Get the most from your Mavic
Join thousands of Mavic owners helping each other
Join Us Now

Mavic zoom crash

Teejag

Member
Joined
Feb 16, 2020
Messages
15
Reaction score
1
Age
25
Location
Zimbabwe
Hi guys had an incident. Had no control over mavic zoom was barrel rolling and spinning down to the ground then recovered and did it again. Sadly hit the ground did some damage recovered and I flew it back.
just been looking at the flight logs and seen that it is showing red cell deviation at some point. Has anyone had this
 

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
4,921
Reaction score
5,686
Age
64
Hi guys had an incident. Had no control over mavic zoom was barrel rolling and spinning down to the ground then recovered and did it again. Sadly hit the ground did some damage recovered and I flew it back.
just been looking at the flight logs and seen that it is showing red cell deviation at some point. Has anyone had this
Small cell deviations aren't a problem and are common.
Large cell deviations might cause the drone to fall from the sky in extreme cases but not barrel rolling in flight.
Post your flight data and some experienced eyes might be able to see what the data is saying.
 

Teejag

Member
Joined
Feb 16, 2020
Messages
15
Reaction score
1
Age
25
Location
Zimbabwe
Small cell deviations aren't a problem and are common.
Large cell deviations might cause the drone to fall from the sky in extreme cases but not barrel rolling in flight.
Post your flight data and some experienced eyes might be able to see what the data is saying.
have posted it up ahead in the other link
 

slup

Well-Known Member
Joined
Oct 18, 2019
Messages
787
Reaction score
849
Age
52
Location
Sweden
That DAT log is the wrong one ... this flight is associated with the DAT that ends with .FLY000

But already a message readout from the TXT gives a hint about what failed ...

Gyro.jpg

And here what's shown in the TXT from the time were the gyros probably failed ...

Alot of graph's but something starts to happen at 243sec ... the downward speed picks up without throttle inputs. Then at 248sec all goes haywire with rapid speed increases in all 3 axis (x,y & z) ... you don't have time there to send so many stick inputs but in relation you put in some short throttle, elevator & aileron inputs

See if you can find the right DAT & post it up.

CsvView.jpg
 
  • Like
Reactions: Thomas B and Teejag

Meta4

Well-Known Member
Joined
Oct 30, 2016
Messages
4,921
Reaction score
5,686
Age
64
here is the flight record. I can't open it up to see the motor speeds and stuff like that just not sure how to do that
The minor cell voltage differences had nothing to do with the incident.
You can see a summary of your flight data here:

You were hovering normally and gave the Mavic a touch of throttle to climb an extra two feet when 175 feet up, at 4:02.8 something happens.
The drone pitched 73° .
At the same time it rolled left 122°, started spinning anti-clockwise and losing altitude.
You left the joysticks centred and as the drone fell further, the falling, pitching, rolling and spinning all became less severe.
At 4:12.6 and 22 feet higher than the launch point, the drone started to gain height. The joysticks were still centred.
The drone reached 51 feet at 4:16.2 and appeared to have almost stabilised when it all started again.
The drone fell tumbling and climbed again, the climb being fairly stable.

The drone stabilised at 4:23.7 at a height of 21 feet.
You flew it until 7:41 when you landed safely.

This incident looks pretty strange.
It's odd that the drone recovered and climbed, particularly without any control inputs from you.
I've not seen anything like it and can't think of a likely cause.
 

slup

Well-Known Member
Joined
Oct 18, 2019
Messages
787
Reaction score
849
Age
52
Location
Sweden
The event as such with a massive flight instability may very well be explained by a failing IMU. When this fails the data needed for a stable flight no longer can be gathered.

But agreed, it's strange that it is recovering ... hopefully we can see the causes once the right DAT is provided.

With the data already provided though... I can't see any problem approaching DJI with this.

Below in short for those eager to learn:

"For a drone to fly perfectly, the IMU, gyro stabilization and flight controller technology is essential. Gyro stabilization technology is one of the most important components, allowing the drone to fly super smooth even in strong winds and gusts. The gyroscope needs to work almost instantly to the forces moving against the drone (gravity, wind etc) to keep it stabilized. In the vast majority of drones, the gyroscope is encompassed or integrated within the IMU.

An IMU unit works by detecting the current rate of acceleration using one or more accelerometers. The IMU detects changes in rotational attributes like pitch, roll and yaw using one or more gyroscopes. Some IMU on drones include a magnetometer, mostly to assist calibration against orientation drift.


On board processors then continually calculate the drones current position. First, it integrates the sensed acceleration, together with an estimate of gravity, to calculate the current velocity. Then it integrates the velocity to calculate the current position.

To fly in any direction, the flight controller gathers the IMU data on present positioning, then sends new data to the motor electronic speed controllers (ESC). These electronic speed controllers signal to the motors the level of thrust and speed required for the drone to fly or hover."
 

Dave Maine

Well-Known Member
Joined
Jan 27, 2018
Messages
437
Reaction score
284
Age
77
An intermittent electrical connection in the flight control logic could create this scenario. Generally these occur at the point where cables mate with connectors, or where the individual wires are crimped into the connector bodies. Unfortunately they can be anywhere, and as intermittents are very difficult to find.
 
  • Like
Reactions: Gindra and Teejag

Teejag

Member
Joined
Feb 16, 2020
Messages
15
Reaction score
1
Age
25
Location
Zimbabwe
here hopefully this is the correct one thanks very much for all the information very interesting. was soo strange I just thought to try keep sticks centred wasn't sure what to do. the first one it recovered so the the next time it was just too low.
 

Attachments

sar104

Dic mihi solum facta, domina.
Premium Pilot
Joined
May 27, 2017
Messages
10,604
Reaction score
13,085
Location
Los Alamos, NM
here hopefully this is the correct one thanks very much for all the information very interesting. was soo strange I just thought to try keep sticks centred wasn't sure what to do. the first one it recovered so the the next time it was just too low.
The two events each look much like a lost prop in terms of attitude and height excursions, but of course the stable flight in between and afterwards rules that out. In fact it's rear left motor failure:

Attitude.png

The rapid pitch, roll and yaw excursions are typical of loss of propulsion at one motor.

unwrapped_yaw.png

The continuous CCW yaw tells you that it has lost propulsion on a CCW motor, i.e. front right or rear left.

motors.png

The motor speed record shows that it is rear left.

pwm.png

And the motor demand PWM data show that the slow rear left motor was not commanded - the FC was requesting nearly full speed from that motor.

That's hardware failure.
 
Last edited:

slup

Well-Known Member
Joined
Oct 18, 2019
Messages
787
Reaction score
849
Age
52
Location
Sweden
The two events each look much like a lost prop in terms of attitude and height excursions, but of course the stable flight in between and afterwards rules that out. In fact it's rear left motor failure:

View attachment 94244

The rapid pitch, roll and yaw excursions are typical of loss of propulsion at one motor.

View attachment 94248

The continuous CCW yaw tells you that it has lost propulsion on a CCW motor, i.e. front right or rear left.

View attachment 94245

The motor speed record shows that it is rear left.

View attachment 94246

And the motor demand PWM data show that the slow rear left motor was not commanded - the FC was requesting nearly full speed from that motor.

That's hardware failure.
@sar104

I can also clearly see that the left motor isn't performing ... but it looks to be more to it.

When the event starts & until the AC lands a lot of errors is indicated, both position, velocity & GPS consistency errors. The position error remains until flight ends.

Hardware failure yes definitely ... but is it the motor or something else that effected it?

Or do I misinterpret thing's ..?
 

sar104

Dic mihi solum facta, domina.
Premium Pilot
Joined
May 27, 2017
Messages
10,604
Reaction score
13,085
Location
Los Alamos, NM
@sar104

I can also clearly see that the left motor isn't performing ... but it looks to be more to it.

When the event starts & until the AC lands a lot of errors is indicated, both position, velocity & GPS consistency errors. The position error remains until flight ends.

Hardware failure yes definitely ... but is it the motor or something else that effected it?

Or do I misinterpret thing's ..?
Those kinds of errors are commonly triggered by such an extreme loss of craft control, so unlikely to be causal or indicative of any other problem. And the fact that the motor slowed almost to a stop despite the PWM holding at nearly 100% clearly indicates mechanical or electrical failure of the ESC/motor. The recorded motor current stayed high too.

amps.png
 
  • Like
Reactions: NedKelly and slup

slup

Well-Known Member
Joined
Oct 18, 2019
Messages
787
Reaction score
849
Age
52
Location
Sweden
...Those kinds of errors are commonly triggered by such an extreme loss of craft control...
...The recorded motor current stayed high too...
OK..
Yeah ... that was seen in the magMod graph also.

Thank's sar104 Thumbswayup:)
 
  • Like
Reactions: sar104

Teejag

Member
Joined
Feb 16, 2020
Messages
15
Reaction score
1
Age
25
Location
Zimbabwe
Those kinds of errors are commonly triggered by such an extreme loss of craft control, so unlikely to be causal or indicative of any other problem. And the fact that the motor slowed almost to a stop despite the PWM holding at nearly 100% clearly indicates mechanical or electrical failure of the ESC/motor. The recorded motor current stayed high too.

View attachment 94255
The only thing I found strange on the flight log I saw that the battery starts showing warnings just after the bird starts to spiral
 

sar104

Dic mihi solum facta, domina.
Premium Pilot
Joined
May 27, 2017
Messages
10,604
Reaction score
13,085
Location
Los Alamos, NM
The only thing I found strange on the flight log I saw that the battery starts showing warnings just after the bird starts to spiral
That was most likely due to the current spikes as the FC attempted to spool up the faulty motor:

current.png
 

Teejag

Member
Joined
Feb 16, 2020
Messages
15
Reaction score
1
Age
25
Location
Zimbabwe
Those kinds of errors are commonly triggered by such an extreme loss of craft control, so unlikely to be causal or indicative of any other problem. And the fact that the motor slowed almost to a stop despite the PWM holding at nearly 100% clearly indicates mechanical or electrical failure of the ESC/motor. The recorded motor current stayed high too.

View attachment 94255
I had an esc failure the day before, after an update and it said restart the drone if the problem persists please contact DJI.
 

Teejag

Member
Joined
Feb 16, 2020
Messages
15
Reaction score
1
Age
25
Location
Zimbabwe
I had an esc failure the day before, after an update and it said restart the drone if the problem persists please contact DJI.
And I did that and the error extinguished. Wonder if that’s all something to do with it.
 

slup

Well-Known Member
Joined
Oct 18, 2019
Messages
787
Reaction score
849
Age
52
Location
Sweden
And I did that and the error extinguished. Wonder if that’s all something to do with it.
I think that it's pretty well established from the logs you provided that this wasn't the pilot to blame, a bird strike shouldn't have lasted over 40sec. The ESC error may very well have something to do with this ... but with that said, I think your next step should be to contact DJI for their point of view.