Mavic 2 Pro disconnected mid-flight and lost

[sar104]

Sorry if I wasn’t clear before, I sent them this error and asked for an explanation in my last exchange along with the file it was retrieved from and that was the response they gave.

I would suggest that you keep escalating the case, because the response you received doesn't address at all the entries in the DAT file - only the fact that the log ends. You haven't reached the necessary level of tech support yet.

 

[Simmo]

Yes. In the case that I assisted with, it took a little to and fro-ing to get the desired result of a replacement AC.
These logs showed the same similar errors as yours.

 

[GrFoss]

DJI did not treat it as a warranty matter. In fact, during the initial analysis, the tech support group communicated they would offer a 30% reduction. Only when I pushed to interact / speak with a supervisor, did I get what I was hoping for. A total replacement.

Hi, how you contact with supervisor? I have the same problem. I lost the drone without any reasons, I lost a signal in the ocean. RTH is did not worked.

 

[BudWalker]

That's the third M2 log I've seen recently with these core recovery errors:

333.765 : {task_e}INFO:i2c3 core recovery||​

333.765 : 21886 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

334.619 : {task_e}INFO:i2c3 core recovery||​

334.619 : 21936 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

334.660 : 21939 [L-COMPASS][mag auto cali(0)] init|​

335.207 : 21971 [L-FMU/LED]action changed. sys error:battery_status_data.communicate_err_status(9)​

335.472 : {task_e}INFO:i2c3 core recovery||​

335.472 : 21986 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

336.325 : {task_e}INFO:i2c3 core recovery||​

336.325 : 22036 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

337.179 : {task_e}INFO:i2c3 core recovery||​

337.179 : 22086 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

338.032 : {task_e}INFO:i2c3 core recovery||​

338.032 : 22136 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

338.885 : {task_e}INFO:i2c3 core recovery||​

338.885 : 22186 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||​

Loss of connection.....loss of drone

Mavic Crash, Can I please get a hand with Analysis?

In the first thread linked, DJI replaced the aircraft based on those errors in the event stream.

@Simmo - there's a pattern here.

@BudWalker - any thoughts on this?

Those messages in the eventLog stream are coincident with the SMART_BATTERY:status taking on the value BatteryCommunicationError. This can be seen by looking at the .txt. Synchronizing the .txt and .DAT is a little difficult since the DatHeader is corrupted - the DatHeader contains the info necessary to know the clock rate. Below the pink area denotes when SMART_BATTERY:status has the value BatteryCommunicationError. Note, also that the battery voltage data seems to flatline for about 5 secs before the error.


You referenced this incident
Flyaway Mavic 2 Pro
the SMART_BATTERY:status BatteryCommunicationError was not apparent in that one. I'm theorizing that it didn't last long enough to be seen in the .txt.

In this incident
Mavic2 Pro Lost :(
the SMART_BATTERY:status BatteryCommunicationError was coincident with the eventLog messages for 410 secs before the M2 lost connection



80.350 : {task_e}INFO:i2c3 core recovery||
80.350 : 6916 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||
:
:
497.514 : {task_e}INFO:i2c3 core recovery||
497.514 : 27816 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||

 

[sar104]

Those messages in the eventLog stream are coincident with the SMART_BATTERY:status taking on the value BatteryCommunicationError. This can be seen by looking at the .txt. Synchronizing the .txt and .DAT is a little difficult since the DatHeader is corrupted - the DatHeader contains the info necessary to know the clock rate. Below the pink area denotes when SMART_BATTERY:status has the value BatteryCommunicationError. Note, also that the battery voltage data seems to flatline for about 5 secs before the error.
View attachment 73481

You referenced this incident
Flyaway Mavic 2 Pro
the SMART_BATTERY:status BatteryCommunicationError was not apparent in that one. I'm theorizing that it didn't last long enough to be seen in the .txt.

In this incident
Mavic2 Pro Lost :(
the SMART_BATTERY:status BatteryCommunicationError was coincident with the eventLog messages for 410 secs before the M2 lost connection

View attachment 73482

80.350 : {task_e}INFO:i2c3 core recovery||
80.350 : 6916 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||
:
:
497.514 : {task_e}INFO:i2c3 core recovery||
497.514 : 27816 [L-BATTERY][ERROR] smbus_read CMD_POWER_STATE error||

Interesting - I forgot to check to see if those correlated. Good spot. That suggests that those event stream entries are just further evidence of a poorly connected battery rather than impending failure of the FC. That would also explain why I've seen a couple of logs containing those events where the aircraft continued to fly without any obvious problems.

 

[BudWalker]

Interesting - I forgot to check to see if those correlated. Good spot. That suggests that those event stream entries are just further evidence of a poorly connected battery rather than impending failure of the FC. That would also explain why I've seen a couple of logs containing those events where the aircraft continued to fly without any obvious problems.

Actually, credit goes to @msinger for seeing the connection. I just fleshed it out.

 

[sar104]

Actually, credit goes to @msinger for seeing the connection. I just fleshed it out.

That guy?

 

[acantril]

Interesting - I forgot to check to see if those correlated. Good spot. That suggests that those event stream entries are just further evidence of a poorly connected battery rather than impending failure of the FC. That would also explain why I've seen a couple of logs containing those events where the aircraft continued to fly without any obvious problems.

Im the sad owner of the 'Mavic2 Pro Lost' thread ... but at least I start to get some understanding of what happened. It was my partner flying at the time so the logs are the only exposure I have to what happened.

When you say 'poorly connected battery' - is there any ongoing though as to the battery not being seated and clicked in correctly - or it being a point of electrical contact issue?

The annoyance from a pilot perspective is no visual indication of a battery issue on the controller ... that could have potentially saved a drone ..

 

[wco81]

So has anyone reported to DJI?

Sounds like bug in one of the recent firmware updates?

 

[austin1124]

Interesting - I forgot to check to see if those correlated. Good spot. That suggests that those event stream entries are just further evidence of a poorly connected battery rather than impending failure of the FC. That would also explain why I've seen a couple of logs containing those events where the aircraft continued to fly without any obvious problems.

This is interesting information, and it is also kind of what they said to me in the most recent message: "The following message would appear when the battery is not installed well before taking off." So I am assuming that they will point to that again and it will mean that it was user error and not replaceable under warranty/the best offer I will get is the 30% off a replacement unit? I just wonder how others with the same issues got the replacement result.

 

[acantril]

Thats the thing. I'm not sure how its possible with the MP2 to not install the battery correctly.. it has a massive click and switching the unit on requires a certain amount of downward force.
is it even possible for the battery to 'work' without it fully clicking.

 

[sar104]

Thats the thing. I'm not sure how its possible with the MP2 to not install the battery correctly.. it has a massive click and switching the unit on requires a certain amount of downward force.
is it even possible for the battery to 'work' without it fully clicking.

Yes - there are numerous cases of that happening.

 

[FormerMember07122019_2]

Thats the thing. I'm not sure how its possible with the MP2 to not install the battery correctly.. it has a massive click and switching the unit on requires a certain amount of downward force.
is it even possible for the battery to 'work' without it fully clicking.

There are actually two clicks, one for each side, that if you push straight down sounds like one bigger click. As SAR104 mentioned there have been a number of cases with batteries not fully seated and then becoming disconnected, or sometime intermittent during flight. Some of the folks reporting problems have had some slight swelling of their battery which restrict it from being fully seated.

 

[Prismatic]

Yes - there are numerous cases of that happening.

One would think it possible to engineer a foolproof battery mount that can't work unless it's 100% positively secured. One would think, wouldn't one?

 

[Thomas B]

One would it possible to engineer a foolproof battery mount that can't work unless it's 100% positively secured. One would think, wouldn't one?

Sounds like something doable, but would require hardware changes... like an interlock on the battery clip seats on each side. Could work like not being able to start your car if you’re not pressing the brake pedal.
Send the idea to DJI

 

[msinger]

Sounds like something doable, but would require hardware changes... like an interlock on the battery clip seats on each side. Could work like not being able to start your car if you’re not pressing the brake pedal.
Send the idea to DJI

It could be done. They added a button on the P4P V2 to detect an improperly inserted battery.

 

[ff22]

One would it possible to engineer a foolproof battery mount that can't work unless it's 100% positively secured. One would think, wouldn't one?

Maybe like the battery mechanism on the Osmo Action Camera - if the orange shows, it is NOT engaged properly.

But maybe just another pilot pre-check necessity - is the battery firmly inserted and locked.

 

[austin1124]

Yes - there are numerous cases of that happening.

Interesting. All this being said, I am still waiting on a response back after sending the "please explain this to me further" email about the errors, if they come back again with it being an issue with the battery locking in and not covering it under warranty would you advise I just accept the 30% off and move on or keep pushing? I'm currently traveling for a few more weeks and would love to try to get a replacement before I return, but it's also a huge amount of money to concede when I'm not even really sure it was my fault.

 

[austin1124]

Here’s the latest update after asking for an explanation for the error, escalating to a supervisor and a readout of the event stream, they are holding firm on 30% so it appears that is the best I can do unless anyone has any other advice. I appreciate all the consideration.

Dear Customer,


Thanks for getting back to us.

We are sorry that the answers are not the one that you are expecting, however, the record ended does not mean the aircraft malfunctioned. As we have been stated in the last email that there are many possibilities will lead to the interruption of the flight record or the connection between the aircraft and the controller. Interference (visible & invisible) or the direction of the antenna would also affect the remote connection.

Yes, the aircraft will start to return to home when the RTH is triggered even it lost connection, provided that the GPS signal is fine and a home point was recorded previously and the flight environment is proper without obstacles and interference, while based on the current data, we could only see the record ended at T=06:37 but we could not make sure the remote signal was also lost at the same time. The loss of the remote signal would cause the disconnection of the flight record; but the disconnection of the record did not mean the remote signal was also lost. If the remote signal was not lost at that time, the RTH would not be triggered. What's more, even if the remote signal was lost, with the record ended, nobody could know what happened after that. But that did not mean the RTH function was failed.

Regarding the "error" message that you've mentioned, if the battery was not installed properly, the hazards will exist during the flight. As the record ended in this case and the aircraft was lost. We are unable to figure out the main cause of the incident. However, the aircraft was responsive to the pilot's command and there is no malfunction found before the record ended. You could also check it with your flight record as well.

We also feel sorry for your loss and we want to try our best to help you if we can, however, a 30% off discount is really the best that we could offer in this case, please consider if you would like to take it. Sorry to tell you that the supervisor doesn't handle customers' cases directly, however, when necessary we do contact them for the solution. Also, the 30% off offer is already the best discount they can offer based on the analysis result while obeying the warranty policy.

Again, we are sorry for your disappointing feelings.

Should you have more questions, please feel free to let me know.

Thanks for your support. Have a nice day.

 

[sar104]

Here’s the latest update after asking for an explanation for the error, escalating to a supervisor and a readout of the event stream, they are holding firm on 30% so it appears that is the best I can do unless anyone has any other advice. I appreciate all the consideration.

In light of the observation that those event stream errors correlate well with the battery communication error, it seems likely that DJI is concluding that an improperly latched battery caused the problem. While they did replace at least one other aircraft, apparently on the strength of those same event stream errors, that's probably a fair conclusion, as there is no other indication of a problem.