EDIT:
After few test flights it appeared that the issue is with the gimbal yaw. I am now looking for a way to fix it if possible. I will keep the first post updated with all relevant data and logs.

Flight 1
================
Log: https://www.phantomhelp.com/LogViewer/0X3KX8PHVQAMVX98HNY0/

Issue started at around minute 3:00 when I aimed down the river and just pushed right stick forward, but the drone started moving to the right and the gimbal tilted a bit. After about 30 seconds it recovered. You can see a short clip of the event here:

Video:

Flight 2
================
Log: https://www.phantomhelp.com/LogViewer/6J6KJAQPHTBUI0AG3PZ8/

Video 1 - footage starts at around 8:30:


Video 2 - footage starts at around 2:37:


Flight 3
================
Log: https://www.phantomhelp.com/LogViewer/7PYIP2N6NK9ANS8W9HWO/

Video:


Flight 4
================
Video:


Flight 5
================
Log: https://www.phantomhelp.com/LogViewer/26T05Y664GRA4NQXPDEP/

Video - problem start at around 00:30:

It looks a bit like how the sensors see sommat and try to bypass
This does show around the time in question though youre only at 30m but i think it was this to blame

AAircraft in Altitude Zone (Altitude limit 60m). Fly with caution).

03:02 - The tilt sensor shows that there is a strong side wind. In strong wind, the drone can accurately hover in one place by GPS. But when flying, it can no longer fly straight.

Drone.Hunter Posted at 1-16 17:47
03:02 - The tilt sensor shows that there is a strong side wind. In strong wind, the drone can accurately hover in one place by GPS. But when flying, it can no longer fly straight.

That makes sense, but it would have to be faster than 22mph to do that and it was only gusting less than 5kph around that time

3:19 PM	10.0 °C	4.6 °C	69 %	NE	1.1 km/h	1.9 km/h	1,014.83 hPa	0.00 mm	0.00 mm	0	35.9 w/m²
3:24 PM	10.0 °C	4.7 °C	69 %	NNE	0.7 km/h	1.7 km/h	1,014.73 hPa	0.00 mm	0.00 mm	0	34.3 w/m²
3:29 PM	9.9 °C	4.7 °C	70 %	NNW	0.5 km/h	1.4 km/h	1,014.83 hPa	0.00 mm	0.00 mm	0	32.2 w/m²
3:34 PM	9.7 °C	4.6 °C	70 %		0.7 km/h	1.5 km/h	1,014.83 hPa	0.00 mm	0.00 mm	0	30.3 w/m²
3:39 PM	9.7 °C	4.5 °C	70 %	NE	1.7 km/h	3.2 km/h	1,014.73 hPa	0.00 mm	0.00 mm	0	26.9 w/m²

The above is 10m winds, the pilot was at about 30m'ish

Thank you for you input.

OA was set to brake and there was no indication in the app that sensors picked something, so I thought that sensors were not the cause. There was some wind on the ground level, but it was against the trajectory of the done. Sudden strong gust that is able to move the drone like so should be visible on the trees to the right in the footage, I guess?

My first thought was that the camera was facing to the left. If you look at the footage you can see that the camera looks down the river, but the arrow on the map shows it was facing more to the shore. Is there a way to see camera angle somewhere in the logs?

Tuxtard Posted at 1-17 05:35
Thank you for you input.

OA was set to brake and there was no indication in the app that sensors picked something, so I thought that sensors were not the cause. There was some wind on the ground level, but it was against the trajectory of the done. Sudden strong gust that is able to move the drone like so should be visible on the trees to the right in the footage, I guess?

Hi,

Cam angle to be found in " GIMBAL.yaw" Field.
Checked in your log. Camera all the time facing forward, so equal as drone heading.

But shortly after take the compass heading moves from 345 to 327.
This move is only indication, drone did not yaw to 327.

Not sure how that is seen on the screen recording irlt cam vieuw.

cheers
JJB

JJB* Posted at 1-17 09:29
Hi,

Cam angle to be found in " GIMBAL.yaw" Field.

@JJB, so does that imply that it was magnetic influence?

TonyPHX Posted at 1-17 09:39
@JJB, so does that imply that it was magnetic influence?

most likely yes.

see this > https://forum.dji.com/forum.php?mod=viewthread&tid=283082

JJB* Posted at 1-17 10:36
most likely yes.

see this > https://forum.dji.com/forum.php?mod=viewthread&tid=283082

I am surprised that happened so far from the home point, this normally happens pretty close to the take-off.

Bashy Posted at 1-17 17:26
I am surprised that happened so far from the home point, this normally happens pretty close to the take-off.

indeed, shortly after takeoff

Sorry, i meant the actual yaw event as these are normally only a couple of seconds after take-off, i.e. raising a few feet from the object that caused the interference, with this flight the drone was at a decent height before this incident took place, unless of course, the object was 10 or 20ft high.

Tuxtard, said it happened around 3 minutes into the flight and it was also a  minimal yaw issue with no yaw/speed warnings in the FL that we normally see, this is why I didn't call it but i did find it interesting that the runway height restriction is pretty much in the exact position where this incident happened thus my thinking that that was the problem but it shouldn't have been due to height being well below the restriction.

Yeah, it's strange that everything was fine in the first 3 minutes and then it started to act weird. Anyway, thank you for your help. I will recalibrate everything and hope for the best.

Tuxtard Posted at 1-19 12:49
Yeah, it's strange that everything was fine in the first 3 minutes and then it started to act weird. Anyway, thank you for your help. I will recalibrate everything and hope for the best.

There is no need to calibrate the drone's compass unless you've already done this and you think that it might be possible that you've created a poor calibration.

Is it possible for you to fly at the same spot again flying the exact same route?

Bashy Posted at 1-19 18:57
There is no need to calibrate the drone's compass unless you've already done this and you think that it might be possible that you've created a poor calibration.

Is it possible for you to fly at the same spot again flying the exact same route?

Yeah, I did a calibration before the flight. I will fly at the same location when the weather settles and see how it goes.

Tuxtard Posted at 1-20 03:55
Yeah, I did a calibration before the flight. I will fly at the same location when the weather settles and see how it goes.

You may have created a poor calibration, best this to now is to find an open area, grass is preferable and do a final calibration and never touch it again.If the app ever says compass error, that just means move it away from the interference, i.e., turn it off, move it away and turn it back on again, shouldn't ever need to calibrate it, its done at the factory and that should be it unless you add something to the drone like a speaker, then it might need to be calibrated...

ANyway, look forward to the results when the weather fairs some...

Bashy Posted at 1-20 04:13
You may have created a poor calibration, best this to now is to find an open area, grass is preferable and do a final calibration and never touch it again.If the app ever says compass error, that just means move it away from the interference, i.e., turn it off, move it away and turn it back on again, shouldn't ever need to calibrate it, its done at the factory and that should be it unless you add something to the drone like a speaker, then it might need to be calibrated...

ANyway, look forward to the results when the weather fairs some...

Yeah, I always perform calibrations away from metal objects or other sources of interference. I did a compass calibration because it similarly misbehaved during couple of previous flights. It tends to turn to the side when I am trying to fly straight and that's why I thought it was a gimbal yaw issue.

Tuxtard Posted at 1-21 01:14
Yeah, I always perform calibrations away from metal objects or other sources of interference. I did a compass calibration because it similarly misbehaved during couple of previous flights. It tends to turn to the side when I am trying to fly straight and that's why I thought it was a gimbal yaw issue.

Hi,

if you fly manually than the compass is not used to keep heading, it will just show heading wich is logged in the log.

cheers
JJB

Another flight, another location, another issue.

https://www.phantomhelp.com/LogViewer/25GI4LOG339K3A87O0XY/

Before the flight I did a compass and camera calibration. It struggles to fly straight.





In these clips you can see that the drone is yawing to the right, even though I am only pushing forward. After I stop it corrects itself without any yaw input, as visible in the footage.

I experienced a similar problem on other DJI models, but on this unit it feels more prominent. At this point I am trying to figure out if this is normal for this model or this unit is faulty.

Can you relate the two videos to flight times in the log?

Tuxtard Posted at 1-21 13:24
Another flight, another location, another issue.

https://www.phantomhelp.com/LogViewer/25GI4LOG339K3A87O0XY/

Where was this location? ive just spent way too long on google maps trying to work it out lol

Its as though its hitting a wall, to me that isn't a compass thing, generally, they shoot off to one side, not only slightly such as the issue you're having, its certainly not normal unless there is a GEO lock, block or whatever they are called

Tuxtard Posted at 1-21 13:24
Another flight, another location, another issue.

https://www.phantomhelp.com/LogViewer/25GI4LOG339K3A87O0XY/

Hi,

Always a ittle difficult to see part of a flight on video, finding out wich part of the video = log data.
But flying only with forward stick (see the chart of your flight), your Mini3pro kept his heading within 0.2 degree.  So straight flight line, heading 333.5 all the time. (first green box)
Move line same (GPS data) same as heading drone, but after relasing stick (speed from 5.9 m/s to zero) drone moved away from heading line.Kind of normal behaviour if there is some wind blowing on the drone, drone needs to get into stable hover with same heading as before, heading might vary during relasing all sticks to zero.

First green box, after releasing forward stick some yaw inut was given, drone follows.

For flying a perfect straight line, to a imaginairy straight line on the ground, you need zero wind.



cheers
JJB

Sean-bumble-bee Posted at 1-21 15:18
Can you relate the two videos to flight times in the log?

The first footage starts at around 8:30 and the second footage starts at around 2:37.

Bashy Posted at 1-21 19:46
Where was this location? ive just spent way too long on google maps trying to work it out lol

Its as though its hitting a wall, to me that isn't a compass thing, generally, they shoot off to one side, not only slightly such as the issue you're having, its certainly not normal unless there is a GEO lock, block or whatever they are called

Building in both videos is "Hotel Jugoslavija":

https://goo.gl/maps/9S2EM4E68Z25wqJw9

JJB* Posted at 1-22 02:13
Hi,

Always a ittle difficult to see part of a flight on video, finding out wich part of the video = log data.

First clip starts at around 8:30 and the second starts at around 2:37.

I kept the drone in LOS the whole time and it looks like it flies perfectly straight. It starts aligned but the more I keep flying straight the more it yaws to the side and on the video it looks like it is flying diagonally. That's why I suspect gimbal is the issue. I can't even fix it with left stick yaw input as it turns but still flies diagonally. I can only apply right stick roll inputs and keep it locked to make it actually go diagonally towards the camera angle.

Tuxtard Posted at 1-22 04:06
The first footage starts at around 8:30 and the second footage starts at around 2:37.

Both lines ; white coloured
To me looks really straight.

cheers
JJB

Tuxtard Posted at 1-22 04:16
First clip starts at around 8:30 and the second starts at around 2:37.

I kept the drone in LOS the whole time and it looks like it flies perfectly straight. It starts aligned but the more I keep flying straight the more it yaws to the side and on the video it looks like it is flying diagonally. That's why I suspect gimbal is the issue. I can't even fix it with left stick yaw input as it turns but still flies diagonally. I can only apply right stick roll inputs and keep it locked to make it actually go diagonally towards the camera angle.

must be indeed that the gimbal is little offset from the straightl ooking forward

Gimbal is hanging in rubbers left and right, any differences in there?

cheers
JJB

Did another flight. This time camera pointed almost completely to the right side.



As before, I was just pushing left stick forward. After I released the stick it stopped and angle recovered. This is how the drone looked liked upon return:





At this point it's rather obvious that the gimbal is the issue. What's strange is that it fixed itself the moment it landed, which means that the drone "knows" that camera is at the wrong angle.

Anyone know what I can do to try to fix this issue? Should I try that firmware refresh procedure or that's just nonsense? The drone is under warranty and I have DJI Care Refresh, but sending drone for repairs from my regions is really painful.

JJB* Posted at 1-22 05:06
must be indeed that the gimbal is little offset from the straightl ooking forward

Gimbal is hanging in rubbers left and right, any differences in there?

Rubbers look fine. You can see how it looks in the post above.

Tuxtard Posted at 1-23 04:17
Rubbers look fine. You can see how it looks in the post above.

Oke, guess the only thing you can do is get it repaired by DJI.
But really weird that at landing it will correct this mis-alignment!

Can you see it goes to one side after take-off ??

JJB* Posted at 1-23 05:12
Oke, guess the only thing you can do is get it repaired by DJI.
But really weird that at landing it will correct this mis-alignment!

No, it is aligned after take-off. It takes some time until it happens, and mostly after flying forward in a straight line. It's strange that:

1. It corrects itself after landing
2. It corrects itself after stopping midair
3. I don't get any gimbal motor overload or similar gimbal errors that I expect to get if the gimbal is damaged and it can't fight incoming airflow
4. Gimbal calibration completes without a problem
5. If I try to push the gimbal to the side while in the air it feels rock solid and I would move the whole drone before the gimbal moves

Tuxtard Posted at 1-23 05:27
No, it is aligned after take-off. It takes some time until it happens, and mostly after flying forward in a straight line. It's strange that:

1. It corrects itself after landing

oke...i am out of any ideas how to help....

Tuxtard Posted at 1-23 05:27
No, it is aligned after take-off. It takes some time until it happens, and mostly after flying forward in a straight line. It's strange that:

1. It corrects itself after landing

Looks like you need to make a warranty claim...

With the mavic 2 pro and zoom ( where the pilot can pan the camera ) a quick method of centring the camera is to use the gimbal up/down button on the controller. I do not know if you can set such a thing on your controller nor do I know if the 'trick' will work with your drone but it might be worth trying.

Sean-bumble-bee Posted at 1-23 09:09
With the mavic 2 pro and zoom ( where the pilot can pan the camera ) a quick method of centring the camera is to use the gimbal up/down button on the controller. I do not know if you can set such a thing on your controller nor do I know if the 'trick' will work with your drone but it might be worth trying.

I don't think this feature exists on this model, but I will try. Thanks.

New firmware update just popped up.



Instead of release notes I got this message.



After the update it stayed on the same version, but it says it's now up to date.



I hope this fixes it. Now I just have to wait for the right weather and test it.

Tuxtard Posted at 1-23 09:33
I don't think this feature exists on this model, but I will try. Thanks.

Its the C1 button underneath if you haven't changed it

Bashy Posted at 1-23 23:01
Its the C1 button underneath if you haven't changed it

Unfortunately, I use DJI RC-N1. FN button switches between 90 degrees and 0, if that's what you had in mind.

Mine started doing the same. Calibrated everything and still will not fly straight and drifts when I let off the stick. What is strange is in C and S mode no issue. Only in normal mode it does this.

Tuxtard Posted at 1-24 13:20
Unfortunately, I use DJI RC-N1. FN button switches between 90 degrees and 0, if that's what you had in mind.

Yes, that's it, I think that's what Sean was on about.

Have you tried refreshing the firmware yet?