Near Fly Away ( Why )

[Drone Chrish]

Hi Flying my Mavic 2 Zoom on Cannock Chase UK I had a near fly away I have been flying in this location before with no problem the drone lost control at point E and would not respond to sticks or RTH only at point G was I able to get control via the sticks but control was very erratic I am very nervous to fly this drone again until I can find out happened Any ideas ? I have attached two pictures Thank You

 

[AMann]

From Meta4: “Follow the instructions there to upload your flight record from your phone or tablet. That will give you a detailed report on the flight data.
Come back and post a link to the report it gives you.”

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

 

[Meta4]

Hi Flying my Mavic 2 Zoom on Cannock Chase UK I had a near fly away I have been flying in this location before with no problem the drone lost control at point E and would not respond to sticks or RTH only at point G was I able to get control via the sticks but control was very erratic I am very nervous to fly this drone again until I can find out happened Any ideas ? I have attached two pictures Thank You

Your drone didn't fly away.
It didn't nearly fly away.
You just had terrible signal connection and kept losing your downlink.
The recorded data has over 100 messages about Downlink Restored.
You had full control except for the poor connection.

 

[Thomas B]

The trees?

 

[Drone Chrish]

From Meta4: “Follow the instructions there to upload your flight record from your phone or tablet. That will give you a detailed report on the flight data.
Come back and post a link to the report it gives you.”

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

I have the link to the log viewer

DJI Flight Log Viewer | Phantom Help

Upload and view DJI Phantom, Mavic, and Spark flight logs online

www.phantomhelp.com

 

[MAvic_South_Oz]

DJI Flight Log Viewer | Phantom Help

Upload and view DJI Phantom, Mavic, and Spark flight logs online

www.phantomhelp.com

Lost it at 6'56.6" when you let the sticks go and it kept rolling, but attitude is messed up opposite to what it should have been ??
Was there a big wind gust or something ?
That problem stopped at about 7', where you remained stationary but took it up some 80 feet, then rolled right again, this time attitude was as it should have been.

It's kinda neat looking at these, some tell me what is the stick yellow, orange, red colour significance, it that 'downlink' ?
Is that ocusync signal ?

 

[Drone Chrish]

The trees?

Not sure about Trees being the problem as the drone was still doing it when only 2ft from the controller

 

[Meta4]

Lost it at 6'56.6" when you let the sticks go and it kept rolling, but attitude is messed up opposite to what it should have been ??
Was there a big wind gust or something ?
That problem stopped at about 7', where you remained stationary but took it up some 80 feet, then rolled right again, this time attitude was as it should have been.

There's nothing at all abnormal for that time period.
He had the right stick forward and went hands off at 6:56.5 with the drone moving NW at 24.5 km/hr.
By 6:59.9 it had slowed to zero as would be expected.

For the whole flight, there are no movements that don't correspond exactly to the joystick movements.
The drone was high enough above the trees for them not to have been a factor.

I'm not sure what caused the poor signal connection which confused the flyer but there's nothing at all suspicious about the flight.

 

[MAvic_South_Oz]

He had the right stick forward and went hands off at 6:56.5 with the drone moving NW at 24.5 km/hr.
By 6:59.9 it had slowed to zero as would be expected.

I'm sorry Meta, I'm lost so hope you can clarify . . . from 6m 53.2s to 6m 56.5s wasn't he rolling right ???
The attitude showed it at the right angle for that roll direction.
Then went hands off.

Then from 6m 56.6s with no stick input, didn't it keep moving in that same NEE (think up is north ?) direction BUT the attitude was the opposite way to what it should have been ???
This stopped at ~ 7m 5s (~ 10 secs), just after that clear carpark area, attitude stabilised,They took it up in altitude a bit, and from there they rolled again to that NEE direction . . .

Then, it appears to do the fly away thing again at 7m 54.5s ???
Hands off, keeps rolling 8m 5s (again ~ 10 secs), again with an opposite attitude to what you'd expect (?) then stabilised again ???

I'll google the downlink thing, see what that's about.

They flew mostly backwards from that point, turning not far from home to come in frontwards.

 

[Meta4]

I'm sorry Meta, I'm lost so hope you can clarify . . . from 6m 53.2s to 6m 56.5s wasn't he rolling right ???
The attitude showed it at the right angle for that roll direction.
Then went hands off.

I think that's just hard braking.

I'm sorry Meta, I'm lost so hope you can clarify . . . from 6m 53.2s to 6m 56.5s wasn't he rolling right ???
The attitude showed it at the right angle for that roll direction.
Then went hands off.

Then from 6m 56.6s with no stick input, didn't it keep moving in that same NEE (think up is north ?) direction BUT the attitude was the opposite way to what it should have been ???
This stopped at ~ 7m 5s (~ 10 secs), just after that clear carpark area, attitude stabilised,They took it up in altitude a bit, and from there they rolled again to that NEE direction . . .

Then, it appears to do the fly away thing again at 7m 54.5s ???
Hands off, keeps rolling 8m 5s (again ~ 10 secs), again with an opposite attitude to what you'd expect (?) then stabilised again ???

From 6:56.5 he went hands off until7:31.6 when he gave it a little left stick.
The drone sat still the whole time, going nowhere.
He took it up 50 feet but still did not touch the right stick until 7:47.4 when he gave it full aileron to move sideways which accounts for the roll numbers suddenly increasing.
He went hands-off again at 7:54.5, the drone rapidly slowed from 29 mph and the roll briefly reversed for braking and went back close to zero.

After 8:13.4 there is a gap of 13 seconds where he had lost the downlink so no data is recorded but he will have had full control as the control signal was not lost.
When the downlink is restored at 8:26.5, it has hardly moved except it is now facing NE.
It's reasonable to assume the operator rotated the drone during that loss of downlink.

I can't see anything to suggest any "flyaway" behaviour.
All movements match the joystick input.
The downlink breaking down so much has just confused the operator.

Then flew mostly backwards from that point, turning not far from home to come in frontwards.

It looks like the drone was always facing the direction of travel with the right stick pushed forwards - ie. the drone was going forwards for the trip home.

 

[MAvic_South_Oz]

I think that's just hard braking.

Braking for 10 secs ?
I'll have to go back and look again, but could understand a sec or two, not sure about longer ??

Will go and cross reference those times etc and check through that flight again, it's interesting stuff, and thanks for the tips.

 

[Meta4]

Braking for 10 secs ?
I'll have to go back and look again, but could understand a sec or two, not sure about longer ??

The braking is only for about 2.5 secs after going hands-off on the sticks.

 

[MAvic_South_Oz]

The braking is only for about 2.5 secs after going hands-off on the sticks.

I'm doing something wrong reading this then . . .

I see hands off sticks (was rolling right) at 6m 56.6s . . . and the drone keep moving until 7m 5s with huge left attitude ?
Is that 9.4 seconds ?

Then same sort of thing again, hands off right roll at 7m 54.5s, drone keeps moving until 8m 5s, again with that weird opposite attitude.
That appears to be about that same 9.4 odd seconds ?

I see the drone continuing to move on the map, with that strange attitude, within those 9 second periods, am I not seeing the times and movement correctly.

This is interesting but frustrating, but I'm keen to learn more.

 

[Meta4]

I'm doing something wrong reading this then . . .

Here's what the numbers look like

 

[MAvic_South_Oz]

Here's what the numbers look like

Thanks Meta, I don't have that spreadsheet and data as you presented, but the GPS data does show it stopped there.
I was looking at the map . . .

DJI Flight Log Viewer | Phantom Help

Upload and view DJI Phantom, Mavic, and Spark flight logs online

www.phantomhelp.com

And scrolling through those times, and it appeared the drone was still moving hands off.

Was this caused by downlink issues ?
Was it guessing the drone continued to move ?

I'm not sure if you look at the map and click through like I did at those 2 timespans mentioned in post #13, or just look at the raw data.

Thanks again, this is a spreadsheet program you flight analysts look at, that one of you has written ?

 

[Meta4]

Was this caused by downlink issues ?
Was it guessing the drone continued to move ?

It's not uncommon to see a lost downlink or even 5 or 6 in a flight but 120 on this flight is really strange.
I'd wonder if there's something wrong with the antennas or the controller for it to be losing signal so much and so close?
The app wasn't doing any guessing - there are blanks in the data each time the downlink is lost.

I'm not sure if you look at the map and click through like I did at those 2 timespans mentioned in post #13, or just look at the raw data.
Thanks again, this is a spreadsheet program you flight analysts look at, that one of you has written ?

I find it easier to read from the numbers in the spreadsheet.
I just download the CSV file straight from the Phantomhelp link.

 

[JeffPerkins]

You have a compass issue.

 

[sar104]

It's not uncommon to see a lost downlink or even 5 or 6 in a flight but 120 on this flight is really strange.
I'd wonder if there's something wrong with the antennas or the controller for it to be losing signal so much and so close?
The app wasn't doing any guessing - there are blanks in the data each time the downlink is lost.


I find it easier to read from the numbers in the spreadsheet.
I just download the CSV file straight from the Phantomhelp link.

I don't think it's an aircraft downlink loss - I think it's more likely an app disconnect problem. It's certainly clear at several points that the uplink is still there even when the telemetry is missing. For example, spot the missing rudder input:



With all the missing data it's difficult to diagnose the other problems, but there do appear to be ignored stick inputs:

 

[sar104]

You have a compass issue.

Why do you think that?