Twitchy Gimbal - Please Help

[srfnmnk]

Hey friends,

I purchased a cheap knock off Gimbal from amazon and I cannot get it to stabilize. Any help would be VERY much appreciated.

Steps taken:
Simple and 6 axis accelerometer calibration
Gyro Calibration
PID Tuning
Verified board orientation as per manual

Product URL: http://www.amazon.com/Brushless-Camera-Gimbal-Controller-Phantom/dp/B00GAY5YBU/ref=cm_cd_al_qh_dp_t

 

[srfnmnk]

Bump

 

[Mostly Harmless]

I haven't yet picked up a gimbal, but have been doing a bit of digging/reading to learn about them before I dive in. I remember seeing something about this, but can't find it, again. I think I saw it on YouTube while searching for BGC 2.2 Gimbal . I can't be sure, but it might have had something to do with the PWM frequency setting on the Advanced Tab.

In general, however, the rocking motion of these sorts of controllers could be more likely related to a too high setting on the P value. Since I haven't set one of these up yet, I can't offer direct insight, but search YouTube for BGC 2.2 Gimbal PID settings... this could lead to useful info.

A good overview of the setup process is seen here: https://www.youtube.com/watch?v=C2S2r5f-AT0

I've also seen discussion about this related to the camera being lighter weight than what the system was designed for... others have added "ballast" to the platform to tame this (which seems like an inelegant solution).

Again, I'm not speaking from any sort of direct experience... but this might get you pointed in the right direction.

Good luck! And let us know how it turns out -- I'll be heading down that road in the near future.

 

[Snarls]

Make sure that there are no wires pulling on the gimbal when it rolls or pitches and that the gimbal is perfectly balanced. Secondly it could be that your motor max PWM% is too low (in BruGi at least). Also check your PIDs again. Otherwise the motor or some part of the controller may not be working right.

 

[RAM]

We just tested one using a laser pointer. Found that the I value was too high. Lowering it while watching the laser dot showed us the sweet spot.

Also, someone recommended that we short out the vibration damping on the multirotor. Too soft and it might resonate with the gimbal swinging. Best to not have too many variables.

 

[srfnmnk]

Thank you so much for your suggestions.

So there seem to have been a few things wrong...

1. Balance is super important. Getting the camera exactly to center of gravity helped quite a bit. I might even add a quarter or some sort of weight in the middle to try and get a ballast...
2. Needed more power to the motors to get it back up the hill to level
3. PIDs needed tuning...

It still has a little trouble getting back to level from extreme angles and kind of does slow twitches somewhere between 40 and 0 degrees...I'm hoping the ballast will fix this.

After all this it works quite well but it's still obviously a cheap gimbal. It seems almost like if the motors had more polls there wouldn't be as big of gaps between angles causing it to twitch a little less...but over all I think it's pretty stable now.

 

[srfnmnk]

We just tested one using a laser pointer. Found that the I value was too high. Lowering it while watching the laser dot showed us the sweet spot.

Also, someone recommended that we short out the vibration damping on the multirotor. Too soft and it might resonate with the gimbal swinging. Best to not have too many variables.

There may be an issue with the gimbal being isolate from the quad. I do have it on some pretty good vibration dampeners...everything I read before mounting it seemed to suggest that gimbals perform best dampened, have you had different experience?

 

[RAM]

There may be an issue with the gimbal being isolate from the quad. I do have it on some pretty good vibration dampeners...everything I read before mounting it seemed to suggest that gimbals perform best dampened, have you had different experience?

We only shorted out the main damping in order to eliminate one possible cause. We didn't keep it that way. It was easier to do that instead of removing the gimbal assembly to bench test it. In our case, shorting out the damping showed no sign of improvement, only the I parameter change showed improvement.