Phantom 4 Pro V2.0 crashed mid-flight

[Andris8888]

Hello Everyone,

I am trying to find out the exact cause of a sudden fall mid-flight. This was not my drone. I repair drones since 2016 (1000+ repaired drone) and I have some experience in analyzing log files with csvview and airdata but I think I can learn more. I cannot find any clue in the log files what would cause the sudden stop of this drone. I suspect something was not right with the battery, maybe undervoltage caused some problems as the drone took off with 40% battery... Motors, ESC communication seems good to me. I also cannot see FC errors. Weather was not bad, not windy, temperature 5C (40F). Pilot told me that the battery was still in the aircraft when it crashed and it was on so not popped out of the shell.

Here is the AirData link:
I uploaded the DAT file if someone more experienced analyst can spot something interesting. I try to get the RC log too as I am curious about the RC side of the story. I couldn't see any abnormal in the Event log too (I can paste if needed, but the DAT file contains it).

Thanks for your help, and I hope I can learn something in the field of black box analysis.

 

[Andris8888]

Actually I couldn't attach the DAT file as it was too big. SO I exported a csv, I hope it will fit. Tell me if it doesn't containg everything needed.

 

[Meta4]

Here is the AirData link:

With mid-flight fall from the sky incidents it's often difficult to identify the cause, because it's a sudden loss of power and the data just stops while everything appears normal.
The numbers coming out of Airdata are odd and hard to make sense of.
It's not letting me download the original .txt file to look at it properly.
If you could post the original .txt file that you uploaded to AIrdata that would be helpful.

Or upload to Phantomhelp and post a link for that.
Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.

 

[Andris8888]

Thenk you @Meta4 for your assistance! I don't have the txt log yet (I suppose I should get it from the RC, I will try tomorrow). In the meantime I uploaded the DAT to Wetransfer and here is the link for it: FLY188.DAT

I found this strange graph in CsvView:

PPMSned and PPMRecv graph is moving together indicating that the ESC communication is fine (I hope I understand this well). But in the last moments even though the controlling signal stays solid, the motor current drops to very minimal values. Also if I I graph the Volts and the PPM signals I can see that the voltage drops fast in the end.

Not sure if I understand everything correctly so please correct me if I am wrong. I think starting a flight with a 40% battery in cold weather was not a good idea. If it was a bettery depleted from the previous flight then the meter reading would be accurate with 40% but the previous flight (according to the previous log file) was done on an other battery. So this battery may have been self discharged without knowing itself and the BMS reported 40% even that the real capacity was maybe lower. I don't know for sure.

I was also thinking about the battery switch too in the V2.0 drones. What happens if the battery sensor switch disengages during mid-flight? I know it won't allow you to take off and warns you in the app but if you are in the air what happens if the battery gets loose and the switch opens the circuit (but battery still powering the drone)? I hope it just warns too and doesn't stop the motors immediately. If would be very illogical and unsafe.

 

[Meta4]

I think starting a flight with a 40% battery in cold weather was not a good idea. If it was a bettery depleted from the previous flight then the meter reading would be accurate with 40% but the previous flight (according to the previous log file) was done on an other battery. So this battery may have been self discharged without knowing itself and the BMS reported 40% even that the real capacity was maybe lower. I don't know for sure.

If the cell voltages I see with Airdata are correct, the cell voltages at the end of the flight record were all>3.5 volts and the % indicator showed 26%.
With cell voltages >3.5 V, the battery was not so run down that it was the cause of the drone crashing.

 

[Andris8888]

Yes, that would be strange for me too. But as you can see the end of the log is missing. I am unsure how much time is missing from the end...

 

[Meta4]

Yes, that would be strange for me too. But as you can see the end of the log is missing. I am unsure how much time is missing from the end...

It's most likely that the power was lost at the time the data stops.

 

[Andy9]

This is an another midflight crash with almost depleted battery and you said that you were not sure for how long the battery was idle? To start with 40% battery is a call for trouble. Was this on purpuse or you didn't motice that?
I'm not an expert on reading these flight reports but it is pretty obvious that the drone lost the power.

 

[Phantomwolf]

This is an another midflight crash with almost depleted battery and you said that you were not sure for how long the battery was idle? To start with 40% battery is a call for trouble. Was this on purpuse or you didn't motice that?
I'm not an expert on reading these flight reports but it is pretty obvious that the drone lost the power.

I don't think the OP was the pilot. He mentions in his first 3 lines he is repairing the drone.

 

[JC in NH]

This is an another midflight crash with almost depleted battery and you said that you were not sure for how long the battery was idle? To start with 40% battery is a call for trouble. Was this on purpuse or you didn't motice that?
I'm not an expert on reading these flight reports but it is pretty obvious that the drone lost the power.

The OP wasn't the pilot. He just wants to gather the best available information for his repair service customer.

 

[Andy9]

Oh yes, I missed this. But anyway, my comment is therefore for the pilot if he is reading this.

 

[Andris8888]

Thanks for the comments, yes, I was not the pilot. I also think that taking off with 40% battery calls for catastrophy. However it would be strange that the drone lost power mid-air but dropped with its battery still in place and turned on. Maybe some power fluctuation that restarted the FC? In the Airdata panel (Power / Volts and Amps) I can see "Low voltage" warning 3 times. The first two while climbing (seems normal) but the third is while floating.

 

[Andy9]

When the voltage of one cell drops under 3,3V or near it the drone shuts down. This is stupid I know. It is for battery protection. LOL

 

[Andris8888]

Do you see in the log if that happened? Is there an official info about this somewhere? Definitely interesting theory.

 

[Pygar70]

Thanks for the comments, yes, I was not the pilot. I also think that taking off with 40% battery calls for catastrophy. However it would be strange that the drone lost power mid-air but dropped with its battery still in place and turned on. Maybe some power fluctuation that restarted the FC? In the Airdata panel (Power / Volts and Amps) I can see "Low voltage" warning 3 times. The first two while climbing (seems normal) but the third is while floating.

It might be strange, but it's not that unusual. In five years of regular flying, I've had my Phantom fall out of the sky (sudden loss of power) four times. It's somewhat disconcerting.
Personally, with my own experience and that of others who've posted online about similar occurrences, I think there's a design flaw. Meta4, however, disputes my theory.

As for taking off with 40% battery leading to catastrophe: Almost every flight I've ever done has my drone in the air, at some point, with a 40% battery charge. So, logically speaking, there must be other issues that precede a launch where the battery is not fully charged that lead to problems - other than the low charge itself.

 

[Meta4]

Do you see in the log if that happened? Is there an official info about this somewhere? Definitely interesting theory.

Except that it's not true.
If your battery cells drop to or below critical low voltage levels, (this one never did), you can get a message about reduced power.
If you keep flying, you see the cell voltages continue to drop and eventually the drone shuts down.
In this case the cell voltages never fell below critical low voltage and were still >3.5 volts when the data ended.

It might be strange, but it's not that unusual. In five years of regular flying, I've had my Phantom fall out of the sky (sudden loss of power) four times. It's somewhat disconcerting.
Personally, with my own experience and that of others who've posted online about similar occurrences, I think there's a design flaw. Meta4, however, disputes my theory.

I'm not sure what your theory is.
Losing power 4 times seems like extraordinarily bad luck or you are doing something wrong.

As for taking off with 40% battery leading to catastrophe: Almost every flight I've ever done has my drone in the air, at some point, with a 40% battery charge. So, logically speaking, there must be other issues that precede a launch where the battery is not fully charged that lead to problems - other than the low charge itself.

Launching with a battery that has been sitting around discharging, is asking for trouble.
There are many cases of fall-from-the-sky incidents from doing this.
The % indicator is not reliable in this situation, as shown by the cell voltages shown in recorded data.

But flying to 40%, landing and flying again shortly after is not a problem.

 

[Pygar70]

I'm not sure what your theory is.
Losing power 4 times seems like extraordinarily bad luck or you are doing something wrong.

My theory is there's something about the design or manufacture of the drone that contributes to the 'sudden loss of power' incidents.

Three different (P4) drones were involved in those four crashes. I can think of four possible causes:

a) Pilot error - not sure what. (I'm alert to the problem of incorrectly seated batteries dislodging during flight. This happened once - in addition to the four crashes I mentioned.)
b) Battery failure (a mixture of old and new batteries, all in regular use and fully charged before flight).
c) A drone design flaw.
d) Random occurrence (over more than 1,000 flights - most things that can go wrong, will go wrong).

When this happens to your drone, you go online and soon find lots of other people that it's happened to as well - all seemingly without pilot error. However, without more data, it's hard to draw conclusions.

 

[MachDiamond]

The reliance on SOC meter is root cause, if cells are recently cycled it reads fine if cells were in storage SOC reads high until aircraft is subjected to load hence sudden voltage drop-out occurs w/o a record of what went wrong. The integrated BMS draws such little power that cells deplete but retain a surface charge and not understanding why.

 

[Andris8888]

Moved forward with the analysis of the logs. I checked 10 logs before the flight and only one flight with the same battery was on 16 november 2022. I think it was just a motor test as the flight data is very short (50s), no GPS, Ultrasonic sensor height 0, and the capacity percent remained 63% during the "flight".

So I suspect that maybe it was the last time that battery was used (at least in this drone) before the crash.

Today I opened the drone and checked for obvious signs for the cause of the crash besides the battery undervoltage. I didn't find anything special and the drone can turn on and work as normal (except the gimbal is torn away in the crash).

However I found this ribbon cable (FC_L) a little bit pulled out of one connector. It is still working normally and I don't think it could be the cause of the accident but maybe got a little pulled out during the crash. What are your opinions?

I also tested the battery sensor switch during "flight" and disengaging the switch with the motors already started and drone reporting "In Flight" it doesn't stop the motors. While the motors are not started it sends a warning "Cannot take off" into the app but if the motors are already started then it doesn't do anything, not even a warning that the battery is not seated perfectly which is interesting logic from DJI.

 

[Andris8888]

Went even further and found another flight with this battery on the same day 16 november 2022. The log shows that after the drone takes off the voltage drops to 12V almost instantly. I plotted the height graph too and it shows rapid descent (maybe a crash too).

Then the drone reports the previously shown 63% capacity in the next flight a few minutes after. And if I assume correctly then this same battery was not charged before the last flight in question in the opening post. In the 10 flights in between november 2022 and march 2023 other batteries was used. Then this same battery reported 40% percent and could fly for about almost 2 minutes before crashing.

Here is the november 2022 flight (I can even see a major cell voltage deviation in this log)


Am I safe to assume that the crash was because of a faulty battery?