Mavic Mini Crash

[JeffJen]

Early morning flight today. 9 satellites, normal take off and hover, home point recorded. Switch to cinematic mode start flying forward, sudden speed up veered off to right uncontrollable and crashed into rocks. Broken quite a few parts, eg arms one motor and props. Have looked at video and dat files. Only abnormal record shows mag compass sudden failure. Only warning on take off was low light vision sensors not working fly with caution. I assume this was the downward looking sensors. I have included info from csv viewer. Any suggestions? I did buy a care refresh package with the drone.
Here is some more info from airdata.

Jeff Jennings | Facebook

www.facebook.com

 

[sar104]

You haven't posted the txt log, and the DAT file that you posted is not the correct one.

 

[MAvic_South_Oz]

JeffJen, good idea to get an analysis done, it could confirm if it should be a warranty claim, rather than use a precious Care Refresh claim.

Go to sar104 sig line above, you'll find quick guide up the top to get your txt log, that should do for the moment.
(I like the Phantom Help page, simples.)

You should look in the sub folder for the dat, as on the mini, it will disappear once it syncs (manual or auto sync).
It any are there grab then and save until you know which one might be needed, though often the txt will reveal what's needed.

 

[JeffJen]

JeffJen, good idea to get an analysis done, it could confirm if it should be a warranty claim, rather than use a precious Care Refresh claim.

Go to sar104 sig line above, you'll find quick guide up the top to get your txt log, that should do for the moment.
(I like the Phantom Help page, simples.)

You should look in the sub folder for the dat, as on the mini, it will disappear once it syncs (manual or auto sync).
It any are there grab then and save until you know which one might be needed, though often the txt will reveal what's needed.

I've looked for all the files and none of them give any more detail than the ones I have already included. The only sign of a problem is the fact I had 9 satellites, indicated as red then it jumped to 10 and the drone seemed to accelerate and veer off to crash. Exactly where do I find the correct files for analysis? I have downloaded the flight records and the MacdatFlight records but none of these are helpful. In AirData under power it says the flight was too short to give information.

 

[slup]

This most likely hadn't anything to do with the GPS, instead this smells like a Yaw error coming from powering on in a magnetic disturbed environment... we need the DAT log for this. The correct one have FLY038.DAT in the end.

At 25,9sec, when you release the elevator & aileron the heading speed increases uncommanded & maxes out to 5,6m/s in the very end.



Both pitch & roll angles becomes progressively more extreme towards the end ... everything much alike a Yaw error making the AC not having the correct heading.

Do you remember if the AC were directed towards south west at take off as the green arrow show in the sat pic above?

 

[sar104]

I've looked for all the files and none of them give any more detail than the ones I have already included. The only sign of a problem is the fact I had 9 satellites, indicated as red then it jumped to 10 and the drone seemed to accelerate and veer off to crash. Exactly where do I find the correct files for analysis? I have downloaded the flight records and the MacdatFlight records but none of these are helpful. In AirData under power it says the flight was too short to give information.

You had already uploaded the correct txt file to AirData - you just hadn't posted a link to it - just a screenshot that was not useful. And the DAT file you posted was the wrong file - the correct file had a name ending FLY038.DAT but you posted FLY040.DAT.

Anyway - as @slup suspected, this was caused by an incorrectly initialized IMU yaw due to magnetic interference at power up. That shows up when comparing position with integrated velocity:



The yaw was initialized at 200° when it fact the actual yaw was around 80°. That is all it takes to cause that kind of uncontrolled flight. If you had looked at the map before takeoff you would have noticed that the aircraft orientation arrow was pointing in the wrong direction. That is the essential test before every takeoff to prevent this kind of problem, even though it is not mentioned in the instruction manuals.

 

[Jeffdj]

This most likely hadn't anything to do with the GPS, instead this smells like a Yaw error coming from powering on in a magnetic disturbed environment... we need the DAT log for this. The correct one have FLY038.DAT in the end.

At 25,9sec, when you release the elevator & aileron the heading speed increases uncommanded & maxes out to 5,6m/s in the very end.

View attachment 98704

Both pitch & roll angles becomes progressively more extreme towards the end ... everything much alike a Yaw error making the AC not having the correct heading.

Do you remember if the AC were directed towards south west at take off as the green arrow show in the sat pic above?

View attachment 98705

No the mini was in fact pointing east as the video shows on take off. I agree compass error but I did not get a compass warning error unless it was masked by the low light, vision sensor not working fly with caution. A lesson to learn always check that direction arrow is correctly orientated. The home point was correctly recorded. I notice the mini often requires a compass calibration at each flight, however it didn’t ask this time and as you can see it initially hovered correctly and started to fly normally. I haven’t had this problem with my three other DJI drones, Mavic pro 2, zoom and phantom 4. Where do I find the correct dat log file with the fly038.dat file? When I tried to use the dat files it said cannot use the file as it was generated by Dji asssistant 2.

 

[sar104]

No the mini was in fact pointing east as the video shows on take off. I agree compass error but I did not get a compass warning error unless it was masked by the low light, vision sensor not working fly with caution. A lesson to learn always check that direction arrow is correctly orientated. The home point was correctly recorded. I notice the mini often requires a compass calibration at each flight, however it didn’t ask this time and as you can see it initially hovered correctly and started to fly normally. I haven’t had this problem with my three other DJI drones, Mavic pro 2, zoom and phantom 4. Where do I find the correct dat log file with the fly038.dat file? When I tried to use the dat files it said cannot use the file as it was generated by Dji asssistant 2.

You will not have got a compass warning because the FC had no way to know that the initialized yaw was incorrect. And you need the mobile device DAT file, not the aircraft DAT file.

 

[Jeffdj]

You had already uploaded the correct txt file to AirData - you just hadn't posted a link to it - just a screenshot that was not useful. And the DAT file you posted was the wrong file - the correct file had a name ending FLY038.DAT but you posted FLY040.DAT.

Anyway - as @slup suspected, this was caused by an incorrectly initialized IMU yaw due to magnetic interference at power up. That shows up when comparing position with integrated velocity:

View attachment 98708

The yaw was initialized at 200° when it fact the actual yaw was around 80°. That is all it takes to cause that kind of uncontrolled flight. If you had looked at the map before takeoff you would have noticed that the aircraft orientation arrow was pointing in the wrong direction. That is the essential test before every takeoff to prevent this kind of problem, even though it is not mentioned in the instruction manuals.

Thank you for that, I suspect this is a weakness of the mini as I have flown many flights with my phantom 4 pro, Mavic, Mavic 2 pro and Mavic 2 zoom from this exact spot with no problems whatsoever. Yes I will in future check the arrow orientation but can’t understand not seeing compass calibration warning unless distracted by low light, vision sensor warning, fly with caution. I assume this was the downward facing sensor which allowed it to hover correctly after take off.

 

[slup]

No the mini was in fact pointing east as the video shows on take off. I agree compass error but I did not get a compass warning error unless it was masked by the low light, vision sensor not working fly with caution. A lesson to learn always check that direction arrow is correctly orientated. The home point was correctly recorded. I notice the mini often requires a compass calibration at each flight, however it didn’t ask this time and as you can see it initially hovered correctly and started to fly normally. I haven’t had this problem with my three other DJI drones, Mavic pro 2, zoom and phantom 4. Where do I find the correct dat log file with the fly038.dat file? When I tried to use the dat files it said cannot use the file as it was generated by Dji asssistant 2.

The mobile device DAT logs are here:

iOS:
Apps » DJI Fly » FlightRecords » MCDatFlightRecords

Android:
DJI » dji.go.v5 » FlightRecords » MCDatFlightRecords

And it doesn't take much to disturb the compass, not always that triggers a calibration prompt ... & as it was a low light situation the VPS sensors most likely weren't holding the AC, you had navhealth 4 (of 5) & locked to 9 sats so that was holding the position.

 

[sar104]

Thank you for that, I suspect this is a weakness of the mini as I have flown many flights with my phantom 4 pro, Mavic, Mavic 2 pro and Mavic 2 zoom from this exact spot with no problems whatsoever. Yes I will in future check the arrow orientation but can’t understand not seeing compass calibration warning unless distracted by low light, vision sensor warning, fly with caution. I assume this was the downward facing sensor which allowed it to hover correctly after take off.

All the DJI aircraft other than the Mavic 2 (with a couple of caveats) are vulnerable to this problem. Magnetic interference can be very short range - just a few cm sometimes, depending on the source. You may have just got unlucky this time. What kind of surface were you launching from? And to elaborate on the lack of a compass error - that is only triggered if the total magnetic field strength or inclination are outside pre-programmed thresholds. If the field is simply distorted from magnetic north then the FC has no way to detect that - it only manifests when the aircraft takes off and is not pointing in the direction it thinks it is.

As for hovering after takeoff - that's just a metastable state. If it doesn't need to correct for drift then the problem it sees when it applies lateral thrust and moves in an unexpected direction simply hasn't yet manifested itself. You can see that if you compare IMU velocity with the time-differential of position - no problem until it starts to move:



By the way - why are you posting from two different accounts?

 

[JeffJen]

All the DJI aircraft other than the Mavic 2 (with a couple of caveats) are vulnerable to this problem. Magnetic interference can be very short range - just a few cm sometimes, depending on the source. You may have just got unlucky this time. What kind of surface were you launching from? And to elaborate on the lack of a compass error - that is only triggered if the total magnetic field strength or inclination are outside pre-programmed thresholds. If the field is simply distorted from magnetic north then the FC has no way to detect that - it only manifests when the aircraft takes off and is not pointing in the direction it thinks it is.

As for hovering after takeoff - that's just a metastable state. If it doesn't need to correct for drift then the problem it sees when it applies lateral thrust and moves in an unexpected direction simply hasn't yet manifested itself. You can see that if you compare IMU velocity with the time-differential of position - no problem until it starts to move:

View attachment 98711

By the way - why are you posting from two different accounts?

Thank you very much for your analysis and help in this matter. I inadvertently started a new account with the mini. I cannot find any FLY038DAT files on the ipad mini, in the MCDATFlightRecords there is no 038 file. This morning I went to the exact launching spot with my Mavic 2 Zoom. No problems, 18 satellites, correct orientation and normal compass. I flew it as per yesterday and no problems whatsover. Is the Mini more susceptible to minor magnetic fluctuations. I launched in both cases from the identical landing mat over a flat sandy surface, rocks nearby. I have included a photo taken just after launch. DJI were prompt in sending me a return postage label and I have sent the mini back. Wonder if it will be covered under warranty or fly care? Thanks again for all your help.

 

[thewishy]

All the DJI aircraft other than the Mavic 2 (with a couple of caveats) are vulnerable to this problem.

Could you point me in the right direction for more information on this. After crashing my mavic mini (Uncontrolled decent), I've seen spending a lot more time trying to understand how to... not do that... with my Mavic 2 Pro (Much better aircraft..)

 

[slup]

Could you point me in the right direction for more information on this. After crashing my mavic mini (Uncontrolled decent), I've seen spending a lot more time trying to understand how to... not do that... with my Mavic 2 Pro (Much better aircraft..)

It's a very easy way to get a "before take off" indication of an upcoming yaw error due to powering on in a magnetic disturbed area ...

Firstly power on the AC were you are going to take off, not elsewhere ... when on, ALWAYS bring up the map in GO4 (or Fly app) & check so the AC symbol is directed in the same way in relation to map objects (like roads, houses) as it is in reality.

If the AC direction in the map isn't correct ... power down & move to another location well away from the first & power up & repeat.

In this way you can be sure that the AC compass have initialized the IMU yaw correctly.

 

[sar104]

Could you point me in the right direction for more information on this. After crashing my mavic mini (Uncontrolled decent), I've seen spending a lot more time trying to understand how to... not do that... with my Mavic 2 Pro (Much better aircraft..)

It's rather a long thread, but worth reading:

Compass error demonstration

With compass errors arising from magnetic interference one of the major causes of problems, I thought it might be useful to show a controlled demonstration of how that works. To recap, when the aircraft is powered up the compass (3-axis magnetometer) reading is used to initialized the IMU yaw...

mavicpilots.com

 

[Jeffdj]

Thank you very much for your analysis and help in this matter. I inadvertently started a new account with the mini. I cannot find any FLY038DAT files on the ipad mini, in the MCDATFlightRecords there is no 038 file. This morning I went to the exact launching spot with my Mavic 2 Zoom. No problems, 18 satellites, correct orientation and normal compass. I flew it as per yesterday and no problems whatsover. Is the Mini more susceptible to minor magnetic fluctuations. I launched in both cases from the identical landing mat over a flat sandy surface, rocks nearby. I have included a photo taken just after launch. DJI were prompt in sending me a return postage label and I have sent the mini back. Wonder if it will be covered under warranty or fly care? Thanks again for all your help.

Pleased to report that my Mini was returned repaired by DJI at no cost to me and refresh intact. Mini needed two new arm assemblies, all new props and main circuit board meaning new serial number, cost would have been $250 A but no charge. Very pleased with quick response and service from DJI.