Low frequency oscillation

[biggestRCEfan]

Maiden flight past weekend on my stock V4. It is flying as expected, but I have a low frequency oscillation on hover and in forward flight. It appears to be a pith/roll combined motion I estimate at around 1-5 Hz.

What I did so far to troubleshoot:
– balanced the stock props – no improvement
– strapped down the 4.25 Ah 3S battery tighter (used velcro only on the battery tray) – no improvement
– flew with 1800 mAh 4s and thought the oscillation was a bit better. This made me think that PIDs may need some adjustment (the Pids are from the pre-config tab on David’s build pages)

What I planned to do next: check board accelerometer response to running (no props) of each individual motor to see if there is a clear culprit causing vibration. Install a data-logger to investigate deeper (will have to order one though).

But I thought maybe someone has already dealt with this problem.

[wynnsquad]

I think I may have the same problem. Also flying stock electronics and gains from the product page. I’d estimate my oscillation to be closer to 5 Hz. I’m thinking that maybe I’ll update to the new Triflight (betaflight) and see if things have improved

[biggestRCEfan]

Update from my side: received logger but did not investigate the logs as yet. One thing I thought about was to fly without the mini canopy conversion I have on the V4, thinking that it may affect the dynamic response/stiffness of the frame. What is your configuration: the plane V4 frame without the canopy? If so, removing the canopy will not solve it.

I also thought notch filters may help, but I believe betaflight would be required. But, would notch filters work at such a low frequency without affecting response to control inputs which could be in that frequency range?

So far I have learnt to live with it and I enjoy flying it. FPV and LOS. However, my hope was to use it to generate smooth HD video, which is not the case yet. Let me know what you find?

I will likely move to betaflight in a month or so, also to access Tara is Lua script PAID control from the TX. And to access betaflight datalog graphics.

[wynnsquad]

Hi RCEfan,

Mine has no canopy–just the plain V4 frame. I’m going to flash betaflight soon and I’ll report on what I find.

[wynnsquad]

Just tried the new Betaflight 0.7 beta 2 and still had a very similar oscillation. With the new Firmware though I’m pretty sure the default PIDs are for the baby tri. Boosting roll/pitch P up to about 45 made it more flyable but still the oscillations. I’m wondering if it has something to do with the loop rate being too high for the dynamics of a larger copter like the V4

[wynnsquad]

Just flew with PID loop rate set down from 4k Hz to 1k Hz. It seems like the oscillation is a little better now but still present. I was noticing during hover though that my tail servo is very twitchy. Even though the tail controls yaw, I’m wondering if the rapid twitchy adjustments are contributing to the “wobble”… Does anyone know if there is a low-pass filter value that can be played with to attempt to smooth out the tail a bit? I’ve completed both ground and in-flight tail tunes.

[wynnsquad]

Switched back to 0.5 Cleanflight for now. The bobble is still there but somehow it seems a little less now?? I don’t know, but for now I think 0.5 seems more solid for the V4

[wynnsquad]

RCEfan,

Any updates from your side? I’m currently tri-ing (haha) to dial in my PIDs to fix the problem but I’m not there yet. I’m also thinking I should try what you were planning to do above where you remove your props and test the vibration from each motor individually.

[biggestRCEfan]

The motor test didn’t reveal much, sorry. I haven’t flown it for a while as I was waiting for props.

[wynnsquad]

Cool cool. I too did a no-prop motor test that didn’t reveal any obvious vibration source. I’ll continue to update here though.

[biggestRCEfan]

I thought I found something this morning. Flew with the new props and no oscillation!! Then tried to reproduce the vibration by stepping back through changes I made. I played with increasing the deadband in yaw and general to 10 microseconds. Nothing. Still no oscillations after resetting them to the lower values of 5/7.

Next, I thought to put back the gopro as the mass change on the battery tray could have done it. I tested that this arvo, but fortunately decided to fly first without gopro (same conf as this morning), but then it was back. Back to the drawing board.

The only difference was the ambient conditions. This morning it was calm (slight, steady breeze), around 9 deg C (freezing winter’s morning 🙂 ). However, this arvo was stronger wind, say 12-15 km/h, around 19 deg C. And the oscillation was back.

So it is not the props (thought maybe the new ones were better balanced), it is not the deadband (oscillations triggered by jitter on stick signals), and it is not the gopro/battery mass changing the shape of battery tray dynamic response.

What it could be: PID gains need adjusting to respond better to disturbances injected by wind. I am still flying with default PIDs. Will have to set aside some time for this, and the conditions need to be right.

Tried to film it with my phone, but should get a mate to do it for me, as flying and using my phone at thes same time didn’t work …

[wynnsquad]

Very interesting finding! All of my flights have been in warm temperatures (20°C and higher) so perhaps I’d have similar results in cooler weather. I have heard that the MPU IMU is fairly temperature sensitive.

Unfortunately I seem to have damaged one of my ESCs so I wasn’t able to do any testing of my own this weekend. I’m going to try and reflash blheli and see if it starts working again and if not, I’ll have to order a new speed controller.

[wynnsquad]

Turns out the ESC problem was actually a bad motor wire. Got it fixed up and now I’m flying again. I had a small sliver of hope that maybe this whole time the wobble has been caused by the bad motor wire connection but…no such luck. The wobble is still the same. Haven’t lost hope yet though!

I still think maybe the problem could be temperature related. It could just be me, but it seems that for the first few seconds of flight, the wobble isn’t present. Maybe as the IMU heats up to temperature, it gets a little off? I stuck my copter in the refrigerator for an hour and then quickly pulled it out and flew to see if I could experience wobble-free flight for longer, but I couldn’t tell a difference.

I’m starting to wonder if I should just get a new F3FC but that’s a little pricey if I’m not certain it will solve the problem.

[biggestRCEfan]

Hi mate, could you capture the behaviour in video? I will try to do the same. Just so we know we are talking about the same problem.

Flew yesterday in calm weather, around 10 deg C, rock solid! Will fly this arvo and see if it the wobble is back. Need to mount a GoPro on something to try and capture it.

I wouldn’t start changing FC without playing with PIDs. But haven’t gotten round to trying that …

[biggestRCEfan]

Ok, here goes. This clip shows what the issue is I have. Am I expecting too much? The slow oscillation makes for not ideal footage.

As said in earlier posts, early mornings when there is no or little steady breeze, it is rock-solid. I believe the oscillation is caused by the system reacting to wind disturbances. I don’t know if these motors, props and FC/ESC combo is able to do better. Before starting to fiddle with PIDs, I am seeking advice from the brains trust out there.

You can also hear the noise variation that is synchronous with the oscillating movement. The extent of wind is visible on the trees.

[Author]

Posts