Mavic just kinda...well fell out of the sky. Here are my logs.

[ToasteyOstey]

DJI Flight Log Viewer - PhantomHelp.com

Hi all, so a few years back I was getting some aerial footage with my Mavic Pro out by the Griffith Observatory. As I was buzzing around I noticed my camera feed started tumbling out of the blue, but I didn't think much of it because I had a similar experience with a camera malfunction. I hit the return to home button...and nothing happened. It wasn't until then that I realized my Mavic was in a free fall for seemingly no reason. I was line of sight and saw nothing near the drone when this happened (not to say I couldn't have missed some sort of small bird I guess). Upon retrieval...which took a good two hours of searching in brush, I found it bad a couple busted props (only the front two if this means anything) and cables, but nothing too serious. Thankfully a bush somewhat softened the fall. I have been meaning to fix it for a while, but have kept putting it off mostly out of fear that the same thing would happen again. Anyway, all the background out of the way, I am mostly curious if there are any obvious signs of what went wrong here? I noticed in the logs there seemed to be some sort of undervoltage in the battery a few seconds prior to the incident starting at 3m 46.8s, but that is the only thing that stick out to me. I want to know if I need to potentially replace more than just the busted cables (such as a motor or battery). I am getting back into insurance photography and really don't want to have a drone stranded on a rooftop. For reference the first sign of tumbling starts at 3m 58.1s in the logs. Thanks!

 

[Deleted member 103366]

I see no logs.

 

[ToasteyOstey]

I see no logs.

My bad, leave it to me to forget the most important part lol. Updated the post to include them, thanks for letting me know.

 

[eEridani]

Airdata gives these notifications. At 2000' doubt you'd see a bird attack. Or a prop may have failed.

 

[ToasteyOstey]

Airdata gives these notifications. At 2000' doubt you'd see a bird attack. Or a prop may have failed.

View attachment 143140

I am curious with the line L what could have happened when it mentioned it detected a side shock. Would this happen from a prop failing?

 

[eEridani]

I am curious with the line L what could have happened when it mentioned it detected a side shock. Would this happen from a prop failing?

Probably. A prop flying off would assert torque on the drone same as if it was hit by something (that action-reaction thing). So if a prop hinge failed, one half flew off, it'd be like the drone is being hammered until the other half flew off.

So the initial shock of the prop failure, loss of lift and the drone is on its way down, single prop hammering the thing until it fails and you get the overrev lost prop warning.

 

[slup]

All that Airdata says regarding "Detected side shock" & "Propeller fell off" is their interpretation coming from the AC sudden & high velocity angular rotation ... not necessarily a true picture to what actual happened.

What's possible to say out from the log is ...

When the AC was flying straight forward by means of only elevator input it suddenly at 238sec into the flight ...

-Pitched down --> indicate thrust loss in either of the 2 front corners

-Rolls to the left --> indicate thrust loss in either of the 2 left corners ... but together with the pitch it narrows it to font left

-Yaw CW --> indicate thrust loss in a CCW torquing corner, which is either front left or rear right ... so with pitch & roll it's confirmed that ...

This incident was due to a thrust loss in the left front corner, either due to a prop or motor failure.

In order to say anything about witch ... the mobile device .DAT log is needed. The correct one for this flight ends with FLY080.DAT. Share that if you want to know for sure.

Here the interesting telemetry ... Pitch, Roll & Yaw, with the height as reference. Have placed the chart marker at 238sec where the incident starts.

(Click on the chart to make it larger)


And here regarding prop torque ...

 

[ToasteyOstey]

Probably. A prop flying off would assert torque on the drone same as if it was hit by something (that action-reaction thing). So if a prop hinge failed, one half flew off, it'd be like the drone is being hammered until the other half flew off.

So the initial shock of the prop failure, loss of lift and the drone is on its way down, single prop hammering the thing until it fails and you get the overrev lost prop warning.

Ah, ok that makes sense

All that Airdata says regarding "Detected side shock" & "Propeller fell off" is their interpretation coming from the AC sudden & high velocity angular rotation ... not necessarily a true picture to what actual happened.

What's possible to say out from the log is ...

When the AC was flying straight forward by means of only elevator input it suddenly at 238sec into the flight ...

-Pitched down --> indicate thrust loss in either of the 2 front corners

-Rolls to the left --> indicate thrust loss in either of the 2 left corners ... but together with the pitch it narrows it to font left

-Yaw CW --> indicate thrust loss in a CCW torquing corner, which is either front left or rear right ... so with pitch & roll it's confirmed that ...

This incident was due to a thrust loss in the left front corner, either due to a prop or motor failure.

In order to say anything about witch ... the mobile device .DAT log is needed. The correct one for this flight ends with FLY080.DAT. Share that if you want to know for sure.

Here the interesting telemetry ... Pitch, Roll & Yaw, with the height as reference. Have placed the chart marker at 238sec where the incident starts.

(Click on the chart to make it larger)
View attachment 143160

And here regarding prop torque ...

View attachment 143161

Seems like a pretty bang on analysis, and definitely makes sense. I caught the roll and yaw, but didn't see the forward pitch. That would make sense as to why only the front two props are busted. Anyway, I have attatched the .dat file to this post.

 

[slup]

...I have attatched the .dat file to this post.

To summarize ... yep, it's the left front motor that's the initial culprit & start off the incident, this with no indications from the accelerometer's that it's due to some collision (...bird or something). But this failed motor recovers 2sec later & the increasing vertical downward velocity decreases ... but then the DAT log message stream indicates a flight control failure ... and the vertical downward velocity increases again.

So starting off with the indicated "blocked" motor ...

The red graph = height, as a reference
The blue graph = left front motor RPM
The purple graph = motor command percentage from the flight controller
The green graph = left front motor current
The black graph = vertical velocity (positive value = descent, negative = ascent)

(Click on the chart to make it larger)


Within the yellow frame ... starting at 326,4sec after craft power on, the FC commands the left front motor to 100% (purple) to counter the craft movement, but we don't see that the motor RPM increases as expected at all (blue) ... at the same time the motor current spikes up (green). All this indicate a motor that have been blocked, fully commanded but no RPM's & rising current.

At 328,3sec all goes back to normal, FC still commands to 100%, the motor revs up & the current falls down. It seems like the motor have become unblocked ... This is also indicated by what's seen within the blue square, the vertical velocity value goes less positive & it seems like the craft will recover.

But here at 328,5sec the flight controller gives up & the downward speed picks up again ...


So ... I think this was a temporary blocked motor (due to some reason) & not a prop failure ... all the rapid movements coming from that made the FC get a brain fart which made it give up as it couldn't stop the rapid movements.

 

[ToasteyOstey]

To summarize ... yep, it's the left front motor that's the initial culprit & start off the incident, this with no indications from the accelerometer's that it's due to some collision (...bird or something). But this failed motor recovers 2sec later & the increasing vertical downward velocity decreases ... but then the DAT log message stream indicates a flight control failure ... and the vertical downward velocity increases again.

So starting off with the indicated "blocked" motor ...

The red graph = height, as a reference
The blue graph = left front motor RPM
The purple graph = motor command percentage from the flight controller
The green graph = left front motor current
The black graph = vertical velocity (positive value = descent, negative = ascent)

(Click on the chart to make it larger)
View attachment 143185

Within the yellow frame ... starting at 326,4sec after craft power on, the FC commands the left front motor to 100% (purple) to counter the craft movement, but we don't see that the motor RPM increases as expected at all (blue) ... at the same time the motor current spikes up (green). All this indicate a motor that have been blocked, fully commanded but no RPM's & rising current.

At 328,3sec all goes back to normal, FC still commands to 100%, the motor revs up & the current falls down. It seems like the motor have become unblocked ... This is also indicated by what's seen within the blue square, the vertical velocity value goes less positive & it seems like the craft will recover.

But here at 328,5sec the flight controller gives up & the downward speed picks up again ...

View attachment 143187
So ... I think this was a temporary blocked motor (due to some reason) & not a prop failure ... all the rapid movements coming from that made the FC get a brain fart which made it give up as it couldn't stop the rapid movements.

Wow, thanks a bunch for the analysis! Glad I finally have more of an idea of what actually happened. So, to prevent further issues like this happening, would it be wise to replace that motor? Or is this likely just one of those unlucky flukes? If you think it's best to replace the motor, do you have a good reference for replacement parts?

 

[BigAl07]

Wow, thanks a bunch for the analysis! Glad I finally have more of an idea of what actually happened. So, to prevent further issues like this happening, would it be wise to replace that motor? Or is this likely just one of those unlucky flukes? If you think it's best to replace the motor, do you have a good reference for replacement parts?

I'd have it looked at for sure but it could have been a physical obstruction not a motor/controller issue.

If you want to get a FREE Evaluation try contacting @DDS I have an airframe there with them right now being repaired.

 

[BigAl07]

DJI Flight Log Viewer - PhantomHelp.com

Hi all, so a few years back I was getting some aerial footage with my Mavic Pro out by the Griffith Observatory. As I was buzzing around I noticed my camera feed started tumbling out of the blue, but I didn't think much of it because I had a similar experience with a camera malfunction. I hit the return to home button...and nothing happened. It wasn't until then that I realized my Mavic was in a free fall for seemingly no reason. I was line of sight and saw nothing near the drone when this happened (not to say I couldn't have missed some sort of small bird I guess). Upon retrieval...which took a good two hours of searching in brush, I found it bad a couple busted props (only the front two if this means anything) and cables, but nothing too serious. Thankfully a bush somewhat softened the fall. I have been meaning to fix it for a while, but have kept putting it off mostly out of fear that the same thing would happen again. Anyway, all the background out of the way, I am mostly curious if there are any obvious signs of what went wrong here? I noticed in the logs there seemed to be some sort of undervoltage in the battery a few seconds prior to the incident starting at 3m 46.8s, but that is the only thing that stick out to me. I want to know if I need to potentially replace more than just the busted cables (such as a motor or battery). I am getting back into insurance photography and really don't want to have a drone stranded on a rooftop. For reference the first sign of tumbling starts at 3m 58.1s in the logs. Thanks!

I meant to ask earlier, didn't I see this exact "scenario" posted on one of the FB forums?

If so, would you be so kind as to mention www.MavicPilots.com on that post and how the members here were able to help you diagnose the issue and hopefully come to a likely reason behind it? It's always good to get some POSITIVE publicity for the forum on the FB groups

 

[ToasteyOstey]

I meant to ask earlier, didn't I see this exact "scenario" posted on one of the FB forums?

If so, would you be so kind as to mention www.MavicPilots.com on that post and how the members here were able to help you diagnose the issue and hopefully come to a likely reason behind it? It's always good to get some POSITIVE publicity for the forum on the FB groups

Actually, it was reddit, and someone from there referenced me here. I did link this thread there, so I am one step ahead of you!

 

[eEridani]

I don't see this as an IMU failure. The IMU is responding to some input, drives the motor PWM higher, the motor current and speed responds, then motor current goes low (no load, total prop failure) the IMU reduces PWM, the motor more or less responds but unloaded.

I'd like to see this plot with pitch roll yaw plot aligned along side it. Looking at the pitch roll yaw plot above, the aircraft had a very abrupt yaw, followed by pitch and roll. Something turned the aircraft almost instantly. Nothing in the drive system would do that except impact or losing a prop. It's simple physics.

 

[ToasteyOstey]

I don't see this as an IMU failure. The IMU is responding to some input, drives the motor PWM higher, the motor current and speed responds, then motor current goes low (no load, total prop failure) the IMU reduces PWM, the motor more or less responds but unloaded.

I'd like to see this plot with pitch roll yaw plot aligned along side it. Looking at the pitch roll yaw plot above, the aircraft had a very abrupt yaw, followed by pitch and roll. Something turned the aircraft almost instantly. Nothing in the drive system would do that except impact or losing a prop. It's simple physics.

If it was a prop fail, would the craft have been able to recover altitude like it did before the control system shut itself off?

 

[MAvic_South_Oz]

I’d have it looked at for sure but it could have been a physical obstruction not a motor/controller issue.

What I thought when @slup mentioned temporary obstruction in his analysis.
Would / could a bird hitting the motor / prop LH front have read like this, or is there too much delay in the timing of the obstruction / recovery and ongoing crash ?
Could the OP see any birds around, near the drone etc, if they were within decent VLOS ?

 

[slup]

... So, to prevent further issues like this happening, would it be wise to replace that motor?

Can't say ... a thorough examination of the motor is needed at least, this could have been debris from this or earlier take-offs or landings or a broken off prop attachment tab, check the motor for any internal scratches, control so it rotates as it should, that the bell isn't touching the magnets, that the motor shaft isn't bent ... and such.

I don't see this as an IMU failure. The IMU is responding to some input, drives the motor PWM higher, the motor current and speed responds, then motor current goes low (no load, total prop failure) the IMU reduces PWM, the motor more or less responds but unloaded.

As I said ... a IMU failure wasn't the root cause & wasn't involved at all, it was a blocked motor with a later failing FC (probably temporary) as it couldn't control the craft movement anymore. And you describe the scenario wrong ... what makes it a blocked motor is that the FC motor command goes high to counter the uncommanded movement, the motor RPM doesn't follow ... & by that the motor current goes high as something prevent it to rotate.

I'd like to see this plot with pitch roll yaw plot aligned along side it. Looking at the pitch roll yaw plot above, the aircraft had a very abrupt yaw, followed by pitch and roll. Something turned the aircraft almost instantly. Nothing in the drive system would do that except impact or losing a prop. It's simple physics.

This yaw change ... either CCW or CW is very typical when one corner stop to produce rotational torque ... and the rotation wasn't instant ... it was well within what's physical, the angular rotation was slightly more than a half turn per second.

You have it all here ... with motor related as dashed graphs.



And if adding in the east & north bound accelerometer sensor readings (black & blue graphs) below ... we see that nothing touches the craft before the left front motor gets blocked.

Would / could a bird hitting the motor / prop LH front have read like this, or is there too much delay in the timing of the obstruction / recovery and ongoing crash ?

The blockage lasted nearly 2,5sec ... if that was a bird stuck in the prop for that long time the MP would have had gore & feathers all over it

 

[ToasteyOstey]

What I thought when @slup mentioned temporary obstruction in his analysis.
Would / could a bird hitting the motor / prop LH front have read like this, or is there too much delay in the timing of the obstruction / recovery and ongoing crash ?
Could the OP see any birds around, near the drone etc, if they were within decent VLOS ?

Weather was clear and a million and I didn't see any birds directly, but to be fair it is something I could has missed since it was just far enough away. Technically I was looking at the camera to line up a shot when it happened, but prior to and after the incident I did not see any birds in the area.

 

[eEridani]

The prop speed doesn't indicate a stuck motor to me - it is spinning the entire time. The FC increases drive to max - the motor speed increases, albeit slowly, as if there is added drag. A flagging prop would do that.

But the timing seems better than I'd expect for the controller to detect a broken prop and respond. Something isn't making sense in the data provided. Probably the log resolution isn't showing the timings clearly enough to make a good estimate of what really came first. Something caused the FC to increase drive to that motor - but we can't see the detail.

What we can see is the FC commanded more lift, the motor responded, then finally red lined. All indicative of a failed prop.

I see no indication of a "stuck" motor.

 

[slup]

The prop speed doesn't indicate a stuck motor to me - it is spinning

That is probably do to that English isn't my native language ... stuck or hard to turn motor, same same together with the additional chart... I thought

Seen cases like this more than I can count ... we have seen motors that have got "stuck" with 0rpm & "hard to turn" with some rpm ... all cases show the same, rpm doesn't obey the command & the current spikes.

When prop fails, then the command goes to 100% & that does the rpm also ... & this together with a lower current than before, have never seen a prop that is "flagging" mimic a hard to turn motor preventing it to obey the FC command.

We have a log that records in 10Hz & we have to do our best to deal with that ... to disregard the obvious & instead suspect that something happened in between the measure points is just ... speculative.

...But the timing seems better than I'd expect for the controller to detect a broken prop and respond.

In an ordinary quad FC, the PID loop runs in 8KHz ... wouldn't expect that the DJI ones is any worse ...