Mavic Air 2 dropped out of sky

[BudWalker]

The reason for this is most probably a seizing left front motor ... already at 137sec into the flight the AC wanted to force land but you cancelled it ... then again at 175sec ... & again at 199sec, but you aborted & continued to fly. This was early warning signs ...

This was the message out from the DAT log event stream at the first force landing ... no apparent reason is given there unfortunately.

View attachment 118814
Then at 243sec the current to that left front motor started to go high ... something makes it hard to turn, see yellow graph.

(Click on the charts to make them larger)
View attachment 118815

At 280sec the flight controller orders full command to the left front to have it keep up the rpm's as needed, see purple graph.

View attachment 118816

At 286sec it's game over & the left front motor seizes & drops the rpm, see purple graph... a couple seconds later the FC orders motor shut down as seen from the DAT log event stream further below ...

View attachment 118817

Here the DAT log event stream that explains why all motors stops ...

View attachment 118818

Hard to say anything about the reason for this ... but you mentioned that you have crashed earlier so a bent motor shaft? Or debris going into the motor at start or earlier landings ... or manufacturing fault?

Contact DJI to get their point of view on this...

I noticed something that is bit odd. All 4 of the motors showed a gradual increase in current and commanded even though the speeds were relatively constant.

This doesn't happen in the other MA2 .DATs that I looked at. Perhaps the problem is not specific to a particular motor.

 

[zeusfl]

Maybe an ESC failure then.

 

[Ralph thompson]

In Canada it’s recommended that in Class G airspace, when operating a drone near aérodromes (land or water operations) you contact the aerodrome manager and monitor the Unicom frequency. Also National parks are off limits and operating near or over wildlife is illegal. Lots of little caveats. Fly safe.

 

[slup]

I noticed something that is bit odd. All 4 of the motors showed a gradual increase in current and commanded even though the speeds were relatively constant.
View attachment 118873
This doesn't happen in the some other MA2 .DATs that I looked at. Perhaps the problem is not specific to a particular motor.

Yeah ... good spot! Hmm, a bit strange ... suggests something else failed further back in the chain.

Maybe an ESC failure then.

That's only one step back from the individual motors ... for all 4 to start to rise the current draw it sounds like even further back as it's several ESC's... isn't it?

But then again ... this is what's indicated in the TXT log ...



And from one force landing occasion from the DAT ..?



And this is the only times ESC's is mentioned in the DAT ...

 

[Kilrah]

Most likely prop icing, dew point was almost equal to temp and temp was freezing. Flying in these conditions is a big no-no.

Nothing wrong with the aircraft.

Guess we now know that the MA2's props increase drag before losing lift when they ice.

 

[David sigglekow]

Hello,

I was flying my Mavic Air early this morning and got a motor warning. I'd twice had to cancel aircraft landing which had come up randomly, and tried the return to home function. After appearing to do nothing for about 30 seconds whilst beeping at me the drone then dropped from the sky. It was still connected to the controller and I could see the lines on the road it had landed on.

I have retrieved the drone which is badly damaged.

I've crashed a fair few which have 100% been my fault but this one was very strange.

Please can someone have a look at the log and see if I did anything wrong? I don't want to be stumping up £5/600 for a repair or new drone if it was a failure from the drone.

It's only around 1 month old, no DJI refresh (I know)


Thanks in advance.

Alex

I had my air 2 drop out with no warning ,Lucky no damage, quoted as motor obstruction which their was none but would not restart on one motor,took it in and was told a warranty will take 4,5 weeks if it gets accepted as I'm travelling i will pay for a new arm which was less than I thought hopefully will take a week,worth checking under warranty for your damage as camera is .ots of cash

 

[JackL]

Well this has certainly been an informative thread. Major takeaway, be thankful I live in South Florida where prop icing will never be a thing. Also, land the moment the drone wants if it starts throwing motor errors.

 

[slup]

Most likely prop icing, dew point was almost equal to temp and temp was freezing. Flying in these conditions is a big no-no.

Nothing wrong with the aircraft.

Guess we now know that the MA2's props increase drag before losing lift when they ice.

View attachment 118911

Yeah, that may very well be a contributing factor. Now, I never have looked into the logs from a prop icing case before, but ... what disturbs me a bit is that the currents wasn't that high so that itself should have caused the left front motor to shut down...

If looking into the timings for different events a bit ...

This is what caused ALL motors to shut down according to the DAT log event stream... mainly the AC is rolling over due to a excessive pitch or roll movement, the stop is initiated at 286.107sec


In below chart the timing for the left front shut down is 285.908sec ... that creates 286.067sec first a roll to -71 degrees & soon after at 286.186sec a pitch to 126 degrees. These excessive movements is more then enough to cause a all over motor shut down.



So still ... the front left motor is first out to stop & the reason isn't that the AC is flipped upside down as for the other 3.

The currents on the motors when the left front stopped at 285.908sec here below. Have seen currents higher then this without the motors stopping ...



If it would have been a lost thrust due to icing we wouldn't have seen a motor stop I think ... instead heavy instability in all axis.

Even though some instability can be seen in roll & pitch when the AC is manually yawing just before the left front motor shuts down ... it's not alarming.

 

[Kilrah]

The currents on the motors when the left front stopped at 285.908sec here below. Have seen currents higher then this without the motors stopping ...

Current doesn't matter, the issue is that the FC was already giving 100% power to the front left motor, so it simply could not command more to maintain the attitude and prevent the flip.

If anything current being lower than you would normally see at full commanded power confirms the motor/prop didn't behave as normal because of the ice.

 

[slup]

Current doesn't matter, the issue is that the FC was already giving 100% power to the front left motor, so it simply could not command more to maintain the attitude and prevent the flip.

If anything current being lower than you would normally see at full commanded power confirms the motor/prop didn't behave as normal because of the ice.


View attachment 118962

When it comes to the Mini for instance ... we saw a lot of 100% commands together with "notEnoughforce" errors when they had those uncommanded descents ... didn't see one case there when the motors shut down due to this... instead instability in all axis & in height.

So don't think the left front motor stopped due to this ... then the flip itself wasn't either due to that the front left was at 100% command, the flip occurred when that motor stopped.

Still odd that the front left stopped ...

 

[Kilrah]

You can see in the graph above that it couldn't maintain yaw or height, and that pitch and roll spike before the motors stop so they all stopped because of the flip.

 

[slup]

You can see in the graph above that it couldn't maintain yaw or height, and that pitch and roll spike before the motors stop so they all stopped because of the flip.

The left front clearly stops before the excessive pitch & roll ... At 285.868sec the motor rpm falls, then the flip occurs ...



... which initiate an all motor shut down at ...




So why shuts the front left down ?

 

[Kilrah]

The ESC was commanded 100% and the motor indeed stopped - I guess that would be becasue the ESC was at 93°C and with the combination of being that hot and commanded full power it triggered overtemp shutdown...

 

[slup]

Yes ... I saw the ESC temps already earlier, all 4 are rising & follow each other pretty equal. Don't think it's that either, it should have been seen in the DAT log event stream.

I actually think it's an easier conclusion that builds on your earlier observation regarding risk for icing ... all 4 currents, all ESC temps are slowly rising ... some slight instability can be seen & then especially the left front needs more command to keep up the rev, but then stops.

I think it wasn't the props that got the ice buildup & it wasn't increased drag from the props that raised the current draw ... it was the motors. The front left was the first to grind to halt due to a seizing motor coming from icing ... that makes the AC flip & game over.

Would be interesting if DJI for once surprise us with a proper explanation ... if the OP files this case to them.

 

[Kilrah]

Yep the final second doesn't matter, it's very clear that things are degrading gradually during the flight and once a motor was at max command it was going to come down in a way or another, how exactly is pretty irrelevant.

Motors don't ice, they're nice little heaters. Props do by the very way they operate.

 

[slup]

... how exactly is pretty irrelevant.
Motors don't ice, they're nice little heaters...

Yeah ... but it's sure interesting to try pinpointing the exact cause

And regarding freezing motors ... well thought that also, but did see ice/frosty buildups on my MA1 in a specific area, marked in red here below ... it was in that case minor but it was building up outwards & down inside, this happened in my early days of flying around 0C degrees in foggy weather. when I tried to get it out it fell inside down to the magnets.

 

[alexgwilliam]

Thanks so much everyone for the detail you’ve gone into. It’s only a 3 month old drone, I lost my original Mavic air into the sea and crashed my old spark. I’m sending it back to DJI so I’ll update when they come back with any information. I’m not overly hopeful based on their track record. Flew my original Mavic air many times in temps far colder so just assumed it would be ok.

 

[Kilrah]

With a bit of luck they don't do a good analysis and it slips through the warranty checks, but if they do they'll see they owe you nothing. Based on your description a repair doesn't seem worth it...

 

[szakeetm]

Exactly the same happened to my MA2, dji sent a new drone for free.

Crashed my MA2. Drone or pilot error? [logs+pics]

Could you help me determine what happened? I flew in a mountaineous area. I started ascending. - Until 2:00 business as usual, I climb. - 2:01, 460m above takeoff point: "motor error, check propellers" message. I immediately started descending. - 2:25, 400m above: the descent rates accelerates...

mavicpilots.com

 

[Jackcutrone]

Hello,

I was flying my Mavic Air early this morning and got a motor warning. I'd twice had to cancel aircraft landing which had come up randomly, and tried the return to home function. After appearing to do nothing for about 30 seconds whilst beeping at me the drone then dropped from the sky. It was still connected to the controller and I could see the lines on the road it had landed on.

I have retrieved the drone which is badly damaged.

I've crashed a fair few which have 100% been my fault but this one was very strange.

Please can someone have a look at the log and see if I did anything wrong? I don't want to be stumping up £5/600 for a repair or new drone if it was a failure from the drone.

It's only around 1 month old, no DJI refresh (I know)


Thanks in advance.

Alex

I had a somewhat similar experience about a year ago with a Mavic Pro. I was flying it over a forest preserve at about 360 feet when the controller reported loss of signal. I had about 85% battery at the time and just waited for the auto RTH, and waited and waited. It never came back so I went to look for it. Unfortunately, it had fallen into a grove of trees and I could not find it. I posted the logs and several people took a look but they could see nothing but an abrupt end to the flight. Sars104 did a mapping of the likely location where it went into the trees. The consensus was that the battery ejected. I searched for weeks and weeks but it was stuck in the trees which were fully leafed out. Some weeks later, we had a tremendous windstorm and so I went to search again thinking the wind might have knocked it out of the trees. After a fruitless ground search, I was just about to give up when I spotted the propeller sticking a little bit out of the water of a creek that ran through the woods. Amazing luck. I retrieved it and lo and behold, the battery was still firmly in place. It also had very little physical damage but since it was in the water for a long time, when I sent it into DJI, they were unable to download the flight log from the aircraft, Not sure whether they even tried, truth be told. So, to this day, the cause of the crash is still a mystery and so I can't help with a suggestion of the possible cause to you. I hope you have better luck.