Had my first ever flyaway today

[Coconut Island Drones]

Ok so I was at Loch Katrine in Scotland doing a bit of filming and it all went well. No issue's, the drone performed as it should. On the way home I was going over an area known as Dukes pass and thought I'd stop and get some pics of a distant Loch. So set the drone up, waited for GPS lock and took off. I only wanted to go straight up to around 50m. As I got to around 40m the drone which was in normal mode suddenly shot off sideways in what seemed like sports mode. It wouldn't respond to my control inputs. It done this for maybe 6-10 seconds before it responded to the controls and stopped. Needless to say I brought it straight back down. I took off from the roof of my car so not sure if that had any effect and did notice it drifting to the left just after take off. I've taken off from the roof before without any issues. I only do that if there's no decent take off spot.
Drone is a mini 2.

 

[Ken1403]

I suggest using a landing pad for takeoff as the metal in your car affects the compass. You should have gotten a compass error notification in the fly app.

 

[GFields]

Ok so I was at Loch Katrine in Scotland doing a bit of filming and it all went well. No issue's, the drone performed as it should. On the way home I was going over an area known as Dukes pass and thought I'd stop and get some pics of a distant Loch. So set the drone up, waited for GPS lock and took off. I only wanted to go straight up to around 50m. As I got to around 40m the drone which was in normal mode suddenly shot off sideways in what seemed like sports mode. It wouldn't respond to my control inputs. It done this for maybe 6-10 seconds before it responded to the controls and stopped. Needless to say I brought it straight back down. I took off from the roof of my car so not sure if that had any effect and did notice it drifting to the left just after take off. I've taken off from the roof before without any issues. I only do that if there's no decent take off spot.
Drone is a mini 2.

Moved your thread to a better section for answers

DJI Flight Log Viewer | Phantom Help

• Post your flight data to get help
• Go to DJI Flight Log Viewer | Phantom Help
• 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 provides.

You are also welcome to just post the txt file here.

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

@Meta4 and @slup are great at offering explanations from the data.

 

[MS Coast]

Did you confirm the home point location and the drone orientation (compass reading) on the map prior to climbing?

This sounds like your compass was biased by ferrous metals, which can create the situation you describe.

 

[msinger]

I took off from the roof of my car

Do you normally take off from the roof of your car?

 

[old man mavic]

the metal roof of the car created its own magnetic field and the compass in the drone used this as it reference point for magnetic North
after it took of and climbed away from that magnetic field the IMU sensed a discrepancy,between what the GPS location was reading as actual magnetic North and what the compass in the drone was telling it ,during that wild dash for freedom it was able to correct the compass to true magnetic North and what the satellites' were telling the IMU was true magnetic north where the drone actually was ,so that meant the anomaly had been sorted and it stopped and waited happily for you to decide where you wanted to fly next
this explanation i am sure will be given in a more scientific language by others on the forum ,
i like to use a simpler (like me ) form

 

[MS Coast]

the metal roof of the car created its own magnetic field and the compass in the drone used this as it reference point for magnetic North
after it took of and climbed away from that magnetic field the IMU sensed a discrepancy,between what the GPS location was reading as actual magnetic North and what the compass in the drone was telling it ,during that wild dash for freedom it was able to correct the compass to true magnetic North and what the satellites' were telling the IMU was true magnetic north where the drone actually was ,so that meant the anomaly had been sorted and it stopped and waited happily for you to decide where you wanted to fly next
this explanation i am sure will be given in a more scientific language by others on the forum ,
i like to use a simpler (like me ) form

It may have not been the roof. I've launched and landed quite a few times from the roof of my Subaru Outback without problems with compass bias. There's no sunroof with a motor and I think the roof is aluminum. It's definitely advisable to confirm the compass reading prior to beginning each flight.

 

[msinger]

the metal roof of the car created its own magnetic field and the compass in the drone used this as it reference point for magnetic North

That's certainly a possibility. Is the roof metal though? Kind of curious since it sounds like the OP might have a routine of taking off from the roof without issue.

 

[MS Coast]

That's certainly a possibility. Is the roof metal though? Kind of curious since it sounds like the OP might have a routine of taking off from the roof without issue.

Only ferrous metals are the problem. I've had issues with the rebar in the patio outside my house. I've located a safe takeoff location among several problem spots. If the drone is placed in the wrong location, over the rebar, the compass can be biased by 45 degrees or more.

 

[Coconut Island Drones]

Thanks for the replies everybody. I'm going to put it down to the roof. I have used the roof of the car to take off a few times but do try to avoid it if I can. When I got home I calibrated the compass & IMU and have flown it since with no issues.

 

[MS Coast]

Thanks for the replies everybody. I'm going to put it down to the roof. I have used the roof of the car to take off a few times but do try to avoid it if I can. When I got home I calibrated the compass & IMU and have flown it since with no issues.

If you had successful takeoffs and flights on other occasions, I'd be reluctant to say the roof is the problem. The real problem may still be out there waiting to bite you. It's worth verifying the compass reading, along with the home point, before each flight.

 

[Coconut Island Drones]

If you had successful takeoffs and flights on other occasions, I'd be reluctant to say the roof is the problem. The real problem may still be out there waiting to bite you. It's worth verifying the compass reading, along with the home point, before each flight.

I'll do that from now on, thanks.

 

[Yorkshire_Pud]

Post the flight log, the answers are quite likely to be in there.

 

[Meta4]

Thanks for the replies everybody. I'm going to put it down to the roof. I have used the roof of the car to take off a few times but do try to avoid it if I can.

Why not check the recorded flight data to confirm whether you actually induced a yaw error by launching from the steel roof rather than assuming without evidence?

Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.
Or .. just post the txt file here.

When I got home I calibrated the compass & IMU and have flown it since with no issues.

Neither of those would have been necessary.

 

[Coconut Island Drones]

Why not check the recorded flight data to confirm whether you actually induced a yaw error by launching from the steel roof rather than assuming without evidence?

Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.
Or .. just post the txt file here.

Neither of those would have been necessary.

I've never looked at the flight logs and to be honest I'm not really sure where to find them.

Edit.....just seen the rest of your post. Will have a look at them tomorrow and if I can't make sense of them I'll post them here to see what others make of it.

 

[Meta4]

I've never looked at the flight logs and to be honest I'm not really sure where to find them.

That's why I posted a link with instructions.

 

[Coconut Island Drones]

That's why I posted a link with instructions.

Seen that after I'd replied, sorry. I'll have a look tomorrow and if I can't make sense of it I'll post it here for those more experienced to have a look.

 

[Tufargon]

Seen that after I'd replied, sorry. I'll have a look tomorrow and if I can't make sense of it I'll post it here for those more experienced to have a look.

You should post it here either way. A few people here are very good at deciphering all that info and will be able to give you detailed information about the flight and what happened.

 

[MAvic_South_Oz]

Thanks for the replies everybody. I'm going to put it down to the roof. I have used the roof of the car to take off a few times but do try to avoid it if I can. When I got home I calibrated the compass & IMU and have flown it since with no issues.

It's really worthwhile getting into at least finding and downloading flight logs, the phantomhelp reports are interesting to look through, and can give you all sorts of good info on the flight, aircraft data, battery data, and in more complex software analysers can decipher motor info and all sorts.

The logs and report can help you prove a flight was legal (if ever needed), in regards to altitude it was (from take off HP), and whether you were still in reasonable VLOS etc.

This does sound like a typical compass / imu error fly away, but the data will confirm.
If so, you were REALLY lucky the aircraft sorted out the discrepancy in 6 - 10 seconds.
Usually the airdraft crashes, or flies like that until it loses signal, keeping on its wayward trajectory until it crashes from power loss.

Seen that after I'd replied, sorry. I'll have a look tomorrow and if I can't make sense of it I'll post it here for those more experienced to have a look.

Look forward to seeing your flight analysis.

 

[MAvic_South_Oz]

So set the drone up, waited for GPS lock and took off.

Oh, meant to add.
To avoid any minor magnetic interference like this (and it often DOESN'T show up as a magnetic interference message), always, always wait for GPS lock before take off, and always check the drone is aligned to the aircraft arrow shown on the app map.
(edit - the nice lady says "HP recorded, please check it on the map" or similar, so do that to ensure mutual alignment every take off.)

With the compass / imu discrepancy problem, they will not align, usually somewhat out showing 90o - 180o or whatever out.
If this occurs, turn off the drone, move 10' or so away, try again, this usually resolves the issue.