Repair / Replace options ?

[RoboticTundra]

I purchased a new MPP from Amazon on 9 April to document the building of our new house. I've had a number of issues with it, primarily a yaw problem (Correcting Yaw w/ Litchi ?) but also when stationary it will slowly rotate. It took some time to determine that it wasn't just operator error. A friend who works with drones flew it and said it's definitely quite messed up.

So, determined it's not functioning correctly but 3 days past Amazon's return period. DJI typically takes 14-30 days on repairs which would mean missing a considerable bit of the building process that we purchased it for. Rather frustrating that it was defective out of the box and I'm in this position.

Ideas?

Thanks,

 

[sar104]

I purchased a new MPP from Amazon on 9 April to document the building of our new house. I've had a number of issues with it, primarily a yaw problem (Correcting Yaw w/ Litchi ?) but also when stationary it will slowly rotate. It took some time to determine that it wasn't just operator error. A friend who works with drones flew it and said it's definitely quite messed up.

So, determined it's not functioning correctly but 3 days past Amazon's return period. DJI typically takes 14-30 days on repairs which would mean missing a considerable bit of the building process that we purchased it for. Rather frustrating that it was defective out of the box and I'm in this position.

Ideas?

Thanks,

I would call Amazon and actually talk to them first - I've never had a problem getting them to replace defective items irrespective of their stated return window. They have always honored 1-year warranties.

 

[Mad Mavic]

Taking into consideration, roll, pitch, and yaw, the latter being mentioned along with rotation; can you provide a distinction between the two characteristics.

I believe it was ThunderDrones who mentioned that a compass error causes large circular rotations while an IMU malfunction can cause an in place (yaw) rotation; not sure which, if either, is relevant.



Since the MPP is new and assuming no crash record. Also assuming all systems have been calibrated and passed (gyros, compass readings, and IMU are all showing good in GO 4 app). Sticks have also been calibrated, all signals are good, with little to no interference, and a close check of the flight parameters has been performed using something such as Airdata.com. Then yes, ship it back to DJI for a NEW replacement, not a refurbished unit.

One word of advice when dealing with DJI. Never assume the service rep understands everything you are saying. Many rely heavily on the recorded conversation. Don’t ship via your choice of carrier. Instead wait on their UPS shipper to arrive. But as you said, it’s a waiting game. They will tell you that they “are dealing with customers globally” therefore it takes time.

If it were me and the drone is in no danger of crashing, and you have purchased DJI Refresh, then I would finish the shoot if time permits.

 

[Former Member]

I would call Amazon and actually talk to them first - I've never had a problem getting them to replace defective items irrespective of their stated return window. They have always honored 1-year warranties.

I agree. It would probably take less time than sending it in to DJI for repair, and they will send the OP a new drone instead of a refurb.

 

[RoboticTundra]

Yes, did RC and gimbal calibrations as well.

Yaw problem example: Flying with a flight path of 0° north. It will fly the path correctly (and verified by Airdata) but maintains a heading of about -10° to -20° or 340° to 350°. This is indicated in both images from the camera (as seen here: Correcting Yaw w/ Litchi ? ) and looking up at the MPP. Results are identical in GO4 and Litchi.

Also, when flying manually it will have a flight path of about +10° to +20° of heading. So heading of 270° but it will take a path of 280° to 300°.

 

[sar104]

Can we see some logs? This discussion is somewhat pointless without actual data.

 

[RoboticTundra]

Yep, let me see what I can come up with.

Should I be able to get logs of MPP? : Downloading .DAT file from MPP ?

 

[RoboticTundra]

I believe I've got the correct Litchi log file and video.

Litchi log file: Dropbox - 2018-04-28_11-20-14_v2.csv

That I believe should include a path of 0° running over the garden below. The path in this garden runs tru south to north.

 

[sar104]

That looks like a compass problem. Looking at the velocities, IMU yaw and gimbal yaw:



The three passes to the north show an accurate track (as defined by the velocities), and indicate that both the aircraft and the gimbal are also pointing due north. Clearly that wasn't really the case.

The two passes to the south are also accurate in terms of track, but now the aircraft and gimbal yaw values are both around -175°. It's not clear why the FC could not achieve the requested 180°.

In any case, it appears that the gimbal and aircraft yaw values are not reflecting it's actual orientation. Based on your photo the gimbal bearing on that northward pass appears to be around -10°.

Have you done a compass calibration?

 

[RoboticTundra]

Thanks. Yep, I've done compass (3x), IMU (2x), RC (2x) and Gimbal (1x).

Here's video of one pass if interested in downloading: Dropbox - DJI_0006.MOV

 

[RoboticTundra]

Am I correct that it thinks it has a correct heading even though reality is -10°?

Likely the same issue that causes it to take a flight path of about +10° of it's heading (so sliding sideways slightly) when it should be going true heading in manual mode?

Thanks,

 

[sar104]

Am I correct that it thinks it has a correct heading even though reality is -10°?

Likely the same issue that causes it to take a flight path of about +10° of it's heading (so sliding sideways slightly) when it should be going true heading in manual mode?

Thanks,

That's correct. I'd request a replacement.

 

[RoboticTundra]

Done. Thanks.

 

[RoboticTundra]

Got a new MPP and ran my simple test course over my neighbors garden and massively better. Only had time to do 1 run w/ GO4 and 3 w/ Litchi. Headings were spot on in 2 of them and very close on 2 others. Entire course seems shifted about 8' west though this may be operator error. More to come.