Hi Guys,

I would really appreciate a little guidance from the resident helpers here as to what may be causing this camera shake.  If anyone can offer some assistance I would be truly grateful.


I attach a video to show the example.  https://we.tl/t-HQQ4dK0MTI (35mb)


Please bear in mind this flight was in a mountainous erea so initially the footage looks tilted.  I do end up flying higher, with the camera pointing forward where you can see the natural geometrey better.

I also attach the flight log for further information. Flight log: https://www.phantomhelp.com/LogViewer/R4GPOXOBMG9FNN828CXT/

When this video was shot I actually had four pins in the cradle, but have since removed two of these.  So currently I have four white dampers in, but only two pins.  I haven't carried out a test flight with this arrangement as yet, as to me it looks like a camera/gimbal problem. I'm hoping this will solve the 'jellying' effect apparent in this video.


Background:
A few months back I unfortunately lost a prop and the aircraft took a 4 meter tumble to the ground, falling onto relatively soft mud/grass.  The casing had cracked (not extensively) and the flat gimble cable was pulled out.  This resulted in me replacing all cables (as the drone was open and the cables came as a kit) and also the shell.  It was clear the gimbal had taken a hit.


I must advise the gimbal arm is leaning slightly, the arm itself is straight but it's bent from the clamp section, only by a couple of degress.  This is currently being corrected through the software as its only slightly out. Could this be a cause perhaps?


Worth noting:
The prop motors are all okay, no signs of burning or discolouration.

As you can see in the attached video the shake only seems to appear when the camera is pointing forward and not when angled down.
IMU and gimble were calibrated before the video was shot.
The yaw arm fixing/joint is sitting, tightly screwed, inbetween the connecting bracket, not too low, not too high.

If you require any further information I am more than happen to provide this.

As this is a Phantom Pro 3 it's unfortunately out of warranty so i wont be sending it in for repair.  Having said I'm quite handy when it comes to mechanics and I'm fairly confident in my ability to carry out repairs.  I'm willing to try anything to get the drone fully operational again.


Once again, I appreciate any guidance and support you can offer.


Drone newbie :-)

If you want people to view the material, it would be best to post it to YouTube.

Looks like the camera is loose on the roll motor spindle.  It is trying to correct itself and remain horizontal while the aircraft is tilting off horizontal.  Being loose, it overshoots, and the motor over corrects in the other direction, making the problem worse.

I believe there is a grub screw or similar to tighten on the motor shaft which should correct the problem.  Fixing this issue should also eliminate the need to adjust the horizon via software.

endotherm Posted at 9-5 02:00
Looks like the camera is loose on the roll motor spindle.  It is trying to correct itself and remain horizontal while the aircraft is tilting off horizontal.  Being loose, it overshoots, and the motor over corrects in the other direction, making the problem worse.

I believe there is a grub screw or similar to tighten on the motor shaft which should correct the problem.  Fixing this issue should also eliminate the need to adjust the horizon via software.

Thanks Endotherm for your reply, it's most appreciated!
I've just got the drone out and had the back plate of the roll arm mechanism off to see if I could find a grub screw of some description which may habe been loose, but theres nothing inside there once I get the back plate off and the printed circuit board unscrewed inside the unit.

One thing i noticed was that the spindle that has a flat side within that unit doesnt sit on the perfect horizontal.  It kind of leans to the right.

This, i noticed, means that when the drone is powered down the camera naturally tilts to that right tilted angle.  Does anyone know if this is normal?


When powered up it seems to correct itself. I'm wondering if the motor and sensors are battling this naturaly seated right tilted position and it is that that's causing the frequent rolling effect.  See the attached image for the flat screw I mention.

That would certainly tie in with your answer above regarding some kind of loose grub screw or leveling screw required to correct it - but I can't see there is such an item.


Can anyone help?



Any help is truly appreciated. Thanks in advance.
db

DavidRBrown Posted at 9-6 12:33
Thanks Endotherm for your reply, it's most appreciated!
I've just got the drone out and had the back plate of the roll arm mechanism off to see if I could find a grub screw of some description which may habe been loose, but theres nothing inside there once I get the back plate off and the printed circuit board unscrewed inside the unit.

I must confess, I've never had to pull mine apart nor seen them up close.  It is just information I have gleaned from reading through various threads discussing faults.

Typically, flats on such a shaft are pushed to mate firmly with a corresponding assembly to eliminate any loose shaking like you are experiencing.  If the mating component doesn't have the exact profile of the shaft, then there should be a grub screw at right angles to the flat on the shaft to press it against the other component.  I'd suspect from the photo you posted that there is a small grub screw inside the shaft, most likely it will have a recessed hexagonal hole to take a small Allen key/wrench.  If not, it should be inline with this hole on the opposing piece.

Hopefully someone who has had to tinker with their gimbal will chime in to elaborate.