HELP Mavic Crash

[BudWalker]

@khooper802 this flight started at time 430 secs which means the battery had been on for 430 secs before this flight started. The problems encountered in this flight likely started before the flight. The .DAT from the tablet will have the data for those 430 secs. If you could retrieve that .DAT maybe more could be determined. Look HERE to see how to retrieve that .DAT.

 

[sar104]

@khooper802 this flight started at time 430 secs which means the battery had been on for 430 secs before this flight started. The problems encountered in this flight likely started before the flight. The .DAT from the tablet will have the data for those 430 secs. If you could retrieve that .DAT maybe more could be determined. Look HERE to see how to retrieve that .DAT.

Good point. The incorrectly initialized yaw probably occurred much earlier. The compass warnings should have been displayed since they were due to actual IMU/compass disagreements rather than just the as-yet-undetected initialization error that happens before takeoff.

 

[khooper802]

@khooper802 this flight started at time 430 secs which means the battery had been on for 430 secs before this flight started. The problems encountered in this flight likely started before the flight. The .DAT from the tablet will have the data for those 430 secs. If you could retrieve that .DAT maybe more could be determined. Look HERE to see how to retrieve that .DAT.

Here is the only .DAT files that are available from that day:
.DAT - Google Drive

 

[BudWalker]

Here is the only .DAT files that are available from that day:
.DAT - Google Drive

Those aren't the correct .DAT files. They came from the FlightRecords folder? The files you want are in the FlightRecords/MCDatFlightRecords folder. Follow these instructions.

 

[khooper802]

Those aren't the correct .DAT files. They came from the FlightRecords folder? The files you want are in the FlightRecords/MCDatFlightRecords folder. Follow these instructions.

Found them, have the right ones uploaded now.

 

[BudWalker]

Found them, have the right ones uploaded now.

Kinda interesting. The incident occurred at the tail end of a much longer flight than what was shown in the .txt that you posted. At 424 secs into the flight the MP was near the ground and the motors were stopped because the MP thought that it had rolled over. This from the event log stream.

-9.398 : 22914 [L-FMU/MOTOR]safe_near_grd:true
-8.718 : 22948 [L-FDI][CTRL]: fault on , tilt_ctrl_fail
-8.700 : 22949 [L-FMU/MOTOR] Stop. reason:roll over or atti fail
And


Not sure exactly why a large Yaw/magYaw discrepancy developed at this point. The resulting compass error is the pink background and green background in the top plot. It doesn't appear that you moved the MP but waited about 9 secs and launched again. There was still a compass error at the second launch causing the problems in this "second" flight.

Maybe you could tell us more about what happened to cause the roll over and abrupt roll, pitch and yaw change. Did one of the legs hit something?

 

[khooper802]

Kinda interesting. The incident occurred at the tail end of a much longer flight than what was shown in the .txt that you posted. At 424 secs into the flight the MP was near the ground and the motors were stopped because the MP thought that it had rolled over. This from the event log stream.

-9.398 : 22914 [L-FMU/MOTOR]safe_near_grd:true
-8.718 : 22948 [L-FDI][CTRL]: fault on , tilt_ctrl_fail
-8.700 : 22949 [L-FMU/MOTOR] Stop. reason:roll over or atti fail
And
View attachment 39844

Not sure exactly why a large Yaw/magYaw discrepancy developed at this point. The resulting compass error is the pink background and green background in the top plot. It doesn't appear that you moved the MP but waited about 9 secs and launched again. There was still a compass error at the second launch causing the problems in this "second" flight.

Maybe you could tell us more about what happened to cause the roll over and abrupt roll, pitch and yaw change. Did one of the legs hit something?

Yes, like I think I said in my original post was that I was flying and it just started to drift back and towards the fall so I just pressed down on the altitude and it lowered a bit. I then hit the doors and the top left leg folded in on itself after making contact with the wall. Just before it hat the ground I grabbed it with both hands stopping the top 2 motors personally then I would assume the back ones stopped due to me just picking it up. Perhaps I stopped all four motors but I don't think so...

 

[BudWalker]

Yes, like I think I said in my original post was that I was flying and it just started to drift back and towards the fall so I just pressed down on the altitude and it lowered a bit. I then hit the doors and the top left leg folded in on itself after making contact with the wall. Just before it hat the ground I grabbed it with both hands stopping the top 2 motors personally then I would assume the back ones stopped due to me just picking it up. Perhaps I stopped all four motors but I don't think so...

Yes, you did say all that. I was confused though because that description didn't match the .txt or video that you submitted in post #1. That .txt and video shows what happened after the re-launch following the crash. Anyway, the problems after the re-launch were due to a confused FC which, in turn, was caused by the abrupt roll, pitch and yaw changes before the re-launch. I suspect that if the MP had sat on the ground for a minute or so the FC would have stabilized. Better yet, powering down then up would have reset the FC.

 

[khooper802]

Yes, you did say all that. I was confused though because that description didn't match the .txt or video that you submitted in post #1. That .txt and video shows what happened after the re-launch following the crash. Anyway, the problems after the re-launch were due to a confused FC which, in turn, was caused by the abrupt roll, pitch and yaw changes before the re-launch. I suspect that if the MP had sat on the ground for a minute or so the FC would have stabilized. Better yet, powering down then up would have reset the FC.

so your saying I should have let it start from the ground opposed to my hand? Also is there an explanation on why it didn't show on my app?

 

[BudWalker]

so your saying I should have let it start from the ground opposed to my hand? Also is there an explanation on why it didn't show on my app?

Let's be sure we're on the same page. The .txt and video you submitted showed a flight that was roughly 20 secs. The .DAT showed 2 flights. First a 424 sec flight ending in a motor stop caused by a roll over. Then the second flight, or re-launch, 9 secs after the motors stopped. The .txt and video you submitted were from this second flight.

I had originally said that it didn't appear that you had moved the MP between the crash/roll over at the end of the first flight and the subsequent re-launch starting the second flight. But, I can see now that you may have - this is probably when you hand launched it.

If the compass error wasn't showing in the App at the start of the second flight it should have. Is it possible you were occupied preparing for the hand launch and didn't notice it?

The single, most effective, most robust method of preventing launch time fly aways is to make sure the Go App map heading indicator is consistent with the actual orientation. That check would have shown there was a problem at the beginning of the second flight. This check can't be automated and done for you by the Go App. It has to be done by the pilot. It's really unfortunate that DJI doesn't include this check in any of it's documentation.

 

[khooper802]

So I rolled the bird to stop the motors on the 424 seconds flight. Then I relaunched from my hand. So the crash happened on the short flight.

 

[BudWalker]

So I rolled the bird to stop the motors on the 424 seconds flight. Then I relaunched from my hand. So the crash happened on the short flight.

I get it now. First time I've seen this. Anyway, my other comments still hold.

 

[BudWalker]

So I rolled the bird to stop the motors on the 424 seconds flight. Then I relaunched from my hand. So the crash happened on the short flight.

I tried the roll-to-make-the-motors-stop maneuver several times in an attempt to duplicate your incident. I was supposing that you had discovered an unknown failure mode. Usually the roll-to-make-the-motors-stop maneuver is followed by turning the battery off and not re-launching. So it wouldn't be noticed if there was a compromised Yaw

I used some pretty extreme roll and pitch movements but the FC was able to keep track of the Yaw. I was surprised how well it did.