Welcome Mavic Pilot!
Jump in and join our free DJI Mavic community today!
Sign up

Sudden Crash New Mavic 2 Pro

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
looks in slow mo like the failure starts with the right rear side of the drone falling.
 

sar104

Well-Known Member
Premium Pilot
Joined
May 27, 2017
Messages
6,178
Reaction score
5,946
Loc
Los Alamos, NM
Okay, but I'm seeing a recovery like when you cut the throttle abruptly and jam it on again (or have an obstruction). Do we have the data for the other three motors? Are they normal all the way to the ground?
I posted the records for all four motors in post #72.
 

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
I think the freeze frame on the post shows the prop not moving.
 

sar104

Well-Known Member
Premium Pilot
Joined
May 27, 2017
Messages
6,178
Reaction score
5,946
Loc
Los Alamos, NM
looks in slow mo like the failure starts with the right rear side of the drone falling.
Okay - a couple of hours of skiing helped figure out how to distinguish these hypotheses.

The rear right motor spins CW, and carries CW angular momentum. If the motor is stopped by an external torque (e.g. supplied by hitting vegetation) then the reaction to that torque exerts a CCW torque on the aircraft and it will turn CCW (the angular equivalent of Newton's third law). On the other hand, if the motor stops by itself, with no external torque applied, then its CW angular momentum is added to the angular momentum of the aircraft (which is zero in a steady hover) and it will turn CW.

Looking at both the rate gyro data and the IMU yaw shows the following:

gyros.png

Positive is CW for the rate gyros, and so you can see that as the motor stops the aircraft immediately acquires CW angular momentum. That means that the motor somehow stopped itself - it was not stopped by an external influence. Either it seized or, possibly, the ESC inadvertently caused a braking effect. I have no idea how that is consistent with the motor current data I posted earlier, but it seems to me to be completely unambiguous.

It's also consistent with the other video that you posted above, where the aircraft executes exactly the same maneuver without hitting anything. I'll try to figure out if that is in one of the other DAT files that you posted.
 

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
Okay - a couple of hours of skiing helped figure out how to distinguish these hypotheses.

The rear right motor spins CW, and carries CW angular momentum. If the motor is stopped by an external torque (e.g. supplied by hitting vegetation) then the reaction to that torque exerts a CCW torque on the aircraft and it will turn CCW (the angular equivalent of Newton's third law). On the other hand, if the motor stops by itself, with no external torque applied, then its CW angular momentum is added to the angular momentum of the aircraft (which is zero in a steady hover) and it will turn CW.

Looking at both the rate gyro data and the IMU yaw shows the following:

View attachment 61762

Positive is CW for the rate gyros, and so you can see that as the motor stops the aircraft immediately acquires CW angular momentum. That means that the motor somehow stopped itself - it was not stopped by an external influence. Either it seized or, possibly, the ESC inadvertently caused a braking effect. I have no idea how that is consistent with the motor current data I posted earlier, but it seems to me to be completely unambiguous.

It's also consistent with the other video that you posted above, where the aircraft executes exactly the same maneuver without hitting anything. I'll try to figure out if that is in one of the other DAT files that you posted.
Thank you so much. Hope you enjoyed the slopes. I used to love to do that but I do not want another fall. Nearly got my neck last time after getting submarined by a runaway snowboard.
 

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
Just went back through the replays and think it is the 07 data

Second flight
 

BudWalker

Well-Known Member
Joined
Oct 24, 2016
Messages
1,215
Reaction score
710
Age
70
Loc
Arroyo Grande, California
Okay - a couple of hours of skiing helped figure out how to distinguish these hypotheses.

The rear right motor spins CW, and carries CW angular momentum. If the motor is stopped by an external torque (e.g. supplied by hitting vegetation) then the reaction to that torque exerts a CCW torque on the aircraft and it will turn CCW (the angular equivalent of Newton's third law). On the other hand, if the motor stops by itself, with no external torque applied, then its CW angular momentum is added to the angular momentum of the aircraft (which is zero in a steady hover) and it will turn CW.

Looking at both the rate gyro data and the IMU yaw shows the following:

View attachment 61762

Positive is CW for the rate gyros, and so you can see that as the motor stops the aircraft immediately acquires CW angular momentum. That means that the motor somehow stopped itself - it was not stopped by an external influence. Either it seized or, possibly, the ESC inadvertently caused a braking effect. I have no idea how that is consistent with the motor current data I posted earlier, but it seems to me to be completely unambiguous.

It's also consistent with the other video that you posted above, where the aircraft executes exactly the same maneuver without hitting anything. I'll try to figure out if that is in one of the other DAT files that you posted.
The current data seems normal to me... When the rightBack starts turning slower the FC commands it to turn faster by increasing the duty cycle from 40% to 90%. Doesn't an increase in duty cycle result in increased measured current?
 

sar104

Well-Known Member
Premium Pilot
Joined
May 27, 2017
Messages
6,178
Reaction score
5,946
Loc
Los Alamos, NM
The current data seems normal to me... When the rightBack starts turning slower the FC commands it to turn faster by increasing the duty cycle from 40% to 90%. Doesn't an increase in duty cycle result in increased measured current?
If it were fighting external resistance, then yes. But if the motor is being slowed down internally then I'd expect the current to drop. Where is the energy going?
 

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
Just reviewed the cctv and it was the second flight......... the 07 data, this matches with the flight path of the drone and my movements relative to itt.
 

DanMan32

Well-Known Member
Premium Pilot
Joined
Sep 1, 2018
Messages
2,109
Reaction score
656
Age
53
How is the surveillance footage moving as if hand-held?

If the motor seized, as if a bad bearing, then the ESC/FC would increase duty cycle and current, and I would think AC would react as if motor lost power.

There should be an alert of motor obstruction though.
 

BudWalker

Well-Known Member
Joined
Oct 24, 2016
Messages
1,215
Reaction score
710
Age
70
Loc
Arroyo Grande, California
If it were fighting external resistance, then yes. But if the motor is being slowed down internally then I'd expect the current to drop. Where is the energy going?
If it seized as you said was a possibility then the energy would be dissipated as heat into the windings and bearing surfaces. There is something kinda odd I didn't mention because I didn't know what it means. After the PWM went to 90% for about a sec there is a transition to a state where the PWM is 60% to 90% but both the current and speed are 0. Like what would be expected if the ESC or the windings had failed.
1549237916354.png
 

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
How is the surveillance footage moving as if hand-held?

If the motor seized, as if a bad bearing, then the ESC/FC would increase duty cycle and current, and I would think AC would react as if motor lost power.

There should be an alert of motor obstruction though.
I used an iPhone to get the video off easily. My system is not that user friendly to offload video. Can do but i have spent enough time on this today.
 
Last edited:

sar104

Well-Known Member
Premium Pilot
Joined
May 27, 2017
Messages
6,178
Reaction score
5,946
Loc
Los Alamos, NM
Just reviewed the cctv and it was the second flight......... the 07 data, this matches with the flight path of the drone and my movements relative to itt.
Unfortunately I cannot find anything resembling that event in FLY007.DAT.
 

Yhodah

Well-Known Member
Joined
Jan 26, 2019
Messages
58
Reaction score
2
I find it reassuring that this seems to be a rare event. Given there are not a lot more recent reports.
 

Mike4884

Well-Known Member
Joined
Nov 25, 2017
Messages
739
Reaction score
336
Age
35
Loc
Bucks county pa
If they choose to do that I will definitely return the drone and give up on drone photography. It is clearly too high risk for my use................... I fix broken humans for a living. I do not want to create any. Unlike your return of items you broke. I did nothing wrong. I would not ask for Walmart to take back an item I broke. I purchased it with my chase platinum. They will make it right if all else fails. They fixed a new lens I broke. Great card. And btw.... I have care refresh. Activated before I turned the drone on. So it will get fixed. Life is good. I thought it might be useful to bring an unusual failure event to the light of day. And if I continue to fly it I will need to be sure I have adequate liability coverage.
I wonder if one of those plastic nubs broke off the underside of the motor/prop combo that had problems. Vegetation doesnt seem plausible to me at that height. A broken nub from the underside of prop does it can get stuck in the windings of the motor. Each prop has three of those plastic "nubs" and thats what locks it onto the motor....question is have u hit any twigs or lots of grass with props that would weaken it? Have u examined your props?
 
  • Like
Reactions: altaic