• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

Uncontrollable wobble on 250 quad with rotated acro naze with cleanflight

#1
Last week I broke two frame arms. Yesterday, I got a new frame, and put everything together again. The same motors, the same ESCs, the same flight controller, the same everything (except for the frame). This time I decided to rotate my flight controller 90 degrees, pointing right, for ease of access to the USB port. To compensate for this, I set the "Yaw adjustment" to 90. This seemed fine as I moved the quad around while looking at the instrumentation in the cleanflight gui. I also made sure to re-calibrate the accelerometer. Just in case something was different.

This morning I took it out to fly, and it was completely unflyabl. I would trye to take off but at about 40% throttle it started wobbling around like crazy. At 50% throttle, where all 4 points of the frame were off the ground, the wobbling got worse. Dropping the throttle made it flip over on its side. I lost 4 propellers before giving up. One time it shot right up into the sky while flipping only to fall down and break propellers.

I tried with pid controllers 1, 2, and 3 (MultiWii 2.2, Lux, and MultiWii 2.3). I tried in acro and in horizon mode. I tried setting the yaw adjustment to 270 (thinking maybe I got the orientation wrong). Nothing helped. It's the same wobbling. I also reset the pids to default. I have an increased yaw rate, which hasn't given me trouble in the past, but since I was only touching the throttle, I doubt that it has an effect. I also tried with different roll and pitch rates. The battery was fully charged, and the propellers were brand new. The motors are, as far as I can tell, connected to the right ports.

I also, technically, disabled current monitoring, but I highly doubt that has anything to do with it.

Nothing has helped, so now I need your help! What could the reason for this be? What am I doing wrong?
 
Last edited:

cranialrectosis

Faster than a speeding faceplant!
Mentor
#2
Rotating a Naze with an improper adjustment makes the copter spin in my experience so I don't think the board rotation is root cause. I have all my Nazes rotated 90* with no worries.

Based on our earlier conversations and the statement you made that you reset the PIDs to default, I am guessing your P gain rate is too high.

Too high a P gain will result in a 'toilet bowl wobble' where the copter will want to raise all by itself and will wobble so hard it is unflyable/unlandable. The copter will also tend to bounce when you land.

Default PIDs tend to be too high for a 250. If I use default Cleanflight PIDs on my WarpQuad, it will wobble like mad.

There are other reasons a quad may wobble. Flex in the booms, badly out of balance motors or motors tilted at odd angles can do this too.

Post video of the problem along with a screenshot of your gains and lets take a look under the hood as it were. :)
 
#3
Rotating a Naze with an improper adjustment makes the copter spin in my experience so I don't think the board rotation is root cause. I have all my Nazes rotated 90* with no worries.

Based on our earlier conversations and the statement you made that you reset the PIDs to default, I am guessing your P gain rate is too high.

Too high a P gain will result in a 'toilet bowl wobble' where the copter will want to raise all by itself and will wobble so hard it is unflyable/unlandable. The copter will also tend to bounce when you land.

Default PIDs tend to be too high for a 250. If I use default Cleanflight PIDs on my WarpQuad, it will wobble like mad.

There are other reasons a quad may wobble. Flex in the booms, badly out of balance motors or motors tilted at odd angles can do this too.

Post video of the problem along with a screenshot of your gains and lets take a look under the hood as it were. :)
If I haven't managed to solve the problem by tomorrow, I'll see about getting some video. My plan of action is to "factory reset" my fc and then restore basic things like AUX channels and yaw rate, and not touch the fucking pids. If they're still too high, down they go... Whoever put the "defaults" command in there knew what they were doing... =)
 

cranialrectosis

Faster than a speeding faceplant!
Mentor
#5
Is it just P that is too high, or would I need to scale I and D down by a similar amount?
Probably. That's the PITA of PID tuning. You never know until you try it and find out.

PIDs are rules for handling fluctuating variables. Every copter is different, the variables are different, so every tune is unique.

Personally, I would start with PIDs from a similar sized machine from a trusted source and adjust from there. I used FGAs quad tune (through Narco) to get my mini-Tricopter flying. It wasn't 'perfect' but it got me in the ballpark so I could fly.
 

cranialrectosis

Faster than a speeding faceplant!
Mentor
#6
If I haven't managed to solve the problem by tomorrow, I'll see about getting some video. My plan of action is to "factory reset" my fc and then restore basic things like AUX channels and yaw rate, and not touch the fucking pids. If they're still too high, down they go... Whoever put the "defaults" command in there knew what they were doing... =)
The 'problem' is that the defaults were built before mini-quads were popular. Some defaults, MW 1 and 2, are designed for 450 sized quads running 1000 kv motors spinning 10" props.

If you put a Naze32 on an Electrohub with 10" booms, you likely have no start up problems. 250 sized quads were considered 'advanced' a year ago. I don't think defaults have caught up to the trend yet.
 

Snarls

Gravity Tester
Mentor
#8
I once was tuning my PIDs and set the P gain too high so that when I took up it jumped in the sky, did a backflip and then landed itself back on the ground safe and sound. If you still can't get it to work and you suspect it's PIDs you can always list your setup and we can try to find you PIDs for a similar build that should at least get the thing flying.
 
#9
This is classic example of a quadcopter with toilet bowl wobble.

That's considerably more flyable than mine, but yeah, the behavior is similar.
The whole reason I'm in this situation in the first place was that I copied someone's pids. Before the crash, with default pids, it was ok. While I was waiting for my new frames, I took yours' (and others') advice and copied pids form people (I think it was boris.b in some youtube video and someone else here on the forums). My guess is I got more than just the pids when I did...

Copying someone's pids might still be solid advice, but caveat emptor!
 
#16
So I did a factory reset last night, restored things like AUX and rates, and went out flying again this morning. The wobble is completely gone! That's the last time I simply copy someone's stuff without checking it carefully first. I'm now running PID Controller 3 (cleanflight), and it seems to be ok. PIDC0 and PIDC1 has slight oscillations with default pids on my frame. PIDC3 has none. None the less, I tuned P down ever so slightly, to be sure.

Now we can start the learning again. If only I could clear the park of the kids, that would be easy... I got 4/8 packs in today before the kids had come close enough that it was no longer safe to fly =(
It seems I have to be out there at the crack of dawn to get the place to myself...