Mavic Pro doesn't fly right, crabs left.

[Danr-62]

Hey guys and gals hope this finds you all well.

Have an issue and I hope perhaps you can give me a hand with getting it corrected.

I started noticing that the drone didn't necessarily follow the cross hair line of travel. It has progressed from something barely noticeable if it occurred, to a condition bad enough that makes flying the drone rather undesirable.

I took a close look at it in flight and it isn't a camera/gimbal problem it is the drone itself. I was able to visually see it crabbing to what looks like some 10degs +/- left.

Issue seems to be effected by direction of travel. If for example I fly a line heading north northeast the drone more often than not flies as it should, body and camera view all lined up as it it should be.
However if I turn around 180deg and follow that same flight line the opposite direction the drone crabs heavy to the left.
Fly a line north or south, both directions the problem is near equal.

I uploaded the files I believe you guys would want below,

Device .txt, csv, .dat AC .dat

Hope I set that right if not let me know and I'll fix it.

Example that may help you visualize the problem.

I take off and fly into the field south of the home point then turn west at 1m 7 odd seconds, I line up pilots view with the crop rows that are running east and west under the drone and fly forward down the field you can see what happens. Me attempting to find where "straight" is!
The flip side though is going east in that same field, the return flight line. By contrast to traveling west east is pretty good, everything lined up reasonably well. Finish up the turn small adjustment to line up then nothing but elevator until nearing the end of the field turning left behind the park.

If it'd just do that any direction I choose to fly it be great again!

 

[MAvic_South_Oz]

It sounds like a controller calibration issue, though I’ve never had the issue myself.
Hopefully someone will be able to tell for ute from the data and you can get it sorted.
If you google Mavic pro controller calibration, you might find an answer there, possibly a YouTube video on it.

 

[Danr-62]

It sounds like a controller calibration issue, though I’ve never had the issue myself.
Hopefully someone will be able to tell for ute from the data and you can get it sorted.
If you google Mavic pro controller calibration, you might find an answer there, possibly a YouTube video on it.

I'm sorry I should have included that in with my OP.

IMU calibrated (flat level surface starting with cold drone)
Compass calibrated (no metal anywhere near it)
Controller calibrated

I re did them just before the listed flight, even though other than the above mentioned issue no outward call for it.

Via the go app no warnings and the readings where good but I re did em anyway in hope of correcting the problem.
I also removed the strobe light so it is as it would be out of the box, no help there either.

Running stock props and in good condition, I tried a different set but no change (MAS) same either set.

Drone is 14 months old.

 

[sar104]

I'm not seeing a significant discrepancy between heading and track in that file:

 

[Danr-62]

I'm not seeing a significant discrepancy between heading and track in that file:

View attachment 84462

First thank you for taking a look at this problem for me, I do appreciate it very much.

I compared early flight records to this one and I don't see any difference, hardly know what I'm looking at too though!
Seems the drone thinks it is flying straight with the line of travel but it is not.

You when flying it you can see it in the live view, looking well to the left to fly a particular line such as following the crop rows in a field. At times thought turn around and fly the reverse course and its dead on straight again as it should be.

No trouble visually verify it looking at the drone itself in flight, it is crabbed to what I'd consider hard to the left.
Using a degree wheel for reference, the drones body is rotated to somewhere near the 10degs mark to its actual line of travel.

I should add that the drone does not rotate in hover, holds position as it should.

The problem is repeatable, following the old railroad bed across the road from the park for example.
I can follow it heading north north east and the drone will directly follow the cross hair on the screen (drone not crabbed left) live view nicely centered as it should be.

But I stop I turn 180degs and fly back the opposite direction? the same is no longer true. I now have to put the cross hair on the screen over on the east side of the road to follow the rail bed (drone crabbed hard to the left) and live view cocked heavily to the left.

I fly over the trees that are between the road and the old rail bed, not over the road and not over the walking path. Just wanted to mention that.....

 

[jimlips]

I'm having the same problem with my MPP. If I start at the end of a straight line, aim the crosshairs on the RC straight ahead, the aircraft does not go the direction it is pointed. It makes it very difficult to fly between narrow objects. I'll try an RC cal and see if that helps.

 

[Danr-62]

I'm having the same problem with my MPP. If I start at the end of a straight line, aim the crosshairs on the RC straight ahead, the aircraft does not go the direction it is pointed. It makes it very difficult to fly between narrow objects. I'll try an RC cal and see if that helps.

Yes it makes it very difficult to fly always wondering where straight is.

 

[Mossiback]

It might need degaussing. If it was stored near a magnet (speaker, etc) it can screw up the internal compass which tends to affect the flight, although most complaints are about turning slowly off-course instead of crabbing.

 

[Danr-62]

It might need degaussing. If it was stored near a magnet (speaker, etc) it can screw up the internal compass which tends to affect the flight, although most complaints are about turning slowly off-course instead of crabbing.

Never stored near a magnet that said I considered buying one just to see if it corrected it.

Lots of people asking similar questions but most of them never posting back with what in the end corrected it, or if never got it corrected either one.

I did come across one where the poster stated it did correct compass error message that wasn't correctable with a calibration, but their problem didn't exactly match mine.

And mine doesn't post a calibration problem.......unless it thinks it is ok but really isn't?

It does place itself accurately on the map.

It does post a compass error here in the house but expect that. Computers monitors my desk is all metal the frig the list goes on. Shows in the red again as one would expect, outside thought it's very low numbers/short green line for both compasses. Never seen it use compass "2", dot always by compass "1". Compass 1 usually has the better reading but not by much if at all.

 

[Danr-62]

I ordered a Cfixer to try on it, be here Friday, rain until then so can't do anything in the way of testing anyway.

I set a compass right at the edge of a table and moved various parts of the mavic near it.

The compass needle would react to the motors (kinda expected that) but also the front and rear of the body. The lower body by the rear compass location, it would lightly move the needle. Its base plate? the screws? something inside? hard to say for sure.

I did the same with my P4 for comparison value. No reaction around anything other than its motors, and very light reaction to them with all four barely moving the needle.

The mavic on the other hand had fairly strong reaction with three of its motors, near nothing with the forth (the forth mirroring reaction of P4 motors).

Three motors of the mavic either pulling or pushing the needle several degrees when put adjacent to it, the fact it wasn't equal strong or weak for all four I found odd.

Not sure if it means anything as I didn't do the same when it did fly correctly and don't have one of the same exact model to compare results with!

 

[sar104]

First thank you for taking a look at this problem for me, I do appreciate it very much.

I compared early flight records to this one and I don't see any difference, hardly know what I'm looking at too though!
Seems the drone thinks it is flying straight with the line of travel but it is not.

You when flying it you can see it in the live view, looking well to the left to fly a particular line such as following the crop rows in a field. At times thought turn around and fly the reverse course and its dead on straight again as it should be.

No trouble visually verify it looking at the drone itself in flight, it is crabbed to what I'd consider hard to the left.
Using a degree wheel for reference, the drones body is rotated to somewhere near the 10degs mark to its actual line of travel.

I should add that the drone does not rotate in hover, holds position as it should.

The problem is repeatable, following the old railroad bed across the road from the park for example.
I can follow it heading north north east and the drone will directly follow the cross hair on the screen (drone not crabbed left) live view nicely centered as it should be.

But I stop I turn 180degs and fly back the opposite direction? the same is no longer true. I now have to put the cross hair on the screen over on the east side of the road to follow the rail bed (drone crabbed hard to the left) and live view cocked heavily to the left.

I fly over the trees that are between the road and the old rail bed, not over the road and not over the walking path. Just wanted to mention that.....

There is some evidence for what you describe there. Looking at the difference between the magnetic yaw and the IMU yaw shows a 2π periodic variation of around 10°:



The error peaks on a southerly heading, and is a minimum on a northerly heading. Is the crabbing flight at its worst when flying south?

This kind of error is due either to an out-of-calibration compass or strong enough magnetization of something on the aircraft to prevent a good calibration.

 

[Danr-62]

The error peaks on a southerly heading, and is a minimum on a northerly heading. Is the crabbing flight at its worst when flying south?

Yes, flying a southerly direction the degree of crab is at it max. Flying a northerly direction reduces it or even eliminating the amount completely.

I got my doubts on the Cfixer but I'm gonna give it a try.

Thank you everyone that took the time to respond and I'll will be sure and post an update here once the issue has been resolved.

 

[sar104]

The error peaks on a southerly heading, and is a minimum on a northerly heading. Is the crabbing flight at its worst when flying south?

Yes, flying a southerly direction the degree of crab is at it max. Flying a northerly direction reduces it or even eliminating the amount completely.

I got my doubts on the Cfixer but I'm gonna give it a try.

Thank you everyone that took the time to respond and I'll will be sure and post an update here once the issue has been resolved.

If you have calibrated the compass and it hasn't cured the problem then demagnetization followed by a new calibration is the correct answer. It does actually work.

 

[Danr-62]

If you have calibrated the compass and it hasn't cured the problem then demagnetization followed by a new calibration is the correct answer. It does actually work.

I really hope it does and yes fully believe the device does as advertised the problem is I have no reason to think the drone has been magnetized based on what I've read here. Present symptoms notwithstanding of course.

Both drones stored and ride in vehicles in the same place etc both well away from any magnetic source.

We do however fly near high voltage power lines, Large lines run across the back of our property.

Myself I always fly high above them but won't claim that I know my son has never been in close proximity to them, unintentionally or otherwise!

If it has been magnetized or otherwise effected I be at a loss for any other likely source.

 

[Danr-62]

Cfixer did not correct the problem.

Any suggestions as what to do next?

 

[sar104]

Cfixer did not correct the problem.

Any suggestions as what to do next?

Post another DAT file from after the demagnetization and recalibration. The problem was completely clear, so the question is why that didn't fix it.

 

[Danr-62]

Post another DAT file from after the demagnetization and recalibration. The problem was completely clear, so the question is why that didn't fix it.

Ok will do.

 

[Danr-62]

A link to the files,

Device .txt, .dat AC .dat

 

[SkyDogII]

Cliffhanger. I’m very interested in what happens in the log files.

 

[sar104]

A link to the files,

Device .txt, .dat AC .dat

That didn't change much - the periodic variation is still there.



Firstly - are you sure that you fully demagnetized the aircraft? I'm very surprised not to see a more significant difference, even if you hadn't properly fixed it.

Secondly - I suggest another flight test. Take off to 50 ft or so and then slowly rotate the aircraft (rudder only), first CW through two revolutions, then CCW. I'd like to see what the resulting DAT file shows.