Mini 2 flew away into some trees - log/video inside

[dutchct]

Basically getting some shots in near Austin Texas.

Mini starts acting a little funny, then starts flying south at a pretty rapid rate.

I have a couple of theories:

1. My wife started up our vehicle and pulled it forward just as it started acting funny. (interference?) Probably unlikely as we've flown from within the vehicle while moving.
2. One of the rotors broke mid-flight? I'm not super careful with take-off locations and the rotors can hit grass from time to time.

Log here:

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

 

[Meta4]

I have a couple of theories:

1. My wife started up our vehicle and pulled it forward just as it started acting funny. (interference?) Probably unlikely as we've flown from within the vehicle while moving.

I'm just starting to look into your flight data .. give me a little while and I'll get back with details.
But we can knock theory 1 on the head immediately.
Moving a car can't possibly affect what your drone is doing.

 

[Meta4]

2. One of the rotors broke mid-flight? I'm not super careful with take-off locations and the rotors can hit grass from time to time.

This did not happen either.
The drone flew straight into the tree without losing altitude.
If there was a problem with a propeller or motor, the drone would have spiralled and lost height.

But I'm having trouble making sense of some of your data.
What joystick mode were you using?

 

[dutchct]

The drone definitely lost altitude when it crashed. My remote is in "sport mode" right now, so that's likely what mode I was in.

I agree with you on the rotors. Thinking about it, I don't think a rotor broke, after the crash, only the rear rotors are damaged, and judging from the way it crashed in the video, it's not possible for it to fly the way it did. Conversely, in the video, it does seem to have a pretty good vibration going before it heads for the ground.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[Phantomrain.org]

From just watching the video it just looks like you flew into the ditch lost GPS and entered atti mode and drifted into some trees. Maybe something more sinister .


Phantomrain.org
Gear to fly your Mini in the Rain.

 

[Meta4]

The drone definitely lost altitude when it crashed. My remote is in "sport mode" right now, so that's likely what mode I was in.

I agree with you on the rotors. Thinking about it, I don't think a rotor broke, after the crash, only the rear rotors are damaged, and judging from the way it crashed in the video, it's not possible for it to fly the way it did. Conversely, in the video, it does seem to have a pretty good vibration going before it heads for the ground.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Here's what the flight data looks like: DJI Flight Log Viewer - PhantomHelp.com

The incident was caused by a very odd fault, possibly hardware, maybe software.
It's hard to tell because some of the data is not believable.
Definitely one for DJI and a warranty claim if the drone is <12 months old.
Here's my assessment

It looks like a very strange and rare malfunction.
The joystick Mode shown the the data is acting up.

It starts as Mode 0 (I have no idea what that is), at 0:51 sec it changes to Mode 2 (the most common mode).
At 2:06 it changes to Mode 1 for 13 seconds, then back to Mode 2.
At 2:50.1 it's back to Mode 0 for 17 seconds and then back to Mode 2.
And so on a few more times throughout the flight.
This is weird and something I've never seen or heard of before.

Towards the end of the flight, you were in Tripod Mode and things start acting up at around 6:07.6, when the drone starts to fly very slowly without corresponding joystick input.

At 6:12.7 the speed indicated in the log jumps from 3.2 to 9.3mph in 1/10th of a second (not believable).
At 6:13.5 it jumps from 7.5 to 16.3 mph (not believable).
And at 6:14.4it jumps from 14.2 to 41.7 mph (extremely unbelievable)
All of this in Tripod mode without corresponding joystick input.

Joystick mode switches to Mode 0 again at 6:17.7 and the Flight Mode switches to Sport.
Indicated speeds go crazy up and down.
The collision happens at 6:23.4 and any messages after that are a result of the collision, not the cause of it.

 

[Meta4]

@slup .. here's an interesting one for you

 

[Meta4]

From just watching the video it just looks like you flew into the ditch lost GPS and entered atti mode and drifted into some trees.

There was no problem with GPS until after the collision when the drone was upside down and affected by the collision.

 

[eEridani]

You were pretty low at that point, descending fast and only 75 feet above home. At about 6 minutes in you're looking down and were pretty close to a tree when you panned back up and spun the drone back to those ramps while descending even more. Then you hit something. My guess is you collided with that tree and damaged a prop.

 

[Deleted member 103366]

One point, if I am interpreting things correctly you left the motors powered for quite some time after the drone got stuck, I think you should have shut the motors down as soon as possible after if got stuck.
If a motor is stalled or blocked and the drone is trying to dry it you may be running the risk of cooking the relevant ESC. I do not know how good the protection is on the Mini 2 but I have a Phantom 3 that clipped a branch and came down, I CSC'd it as soon as it started to fall but for maybe less than 2 seconds on the ground the motors were trying to start.
After picking it up I saw an ESC failure or overheat warning. I was lucky I think, I switched the drone off and let it sit and cool down and everything was ok after it had cooled but I am fairly sure I have read of cooked ESC's in Phantoms.

 

[slup]

@slup .. here's an interesting one for you

As this is a Mini 2 we will never be able to definitely tell if this was a IMU failure or if it was a Yaw error from powering on the drone in a magnetically interfered spot, as the mobile device DAT log is encrypted ... but I'm very sure it was either of these 2 reasons.

The very first seconds in the flight show a large & abrupt uncommanded Yaw angle rotation ...

Here at 2,8sec into the flight ... the blue graph reports a yaw angle of 151 degrees ... & the corresponding yellow bar in the sat pic. depicts that.

(Click on all charts below to make them larger)


Here approx 1,5sec later ... without any rudder input (dashed red graph, value 1024=neutral), the IMU reports a yaw angle of -30,9 degrees. That's a uncommanded yaw change a tad over 180 degrees.



If then going forward to where the incident starts at about 371sec into the flight ... and compare the yaw directions with the crash clip.

Below the clip starts when the incident starts ... first, the yaw movement seen there initially in the clip doesn't get reported in the log by the IMU ... secondly, when the Mini 2 goes down in the trees it does it with nose first ... but strangely in the log the Mini 2 goes into the trees left side first, meaning a deviation between reality & log with +90 degrees.

The clip further below shall start where the red cross is ... then follow the hand-drawn yellow direction arrow. As seen here in the sat. pic with values from the log, it goes left side first.



And here the clip that should start about where the red cross is located in the sat. pic. above ...

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

If we then compare the GPS & IMU velocities, which normally shouldn't deviate, instead be very near 0mph we find this ... which also points to either of the 2 reasons I mentioned in the beginning of the post. The velocity deviations reach during the incident maximum speeds of 38mph (17m/s) in a northerly direction ... & 33mph (15m/s) in a easterly direction.

 

[Meta4]

As this is a Mini 2 we will never be able to definitely tell if this was a IMU failure or if it was a Yaw error from powering on the drone in a magnetically interfered spot, as the mobile device DAT log is encrypted ... but I'm very sure it was either of these 2 reasons.

And the joystick mode ??

 

[slup]

And the joystick mode ??

I don't see that when bringing up the PhantomHelp's CSV in CsvView ...

Stick mode looks to be 2 & below in the chart the different flight modes are clearly spelled out ...

Blue says P-GPS
Pink Says Tripod
Yellow says Sport
And greenish in the end says ATTI

 

[Meta4]

I don't see that when bringing up the PhantomHelp's CSV in CsvView ...

Stick mode looks to be 2 & below in the chart the different flight modes are clearly spelled out ...

Not the flight mode .. the joystick mode.
In the Phantomhelp CSV it's all over the place as I described in post #6.

 

[slup]

Not the flight mode .. the joystick mode.
In the Phantomhelp CSV it's all over the place as I described in post #6.

Aha ... you mean RC-mode.

0 & 1 is used in both P-GPS+Sport+ATTI
2 only in Tripod

The RC-mode change when the flight mode changes ...

Can't really say if it's relevant ... the initial uncommanded yaw movement didn't involve any input , neither did the uncommanded yaw turn just before the flyaway.

In the incident it's 3 clear deviations between either the clip & what's shown in the log or between yaw changes and recorded rudder inputs.

1.What happens to yaw just seconds after take off when it abruptly changes 180 degrees without rudder input
2.The CCW 90 degree yaw movement just before the flyaway that doesn't get recorded in the log
3.The yaw difference between the log & what's seen in the clip when the Mini 2 goes into the trees.

 

[careless]

I'm no expert but could it have lost control from flying it from inside the vehicle?

 

[slup]

I'm no expert but could it have lost control from flying it from inside the vehicle?

Yes ... control in the sense that the RC-Mini 2 control connection could have been broken. But this doesn't make your drone fly away ... it just initiate the failsafe action, which in your case was RTH. So your drone had only gone up to the set RTH height (in your case 144m) & flown back to the recorded HP & landed.

 

[Skwirl]

Just a question about the short crash video.
At 44/45 seconds there is something brown that appears top left of center for those 2 seconds. Anyway to identify what that is?
The drone goes down immediately after that blip of brown.
Maybe a wire/ tree branch?

 

[slup]

Just a question about the short crash video.
At 44/45 seconds there is something brown that appears top left of center for those 2 seconds. Anyway to identify what that is?
The drone goes down immediately after that blip of brown.
Maybe a wire/ tree branch?

It's totally unrelated unfortunately ... if the drone gets either of the 4 props damaged by something it will go down spiraling ... this as 2 props applies rotational torque CW & 2 CCW, if one prop gets damaged it will be a torque imbalance & the drone will start to rotate & lose height. This wasn't the case here.

 

[careless]

Yes ... control in the sense that the RC-Mini 2 control connection could have been broken. But this doesn't make your drone fly away ... it just initiate the failsafe action, which in your case was RTH. So your drone had only gone up to the set RTH height (in your case 144m) & flown back to the recorded HP & landed.

Thanks, it's 17F or -8C here in Kentucky USA and I love to fly from inside my truck.