Battery data communication error

[sbonev]

hello, the other day after having zero issues so far in the last month of flying my mini i had the first major one, expectedly. long story short some 5 min in the flight unexpectedly the drone went from some 14 or 11 to 0 satelites, which got accompanied with a battery data communication error....This resulted of course in an ATTI mode switch. Some seconds later it acquired briefly 15 satelites, after that entered permanent ATTI mode! I had to return it to home and land as fast as i could. The lost of gps is not strange, but it is not explicable as i was above a lake some 80 m above launch altitude, no clouds or anything. it should not be solely a gps issue! the battery data communication error most definitely is a big issue and my suspicion is that it was the main culprit, as i had it for the whole resulting flight and the craft never recovered any gps afterwards until landing and even after that. This points to some possible hardware/software issue that is quite serious and needs to be addressed.
Luckily i never screen capture my flights, but today for some strange reason decided to do so, and i got all of it on video, here is the video:

I am attaching the logfile - I am using mode 3 for controls


https://www.phantomhelp.com/LogViewer/XR8BUKGWS220S91NEZPF

and dat files

20-03-15-03-40-50_FLY029.DAT

drive.google.com

20-03-15-03-36-53_FLY029.DAT

drive.google.com

20-03-15-03-35-35_FLY029.DAT

drive.google.com

There is nothing much on first glance in the log apart from the battery error and unidentified battery error.

After i landed i switched to a different battery and got another flight without issues, no gps issue at all, stable 14 satelites and 5 bars. Which makes me think issue is a combination of factors, most definitely software and possible hardware, and appear in certain scenarios. This should definitely be addressed by DJI as it is clearly a big and serious issue, drone would have never acquired gps with that error, most definitely gps was not the issue but something with internal flight controller. If there was wind it would have been an issue to recover it, even like that it was tricky to land as there were a lot of people passing, so i was relying on visual system and on having enough light, otherwise i would have had harder time for sure...
had i lost RC or video connection this drone was lost definitely and this is really frightening...not sure i can fly it confidently anymore.
I hope somebody can shed some other insight or idea as to what might have been the issue. I have tried again with the same battery, no issues, so it is definitely no battery malfunction. I also did 5 or 6 flight after that one without an issue, so i am almost certain it is some kind of firmware issue at this point.

 

[sar104]

It's difficult to know exactly what was happening here. All those files are just fragments of this flight, and the DAT files don't include any data past 287 seconds, which is just before the battery and GPS problems started. The txt file does show the satellite count dropping to zero and the battery ceasing to report data at 289 seconds.

 

[BudWalker]

I had looked at this some when it was posted over on the DJI forum. As you can see the .DAT was split into 3 separate .DATs which I concatenated back together. For anyone interested the attached FLY029.DAT can be used.

Unfortunately, the incident is coincident with the end of FLY029. I didn't find anything relevant in FLY029.

 

[sar104]

I had looked at this some when it was posted over on the DJI forum. As you can see the .DAT was split into 3 separate .DATs which I concatenated back together. For anyone interested the attached FLY029.DAT can be used.

Unfortunately, the incident is coincident with the end of FLY029. I didn't find anything relevant in FLY029.

I was trying to figure out whether the generation of multiple files had anything directly to do with the main issue, but came to no useful conclusions. Any thoughts?

 

[BudWalker]

I was trying to figure out whether the generation of multiple files had anything directly to do with the main issue, but came to no useful conclusions. Any thoughts?

Over on the DJI forum there was speculation that the battery connection was the root cause. This doesn't seem right to me. At the same time there were problems with the GPS and the .DAT stopped recording. This suggests some underlying problem that would affect all three of these systems. The flight continued for another 200 secs indicating there were no physical battery connection problems.

 

[sbonev]

I was trying to figure out whether the generation of multiple files had anything directly to do with the main issue, but came to no useful conclusions. Any thoughts?

no, this is because of the app restart, if you close the app it starts a new flight log and dat file, due to the issues with the app i had to restart it as seen in the video. my theory this could be what cause the issue at the first place...

 

[sbonev]

It's difficult to know exactly what was happening here. All those files are just fragments of this flight, and the DAT files don't include any data past 287 seconds, which is just before the battery and GPS problems started. The txt file does show the satellite count dropping to zero and the battery ceasing to report data at 289 seconds.

battery error appeared, then count of gps sats dropped to zero...then craft entered atti, and never got out of it until end of flight, also battery in the app stayed at 79% up till the end of the flight. i can only guess the issue with the battery turned off gps for some reason, or caused it to stop. at least there can't be any other explanation as it was a sunny day and at different height and location it was not getting even one sat up until i landed, and all got fixed when changed the battery - literally the minute i switched on the drone it got stable gps connection and never lost it for the whole second flight.

 

[sbonev]

Over on the DJI forum there was speculation that the battery connection was the root cause. This doesn't seem right to me. At the same time there were problems with the GPS and the .DAT stopped recording. This suggests some underlying problem that would affect all three of these systems. The flight continued for another 200 secs indicating there were no physical battery connection problems.

yes, i can only speculate. overall it is really strange problem, or at least it appears there was no issue with battery , as flight finished well, though in atti mode. and battery also is fine as i did two other flights after that day without a single issue. it has to be some more serous issue, i'd bet it is something with firmware and the dji fly app...

 

[BudWalker]

yes, i can only speculate. overall it is really strange problem, or at least it appears there was no issue with battery , as flight finished well, though in atti mode. and battery also is fine as i did two other flights after that day without a single issue. it has to be some more serous issue, i'd bet it is something with framework and the dji fly app...

Why would you think the Fly App was a contributing factor? The GPS and battery are on board systems. It's difficult to imagine how the Fly App would affect these.

What do you mean by "framework"?

 

[sbonev]

Why would you think the Fly App was a contributing factor? The GPS and battery are on board systems. It's difficult to imagine how the Fly App would affect these.

What do you mean by "framework"?

sorry, my bad, corrector issue - actually i wanted to say firmware....
well the issue happened after a restart of the app, and in it battery shows 79% constantly after error appeared. of course it doesn't make sense the app could affect the board systems, on theory should not be that. but for some reason it stops getting data from the battery, or it could be it got confusing data, not sure how is the connection to the dji fly done and what kind of data exchanged. that is why i can only speculate. and if it is an issue with the hardware or board systems it should be definitely replicable, so far didn't happen again, fortunately.

 

[sar104]

no, this is because of the app restart, if you close the app it starts a new flight log and dat file, due to the issues with the app i had to restart it as seen in the video. my theory this could be what cause the issue at the first place...

Restarting the app has no effect on aircraft control or its systems.

 

[BudWalker]

sorry, my bad, corrector issue - actually i wanted to say firmware....
well the issue happened after a restart of the app, and in it battery shows 79% constantly after error appeared. of course it doesn't make sense the app could affect the board systems, on theory should not be that. but for some reason it stops getting data from the battery, or it could be it got confusing data, not sure how is the connection to the dji fly done and what kind of data exchanged. that is why i can only speculate. and if it is an issue with the hardware or board systems it should be definitely replicable, so far didn't happen again, fortunately.

The battery being stuck at 79% is probably because the battery stopped communicating with the rest of the AC.

With the Mavic the GPS and Battery data are both less than 10 Hz. The .DAT recorded on the tablet .DAT is 10 Hz. It's possible that the data for these all move over the same bus. A problem with this bus may be the cause of the problems seen here.

I managed to find the performance data for the 7 Uarts on the MM (they will be in the next version). Nothing unusual, but the problem occurred after the .DAT stopped.