Short flight in mountains drone drops, wondering why?

[Craig Smith]

I was flying my Mavic Pro in the mountains. Had done many flights from this area. Drone flew a short distance with a new battery. It apparently stopped and fell from the sky. I don't know how to read logs, can someone help? I would like to learn if this was a machine failure or if I did something wrong.

I had been using DJI goggles at the time, and there were others with me who saw the drone drop. I saw no error on the goggles except connection broken which I think occurred at the time the problem happened.

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

Thanks for the help! - Craig

 

[Meta4]

Drone flew a short distance with a new battery. It apparently stopped and fell from the sky. I don't know how to read logs, can someone help?
I would like to learn if this was a machine failure or if I did something wrong.

There isn't much to go with in the flight data
You had plenty battery, that's not a factor.
The data does show you pushing the left and right sticks hard until 0:14.1.
The last data point is at 0:14.2 and shows the sticks centred.
Having the data stop immediately after a sudden braking manoeuvre suggests a loosely fitted battery as one possibility.

 

[Craig Smith]

There isn't much to go with in the flight data
You had plenty battery, that's not a factor.
The data does show you pushing the left and right sticks hard until 0:14.1.
The last data point is at 0:14.2 and shows the sticks centred.
Having the data stop immediately after a sudden braking manoeuvre suggests a loosely fitted battery as one possibility.

The battery had separated from the mavic at some point. I assumed it was knocked off after impact or when falling, but that could be wrong. Is there a way to eliminate all other causes?

 

[Meta4]

The battery had separated from the mavic at some point. I assumed it was knocked off after impact or when falling, but that could be wrong. Is there a way to eliminate all other causes?

It's common for the battery to be dislodge in an impact so that's not conclusive.
@sar104 might be able to dig deeper to find something.

 

[sar104]

It's common for the battery to be dislodge in an impact so that's not conclusive.
@sar104 might be able to dig deeper to find something.

Not much to go on there - your hypothesis looks very likely:



The failure occurred on rapid deceleration taking the elevator from full forwards to centered (stop) in less than 1/10th of a second. That maneuver has a tendency to flick out an unclipped battery, as we've seen before.

The aircraft DAT file (FLY182.DAT), if accessible, should show the voltage drop as the battery disconnected:

Aircraft V3 DAT file: How to retrieve a V3 .DAT File from the AC

 

[Craig Smith]

Because the .dat file was too large I am attaching the csv derived using datcon. Thanks for your help on this!

 

[Craig Smith]

and here is the event log also derived using datcon...

 

[sar104]

Because the .dat file was too large I am attaching the csv derived using datcon. Thanks for your help on this!

I really need the original in order to extract the data at the maximum data rate. Can you put it on Dropbox or equivalent?

 

[Craig Smith]

Here is a link to the whole dat file. Thanks. FLY182.DAT

 

[Craig Smith]

Using CSVView I see the SMART_BATT values start rising from 0 to significant values1 sec before the drone dropped? goHome%, land%, goHometime, landtime?

 

[sar104]

Using CSVView I see the SMART_BATT values start rising from 0 to significant values1 sec before the drone dropped? goHome%, land%, goHometime, landtime?

The aircraft DAT file doesn't show any voltage drop at the end - unusual if it were a battery disconnect. It also doesn't show the elevator going back to zero - that log ends with full forward elevator, which is strange.



Need a sanity check here from @BudWalker.

 

[BudWalker]

The aircraft DAT file doesn't show any voltage drop at the end - unusual if it were a battery disconnect. It also doesn't show the elevator going back to zero - that log ends with full forward elevator, which is strange.

View attachment 81095

Need a sanity check here from @BudWalker.

"Need a sanity check here from @BudWalker."
Good luck with that..

The eventLog stream is showing a lot of errors that I have no idea what they mean. Maybe they are a precursor to an FC crash.

@Craig Smith the SMART_BATT data you were asking about is normal. E.G. goHome% is the battery reserve (expressed as percentage) necessary for the AC to make it back to the home point. It starts out as 0.0 because no reserve is required to get back to the home point. After the AC moves a away from the home point more reserve is required and the goHome% increases. I suspect the goHome% value is also based on the wind field estimation.

 

[sar104]

"Need a sanity check here from @BudWalker."
Good luck with that..

The eventLog stream is showing a lot of errors that I have no idea what they mean. Maybe they are a precursor to an FC crash.

I've seen those errors before, at least once, but I haven't been able to find the event, and I can't remember if they turned out to be significant.

Any thoughts on why the aircraft DAT file would terminate before the txt log, or do you think that the sudden elevator going to zero in the txt log might just be an artifact of the log stopping, rather than real?

@Craig Smith - do you recall if you suddenly centered the right stick to stop the aircraft just before it disconnected, or were you still flying full speed forwards?

 

[BudWalker]

I've seen those errors before, at least once, but I haven't been able to find the event, and I can't remember if they turned out to be significant.

Any thoughts on why the aircraft DAT file would terminate before the txt log, or do you think that the sudden elevator going to zero in the txt log might just be an artifact of the log stopping, rather than real?

@Craig Smith - do you recall if you suddenly centered the right stick to stop the aircraft just before it disconnected, or were you still flying full speed forwards?

I didn't realize that the .DAT stops before the .txt. First time I've seen that. I checked and there isn't anything unusual about the end of the .DAT. It appears to have closed properly. @Craig Smith can you retrieve FLY183.DAT? It may actually be a continuation of FLY182.

 

[sar104]

I didn't realize that the .DAT stops before the .txt. First time I've seen that. I checked and there isn't anything unusual about the end of the .DAT. It appears to have closed properly. @Craig Smith can you retrieve FLY183.DAT? It may actually be a continuation of FLY182.

It actually appears to end around 0.4 seconds before the txt log, based on the duration of the full elevator input. That seems more than would be accounted for by an errant last data point in the txt log.

 

[Meta4]

or do you think that the sudden elevator going to zero in the txt log might just be an artifact of the log stopping, rather than real?

The last data line in the txt file has updated values for Position, Altitude, Distance, Velocity Z, Pitch, Roll and Yaw.
Does that seem more real than an artifact?

 

[sar104]

The last data line in the txt file has updated values for Position, Altitude, Distance, Velocity Z, Pitch, Roll and Yaw.
Does that seem more real than an artifact?

Yes - it seems real. I think the DAT has ended early - maybe opening a new one.

 

[Craig Smith]

FLY183.DAT

DAT File

1drv.ms

this is the next DAT file. According to dates/times visible during export it was created while AC was on as I was downloading the DAT files (the first time I downloaded). I noticed during the export that there were two logs (one of which was a vision log) exported for each earlier flight, but no vision export for the crash. I was also using goggles on that flight. Is it because I wasn't using the goggles for controlling the AC that there was no Vision log generated?

 

[BudWalker]

FLY183.DAT

DAT File

1drv.ms

this is the next DAT file. According to dates/times visible during export it was created while AC was on as I was downloading the DAT files (the first time I downloaded). I noticed during the export that there were two logs (one of which was a vision log) exported for each earlier flight, but no vision export for the crash. I was also using goggles on that flight. Is it because I wasn't using the goggles for controlling the AC that there was no Vision log generated?

Thanks, but as you determined FLY183 isn't a continuation of FLY182. So, we are left with the mystery of why FLY182 terminated before the .txt

I suspect the vision logs have to do with the vision system and are unrelated to the goggles.

 

[Craig Smith]

I have no recollection of suddenly stopping the AC. Certainly there were no obstacles in the area, nothing of interest I was trying to film right there, etc. I was on a mission to go up a mountain and film some family/friends that were hiking up there, and was trying to get there as fast as possible. I just remember that there was a connection lost error displayed in the goggles, the image froze for a while then went to black/white I think - at the same time people who were with me reported hearing the sound of the drone stop/see it fall (it was still fairly close to us). We were quickly able to retrieve it, no damage to the props like you would expect from collision, one front leg busted and the camera/gimbal a mess, and the battery elsewhere but in good-looking condition. Pretty amazing for falling a bit more than 100 feet I think. Also wondering about thoughts on whether I should trust this battery? I haven't recharged it since the accident, its still at a bit more than 50 percent.