Can someone help me how this happens?

vncntybnz · Sep 9, 2020

I lost control on my mavic air.
I got full batteries on both aircraft and controller as well as my iphone. Weather is clear and not windy. Airmap app says I'm clear to fly and no interference. Line of sight and can still see my drone hovering. And it suddenly fly away. Got lucky it didnt hit the trees. Can someone help how to fix this? I appreciate it. Im scared to fly now.

Meta4 · Sep 9, 2020

vncntybnz said:
I lost control on my mavic air.
I got full batteries on both aircraft and controller as well as my iphone. Weather is clear and not windy. Airmap app says I'm clear to fly and no interference. Line of sight and can still see my drone hovering. And it suddenly fly away. Got lucky it didnt hit the trees. Can someone help how to fix this? I appreciate it. Im scared to fly now.

As usual, flight data is needed to be able to provide useful comment.
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 gives you.

Also it might be helpful if you can describe your launch spot.

vncntybnz · Sep 9, 2020

Oh I'm sorry, here is my flight log. It starts going away on it's own around 0:58 - 1:20

DJI Flight Log Viewer - PhantomHelp.com

Here's how it looks on my controls

Thank you so much

Meta4 · Sep 9, 2020

vncntybnz said:
Oh I'm sorry, here is my flight log. It starts going away on it's own around 0:58 - 1:20
DJI Flight Log Viewer - PhantomHelp.com

I'm sorry ... it's difficult to interpret your flight data because you aren't using Mode 2 controls and trying to work out what joystick controls do what is confusing.
I have a suspicion that the issue is a small yaw error situation caused by your choice of launch spot, because of the curved flight and your description, but I can't confirm that.
It would also be helpful if you could describe your launch spot.
The flight data doesn't even tell where it was because you started flying before you had GPS and could record a launch point.

Perhaps @Slip can tell you more .. he likes a challenge.

boblui · Sep 9, 2020

Meta4 said:
Perhaps @Slip can tell you more .. he likes a challenge.

You meant @slup ?

Meta4 · Sep 9, 2020

boblui said:
You meant @slup ?

Yes .. it was a tipo

slup · Sep 10, 2020

The stick mode used is 1 (left/right stick forward & backwards is shifted ... so throttle & elevator shifted but rudder & aileron remains as mode 2).

But yeah ... we see a yaw error, pretty easy to see now when we have the video.

The AC is pointing towards the center of the Garden circle in the vid. just before the flyaway starts ...

But the log shows that the imuYaw is about 90 degrees wrong (Green bar shows yaw angle) ...

So once up there on height at 53sec into the flight, when wind starts to affect the AC & it needs to apply thrust to maintain position it does it with the wrong motors due to not knowing the true aircraft heading ... loosing position even more, tries again ... all this in a lightning speed & in each try the positional error just becomes larger & the flyaway is a fact.

The interesting here is that the AC seems to get everything right after that the circular flyaway have made a half lap at 72sec ...

Here the imuYaw in the log shows this ...

And the vid. this ... which is in agreement with the imuYaw.

In order to pinpoint the reasons for the initial yaw error & the later correction, the mobile device .DAT log is needed. The correct log ends with FLY051.DAT & you find it in a subfolder, MCDatFlightRecords, in the folder that contains the TXT logs. Attach that in a new post here & we can dig a bit deeper.

baristabob · Sep 10, 2020

It is amazing to see the detailed analysis you provided.

vncntybnz · Sep 10, 2020

slup said:
The stick mode used is 1 (left/right stick forward & backwards is shifted ... so throttle & elevator shifted but rudder & aileron remains as mode 2).

But yeah ... we see a yaw error, pretty easy to see now when we have the video.

The AC is pointing towards the center of the Garden circle in the vid. just before the flyaway starts ...
View attachment 112755

But the log shows that the imuYaw is about 90 degrees wrong (Green bar shows yaw angle) ...
View attachment 112756

So once up there on height at 53sec into the flight, when wind starts to affect the AC & it needs to apply thrust to maintain position it does it with the wrong motors due to not knowing the true aircraft heading ... loosing position even more, tries again ... all this in a lightning speed & in each try the positional error just becomes larger & the flyaway is a fact.

The interesting here is that the AC seems to get everything right after that the circular flyaway have made a half lap at 72sec ...

Here the imuYaw in the log shows this ...
View attachment 112757

And the vid. this ... which is in agreement with the imuYaw.
View attachment 112758

In order to pinpoint the reasons for the initial yaw error & the later correction, the mobile device .DAT log is needed. The correct log ends with FLY051.DAT & you find it in a subfolder, MCDatFlightRecords, in the folder that contains the TXT logs. Attach that in a new post here & we can dig a bit deeper.

WOW!!! That's so amazing how you see everything! Thank you so much and I appreciate you looking into my AC issue. I have attached my DAT file in case you need more info. I wish I know how to read all these ? Thanks so so much again!

Meta4 · Sep 10, 2020

vncntybnz said:
WOW!!! That's so amazing how you see everything! Thank you so much and I appreciate you looking into my AC issue.

I don't mean to nag but is there any chance you could describe your launch spot.
What was the surface you launched from?

vncntybnz · Sep 10, 2020

Meta4 said:
I don't mean to nag but is there any chance you could describe your launch spot.

Hi, yeah, I launched my mavic in an unpaved pathway with a little rocks below it. I don't have a launch pad. ?

Meta4 · Sep 10, 2020

vncntybnz said:
Hi, yeah, I launched my mavic in an unpaved pathway with a little rocks below it. I don't have a launch pad. ?

Since the issue is one of the compass being deflected at the initialisation stage, I was wondering what steel items might have been close by.
Anything in the case?
Was the case on reinforced concrete or was there any steel object near?

vncntybnz · Sep 10, 2020

Meta4 said:
Since the issue is one of the compass being deflected at the initialisation stage, I was wondering what steel items might have been close by.
Anything in the case?
Was the case on reinforced concrete or was there any steel object near?

All I can think of is the metal light pole in front of me. Actually, this is the second time. The first one was I launched it in front of our garage and an open area no obstruction around and it suddenly did a little loop (like in the video) and it's scary coz it goes by itself.

Meta4 · Sep 10, 2020

vncntybnz said:
All I can think of is the metal light pole in front of me. Actually, this is the second time. The first one was I launched it in front of our garage and an open area no obstruction around and it suddenly did a little loop (like in the video) and it's scary coz it goes by itself.

In front of your garage ... was it a reinforced concrete surface?
That's the most common source of yaw errors.

vncntybnz · Sep 10, 2020

Meta4 said:
In front of your garage ... was it a reinforced concrete surface?
That's the most common source of yaw errors.

It's a reinforced concrete. I think you are right on that. I flew it everywhere except and it was fine except launching in reinforced concrete.

Clvrtrvr · Sep 14, 2020

Im replying cause i want to see Slurp's response, haha. Me being a newbie, would doing a calibration for Compass and IMUI help prevent this?

Meta4 · Sep 14, 2020

Clvrtrvr said:
Im replying cause i want to see Slurp's response, haha. Me being a newbie, would doing a calibration for Compass and IMUI help prevent this?

No
Recalibrating your compass cannot ever help overcome any magnetic issues that are outside the drone.
And the IMU is working properly.
Not launching from a magnetically compromised location is the only way to avoid this.

Fred Garvin · Sep 14, 2020

Clvrtrvr said:
Im replying cause i want to see Slurp's response, haha. Me being a newbie, would doing a calibration for Compass and IMUI help prevent this?

You don't have to....just click the "Watch" button top right:

Clvrtrvr · Sep 14, 2020

Fred Garvin said:
You don't have to....just click the "Watch" button top right:

View attachment 112996

thanks tips.

Clvrtrvr · Sep 14, 2020

Meta4 said:
No
Recalibrating your compass cannot ever help overcome any magnetic issues that are outside the drone.
And the IMU is working properly.
Not launching from a magnetically compromised location is the only way to avoid this.

ok, but! if you did a IMU calibration where he took off, wouldn't it come up with Magnetic interference there?