Hi all, I'd like some insight as to why my quad wobbles when increasing altitude/throttle. I have lowered the PIDs in stabilize mode (now as low as 3.2 for Roll and Pitch) which helped but I 'm still getting some wobble when increasing thrust apparently. I have attached the flashlog (I think I'm getting too much vibrations on accZ too) and a pic of the drone. I have the APM on a rubber stabilizer which I hope is not the issue (I have not tried removing it and attaching the APM to the F450 directly with some silicon rubber biadhesive). Any feedback is appreciated!
Thanks!
Replies
EKF is not enabled, so only looking at standard vibration levels, the Z axis is quite high. I know from experience that alone can cause wobbling.
Thanks for the analysis: yes I noticed this Z axis vibration too in the logs but it didn't seem to me a major problem. My APM is mounted on a "damped platform" with it's 4 silicon supports. I tried removing even them and taping it with a particular thick (2mm) silicon biadhesive directly to the frame but it didn't change the wobble problem. The problem was fixed as stated above by reducing the P valie to 0.05 under the suggested 0.08 level, it was way too overreactive. Now it's flying fine!
Hi all,
just wanted to update you on my "wobbly" issue, and I finally managed to stabilize the damn thing. It wasn't any hardware problem (though with the 4 new Sunnysky 980KV motors it runs as smooth as ever) but simply PID tuning: Missionplanner wouldn't let me lower the P values under 0.08 for some reason and so I always stayed at the minimum which was way too over reactive for my setup. I had to enter the full parameter list and force it down to 0.05 and acknowledge the Miissionplanner non standard value warning to set it. Now it's as smooth and stable as a "knife in warm butter", with and without my gimbal. It's impressive how the you can shake the quad left or right and it appears as if it is a dense liquid which slows it down dampening all oscillations. Now that's what I was expecting! After a couple of "missions" I'm now just wishing I could get more flight time as with my 3S Multistar 5200mAh pack I get barely 12-13mins to stay on the safe side. I could get more but I'd be dangerously risking an ESC failsafe (which these ESCs still have while the SimonK based OPTO ones - ZTW Spider series - I ordered should not). I got DCM bad heading once and the mission "Failsafed" but I think it's because I routed 2 of the 4 ESC control wires OVER the FC instead of under where they were before. Anyhow, I'm pretty satisfied, for now;).
Glad to hear you're up and flying again! I'll maybe try a lower P on my drone just to see what it can do.
I have a very similar setup. Same props, motors, FC anti-vibration mount, gear, frame ....., and don't have any inordinate vibrations or wobbles while ascending. Descending is a very different story, but I digress.
My only advice would be to ensure that the 1mm hex on the motor shaft is tight. Sometimes they seem tight, but come loose at high speed / thrust.
I never did any PID tuning or Autotune. Never had the need.
Good luck.
David, thanks for the heads up. Yes I double check that nut everytime and I actually add always some cyanoacrylate on the shaft before placing the prop adapters on and then screw the nut on tight. To be honest, the adapters that were shipped with the motors really sucked, they were the classic "X" cut with a conic washer but they are never centered and the more you tighten the more you bring the center of out alignment due to the unsimmetrical tightening on soft aluminum (and, made in China:). So you have 2 problems with those adapters, the first is that the machined part somehow has the center of the shaft entrance not in axis with the out (happens but rare, I guess they work them the opposite way!) and second and more problematic is the conical prop washer that when tightened on the conical matching shaft adapter never stays parallel with the rotation plane so your props wobble. I had my friend machine me some "straight" adapters with no conical weirdness precise for my motors' shafts and just a nut to tighten. Of course I add always as I said cyanoacrylate on the shaft and quickly place the adapter on and tighten the nut. That gives a pretty firm and balanced fix (it is a good gap filler though sensitive to extreme heat). No more props flying off since that! Attached you can see the photo of the original adpater (left) and the one I had my friend machine for me (right) - simpler and more effective and perfectly balanced.
shaft_adapters.jpg
Hope that helps!
Clear Skies,
Khoa
Hi Khoa, thanks for the quick reply. My Frame is a DJI Flame 450 (it's written in the subject:). Yes I did go through basic tuning, and also fiddled with the stabilize PIDs: decreasing them from the standard values to under 3.6 reduces the wobble but it is still present when accelerating/increasing thrust/increasing altitude. I know that decreasing and passing through your own prop wake(s) makes the copter wobble, but not when rising. Everything seems fine when hovering or in altitude hold or position hold but I guess this is because the copter doesn't move up and down a lot (luckily). I suspect a glitchy ESC because only one of my 4 brushless motors heats up a lot more than the other 3 and I have replaced the motor already (A2212 1000KV) but it still heats up more than the other 3 and there is no reason it should as the copter is well balanced (the CG is centered). I think at this point I will try replacing the ESC and see. These ESCs have only 3 basic options (30Amps, but 2 years old), NimH, LiPo, Brake On/Off - that's it, no other settings (no ramp speed, no frequency settings) which I can tune. I had them laying around from previous projects so I decided to use them - they should be identical (ok one of the 4 I repaired as a FET decided to short, but it's not the one apparently causing the overheat).
Cheers!
No probs - I have no idea what they are, they came in a kit which my friend bought over 2 years ago with the motors and are unbranded:) Definitely not SimonK firmware;)
-
1
-
2
of 2 Next