exhausted all other possibilities that I can think of.

    I've been so desperate that I've been playing around with things like how well centered the prop hubs are on the motor shafts, and trying to eliminate any play. I've been screwing around with things that other people don't even need to worry about in an effort to understand an solve this issue.

    For reference my motor/prop balancing procedure:

    Dynamically balance motor alone, with a cell phone accelerometer and tape. I can get all motors to around +/- 0.2m/s^2.

    Statically balance props on a Du-Bro True Spin, then on a magnetic balancer for fine-tuning. I also balance across the hub.

    Attach prop to motor and dynamically balance to find the best prop rotation on the motor shaft.

    Dynamically balance motor/prop combination to trim tape amounts for lowest vibration.

    In most dynamic tests I run the throttle on one motor at a time up to 35%. This is about the max I can get without having the aircraft strapped down. I don't like to strap it down due to the impact on vibration readings.

    The weird part about this is that I didn't have this jitter issue when the motors and props were more poorly balanced than they are now. Whether or not that change has anything to do with the problem, I don't know. However I hesitate to try to fly with worse balance than what I have. I have had some scary crashes and close-calls due to the Pixhawk's vibration intolerance (in earlier projects).

    Sorry for the verbose reply. I hope that gives you a little more background on the state of the troubleshooting efforts.

    I'll see if I can get an answer for you on prop wobble. It will probably be after the holiday before I can get in a test flight.

    Happy Holidays!

    Chris

  • Moderator

    As you've tried taking off the VTx extension the only thing I can think of is checking the prop tracking, often a cause of high Z vibs, a slightly bent shaft could also cause this. Run up the motors slowly and from the side watch for the tips not aligning to each other.

    Can you hover on a calm day and feel each arm just near the motor individually? If one is vibrating you should be able to feel it.

    Also, anything 'lengthy' like the antennas/long gimbal arms/etc can exacerbate any minor vibrations.

  • In case anyone is interested in this issue, I have continued to work on it.

    Additional actions taken are:

    Swap arms for original fixed arms

    Reverted to early gimbal mount

    Removed VTx antenna extension

    Disabled camera gimbal, now fixed in place with tape

    Reflashed Pixhawk with Rover firmware, then AC 3.4.3

    Formatted microSD

    Still having vibration problems like the video above. The only time a noticed an improvement was after firmware reflash, in a short flight before Autotune. I would be tempted to leave the PIDs in that state, except for the controls being really mushy.

    If anyone has any thoughts on this issue, I would be most grateful.

This reply was deleted.

Activity