Mavic 2 Pro odd yaw behavior

[mintakax]

Since this is my first post... Hello !
I am a new M2P pilot. I haven't flown for 5 years but had a fair bit of experience in the past with Phantom 1, home built F450, F550 and a large Wookong X-8 with FPV and first Zenmuse gimbal. Amazing advancements in drone technology since then.

I've been flying the M2P for a couple of days with the smart controller and DJI goggles. Yesterday on a close flight, the drone began to endlessly spin around in yaw. I pushed RTH, but spinning still continued so I cancelled. I noticed that if I pushed the yaw stick to one side the spinning stopped, so I was able to land. Obviously the drone was getting an input for 100% yaw in the other direction.
Restarted and everything was fine. I'm wondering what might be the source of the problem ? If this had happened to the roll axis instead of yaw it probably would have been a crash by the time I figured out how to correct. I'm finding that the goggles have a steep learning curve and at this point are pretty awkward for me as far as changing and setting various features. I wonder if some accidental goggle input could have caused this ?

 

[Paul2660]

I would recommend you post the flight log, it will show what was commanded when the continuous yaw was happening.

Paul C

 

[old man mavic]

welcome to the forum could be some magnetic interference affecting the stick that controls yaw from something magnetic close to the controller

 

[mintakax]

welcome to the forum could be some magnetic interference affecting the stick that controls yaw from something magnetic close to the controller

Thanks
Very unlikely since the drone was always close to me in my backyard in the exact spot where I have made dozens of flights lately. Unless it was some kind of transient magnetic interference.

 

[mintakax]

I would recommend you post the flight log, it will show what was commanded when the continuous yaw was happening.

Paul C

Thank you-- I've attached both the *.dat and *.txt files from the SC. I have no idea how to read either. I'm on Mac OS.

 

[sar104]

That's unusual behavior.



The uncommanded rotation is clear, and it wasn't due to spurious stick inputs. This is going to require the flight DAT file ending FLY045.DAT - you posted the wrong file (FLY044.DAT) above.

 

[mintakax]

That's unusual behavior.

View attachment 98207

The uncommanded rotation is clear, and it wasn't due to spurious stick inputs. This is going to require the flight DAT file ending FLY045.DAT - you posted the wrong file (FLY044.DAT) above.

Oh thank you kind sir and sorry for the wrong file. The correct file is too large to attach, so here is the dropbox link for download. I really appreciate your help !

20-04-06-05-21-30_FLY045.DAT

Shared with Dropbox

www.dropbox.com

 

[sar104]

Oh thank you kind sir and sorry for the wrong file. The correct file is too large to attach, so here is the dropbox link for download. I really appreciate your help !

20-04-06-05-21-30_FLY045.DAT

Shared with Dropbox

www.dropbox.com

That's the file. When you said that it was spinning - how much are you talking about? Multiple rotations or less than a single rotation?

 

[mintakax]

That's the file. When you said that it was spinning - how much are you talking about? Multiple rotations or less than a single rotation?

Thanks. Multiple, as if the yaw stick was pushed all the way to the left. The only way to stop the rotation was to push yaw all the way to the right.
The craft was stable except for spinning so I experimented for a minute or two before landing.

 

[sar104]

Thanks. Multiple, as if the yaw stick was pushed all the way to the left. The only way to stop the rotation was to push yaw all the way to the right.
The craft was stable except for spinning so I experimented for a minute or two before landing.

So that event is not in the txt file that you posted - it only shows some limited rotation without stick input. That said, the DAT file does show stick input - i.e. disagrees with the txt log - something I haven't seen before:



The event that you describe, with approximately 16 full CCW revolutions, appears in the DAT file after you landed and took off again (that will be in a separate txt log). For now there is no txt log to compare with, but the DAT file indicates that, as before, the aircraft was detecting and responding to stick inputs:



Since you presumably did not actually make those stick inputs, that's very curious. Assuming that the txt log is accurately showing stick inputs it would be good to see the one that covers the latter part of this flight.

 

[mintakax]

So that event is not in the txt file that you posted - it only shows some limited rotation without stick input. That said, the DAT file does show stick input - i.e. disagrees with the txt log - something I haven't seen before:

View attachment 98235

The event that you describe, with approximately 16 full CCW revolutions, appears in the DAT file after you landed and took off again (that will be in a separate txt log). For now there is no txt log to compare with, but the DAT file indicates that, as before, the aircraft was detecting and responding to stick inputs:

View attachment 98236

Since you presumably did not actually make those stick inputs, that's very curious. Assuming that the txt log is accurately showing stick inputs it would be good to see the one that covers the latter part of this flight.

Thanks. I cant remember exactly what transpired as far as landing and taking off again, turning off the drone, etc. I only made two flights the entire day and they came one after the other, so here is the second *.txt file. But you are correct that I did not make those stick inputs ! I'm really suspecting that it was related to something I did with the goggles, but I have no idea what that might be.

 

[sar104]

Thanks. I cant remember exactly what transpired as far as landing and taking off again, turning off the drone, etc. I only made two flights the entire day and they came one after the other, so here is the second *.txt file. But you are correct that I did not make those stick inputs ! I'm really suspecting that it was related to something I did with the goggles, but I have no idea what that might be.

Okay - same pathology - the DAT file shows rudder inputs that are not present in the txt log:



It would appear that the DAT rudder data are the linear sum of the txt rudder data and some other input source - presumably the goggles - that is basically commanding full CCW rudder during the period shown above up until 665 seconds, when it went to zero.

 

[mintakax]

Okay - same pathology - the DAT file shows rudder inputs that are not present in the txt log:

View attachment 98240

It would appear that the DAT rudder data are the linear sum of the txt rudder data and some other input source - presumably the goggles - that is basically commanding full CCW rudder during the period shown above up until 665 seconds, when it went to zero.

Thanks so much for the analysis. I'm OK with assuming that it was the goggles since I was fumbling around quite a bit with them at that time. I've made many flights since, just using the goggles for viewing, and have not had any issues.