MP spins without stick input, camera view wobbles - intermittently

[erkme73]

I'm trying really hard not to be unfair, but... After taking nearly 2 weeks to get my MP 'repaired' for the compass noise/TBE problem, they sent me a replacement. The paperwork says repaired, but it has a different SN, and all the stickies were on the arms and sensors.

I've made several flights - none of which were able to produce an excessively noisy rear compass. So that's good. Precision landing worked EVERY TIME. All was going well until my last flight today. I nearly had a heart attack when the video feed showed spinning and shaking like it was on its way to the ground after a bird strike. All the telemetry showed it was holding altitude and location, but was spinning clockwise without any input.

I let it spin a few revolutions while I played with the camera pitch. I could not get the camera to move up or down with the wheel, and the snap up/down only changed the pitch by about 20° - and then caused the horizon to tilt about 45°.

I then gave counter-clockwise yaw input, and it did nothing. I tried the other direction and it sped up. But nothing else worked. I ended up initiating an RTH, which fixed everything. Rock-solid, controllable gimbal, flight controls all normal. VERY frustrating as STILL this thing has left me with very shaken confidence. I've asked @BudWalker to have a look at the DAT, but fear I may be Mavic-less again once DJI gets wind.

Bummer

And this is a screen recording of the flight record, showing telemetry and stick input. Notice how the rotation is happening without any input, and does not stop or slow when full left yaw is applied.

 

[wnfaknd]

This looks horrible. The wobbliness should not happen at all. Can you confirm if the camera gimball is securely attached by the 3 little holders that it has? Maybe its hanging by just 2 of them.

And that wouldn't affect the rotation of the drone itself. something is definitely wrong with it.

 

[james_mav]

I'm trying really hard not to be unfair, but... After taking nearly 2 weeks to get my MP 'repaired' for the compass noise/TBE problem, they sent me a replacement. The paperwork says repaired, but it has a different SN, and all the stickies were on the arms and sensors.

I've made several flights - none of which were able to produce an excessively noisy rear compass. So that's good. Precision landing worked EVERY TIME. All was going well until my last flight today. I nearly had a heart attack when the video feed showed spinning and shaking like it was on its way to the ground after a bird strike. All the telemetry showed it was holding altitude and location, but was spinning clockwise without any input.

I let it spin a few revolutions while I played with the camera pitch. I could not get the camera to move up or down with the wheel, and the snap up/down only changed the pitch by about 20° - and then caused the horizon to tilt about 45°.

I then gave counter-clockwise yaw input, and it did nothing. I tried the other direction and it sped up. But nothing else worked. I ended up initiating an RTH, which fixed everything. Rock-solid, controllable gimbal, flight controls all normal. VERY frustrating as STILL this thing has left me with very shaken confidence. I've asked @BudWalker to have a look at the DAT, but fear I may be Mavic-less again once DJI gets wind.

Bummer

And this is a screen recording of the flight record, showing telemetry and stick input. Notice how the rotation is happening without any input, and does not stop or slow when full left yaw is applied.

I would stick that thing back in the box & tell them to send you a new one.

 

[erkme73]

I would stick that thing back in the box & tell them to send you a new one.

I have to assume what they sent me was a new one. It had a new SN and all the sticky covers on the sensors and arms. Though I do have to wonder how they 'flight tested' (which PASSED) if it had those on there. It's turning out to be a crap-shoot, I think. Stuff happens, I just hate having to wait another two weeks for turn around.

 

[erkme73]

This looks horrible. The wobbliness should not happen at all. Can you confirm if the camera gimball is securely attached by the 3 little holders that it has? Maybe its hanging by just 2 of them.

And that wouldn't affect the rotation of the drone itself. something is definitely wrong with it.

When I took it out of the box today, the gimbal lock/clamp was not attached (or barely). I checked to make sure the gimbal plate floated freely and hadn't hopped out of the air frame. What makes me thing this is an intermittent electrical problem is that it goes from rock solid/stable to spastic, back to stable. If it was a binding issue, one would think it would be constant - or at least repeatable under the same angle/conditions. That, and the spinning. That was gut wrenching.

 

[james_mav]

I have to assume what they sent me was a new one. It had a new SN and all the sticky covers on the sensors and arms. Though I do have to wonder how they 'flight tested' (which PASSED) if it had those on there. It's turning out to be a crap-shoot, I think. Stuff happens, I just hate having to wait another two weeks for turn around.

That thing can't be trusted. I'd talk to their rep & find out if they sent you a new one or refurbished. Tell them your problems & you want a new one.

 

[BudWalker]

I'm trying really hard not to be unfair, but... After taking nearly 2 weeks to get my MP 'repaired' for the compass noise/TBE problem, they sent me a replacement. The paperwork says repaired, but it has a different SN, and all the stickies were on the arms and sensors.

I've made several flights - none of which were able to produce an excessively noisy rear compass. So that's good. Precision landing worked EVERY TIME. All was going well until my last flight today. I nearly had a heart attack when the video feed showed spinning and shaking like it was on its way to the ground after a bird strike. All the telemetry showed it was holding altitude and location, but was spinning clockwise without any input.

I let it spin a few revolutions while I played with the camera pitch. I could not get the camera to move up or down with the wheel, and the snap up/down only changed the pitch by about 20° - and then caused the horizon to tilt about 45°.

I then gave counter-clockwise yaw input, and it did nothing. I tried the other direction and it sped up. But nothing else worked. I ended up initiating an RTH, which fixed everything. Rock-solid, controllable gimbal, flight controls all normal. VERY frustrating as STILL this thing has left me with very shaken confidence. I've asked @BudWalker to have a look at the DAT, but fear I may be Mavic-less again once DJI gets wind.

Bummer

And this is a screen recording of the flight record, showing telemetry and stick input. Notice how the rotation is happening without any input, and does not stop or slow when full left yaw is applied.

I know one thing for sure. I ain't going flying with you.

I can't add much. I did confirm that the Mavic had the same stick input data. And, there were no compass or gyro issues. Interesting that the RTH got it's attention.

 

[erkme73]

I just wonder what else could cause the MP to spin. Unlike the TBE, where the entire aircraft followed a spiral pattern, the craft stayed locked at a specific point, and then just yawed like the stick was stuck applying right yaw.

If it had been an issue with the remote not being calibrated properly, would that not have shown in the DAT or the stick inputs on the flight record? I presume if the stick input is showing ZERO, then it can't be that the remote is sending some kind of phantom input signal (like one with a stick that doesn't zero out), right?

 

[BD0G]

Fly it in WiFi mode to "exclude" the actual physical remote control. Use phone only and keep it close by. That will eliminate the remote controller as influencing the behavior exhibited.

 

[BudWalker]

I just wonder what else could cause the MP to spin. Unlike the TBE, where the entire aircraft followed a spiral pattern, the craft stayed locked at a specific point, and then just yawed like the stick was stuck applying right yaw.

If it had been an issue with the remote not being calibrated properly, would that not have shown in the DAT or the stick inputs on the flight record? I presume if the stick input is showing ZERO, then it can't be that the remote is sending some kind of phantom input signal (like one with a stick that doesn't zero out), right?

That's right. The .DAT is showing no stick input as the Mavic was spinning. So it was the FC telling it to spin. I'll speculate when RTH was initiated the FC terminated whatever it was trying to do and that's why the spinning stopped. Just spit balling here.

 

[james_mav]

Fly it in WiFi mode to "exclude" the actual physical remote control. Use phone only and keep it close by. That will eliminate the remote controller as influencing the behavior exhibited.

Even if you figure out the problem is the remote then DJI will want you to send everything back. I'd talk to DJI & see what they advise before wrecking it trying to identify the problem.