It seems i am encountering a high Z vibration on a 1400mm Hexacopter with 22inch props. https://plot.dron.ee/BDzB
I have been trying to figure out solutions to solve, since as its noticable that three minutes into the flight clip0 jumps to 40 and 5 minutes in, increases to 900+
Clip1, 5 minutes in jumps to 32 and clip2 jumps to 300+
At the same time X-Y vibes peak at 30 while Z goes crazy and accelZ is pretty close to max values in many occasions.
Is there some “soft” dampening solution, since frame has been checked over and over for flex and vibrations.
The whole issue causes the HEXA to start rolling (5minutes in) at extreme angles and after about a minute it recovers and starts flight normally again.
Any assistance would be appreciated
(Upon investigation under the Cube, i found that the actual base plate has flex and in the meantime i will switch it with a 2.5mm carbon to verify if this will lower the issue, besides this structural fix, since flex is almost zero for the size, would ins_accel filtering help?)
Hi, this is exactly what my medium size Hexa did. I have a Pixhawk 1 with a wooden prototype frame that I imagine isn’t the sturdiest. (32”/around 800mm with 15” props). I tried everything from tuning Accel filter, to adjusting accelerometer weighting on altitude and position, with no real luck. I had the same Clipping issues in the data logs.
My filter flew fine until I switched to ALTHold or Loiter then the drone would fall out of the sky in a bad series of twitches and bumps.
In the end, the fix was buying a pixhawk vibration dampener (Amazon around $25).
I had tried bubble insulation mounting with no luck, And I also tried a new pixhawk board because my old board had been involved in a crash, so I had wondered if the Accels had been damaged. I will try to test it again with the old pixhawk to complete the experiment but I would highly suggest those fiberglass dampeners. Now I can sit and drink tea while letting the drone hover at a standstill. No clipping whatsoever.
Vibrations are coming from a faulty or unbalanced mechanical part, without fixing that you will just hide the issue with more dampening.
Option 1 : Fix the faulty part, balance your prop, use quality part.
Option 2 : If the vibration is amplified by the frame structure then reinforce/redesign frame.
Option 3 : Do a FFT analysis, if vibration is only at one specific frequency then set the notch filter… (note: spread vibrations cannot be filtered.) ardupilot.org/copter/docs/common-imu-batchsampling.html
Option 4: IF you don’t want to do any of the options above, then as a last resort you can weight down the Pixhawk1, put it on a vibration dampener mount (https://www.thingiverse.com/thing:163472) and put a 20-30g weight (Fisher lead) under the top plate. Remember this just masks the issue… but in many cases makes the copter flyable. (For the cube it is not a solution, since the IMU in the cube is already weighted and isolated).
Thank you both for your replies. @Eosbandi
Option 1 checked even before flight.
Option 2 Frame it self checked both on simulator and actual flight and at the moment all point to an additional payload not build by my self and its ancoring.
Option 3 Great suggestion and at the moment i am designing a test timetable with or without payloads in order to pin it down, since on my logs with dummy payloads i have not had anything high on vibrations and all clips were flat.
Option 4, i will keep in mind, but cubes performance is always superior.
As about the quality of parts, they are all high end since only on my test builds i only use ali stuff. This is all around T motor u7 and Tm props and ESC’s, good cables, Vulcan PDB.
In any case i will follow your lead and change a couple of things and run FFT test.