MA battery failure scare

[Prismatic]

TLDR: Rapid loss of power due to battery failure (?) in a Mavic Air, auto-landed in tree; got lucky, easy recovery, minimal damage, but still wondering what-the-devil really happened, and implications regarding same-age batteries. Airdata link to the ill-fated flight: https://app.airdata.com/share/orjCoJ.
--------------------------------------
While camping in September, I took a short morning flight, to get an aerial view of our camp site in the mountains of Colorado. I took off with a freshly charged battery. It was one of my first flights taken with the FT Aviator v1 software/hardware as the controller, with an iPad Mini 4 for display. (Thus I'm unsure of the value of the FTAviator flight TXT file; it looks encrypted, but then so do the TXT files from DJI Go4, which do get used for analysis, so ...)

It took a while to get a good GPS fix, so I took off with 96% battery. Seconds into the flight I began getting a "Critical power level! Landing now!" message (I forget the exact wording, but that was the message, in red.), though I'd inadvertently muted the iPad so I think I failed to notice at the first instant. When it began autolanding, I still had minimal control, and cleverly managed to get the bird centered over one of the nearby scattered trees before it settled down for good. Fortunately, it was a young fir with soft branches & needles, and the drone gently tumbled down through it to settle on a branch within easy reach. One prop was nicked--on the rear edge (?)--but otherwise there was no hint of damage. Subsequent flights with a replacement prop have all been normal.

I'd recently read here about how the collapse of a battery cell could result in this sort of behavior. So I put the battery from this flight away for the rest of the trip, and am just now getting around to asking if an expert can offer any insight. I uploaded the flight to Airdata; there's a link to that above. I've also attached the FT Aviator TXT file, in case that's of use to anyone.

But I don't know what to make of the Airdata info. The "Cell Voltage" graph in the Power section (shown) is alarming, but how/why does the battery "recover" near the end of the flight? Furthermore, the "Efficiency" map in the Power section shows initial problems but "full power" at the end, when the MA was most definitely autolanding on account of power issues. Airdata tells me I had 94% power at the end, but that drone was not going to fly a minute more.

Finally ... this is one of three batteries from the FlyMore package I bought in Feb. 2018. It has 73 flights on it, and 79 charge cycles (static use would be downloading data and similar). I don't see any indication of impending doom in its prior history (via Airdata), certainly not in the prior half-dozen flights I closely examined. But I wonder now if this sort of failure is on the horizon for the other 2 "sister" batteries from that purchase.

Oh, and FWIW, I'd recently run a "battery maintenance" procedure against this battery (and the other five in my kit): see PPS below for details. This was the first flight subsequent to that.

==> Any thoughts will be of interest, particularly regarding the apparent inconsistencies (?) in what I see on Airdata and/or the prognosis for my other batteries.

PS: I flew this battery once since the incident, just around the back yard. The behavior was "normal", as was the "Cell Voltage" graph and other parameters seen in Airdata. But I daresay I won't be flying this battery again for anything other than test purposes! (And I definitely got lucky with the 'bad' flight; my next destination after the campsite selfie would have put me over the wetlands at the head end of a lake!)

PPS: This is the battery maintenance procedure I ran just before the mishap, taken--for better or worse--from Airdata recommendations:

1. Fly until battery level reaches 25%-30%
2. Allow battery to cool completely to room temperature
3. Put battery back in and turn on aircraft (optionally starting motors with no propellers) and allow battery to discharge until it gets down to 8%, or until the battery can no longer be turned on. Launch the DJI GO app to check battery levels.
4. Allow the battery to cool completely again to room temperature
5. Recharge battery normally

 

[old man mavic]

View attachment 85827
TLDR: Rapid loss of power due to battery failure (?) in a Mavic Air, auto-landed in tree; got lucky, easy recovery, minimal damage, but still wondering what-the-devil really happened, and implications regarding same-age batteries. Airdata link to the ill-fated flight: https://app.airdata.com/share/orjCoJ.
--------------------------------------
While camping in September, I took a short morning flight, to get an aerial view of our camp site in the mountains of Colorado. I took off with a freshly charged battery. It was one of my first flights taken with the FT Aviator v1 software/hardware as the controller, with an iPad Mini 4 for display. (Thus I'm unsure of the value of the FTAviator flight TXT file; it looks encrypted, but then so do the TXT files from DJI Go4, which do get used for analysis, so ...)

It took a while to get a good GPS fix, so I took off with 96% battery. Seconds into the flight I began getting a "Critical power level! Landing now!" message (I forget the exact wording, but that was the message, in red.), though I'd inadvertently muted the iPad so I think I failed to notice at the first instant. When it began autolanding, I still had minimal control, and cleverly managed to get the bird centered over one of the nearby scattered trees before it settled down for good. Fortunately, it was a young fir with soft branches & needles, and the drone gently tumbled down through it to settle on a branch within easy reach. One prop was nicked--on the rear edge (?)--but otherwise there was no hint of damage. Subsequent flights with a replacement prop have all been normal.

I'd recently read here about how the collapse of a battery cell could result in this sort of behavior. So I put the battery from this flight away for the rest of the trip, and am just now getting around to asking if an expert can offer any insight. I uploaded the flight to Airdata; there's a link to that above. I've also attached the FT Aviator TXT file, in case that's of use to anyone.

But I don't know what to make of the Airdata info. The "Cell Voltage" graph in the Power section (shown) is alarming, but how/why does the battery "recover" near the end of the flight? Furthermore, the "Efficiency" map in the Power section shows initial problems but "full power" at the end, when the MA was most definitely autolanding on account of power issues. Airdata tells me I had 94% power at the end, but that drone was not going to fly a minute more.

Finally ... this is one of three batteries from the FlyMore package I bought in Feb. 2018. It has 73 flights on it, and 79 charge cycles (static use would be downloading data and similar). I don't see any indication of impending doom in its prior history (via Airdata), certainly not in the prior half-dozen flights I closely examined. But I wonder now if this sort of failure is on the horizon for the other 2 "sister" batteries from that purchase.

Oh, and FWIW, I'd recently run a "battery maintenance" procedure against this battery (and the other five in my kit): see PPS below for details. This was the first flight subsequent to that.

==> Any thoughts will be of interest, particularly regarding the apparent inconsistencies (?) in what I see on Airdata and/or the prognosis for my other batteries.

PS: I flew this battery once since the incident, just around the back yard. The behavior was "normal", as was the "Cell Voltage" graph and other parameters seen in Airdata. But I daresay I won't be flying this battery again for anything other than test purposes! (And I definitely got lucky with the 'bad' flight; my next destination after the campsite selfie would have put me over the wetlands at the head end of a lake!)

PPS: This is the battery maintenance procedure I ran just before the mishap, taken--for better or worse--from Airdata recommendations:

1. Fly until battery level reaches 25%-30%
2. Allow battery to cool completely to room temperature
3. Put battery back in and turn on aircraft (optionally starting motors with no propellers) and allow battery to discharge until it gets down to 8%, or until the battery can no longer be turned on. Launch the DJI GO app to check battery levels.
4. Allow the battery to cool completely again to room temperature
5. Recharge battery normally

when this problem happens the drone reacts as if the battery has reached a critical low battery level and as you saw tries to land all lipos have a,voltage recovery period after use as they cool you could land at a certain voltage remaining and then after recovery you will actually have more charge left i have landed at a stated in the app 30% battery with the first light solid and the second one flashing only to get home some time later and found that i have two solid lights showing as the cells have recovered

 

[sar104]

If that was a freshly charged battery then it's a dead battery - the voltage on all cells collapses badly as soon as significant current is pulled:

 

[Prismatic]

@sar104: I guess I'm not surprised to hear that, yet as I mention late in my rambling, I have flown that battery since this incident. Not only were there no issues, the retrieved data--which I can only get via Airdata--looked 100% normal, just like all prior flights with this battery.

But on further review, I discovered that battery—indeed that cell of that battery—had suffered voltage issues on prior flights. And it’s strangely inconsistent. One flight shows voltage issues, but the next four don’t. Then another sketchy flight, and back to normal.

Is that common behavior wrt the Mavic Air battery, or I suppose, lipo batteries in general?

Also, does anyone know of a convenient way to discover—before your drone unexpectedly loses power over a swamp ?—that a battery is intermittently experiencing an issue with one or more cells? (The answer is not DJI Go4, it doesn‘t pass the ‘convenience’ test! ?)

 

[ob1quixote]

Every time I read a new thread, I realize I dont know much about this thing! Hope it survives my learning curve!

More reading to do...

 

[jeffcutler]

OK, let's jump into this. First, I only know what has happened to me. Second, I've had three battery failures - all from the Fly-More MA package purchased in Nov. 2018.

The first two failures happened after I put them through normal flying, but ran them down to around 10%. These two didn't give critical errors, but both of them - one in March and one in July - wouldn't take a charge. They just stayed with all lights off no matter what I did. **If I researched, I see that maybe there was a way to recover them, but it's just a MAYBE.

The THIRD was in September. I had done a bunch of flying and used all three batteries in order (***At this point two of the batteries had be replaced by my local retail outlet - just a swap in and new battery out to me. They and I believed warranty period was a year, so all the batteries fell into that period. AND they were from the 2018 batches mentioned in other threads about bad MA batteries.)

...so using all three in order, I saw a farm I wanted to fly on the way home from my paid mission and I stopped there and asked the owner. I'm 107 and the farm was NOT in controlled airspace, so I was good to go. I charged ONE battery in the car...and patiently waited for it to fully charge. Once it did I took to the sky.

Flight went fine until I was thinking of wrapping up around 53% of the battery - WHEN IT ALL WENT RED AND BEEPY and screamed at me about critical level, etc. etc.

Suffice it to say, I was able to control the UAV like a glider until the props stopped and it dropped from about nine feet onto some hay....MERE INCHES from a jagged stone wall.

I took the battery home and tried everything to get it to charge, but it wouldn't even show any lights. Store took it back and swapped it out. Now I have three newish batteries, but don't want to experience that heart failure scenario again.

Any advice - other than ramping up the percentage warning and not doing dual flights on the same battery in the same day?

Thanks!

Jeff

 

[Prismatic]

... Any advice - other than ramping up the percentage warning and not doing dual flights on the same battery in the same day? ...

I doubt flying a properly cooled and recharged battery twice in one day would b a problem, unless the battery is already on the verge of failure.
I subscribe to the Airdata service. After the event that spurred this thread, I examined the history of all six of my batteries. Besides the battery that caused this incident, one of the other batteries from the original Fly More package showed multiple instances of major voltage deviations in its recent flights. I preemptively retired that one, too.
I replaced both with new ones, to keep a sextet of flight-safe batteries in my kit.

 

[Deleted member 94047]

This is why I review battery performance after each flight. Thankfully, Airdata makes it easy to do that; have your logs sync with Airdata and review the flight data as soon as you can; make notes on any notable battery performance issues, and rank your batteries accordingly. I have one battery that often shows some significant cell deviations and I only use it for flights I designate "safe". For a flight that will take my AC over water or out of visual range, I only use the safest of my batteries.

To the OP, would the battery in question happen to be made by SUNWUDA by chance?

 

[Prismatic]

... To the OP, would the battery in question happen to be made by SUNWUDA by chance?

I have no idea, nor the first clue how to determine the manufacturer. Both retired batteries were in the MA Fly More bundle I purchased in February 2018.

 

[Deleted member 94047]

I have no idea, nor the first clue how to determine the manufacturer. Both retired batteries were in the MA Fly More bundle I purchased in February 2018.

It is written at the bottom end of the text on the reverse side of the battery. This is mine as an example

 

[Prismatic]

All my batteries, old or new, were made by Dongguan Amprex Technologies.

 

[Deleted member 94047]

Thanks. I was just curious because the batteries I have that are not doing so well were made by Sunwoda. The one battery I have made by Dongguan Amprex is still doing great though. So may be manufacturer doesn't matter after all.