I had a 'successful day', yesterday, trying to trigger the shaking that I experienced with different firmwares and under different app versions
Looks like the gimbal hooks sometimes which is resulting in jumps (to the top or bottom, if the wheel is used), or in video shaking, if this happens during a flight in video mode.
Also I am relatively certain, some of the timeouts during panos are related to this, too

And it is not so easy to find a solution.
I could trigger the shaking in tripod mode every time, yesterday.
Only not every time it 'moved' forward to P mode, resulting in video 'nodding' if flying forward
Sometimes flying backwards (gimbal moves down) is clearing this issue, somtimes not.
Same with the customizable button, gimble up and down, if it follows, down stops shaking.
Unlucky, DJI did neither find this issue, nor the one i call 'bumps to the ground starts', I have experienced some times already, so also yesterday.
So, whats next, DJI?
How we finding a solution for this?

Yesterday:




2017:

Hi 7, there is one post late last year about one of the spark owners having the gimbal bracket (not sure what the term is) is not secured properly to the AC. But you did send the AC back earlier to the DJI guys to check and they found nothing.

But the fact that you are only having this issue with one AC and not with the other AC,surely points to to hardware issue doesn't it?

Fingers crossed hope you get resolution at the earliest.

rimza Posted at 2018-2-1 18:42
Hi 7, there is one post late last year about one of the spark owners having the gimbal bracket (not sure what the term is) is not secured properly to the AC. But you did send the AC back earlier to the DJI guys to check and they found nothing.

But the fact that you are only having this issue with one AC and not with the other AC,surely points to to hardware issue doesn't it?

Na, the gimbal is fine (attached ;-). It is only sometimes not moving, get stucked for some reasons.
I did today a 'side by side' comparsion,  the 'bad' with the 'good' Spark.
It looks more and more, that somehow the gimbal is not always doing, what it is supposed to do.
On the good one, there is no 'nodding' on forward moving in 'P'
Also there is no "the white dot still moves up, but the gimbal stucks and starts shaking" (0:52/2:24 in top video).
I think, I will edit a video later about this.

Surprise that DJI did not find any problem with it. Coz with the AC on, you can to some extend see the gimbal movement on the duo axis by turning and yawing the AC. But as you said if the problem is intermittent it does pose as a problem to diagnose possibly.

But still they could fly the AC and see the resultant video result.

rimza Posted at 2018-2-3 01:00
Surprise that DJI did not find any problem with it. Coz with the AC on, you can to some extend see the gimbal movement on the duo axis by turning and yawing the AC. But as you said if the problem is intermittent it does pose as a problem to diagnose possibly.

But still they could fly the AC and see the resultant video result.

Seems more and more not so easy to find this malfunctions.
And i don't think, DJI will pay a techician, to fly a spark for an h or so, 5, 6 batteries, in the open, to check on a drone. And in the records was NOTHING of it. Only the videos showing there IS something.

The one day I did fly 5 1/2 batteries and didn't get any misbehavior.
The other day it started with the first batterie, with the on start bumping on to the ground, and after that every batterie shaking like hell.
Today in the morning, I used both Sparks, 3 batteries each aside each other, and could trigger this shaking every time.
And in the afternoon, I flew again 5 batteries with 'the one', only used this time the latest android app and wifi to RC, to verify that is is NOT a "old app" plus OTG thing, and:
Just on the last batterie, ~30%  left in it, I got the 'nodding' effect, nothing else.
Aside that 'image transmission is weak' signal, of course.
Which I never saw with OTG.

But: I am certain, there is somwething, that let the gimbal stuck every now and then. In about just under the middle position, I think.

I doubt that I get the editing done today, but I will create a video about it
Probably 2morrow ;-)

I think the video shows clearly, that the gimbal stucked sometimes.
And with the 'right' wind conditions, that gets a serious moving of the Spark.
Which will be visible on the video, then.
I assume that the digital stabilisation will even try to cover, but it is simple not always possible.

Also this may explain that the service technicians didn't get any result.
Even if it is obvious when happening, every forward move ends in a 'nodding' in the video-footage,
the shaking from a stabil (no wind) hovering bird will be little!
Digital stabilized, perhaps?

Are there some dust o-rings in the gimbal axis, which are not 100% aligned?