Mavic 2 Pro Flyaway

[Anndybo]

Magnetic interference is always red mark! That's mean compass may see wrong geographic direction. North is not were should be for the drone. When you hear "Home Point..." it also says "check it on the map" - it tells you to check map on your device if the drone is pointing correct direction - if drone front is pointed North - on the map you have to see is pointing North. It is very important! If not - drone see wrong direction and following GPS is trying to compensate wrong movement - and you have fly away. Only way to save any GPS drone in this case - switch to manual mode. But we don't have it in our Mavic.

 

[Meta4]

Sorry folks ... there's nothing in the flight data to indicate any compass issues, just a loss of signal and failure to RTH.
If the compass was the issue, there would be lots of evidence in the flight data.

It's a lot more productive to work with the evidence rather than listing all the things you can guess may have happened.

 

[CRCs Reality]

Sorry folks ... there's nothing in the flight data to indicate any compass issues, just a loss of signal and failure to RTH.
If the compass was the issue, there would be lots of evidence in the flight data.

It's a lot more productive to work with the evidence rather than listing all the things you can guess may have happened.

FWIW, my comments weren't working off of things I "guess may have happened" , I was working off of his statement that he got a "Mag interference" error, no guesswork here, that's what he said. Since I've had personal experience of a near-flyaway after the same error I offered my thoughts on the subject. When I made my post, the flight data hadn't been uploaded yet, so I couldn't see if it showed compass errors or not

 

[Meta4]

FWIW, my comments weren't working off of things I "guess may have happened" , I was working off of his statement that he got a "Mag interference" error, no guesswork here, that's what he said. Since I've had personal experience of a near-flyaway after the same error I offered my thoughts on the subject. When I made my post, the flight data hadn't been uploaded yet, so I couldn't see if it showed compass errors or not

There's also no hint that the drone "flew away".
The data only shows that it went exactly where he flew it and then the data just stops because the signal was lost.
For all we can see in the data, it may have been abducted by a passing flying saucer.

And the flight data was there in the post #1.

 

[CRCs Reality]

There's also no hint that the drone "flew away".
The data only shows that it went exactly where he flew it and then the data just stops because the signal was lost.
For all we can see in the data, it may have been abducted by a passing flying saucer.

LOL, agreed, but like I said when I was typing my post the data hadn't been uploaded yet. I was working with the information I had available at the time.
And yeah, looking at the uploaded data, something else must have been up as it was nowhere near far enough away to lose signal.

 

[seanmclean]

I have always been under the impression that it was good practice to do so since the last place it was done was, well, China. No different than the recommendation if you travel more than approx 50 miles from the last place it was calibrated, you should calibrate the compass it again.

This is terrible outdated advice. If the drone requires calibration, it will tell you. Frequent calibrations only increase the likelihood of doing one in a bad environment.

 

[sar104]

Sorry folks ... there's nothing in the flight data to indicate any compass issues, just a loss of signal and failure to RTH.
If the compass was the issue, there would be lots of evidence in the flight data.

It's a lot more productive to work with the evidence rather than listing all the things you can guess may have happened.

Agreed. I don't see much in the data other than a weak signal warning in the last line of data and an unphysical RC elevator value in the same line. There was no data loss prior to the end of the log and so I would probably discount loss of downlink as the problem which, in any case, would eventually have resulted in an RTH.

It looks like either sudden power loss or FC failure. If it were a Phantom I'd strongly lean towards power failure but among the few cases of M2 losses there have been no clear power failures but some identifiable FC failures, and so that would be the most likely candidate in my opinion.

 

[Macabre]

What a freak situation.
Although it reminds me of another user issue where the battery popped out.
Still you say that you checked the last known location.

 

[ChopperLinc]

An urban area like this isn't a great flying location, particularly if you are a new flyer.
A large, open area would be a much safer place to learn.

Here's what your flight data looks like:
DJI Flight Log Viewer - PhantomHelp.com
But unfortunately it's not much help as it shows everything normal until you lost the connection.
I've scanned all the data and there's nothing I can pick out that indicates a problem.
That leaves us with two issues, neither of which has an obvious solution.
1. Why did it lose signal?
2. Why did it fail to RTH?

1. Did you have a clear line of sight to the drone?
Were you close to the house and did it block signal?
Did you move out across the street to get a better chance of connecting?

2. You had a good homepoint and GPS.
Unless you changed the Loss of Signal action from the default (no-one ever does) it should have initiated RTH
What was the wind like?
????

There was a clear line of sight, but may have gotten obscures right when it lost signal. I moved across the street, then walked directly to last known location, leaving someone at home point. My feeling is that it lost power, because the odds of it just dropping without any intermittent signal seems unlikely.

I am new, but I’ve flown it in the area several times with no problem, including following my son on his skateboard, admittedly, not in line of sight on the street over.

Did not change loss of signal action.

 

[ChopperLinc]

What surface did you take off from? When I had my Phantom3 Pro, I "almost" had a flyaway after launching from the roof of a parking garage (yeah, I know).. I had the same "Mag Interference" initially, but then it went away.
Once it was airborne, the compass and GPS disagreed on where north was, and it took off for parts unknown. Luckily I caught what was going on and switched to manual mode to bring it back, but it was a scary few moments.

If you launched from something with metal in it, I could see where it might cause confusion with the guidance system, and once it's in that state even if it tries to RTH, with the compass off it won't find home.

Took off from the middle of my street.

 

[ChopperLinc]

What a freak situation.
Although it reminds me of another user issue where the battery popped out.
Still you say that you checked the last known location.

Possible. There was one instance when I put the battery in a couple days after I got it and I noticed it didn’t firmly click.

 

[ChopperLinc]

If you follow the instructions and upload your log here then we might be able to help find it, or at least tell you what happened ;

DJI Flight Log Viewer - Phantom Help

Log attached in first post

 

[ChopperLinc]

What was your signal loss setting set to? If it were set to hover instead of RTH, it could've just been hovering there somewhere beyond its last know location waiting for signal re-aquisition.

RTH. I thought perhaps it stopped and hovered, but it wasn’t anywhere to be seen when I walked to location

 

[ChopperLinc]

Was this the first flight or did you have a few successful flights before with it.
If it effectively just disappeared there's always a chance of bird strike, or other (human) attack, or perhaps a prop or motor failed.
In some ways a motor failure is more likely in the first hour or two of use.
It is a vanishingly rare occurrence though.

It's highly unlikely to have literally flown away.
The term was coined with early phantoms which only used GPS rather than GPS and GloNass - and would struggle to hold a fix, they would go into atti mode through the flight and drift with the wind - difficult to deal with if you could barely see it and were at the limit of control range.

It was 300 feet away and the controller got 0 response either video or RC.

I’d flown it several times, I think logs say 50, but that counts every time I took off.

 

[ChopperLinc]

It has all the characteristics of a sudden total power failure. But if that's the case you should be able to find the crashed M2P somewhere at the end of the (in phone) recorded flight path.


Since you haven't, it's a mystery, maybe someone picked it up who knows. No contact details on the drone?
(reminds me I still have to do that on my MP...)

No tile, no contact details....stupid.

 

[Chopstix Kid]

Loss of GPS (something suggested in forums much, much more often than it actually occurs) won't make the drone fly anywhere.
Without GPS, the drone has no way of knowing where it is or how to get to anywhere else.
It would just hover without position holding and drift with the wind.

You're right. I was thinking malfunction, not complete failure.

 

[Chopstix Kid]

The compass is the most misunderstood aspect of flying.
Forums (and even some DJI manuals) are full of myths.
It's completely unnecessary to recalibrate the compass or anything else before flying.
Whether you travel 10 miles or 2000 miles makes no difference at all.

Yep I did some searching and looks like even DJI has changed their position on it. Good to know. My M2P did however pop up with IMU and compass errors right out of the box so for me, it was a given must-do.

 

[Chopstix Kid]

This is terrible outdated advice. If the drone requires calibration, it will tell you. Frequent calibrations only increase the likelihood of doing one in a bad environment.

I'm still living in 2016 apparently. This was gospel back when the P4 was first released

 

[sar104]

There was a clear line of sight, but may have gotten obscures right when it lost signal. I moved across the street, then walked directly to last known location, leaving someone at home point. My feeling is that it lost power, because the odds of it just dropping without any intermittent signal seems unlikely.

I am new, but I’ve flown it in the area several times with no problem, including following my son on his skateboard, admittedly, not in line of sight on the street over.

Did not change loss of signal action.

Bear in mind that it was at 45 m altitude and moving at 12 m/s on a track of -133° when the downlink stopped:



Estimating the resulting ballistic trajectory from that point indicates that it could have travelled another 20 - 40 m while falling which would likely have cleared the house that it was approaching, and it certainly would not have ended up on the street:



Have you checked with the owners to see if it landed on the roof or in their swimming pool?

 

[ChopperLinc]

Bear in mind that it was at 45 m altitude and moving at 12 m/s on a track of -133° when the downlink stopped:

View attachment 46462

Estimating the resulting ballistic trajectory from that point indicates that it could have travelled another 20 - 40 m while falling which would likely have cleared the house that it was approaching, and it certainly would not have ended up on the street:

View attachment 46463

Have you checked with the owners to see if it landed on the roof or in their swimming pool?