For the life of me I can’t seem to get my quad to be stable in any wind. It holds position decently well, but even the slightest wind causes it to twitch in roll and pitch. VIBE values seem to be ok. I have no notch filtering applied as there were no obvious peaks in my logs. Quad is a Tarot 650 with 15in props. Any assistance would be appreciated!
Compressed BIN and params:
https://drive.google.com/file/d/1Z1FCHaxFrWxQOEna46cVPtlSoZxhaymT/view?usp=sharing
Fix above and do another hover flight. With a fully charged battery…
Out of curiosity what tuning guide did you use?
You have 2nd notch configured, but your gyro log looks very strange, you should see some motor vibration (if your setup is not pusher), I believe it’s some kind of logging issue, please switch to batch logging, your CPU can’t handle current load.
That’s what you will see with Raw IMU and Raw Log Option enabled. Neither should be enabled on this FC and not both on anything.
Ok, I fixed the logging, PSC, and MOT_SPIN values. My MOT_SPIN_MIN was correct, but I changed MOT_SPIN_ARM to proper value.
Will working on notches now: https://drive.google.com/file/d/1U5t4LhE_SLJk3GvwUpsJT5oYt3b5JA5I/view?usp=sharing
At 22% throttle? Very high, what ESC’s are they? Calibrated I suppose?
I would work on lowering the Z-vibe levels before advancing. But, at least now you have something to see in the FFT graph.
Yes, they are calibrated. XRotor Pro ESCs
I tried working on ZVibe levels in the past by gummy mounting the FC, but it ended up making things worse.
Not isolation, the mechanics of the vibration. Props balanced, nothing shaking around like landing gear, GPS stand, crappy arm folding mechanism, etc.
Now you have a lot of peaks, but I never saw so much harmonics, what propellers do you use?
Try filtering like that.
also disable FFT calculation, you CPU can’t handle it.
FFT_ENABLE,0
I will try that. The quad has 15in CF props. They are not plastic composites. The drone has a CF mast with mass flow sensor mounted on top. Its fairly rigid.
Still pretty twitchy: https://drive.google.com/file/d/1ToXzKoKGWnnRMFxNMHPLWTADeF9Ip8k5/view?usp=sharing
Probably not rigid enough. I suppose there are several modes of vibration on that tower (torsional, bending)
Just to be clear, the twitching and instability is only in wind. When I autotune at night it’s very stable
Just to be clear, if it is built and tuned correctly it will not twitch with our without wind.
So it must either be the build and/or the configuration.
Your D oscillating very much, the main reason is all the noise that coming in, than your PID values looks wrong - too much D, maybe reset all your PIDs to default, I made 15” that flies perfectly with default PID values.
Then your noise is very bad, especially in Z axis, IMO the frame is not rigid enough or/and the AUW is too big for this frame. You cannot start with autotune with such vibrations, ideally post filter noise should be below -50db and delay does not cross -180 degree.
Next problem is your FC load, more then 70%,
it can’t keep up, loop rate struggles
Try lowering the amount of data by defaulting LOG_BITMASK and disabling all imu logging, 1 is enough.
LOG_BITMASK, 180222
INS_LOG_BAT_MASK, 1
We have seen this before with that Flight Controller. I would not suggest any products from Radiolink. While this doesn’t apply specifically to this flight controller it’s damning enough to shun their products:
RadioLink are in violation of ArduPilot’s GPLv3 license in that they refuse to release the source code for their modifications to ArduPilot.
Why would you not see motor vibration if you setup isn’t a pusher? I did I misinterpret your response?
On pusher air stream from propellers does not cross arms, which is main source of motor noise.
Interesting, and the ”sucking” of air from top doesn’t affect as much? Is there some paper/measurements on this topic?
