Question about a crash... a little help please...

[slup]

Something odd is going on here ...

At 767sec the MA2 are tracking with a 55 degree yaw angle, but there the log indicates that the AC starts to rotate CW & have made a half turn by 769sec & continue to rotate CW and quits the tracking mode 772sec ... a full turn is made at 773,2sec & the crash occurs 773,8sec. None of this rotation is shown in your vid. & apparently didn't happen.

In below chart the rotating yaw is shown by the magenta colored graph ... and the background color change indicate a change in flight mode (where the tracking mode stopped). And the OA was on ... and no stick inputs is shown in the log until after the crash & then only negative throttle to land the AC.



If we take a look at the IMU vs. GPS velocities in East, North direction ... these should be very close to zero if everything is as it should ... but here they start to show big deviations at 767sec. This usually indicate either IMU or compass failures.

I really would like to see the mobile DAT log for this ... it's the one that ends with FLY092.DAT

 

[srnolten]

Something odd is going on here ...

At 767sec the MA2 are tracking with a 55 degree yaw angle, but there the log indicates that the AC starts to rotate CW & have made a half turn by 769sec & continue to rotate CW and quits the tracking mode 772sec ... a full turn is made at 773,2sec & the crash occurs 773,8sec. None of this rotation is shown in your vid. & apparently didn't happen.

In below chart the rotating yaw is shown by the magenta colored graph ... and the background color change indicate a change in flight mode (where the tracking mode stopped). And the OA was on ... and no stick inputs is shown in the log until after the crash & then only negative throttle to land the AC.

View attachment 106504

If we take a look at the IMU vs. GPS velocities in East, North direction ... these should be very close to zero if everything is as it should ... but here they start to show big deviations at 767sec. This usually indicate either IMU or compass failures.

I really would like to see the mobile DAT log for this ... it's the one that ends with FLY092.DAT

View attachment 106510

I am searching for it but my DJI \ dji.go.v5 \ flightrecord \ MCdatFlightRecords is empty ? Where do I look ?

 

[BadWolf1]

at 12:51.9s joystick activated
I am not a log reader by far, others here are the experts at that
but did you inadvertently mess the control while scooting along?
It seems to have happened the veer off at exactly that point

It's strange. The "fly state" is NaviSubMode_Tracking up to that point. There is nothing from the controls themselves but the mode changes to "gentle GPS" (going through GPS_ATTI and Joystick) and the drone turns left and goes into the tree.

I don't understand why it didn't go into a hover when it stopped tracking

 

[Torcan]

It's strange. The "fly state" is NaviSubMode_Tracking up to that point. There is nothing from the controls themselves but the mode changes to "gentle GPS" (going through GPS_ATTI and Joystick) and the drone turns left and goes into the tree.

I don't understand why it didn't go into a hover when it stopped tracking

as I said before, I am far from reading or understanding the files. Way beyond my abilities, I am just a take her up, play with it and enjoy the day. Past my expiry date for that, however, that being said, I am intrigued and amazed how some of the members here are able to disect and pinpoint with accuracy so many incidents from reading the dat files.
Hats off to all of them, they have my respect.

 

[itsneedtokno]

OP: do you have auto sync enabled in your DJI FLY app?

Also, what OS on your phone?

 

[slup]

I am searching for it but my DJI \ dji.go.v5 \ flightrecord \ MCdatFlightRecords is empty ? Where do I look ?

Then I'm afraid, it's gone ... if you had autosync activated in the DJI Fly app it probably got deleted during sync to DJI servers, have been like this since the Mini (& Fly app) was released.

... I don't understand why it didn't go into a hover when it stopped tracking

For that we need the mobile device DAT ... but it seems it's gone as usual when it comes to the DJI Fly app with autosync activated.

 

[itsneedtokno]

Then I'm afraid, it's gone ... if you had autosync activated in the DJI Fly app it probably got deleted during sync to DJI servers, have been like this since the Mini (& Fly app) was released.



For that we need the mobile device DAT ... but it seems it's gone as usual when it comes to the DJI Fly app with autosync activated.

I just now turned mine off.

Kept forgetting, and this is why it's so important.

 

[srnolten]

I have Android, latest version I guess. I think autosync is enabled indeed. I will turn it off from now on..


Thanks everyone for their time & energy

 

[srnolten]

It's strange. The "fly state" is NaviSubMode_Tracking up to that point. There is nothing from the controls themselves but the mode changes to "gentle GPS" (going through GPS_ATTI and Joystick) and the drone turns left and goes into the tree.

I don't understand why it didn't go into a hover when it stopped tracking

I did go into hover, I ran towards the drone cause I tought it landed somewhere

It's strange. The "fly state" is NaviSubMode_Tracking up to that point. There is nothing from the controls themselves but the mode changes to "gentle GPS" (going through GPS_ATTI and Joystick) and the drone turns left and goes into the tree.

I don't understand why it didn't go into a hover when it stopped tracking

I did go into hover, I ran towards the drone cause I tought it landed somewhere high in a tree. I panicked a little bit cause I saw it filmed its own leg. So I thought the leg was broken or something, but it was just the camera sideways.... Once I saw the drone hovering I landed it at once....

 

[itsneedtokno]

It seems like it also increases speed

 

[BadWolf1]

I did go into hover, I ran towards the drone cause I tought it landed somewhere

I did go into hover, I ran towards the drone cause I tought it landed somewhere high in a tree. I panicked a little bit cause I saw it filmed its own leg. So I thought the leg was broken or something, but it was just the camera sideways.... Once I saw the drone hovering I landed it at once....

Yes - we can see it hovers after the crash, and the logs show you landing - the poor thing was stunned and I assumed the camera mount was broken from the video of its leg. But just at the moment we see it turn in the video, it dropped out of tracking mode. So if it is no longer tracking and you aren't giving it instructions to keep flying from the remote, so it should have gone into a hover and not flown into the tree. As @slup said we'd need to the DAT file to see if there is any explanation for why it did what did.

 

[Motorcycle Rider]

It definitely shouldn't have flown into the tree, that is sensor failure, or they were turned off.
How much damage does it have?

 

[slup]

It definitely shouldn't have flown into the tree, that is sensor failure, or they were turned off.
How much damage does it have?

The OD was active according to the log ... that they failed was most probably not the root cause here. The very large difference in IMU & GPS velocity together with the video clip instead points at the IMU as the main cause ... unfortunately the mobile device DAT log is needed to confirm that.

The red path is based on GPS, but the green one is based on the IMU ... & from the video clip we know which one the AC took ...

 

[srnolten]

There is no damage. I used the drone a few times after this accident Though I am careful.
I am a bit scared now that it will fail again ... Yesterday I did a full compass & VP sensor calibration.
Maybe I have time this afternoon to test fly again...

 

[srnolten]

@slup I am going tot calibrate the IMU asap, thanks for all your help

 

[Peio64270]

Perhaps you are too thin and look like the tree trunk! It looks like the video got dark a little and lost track of you and picked up the tree instead. Not sure about the accident ovoidance sensors though! Maybe next time wear a bright orange vest?

Why did the video got dark for 1 second ?
This is not obvious.

 

[boblui]

The OD was active according to the log ... that they failed was most probably not the root cause here. The very large difference in IMU & GPS velocity together with the video clip instead points at the IMU as the main cause ... unfortunately the mobile device DAT log is needed to confirm that.

The red path is based on GPS, but the green one is based on the IMU ... & from the video clip we know which one the AC took ...

View attachment 106560

Is this the new version of CsvView ? Mine ( V3.7.6 ) doesn't have the ImuCalcs track :

 

[BudWalker]

Is this the new version of CsvView ? Mine doesn't have the ImuCalcs track. Mine :

View attachment 106760

CsvView needs to be told to do the ImuCalcs.


Selecting this option adds a small increase in processing time. Not really enough to matter IMHO. This will change in CsvView/DatCon 4.0.0 where ImuCalcs will always be done.

 

[boblui]

CsvView needs to be told to do the ImuCalcs.

Have checked the ImuCalcs box but still no display of the corresponding track ....

 

[boblui]

CsvView needs to be told to do the ImuCalcs.
View attachment 106763

Selecting this option adds a small increase in processing time. Not really enough to matter IMHO. This will change in CsvView/DatCon 4.0.0 where ImuCalcs will always be done.

OK, I got it after re-launching CsvView. This is a really useful feature !