Dear DJI community

I hope you are all doing well.
I have a mavic mini drone. Recently, when it hovers outdoors, with plenty of satellites connected, when the push the sticks forward, the drone goes forward and slowly to the left or forwards and to the right.

Most of the times, when it hovers, it goes slowly to the left slowly and more and more until it wants to crash. Or the other direction (right)

Being the Electrical and Electronic engineer that I am, I have tried a cold IMU calibration, compass calibration, controller calibration, changed the propellers to brand new ones... But the issue still persists. I have uploaded a video fr you guys to see (this example when is the push the stick forward and then backwards,.. The mavic goes to the left.
I have uploaded the dat file of the last flight performed outdoors where you can see the drone hovering to the left when no sticks are pressed.
Can somebody please look at rhe dat file or help me?

Thanks in advance my fellow DJI community

https://we.tl/t-xEDp0ptW6h

Check that your props are not bent or damaged, especially the rear ones. Bad props can cause unusual flight characteristics.  If you have the Fly More Combo, the rears can get bent in the case very easily. I see you changed the props, is it possible the the new ones are bent or distorted. Are these after market props?

Hamiltom Posted at 4-18 13:59
Check that your props are not bent or damaged, especially the rear ones. Bad props can cause unusual flight characteristics.  If you have the Fly More Combo, the rears can get bent in the case very easily.

Thanks alot for your comment my friend. As I mentioned in the description, I've changed the propellers to brand new ones for all 4 sides. Right before I filmed that video and uploaded the dat files.

Is there much play in the screw shafts where the screws go through the props? I've read some people complain about props striking the side of the MM, leaving marks. The other thing I noticed myself was I had instability when updating firmware and I went back to the previous version and reflashed the new one again. It helped resolve sudden drops to the left on rapid deceleration.

Hamiltom Posted at 4-18 14:17
Is there much play in the screw shafts where the screws go through the props? I've read some people complain about props striking the side of the MM, leaving marks. The other thing I noticed myself was I had instability when updating firmware and I went back to the previous version and reflashed the new one again. It helped resolve sudden drops to the left on rapid deceleration.

Hi. Yes my propeller screws are nice and tight and thankfully it doesn't hit the inside of my drone. My firmware on the mini is also stock. From experience, I usually don't update.. Unless I have to. Thanks for replying

It seems to me a very challenging environment, close to buildings and trees. That can be attributed to inaccurate GPS guidance.

Try flying in an open field and see if you're getting a more linear path.

Can somebody please look at the dat file or help me?

I can't see a problem in your video.
How much is your Mini drifting?
Can you put some numbers on it?

How about posting the .txt file rather than the .dat file?

I have included both a txt and dot file in the wetransfer link my friend. From the video. I've just pressed forward and backwards. The mavic shouldnt go to the left.

If you can please have a look at the txt file, have a look at the part where I hover. And the mini drift to the left and I have to enetually bring it back or it would crash by going left so much.

Thanks in advance

I am no expert by any means but your file shows the battery charge is very low, starting at only 19% and ending at 9%.
And yes, both files are available using that link.

GaryDoug Posted at 4-18 18:16
I am no expert by any means but your file shows the battery charge is very low, starting at only 19% and ending at 9%.
And yes, both files are available using that link.

Hi. The drift also persists irrespective of the battery value. The video was filmed at 50-60 percent battery but sure. I will test it again today in the lovely weather we are having in London after a few days of shower

GaryDoug Posted at 4-18 18:16
I am no expert by any means but your file shows the battery charge is very low, starting at only 19% and ending at 9%.
And yes, both files are available using that link.

Hi again.

I tested it again just going forward and back. The mavic can't go forward in a straight line. It moves to the left and right as you can see in this video:

https://youtu.be/D-obOEnN4I4

Michaelsalt Posted at 4-19 01:24
Hi again.

I tested it again just going forward and back. The mavic can't go forward in a straight line. It moves to the left and right as you can see in this video:

Hi Michaelsalt, please check the RC input when you calibrate the controller and see if there is any abnormal input as well, keep us updated about the result.

DJI Thor Posted at 4-19 02:37
Hi Michaelsalt, please check the RC input when you calibrate the controller and see if there is any abnormal input as well, keep us updated about the result.

Hi DJI Thor.

Thanks for your advice.
I once again calibrated the remote controller through the DJI Fly app. The calibration was successful and the remote inputs function correctly and accurately.

I purchased the Drone brand new from Currys Pc world a month and a half ago so I find it hard to determine what could be causing the problem.

Thanks in advance.

I certainly think that this is because of GPS issue. You are flying between two tall buildings which will cause that drift right and left like you are flying indoor with no GPS. I noticed that when I was flying indoors. Try to do it on an open field. This test is not good in terms of conditions.

Burak C. Posted at 4-19 03:38
I certainly think that this is because of GPS issue. You are flying between two tall buildings which will cause that drift right and left like you are flying indoor with no GPS. I noticed that when I was flying indoors. Try to do it on an open field. This test is not good in terms of conditions.

Hi Burak

Thanks for your advice.

As you can see in the picture, I had about 13 satellites locked whilst I was flying the drone. I also flew my drone in a a big open park a few days ago which is when I first started noticing the problem.

I'm hoping somebody can have a look at the txt and dat files and see if it's possible to spot the fault from there. I've really looked after my drone from the day I purchased it.

In the first video I could estimate the number of satellites locked and in unobstructed view to a couple or handful at best. Small wonder it drifted.

If you'd ask me actually the MM flies so much better indoor in ATTI because no spurious GPS corrections take place. Outside w/out perfect coverage would be best not to fly. 99% of the times the FC manages to acknowledge and compensate for position errors pretty good (like in this case) but it's not invincible. I've seen cases with upscale models that made them rocket to a side with very little chance of recovery.

Playing with fire is a dangerous business. The less you play with it the better...

Michaelsalt Posted at 4-19 03:52
Hi Burak

Thanks for your advice.

Hi again,

Can you please upload your logs to https://www.phantomhelp.com/LogViewer/Upload/ so that everyone can easily check it out.

Burak C. Posted at 4-19 08:04
Hi again,

Can you please upload your logs to https://www.phantomhelp.com/LogViewer/Upload/ so that everyone can easily check it out.

I have just done another test with it after multiple IMU calibrations and compass calibrations. The mavic mini is COMPLETELY UNSTABLE. It randomly moves around like it has a mind of its own. I have uploaded the. Txt file and dat file in the wetransfer link below:

https://we.tl/t-WqGcuyVut6

Somebody please help?

We have been looking at the files but nothing odd appears except the instability of position. I hesitate to suggest this as it may not end well if there is a fault, but have you tried to fly it indoors, no GPS and no wind? Mine hovers very stable in that environment. Perhaps there is a large room somewhere? It must be well lit and have some kind of texture or random pattern on the floor.

You have a log with GPS PositionNoMatch, you've been warned with GPS low signal notice and was operating on 4 satellites most of the flight, I guess the only ones that were somehow still directly visibile above the Aircraft.

There's no reason for me to think you have a defective unit since operating under those conditions instability is to be expected.  Get yourself FRAP, the program is on this forum and you can read yourself your log.data.

m80116 Posted at 4-19 14:06
You have a log with GPS PositionNoMatch, you've been warned with GPS low signal notice and was operating on 4 satellites most of the flight, I guess the only ones that were somehow still directly visibile above the Aircraft.

There's no reason for me to think you have a defective unit since operating under those conditions instability is to be expected.  Get yourself FRAP, the program is on this forum and you can read yourself your log.data.

Where are you seeing only 4 sats? I don't see any less than 10 at any time in the two logs I have viewed.

GPS r in FRAP... but, looking at my logs it seems something related to GPS health and not a count. Although I see it dropping to 4 in my partly-enclosed flights.

Checked the .csv file (FRAP generated) and is being reported as a more understandable GPS Level.  No idea what was the actual count then... 10 seems a way too high number in those conditions. I am sure somebody could look up and see how many were actually in view in that position and time, but meh... GPS PositionNoMatch is already a sentence. The last time I've seen this error mentioned in a forum was in a fly away (probably because finally the FC was outsmarted by spurious GPS signals all over the place).

m80116 Posted at 4-19 15:29
GPS r in FRAP... but, looking at my logs it seems something related to GPS health and not a count. Although I see it dropping to 4 in my partly-enclosed flights.

Checked the .csv file (FRAP generated) and is being reported as a more understandable GPS Level.  No idea what was the actual count then... 10 seems a way too high number in those conditions. I am sure somebody could look up and see how many were actually in view in that position and time, but meh... GPS PositionNoMatch is already a sentence. The last time I've seen this error mentioned in a forum was in a fly away (probably because finally the FC was outsmarted by spurious GPS signals all over the place).

FYI the actual number of satellites is right there in the flight data and it's >10 the whole time.
And GPS PositionNoMatch is a very common false alarm.
It means nothing at all.
There is no hint of any GPS problem in the flight data.

Michaelsalt Posted at 4-19 13:11
I have just done another test with it after multiple IMU calibrations and compass calibrations. The mavic mini is COMPLETELY UNSTABLE. It randomly moves around like it has a mind of its own. I have uploaded the. Txt file and dat file in the wetransfer link below:

https://we.tl/t-WqGcuyVut6

I've never seen anyone making such slow flights.
I suspect if I tried to make mine crawl at that speed it would also look wobbly.
How does your Mavic perform when you fly it faster than half a mile per hour?
Go to a nice open area, put it >30 ft up and fly at >10 mph and see how that looks.

I have flown slowly along sidewalk with trees before and you have to be dead on in your heading or you will find yourself having to correct. For the most part, my mini hovers in one spot very well, even at 400 feet up. But I have noticed it drift a couple times in confined spaces. Btw, I don't think hovering and moving forward in a straight line are related. I mean, I don't think the mini plots a straight line using GPS coordinates and follows it. It just flies in the direction it is pointed.

Quadcopters are not camera dolly. There is always cap corrections to make. The less the speed, the more corrections to make. You can't see the difference between 0° angle and 0,1° when you start but as the distance increase, the minor error become visible.

Michaelsalt Posted at 4-19 13:11
I have just done another test with it after multiple IMU calibrations and compass calibrations. The mavic mini is COMPLETELY UNSTABLE. It randomly moves around like it has a mind of its own. I have uploaded the. Txt file and dat file in the wetransfer link below:

https://we.tl/t-WqGcuyVut6

I looked at both of your .DATs. I suspect that the cause is a "challenging" geomagnetic site. Not bad enough to cause a switch to ATTI mode or a fly away but enough to cause the instabilities you noticed. There are buildings close by and the height brought the MM close to any roadway ferrous material.
Looking at the 2nd flight it can be seen that the geomagnetic field strength varies depending on location but is mostly not dependent on height.



It looks like you powered up the MM, then started the Fly App, and then moved the MM to the launch site? Further, the power up location was geomagnetically distorted. For your next test could you power up at the launch site and not move it prior to launch. This should make it easier to understand if the problem is caused by a geomagnetically distorted launch site.

There were no problems with the number of GPS sats or the dilution of precision. There were some questions about navHealth (or gpsLevel). The label gpsLevel is confusing (I think) which is why it was renamed to navHealth. The FC computes a location based on GPS, accelerometer, gyro and magnetometer data. NavHealth is a measure of the confidence the FC has in that computed location.  NavHealth ranges from 0 (no confidence) to 5 (full confidence). In this flight navHealth was 4 or 5.

Compass calibration at the launch site or anywharer else can not fix a geomagnetically distorted launch site. A compass calibration compensates for the AC's magnetic anomalies not geomagnetic distortions external to the AC.

My advice is to try a test away from buildings, roadways, sidewalks, etc. Higher would also help.

Are you absolutely sure you are not inducing any left input while you're going forward, and reverse?
I saw the same effect when you were flying it backwards.
Just mt 2 cents, but that's why I "pinch" the sticks, not my thumbs on top of the sticks.

Flycaster Posted at 4-20 10:26
Are you absolutely sure you are not inducing any left input while you're going forward, and reverse?
I saw the same effect when you were flying it backwards.
Just mt 2 cents, but that's why I "pinch" the sticks, not my thumbs on top of the sticks.

There was no aileron or rudder input recorded while flying forward or backwards. In this plot the elevator is shown in blue. There was some aileron input (red) at the two ends of the traversal - I'm supposing to reposition after drifting to the side.

BudWalker Posted at 4-20 08:57
I looked at both of your .DATs. I suspect that the cause is a "challenging" geomagnetic site. Not bad enough to cause a switch to ATTI mode or a fly away but enough to cause the instabilities you noticed. There are buildings close by and the height brought the MM close to any roadway ferrous material.
Looking at the 2nd flight it can be seen that the geomagnetic field strength varies depending on location but is mostly not dependent on height.

[Image]

Hi BudWalker.

Thank you very, very much for your detailed flight log analysis and explanation. I am intreged by your flight log knowledge as your possesses a strong understanding in this field than most of my fellow university friends in the Electrical and Electronic Engineering field.

I will absolutely conduct the test again in an open park next to my house with no buildings, pathways and trees around and I will get back to you very very soon!

Thanks again in advance and see you tomorrow

Labroides Posted at 4-19 16:32
FYI the actual number of satellites is right there in the flight data and it's >10 the whole time.
And GPS PositionNoMatch is a very common false alarm.
It means nothing at all.

Yes and you know it's just the satellite number count... it's not the actual number with good lock and used for guidance. That's what I was looking for... but N/A for what I see, not sure if it's not in the .txt

GPS PositionNoMatch is a very worrisome error, I have a hard time finding one in my logs. Should the FC not realize the no match it would probably match a position with a completely different and spurious one... and a fly-away would take place.
How many GPS PositionNoMatch does one need to call it quits ? I don't like playing with fire... you better do quickly if you see that error. The GPS PositionMatch with invalid data is just a bit away.

m80116 Posted at 4-20 14:37
Yes and you know it's just the satellite number count... it's not the actual number with good lock and used for guidance. That's what I was looking for... but N/A for what I see, not sure if it's not in the .txt

GPS PositionNoMatch is a very worrisome error, I have a hard time finding one in my logs. Should the FC not realize the no match it would probably match a position with a completely different and spurious one... and a fly-away would take place.

You are making guesses about things you don't understand.

Try and go to the Mavic Pilots and see what happens to those who repeatedly ignore GPS warnings. If you know something we don't why don't you tell us ? Is it perhaps because you have no idea yourself and you don't know what you're talking about ?
Thanks for reminding I assumed wrongly out of a free program that labeled the GPS Health status as GPS r.... how could I knew !? I was aware of it... but never entered into details to verify where it was in the software.

There was a debate exactly on satellites and GPS Health, and I don't remember exactly where it was suggested a number. Can you find it on the .txt log file ? Is it on the .dat file ? Was it just a guess ? Since your statu of emeritus professor you should know, why don't you share your honorable knowledge with us ? Why don't you enlighten us ?

By the way... I understand far better than you that being SO CLOSE TO A WALL implies that almost HALF of the sky was blocked from view... and a good part of the horizon was blocked on the other side by trees. How do you think the GPS positioning works!? You need satellite signals coming from different angles to have a reliable position. But evidently you clearly IGNORE EVEN the basics of GPS positioning.

Go back to school and come back when you've actually learned something...

Does anyone know if it is possible to switch from GPS to ATTI mode manually? I noticed that when I fly indoors with ATTI mode, hovering was good but when I was doing tests at my balcony (MM automatically switched to GPS), the MM drift all over the place.

Bloodied_Mini Posted at 4-22 22:40
Does anyone know if it is possible to switch from GPS to ATTI mode manually? I noticed that when I fly indoors with ATTI mode, hovering was good but when I was doing tests at my balcony (MM automatically switched to GPS), the MM drift all over the place.

Does anyone know if it is possible to switch from GPS to ATTI mode manually?
It's easy ... with a drone that has a switch to allow it but impossible with the Mini or other Mavics.
Your balcony is not a good place to test.
It's likely that there's nothing wrong with your GPS but you need to be out in the open with a good view of most of the sky and away from magnetic interference for things to work properly.

m80116 Posted at 4-21 13:57
Try and go to the Mavic Pilots and see what happens to those who repeatedly ignore GPS warnings. If you know something we don't why don't you tell us ? Is it perhaps because you have no idea yourself and you don't know what you're talking about ?
Thanks for reminding I assumed wrongly out of a free program that labeled the GPS Health status as GPS r.... how could I knew !? I was aware of it... but never entered into details to verify where it was in the software.

Since your statu of emeritus professor you should know, why don't you share your honorable knowledge with us ? Why don't you enlighten us ?
I've tried multiple times, but very little gets through to you.
It's rare to find someone that knows so little but completely fails to recognise how that limits their ability to understand a subject.

But evidently you clearly IGNORE EVEN the basics of GPS positioning.
Go back to school and come back when you've actually learned something...
FYI .. I've used GPS professionally ever since affordable receivers were available (>20 years ago)
I'm fairly well acquainted with the basics of GPS (and plenty more).

I would try to address the rest of your rant but I'm sorry, I couldn't make any sense of it.

m80116 Posted at 4-21 13:57
Try and go to the Mavic Pilots and see what happens to those who repeatedly ignore GPS warnings. If you know something we don't why don't you tell us ? Is it perhaps because you have no idea yourself and you don't know what you're talking about ?
Thanks for reminding I assumed wrongly out of a free program that labeled the GPS Health status as GPS r.... how could I knew !? I was aware of it... but never entered into details to verify where it was in the software.

Try and go to the Mavic Pilots and see what happens to those who repeatedly ignore GPS warnings. If you know something we don't why don't you tell us ? Is it perhaps because you have no idea yourself and you don't know what you're talking about ?
OK Dolt .. since you asked, I'll try to explain it to you.
First it has been shown clearly in posts #32 & 34 that the issue was not GPS related.
That seems to have escaped your attention, but nevertheless we'll address the spurious points you've raised.

Here are relevant parts of the garbage you've been posting on this thread.
They show that you know little or nothing about reading flight data and are an unreliable authority on the subject.

Post #6
It seems to me a very challenging environment, close to buildings and trees. That can be attributed to inaccurate GPS guidance.
#18
I could estimate the number of satellites locked and in unobstructed view to a couple or handful at best.
#24
No idea what was the actual count then... 10 seems a way too high number in those conditions.

#22
You have a log with GPS PositionNoMatch, you've been warned with  GPS low signal notice and was operating on 4 satellites most of the  flight, I guess the only ones that were somehow still directly visibile  above the Aircraft
#36
Yes and you know it's just the satellite number count... it's not the actual number with good lock and used for guidance. That's what I was looking for... but N/A for what I see, not sure if it's not in the .txt
You don't read flight data, you don't know what's in the flight data but you want to argue with someone who does about what's in the data.
Smart boy  !!
That's a real indication of your stupidity.

The recorded flight data clearly shows the number of satellites connected for every 1/10th of a second of the flight.
The lowest number was 10 and that was only for 1.5% of the flight.
The rest of the flight the numbers went as high as 15 with a mean number of 12.3.
The flight controller looks at much more than just satellite numbersand assigns a GPS health value to indicate the confidence in the GPS positioning.
For the entire flight is was 4 which is very good.
The flight controller is telling anyone that looks that GPS data for this flight was fine.

If you look at the flying environment rather than making up something and bleating about what you imagine it was, you'd see that the environment was nowhere near as constrained as you believe.
He was flying near two 4-storey buildings but still had a reasonably good skyview (as shown by satellite numbers and GPS health data)

#36
GPS PositionNoMatch is a very worrisome error, I have a hard time finding one in my logs. Should the FC not realize the no match it would probably match a position with a completely different and spurious one... and a fly-away would take place.
How many GPS PositionNoMatch does one need to call it quits ? I don't like playing with fire... you better do quickly if you see that error. The GPS PositionMatch with invalid data is just a bit away.
GPS PositionNoMatch is a meaningless false alarm.
It does not indicate any problem at all with GPS reliability.
You are skwarking about nothing (that's nothing new).

How about posting the .txt file rather than the .dat file?

ro_flyer Posted at 4-23 01:05
How about posting the .txt file rather than the .dat file?

The link he posted in #1 does include the .txt file as well as the .dat file.

Labroides Posted at 4-23 00:28
Try and go to the Mavic Pilots and see what happens to those who repeatedly ignore GPS warnings. If you know something we don't why don't you tell us ? Is it perhaps because you have no idea yourself and you don't know what you're talking about ?
OK Dolt .. since you asked, I'll try to explain it to you.
First it has been shown clearly in posts #32 & 34 that the issue was not GPS related.

Yes...12 sats can be usually present in unobstructed sky. Guess what happens when you're next to a wall !? That drone was probably in view of less than half of the sky considering not only the left part was blocked but also a considerable angle at starboard.

Anyway... I already admitted the mistake of confusing the GPS Health tag labeled in the program GPS r. What are you bragging about ?

You wanted to give a lesson and you've clearly shown you have no idea how GPS positioning works, it now remains marked in your words. Go invent some other random stuff to put into my mouth... perhaps we can have a laugh.

Your pathetic reiterated and gratuitous personal offenses denote just your lack understanding and arrogance of the small minded person you are. I never begin offending first and but since you asked I am calling you the way you are, which it can't be an offense.

I'm really having fun seeing your scramble... I won't answer further than this as I have better things to do, I let you cook on your broth since you're surely past your expiry date.