Mavic Pro 2 Motor Failure : crash and lost. What to do ?

[slup]

@slup I'm not convinced that this was a motor issue. Both of the "impacts" are similar in response. If you look at only the roll values for both, they are opposite. If this was ( as it appears to be ) a bird impact, the second one was from the opposite side.

View attachment 90890

Sorry, was perhaps a bit unclear ... I also think the first impact was coming from outside (a bird?), but the last in the end, as a hypothesis, a consequence of the damage sustained from the first.

I'm far from an expert at interpret all data ... but the left rear engine doesn't look like it gives the commanded rpm's as it should (both rear actually?) ... damaged engine, or prop that falls off ... flips the drone ... end of log (story)

*EDIT*
... or should we ask us the question why the log ends mid air after all curves goes haywire, sudden uncommanded drop of altitude ... battery falls out? Or what can cause the sudden log ending?

 

[Fly Dawg]

Or what can cause the sudden log ending?

Battery ejection or just plain downlink loss. No way to know which.

 

[Tristan29photography]

@slup I'm not convinced that this was a motor issue. Both of the "impacts" are similar in response. If you look at only the roll values for both, they are opposite. If this was ( as it appears to be ) a bird impact, the second one was from the opposite side.

Yes @Fly Dawg this is a clever observation, and that totally make sense.

So a bird impact would be the most possible hypothesis.

As @slup says, one last thing weird is why the GPS and data tracking stops before technically hitting the ground.

As you can see on the image below, that fall started at 272m of altitude, and ended 5 seconds later a 191m

Ground elevation is 30 meters + 30 meters high trees

The signal was lost in mid-air before hitting any trees 5 secs after impact...

The battery would have been ejected during the fall 4 or 5 sec after impact ?

This is weird to loose a clipped battery for bird impact, even more 5 secs later. This is more likely to happen when hitting solid obstacles like trees.

But loosing the signal because of hitting trees don't seem to be possible either , because the image show it happened far above ground level.

So I am still confused.

Sudden power issue ?

At 14m30s.5 there is a sudden change in the Cell Deviation values, which gets marked in red in the Log report , just at the end of the first fall. I don't know what it means...
Battery voltage also shows a constant 84% during all flight according to the log, while it had normal values in reality...

So this would mean there is an issue with the report and some values may not be reported correctly?

 

[Fly Dawg]

But loosing the signal because of hitting trees don't seem to be possible either , because the image show it happened far above ground level.

That is true, however, assuming that this was not a battery ejection scenario, it is easy to see why you would lose signal. If you look at the GE elevation profile in relation to your takeoff point, there is a very large rise, that you would have been flying over. So again assuming a signal loss scenario, as the aircraft began to fall, it fell on the back side of the rise which in turn blocked your signal. End of Data.

 

[slup]

Seems unfortunately that this event possibly is one of those that can't be fully (well closer to reality anyway) explained from the log ... something probably hit it once, then perhaps later a second time, or it suffered from the damages from the first & didn't make it, values goes haywire, errors are thrown & logs ends in mid air due to perhaps battery ejection or signal loss (but shouldn't that at least show a glimpse of down link degradation in the log?).

It's to many perhaps & possible in this to say, this is how ...

Really unsatisfying

Regarding battery ejection @Tristan29photography ... don't know how much wear your battery had on the latches & if it was a bit swollen, have seen M2P batteries come lose surprisingly easy by just a light squeeze/pull on the rear sides of the battery ... if batteries swollen & the latches are somewhat worn. The treacherous here is that the battery is just barely loose, still no error & drone works, but its partly detached & can break the power anytime. I don't know how big birds you have there in the jungle but it can happen.

 

[Fly Dawg]

Seems unfortunately that this event possibly is one of those that can't be fully (well closer to reality anyway) explained from the log ... something probably hit it once, then perhaps later a second time, or it suffered from the damages from the first & didn't make it, values goes haywire, errors are thrown & logs ends in mid air due to perhaps battery ejection or signal loss (but shouldn't that at least show a glimpse of down link degradation in the log?).

I suggest a second impact due to the time between the two incidents. The aircraft recovered from the first and was responding to all RC inputs as commanded. Approx a full 60 seconds of flight between the two. It did not survive the second.

 

[slup]

I suggest a second impact due to the time between the two incidents. The aircraft recovered from the first and was responding to all RC inputs as commanded. Approx a full 60 seconds of flight between the two. It did not survive the second.

If it were a second impact, it killed the M2P pretty instantly, that itself can be an explanation to the abrupt ending in the log ...

 

[Fly Dawg]

that itself can be an explanation to the abrupt ending in the log ...

Correct. Just no way to know exactly in this case.

 

[Tristan29photography]

Yes you are totally true, I have tried to draw this straight line in google earth
(btw it was not easy to figure out how ;-)

As you said, this images shows neatly that the loss of signal is due to the rise blocking signal

 

[Tristan29photography]

Thank you so much @Fly Dawg and @slup , I now have an idea of how it happened.

Speculation have ended, I will get back hiking there tomorrow with a local guide.

The first time I was searching too close around of the last gps trace.

Drawing a straight line in Google earth, it shows me an area where it could be at 80 meters ahead from the last signal. we will search mostly in the orange area


First time I was looking in a wrong area, because I didn't have a precise 3D idea of its trajectory, so I was going up in the hill in the red area...


So I guess it will never work again, but if can get the micro sd card, that would be awesome.

I will let you informed if a miracle happens ;-)
Best regards and thanks again

 

[sar104]

The first event was a front right impact in the air. The recorded accelerations (over 6 g backwards) and motor response are conclusive on that - the front right motor stops while the motor demand climbs - and the FC concludes the same in the event stream.



The aircraft recovers, and the second issue, around 60 seconds later, is less clear:



This time the motor deviations precede the acceleration and attitude excursions, and the accelerations don't look like an impact. This is probably propulsion failure. Although the details are not definitive, there is clearly a problem with the rear left prop, and the subsequent attitude excursions and altitude loss are consistent with a lost prop.

As noted by @Fly Dawg in post #24, the disconnect is due to the aircraft dropping out of line of sight of the home point behind the hill rather than power loss:



In terms of crash location, the rather chaotic descent in 12 mph wind out of the SE is very hard to predict. The best that can be done is to extrapolate the descent path, and the OP's estimate in post #30 is probably as good as can be done.

 

[Deleted member 94047]

@sar104 have you looked at this at the end of the flight?

927.500 : 48568 [L-FDI]NS(0) FUSION(0): fault on , disagree_with_mini_filter
927.520 : 48569 [L-FDI]NS(0),fdi req switch to mini filter
927.540 : 48570 [L-IMU]usr_id[0] idx[0] mini_filter_used state:1
927.760 : 48581 [L-RC]craft ctrl failed!!!

Could the second event have been a flight control issue rather than a propulsion problem? And considering the force of the first impact, do you think this could have been a bird?

 

[sar104]

@sar104 have you looked at this at the end of the flight?

Could the second event have been a flight control issue rather than a propulsion problem? And considering the force of the first impact, do you think this could have been a bird?

Yes - those are consistent with propulsion loss. The first impact was likely a bird.

 

[Tristan29photography]

Hi Sherlock @sar104 !

I am amazed by your analysis. I don't how you figured out this, but you are the specialist so I trust you ;-)

It is weird that the impact affects the front right propeller the first time, and then it is the left back propeller that fails the second time...

from 920 to 924 sec the motor PWM is a bit unstable, isn't it ?... I confirm that on my screen the aircraft was completely disturbed... I am wondering why. A Bird impact should not affect the aircraft this way...

 

[sar104]

Hi Sherlock @sar104 !

I am amazed by your analysis. I don't how you figured out this, but you are the specialist so I trust you ;-)

It is weird that the impact affects the front right propeller the first time, and then it is the left back propeller that fails the second time...

from 920 to 924 sec the motor PWM is a bit unstable, isn't it ?... I confirm that on my screen the aircraft was completely disturbed... I am wondering why. A Bird impact should not affect the aircraft this way...

It only took a couple of seconds for the aircraft to recover attitude after the first impact. That seems about right, given that it was completely flipped by the impact.

Back left failure is not the obvious result of a front right impact, but it's impossible to determine from the data how the impact progressed.

The PWM fluctuations prior to failure look fairly normal to me:

 

[BudWalker]

The first event was a front right impact in the air. The recorded accelerations (over 6 g backwards) and motor response are conclusive on that - the front right motor stops while the motor demand climbs - and the FC concludes the same in the event stream.

View attachment 90935

The aircraft recovers, and the second issue, around 60 seconds later, is less clear:

View attachment 90938

This time the motor deviations precede the acceleration and attitude excursions, and the accelerations don't look like an impact. This is probably propulsion failure. Although the details are not definitive, there is clearly a problem with the rear left prop, and the subsequent attitude excursions and altitude loss are consistent with a lost prop.

As noted by @Fly Dawg in post #24, the disconnect is due to the aircraft dropping out of line of sight of the home point behind the hill rather than power loss:

View attachment 90939

In terms of crash location, the rather chaotic descent in 12 mph wind out of the SE is very hard to predict. The best that can be done is to extrapolate the descent path, and the OP's estimate in post #30 is probably as good as can be done.

Doesn't the acceleration data show that the M2 was struck from the left, back and above? I believe that the negative accelX means the AC accelerated forward along the X axis.

Also, the M2 was stationary when it was hit - almost certainly a bird strike.

But, the impact point was the right front quadrant as can be seen from the pitch and roll data.

When the M2 was hit from above in the right front quadrant it rolled CW and pitched down.

 

[slup]

Doesn't the acceleration data show that the M2 was struck from the left, back and above? I believe that the negative accelX means the AC accelerated forward along the X axis.
View attachment 90949

Think I agree with @BudWalker regarding accelX direction ... forward movement creates minus values.

Beside that modest comment from me ... it's really intriguing to see how you experienced guys take on the mystery & what values you choose to look at. I'm following your angle (@sar104 & @BudWalker ) on my side & it's not far from that it's possible to visualize the event with the inner vision.

Reading & learning ... thank's guys

 

[Deleted member 94047]

Sorry for the OT remark but I really wanted to say this @slup; you are not far off their level yourself. It has been a marvel watching you get consistently better at this the last few months. Another addition to the awesome flight analysis team! Well done mate! Written with eyes absolutely green with envy.

 

[Deleted member 113944]

Hi there !

Sad day for me today. My Drone crashed after a short 6 month life, and I have no money to buy another one…

There were no wind, no obstacles, It had still enough batteries.

The Drone was flying normally, and after completing a 360 panorama, it suddenly falled of around 15 meters in the air, like if it was doing some kind of pirouettes and it miraculously stabilized.

Scared, I flew it directly to the home point, to get him back as soon as possible. And then all of a sudden, he fell again from the air at 270 meters of altitude and lost his signal

Surprisingly, the signal indicates that he lost connection at 191 meters high, while the ground rises to a 30 meters at that point. This is a jungle area on the north coast of Colombia.

I walked up the river for one hour, then I searched for it for two hours on the very uneven ground, in the very dense forest. Mosquitoes, spiders and ticks were very happy to find me there.

Despite the desperate search with Find My Drone function, nothing to do, it did not seem to be where the GPS had stopped.

I’d really like to know why he crashed… I guess the one motor stopped running

Is there any point in writing to DJI support? Is there anything to do?

In advance many thanks for your help

Here is the Log File if somebody would like to see how it crashed :

DJI Flight Log Viewer - PhantomHelp.com

I feel sadly for your loss. I hope things will look up soon with a new drone for you. ?

 

[Deleted member 113944]

Seems unfortunately that this event possibly is one of those that can't be fully (well closer to reality anyway) explained from the log ... something probably hit it once, then perhaps later a second time, or it suffered from the damages from the first & didn't make it, values goes haywire, errors are thrown & logs ends in mid air due to perhaps battery ejection or signal loss (but shouldn't that at least show a glimpse of down link degradation in the log?).

It's to many perhaps & possible in this to say, this is how ...

Really unsatisfying

Regarding battery ejection @Tristan29photography ... don't know how much wear your battery had on the latches & if it was a bit swollen, have seen M2P batteries come lose surprisingly easy by just a light squeeze/pull on the rear sides of the battery ... if batteries swollen & the latches are somewhat worn. The treacherous here is that the battery is just barely loose, still no error & drone works, but its partly detached & can break the power anytime. I don't know how big birds you have there in the jungle but it can happen.

Your last sentence. I'm reading all this and thinking the same. How big the birds.
I'm intrigued by how knowledgeable members here in the forum are. ??