Fixing DJI's Compass Problem

[IEATSRT]

IEATSRT…question for you…which way did you rotate it? Does it matter? I am in New York and getting J-hook to the right and TBE also…been thinking of trying to rotate it but wondering how much it helped and if you feel better with it like that?

Sorry for the late reply, I think Ian pretty much answered much better than I could. But yes if it is Toilet bowling counter clockwise go clockwise and vise versa. I never measured any degrees just took a wild guess. Naza V2 manual says to go between 10-30 degrees. Anyway it is better but still not pefect, think it still J hooks a bit on my first run out. Been flying FPV so not paying all that much attention, will test some more and report back. As Ian has mentioned it is not best solution. Had DJI designed the legs and compass just a bit different so it could be rotated none of this would be necessary. Why they didn't I really don't know, as their fix for their rotors such as 450 and 550 is to simply turn the puck. They obviously know it would need to be done on this craft also? Did the original phantoms not exhibit this at all?

 

[IEATSRT]

If there is going to be a firmware fix for it - can it be separate from other firmware updates please - not to correct the fly restriction zones (I actually will have a job photographing for a small airport! so, if I am blocked - it will make no sense. and I am a big boy over 56 years old, and can decide what is good or bad idea when flying).

I also hope it would not add the OEM battery warning, which is simply not a fair play. I am not updating firmware because of this issue. I should be able to make my own decision as for supplies, not being forced to pay DJI so much more money than other batteries cost. (just saying!)

If it is just a small airport chances are it wont be in the no fly zone, but worth a double check. I upgraded to 3.0 primarily so that I could use FPV mode and slow down my tilt via gain setting, unfortunately it appears that so far it is only working with the 3D and not the 2D. All that said, once you go to 3.0 it is possible to go back, at least for the time being. I don't know if you will ever get a fix without upgrading, it's going to come via firmware and pretty sure it wont be a separate firmware fix just for that.

 

[Fximaging]

Hello Everyone...

So here are the instructions DJI gave us beta testers for fixing the firmware... Seems odd they don't want us to do anything but hover.




Dear Buddies,

Good day!

We sent you IOSD MARK II several days ago and probably they’ve already reached you.

As far as I know, there are two firmware versions – V1.08 & V2.00. After installing IOSD MARK II, please follow the procedure below to go on with the tests.

- Firmware version V1.08

Test 1

1. Please recover the position of the compass on the landing gear the same as that when they have been out of the factory.
2. Place the phantom on a wide-open area and carefully calibrate the compass before the flight.
3. Take off and then keep the aircraft hover at a certain height.

Test 2

1. Move to another testing environment and calibrate the compass once again.
2. Take off and then keep the aircraft hover at a certain height.

- Firmware version V2.00

Test 1

1. Please recover the position of the compass on the aircraft the same as that when they are out of the factory.
2. Place the phantom on a wide-open area and calibrate the compass carefully before the flight.
3. Take off and then keep the aircraft hover at a certain height.


Test 2

1. Move to another testing environment and calibrate the compass once again.
2. Take off and then keep the aircraft hover at a certain height.

BTW, please name the data file captured from each test as “model of the aircraft + version number + city

For instance:” phantom2, v1.08, new york”.

Meanwhile, below are some tips on installing IOSD MARK II.

- For phantom 2, weld the BATT cable to the corresponding positive and negative terminals on the center board(red for positive and black for negative). After that, connect one end of the CAN cable on corresponding interface of IOSD and the other end to the CAN interface on the undercarriage. Please refer to the enclosed user manual to get needed instructions.

After all the tests, please kindly send the testing data to me via network disk.

Looking forward to hearing from you!

If you have any questions, please feel free to contact with me separately !

Thanks & Best regards!


Ryan
DJI Innovations
http://www.dji.com

 

[PhotomanJ]

Re: Fixing DJI's Compass Problem AND FLYAWAY PLUS CRASH.

I still have the yah to the by 10-12 degrees in the northwest US in stick forward. Hasn't been anything I can't adjust for when I am flying FPV. I have been flying almost everyday for a couple of months now. Alway do a complete check and a test flight(10ft) to check stability and all functions. Also always make sure I wait 30 seconds to lock gps plus check screen to have more than 8 satellites.
Then, today a was doing a normal pattern flight and was flying it back to land.(always with FPV monitor to verify direction and all iosd info). About 90 seconds from touchdown the P2 lost forward momentum with stick forward. With new batteries that were double checked before flight in controller. The P2 started positive roll to the right and started its flyaway, but then it flipped over and dropped right into the trees. All on video.
With the yah issues and now the flyaway, flip over and crash. I will not be able to fly it like I always do because I can't have that happen when there are other people around. None of my videos have been up uploaded, but I am going to try and get this one up due to the crazy turn of events. Gave no precursor that anything like this could happen and on a flight path I have flown more time than I can put a number too with not one issue in previous flight.

Just want to see if anyone else has had this happen in conjunction with this yah issue. I am addicted to flying and need to understand how to avoid this ever happening again.

 

[Lenny7118]

Hello Everyone....

I have been selected to participate in the beta program to help fix the compass issues here on this forum. I just thought you might want to know that yesterday, I downgraded my firmware to 1.08 (from 2.0), and my phantom flight is much improved. The j-hook is still there in the first few seconds of flight, but it disapears quickly. My altitude is more consistent, and my spins are much more centered. I have the iosd-mark2 on my phantom, so i am going to record some flights and export them into google maps (i think this is the program) where we should be able to see a graphical view of the flight. I am then going to put firmware 2.0 back on, and do the same thing. My only issue is i installed my iosd inside my phantom, so i dont have access to the usb port. I never planned on using the recording feature, but now..... I am going to try and install an extension cable out and down one leg to make it easy to access. Anyhow, you may want to try downgrading.... It is a major improvement as far as i am concerned.

David

Hi, I'm just going to be setting up my P2 that a got a few months ago. I didn't set it up back than because of the bad winter I had in my area.
My question is should I leave it on 1.8 if it's on that firmware because if the current Compass problems ?

 

[tom3holer]

I have a new P2Z3 to replace my P2Z2 and after upgrading to the latest S/W I did some tests and am cautiously optimistic with the results.
I was seeing all the classic problems, Toilet Bowl, J hooking and not flying in a straight line.

I tried this on the old P2 by offsetting the compass and had marginal results.
This time I offset the compass about double the variation in my area. You offset the compass to the left if toilet bowl is clockwise and visa versa. I offset it about 25-30 deg. After that I did a compass calibration which I had not dome before and this is again with the newest S/W. Then much to my surprise it was operating perfect. Normally when going forward at a good clip if I centered the pitch stick to stop it the Phantom would stop but make a rather J hook to the right and to the left if backing up. Now it just stops and no hook. The toilet bowl effect is also completely gone.

Hope this helps for those that live in areas of high variation.

Tom

Finally got a chance to fly in an open area to test the compass offset.

Although it was still windy gusts to maybe 20Kts I had a short flight of 5 min.
The Phantom tracked straight and even after a few minutes of flying it had no Jhook when coming to a stop after a fast forward stretch. I am very pleased and hopefully tomorrow will get more time to fully explore the change.

Tom

 

[Bigbells]

Hi, I'm just going to be setting up my P2 that a got a few months ago. I didn't set it up back than because of the bad winter I had in my area.
My question is should I leave it on 1.8 if it's on that firmware because if the current Compass problems ?

You mean 1.08, right?

If I were you I'd definitely leave it with firmware version 1.08.

I've "upgraded" my firmware to version 1.08 after experiencing the difficulties of versions 2.0 and 3.0.

I don't know of any good reason to deal with the hooking and Toilet Bowl Effect caused by a programming error in 2.0 and 3.0. Until that error is corrected, it's a lot easier for me to use the last version that did NOT have that problem, rather than messing with rotating the compass.

The only thing I miss about the flawed versions is that my maximum descent speed is now back to 6 meters per second instead of the safer 3 meters per second that was programmed into 2.0 and 3.0.

 

[Lenny7118]

Hi, I'm just going to be setting up my P2 that a got a few months ago. I didn't set it up back than because of the bad winter I had in my area.
My question is should I leave it on 1.8 if it's on that firmware because if the current Compass problems ?

You mean 1.08, right?

If I were you I'd definitely leave it with firmware version 1.08.

I've "upgraded" my firmware to version 1.08 after experiencing the difficulties of versions 2.0 and 3.0.

I don't know of any good reason to deal with the hooking and Toilet Bowl Effect caused by a programming error in 2.0 and 3.0. Until that error is corrected, it's a lot easier for me to use the last version that did NOT have that problem, rather than messing with rotating the compass.

The only thing I miss about the flawed versions is that my maximum descent speed is now back to 6 meters per second instead of the safer 3 meters per second that was programmed into 2.0 and 3.0.

Yes 1.08 firmware

 

[xtonex]

ive had this issue with 1.08, 2.0 and 3.0 its not the new firm causing whats being talked about here. it may have gotten worse for some but mine is the same as it was day 1

 

[landonkk]

ive had this issue with 1.08, 2.0 and 3.0 its not the new firm causing whats being talked about here. it may have gotten worse for some but mine is the same as it was day 1

Ditto. I think it may have gotten worse, but I went back and checked the video of my first flight with 1.08 and it is evident the j-hook issue is there. I didn't notice it right away because this was my very first time flying a quad and i was concentrating on other things, but after checking the video I am convinced the issue is across all firmware versions.

 

[ianwood]

Ditto. I think it may have gotten worse, but I went back and checked the video of my first flight with 1.08 and it is evident the j-hook issue is there. I didn't notice it right away because this was my very first time flying a quad, but after checking the I am convinced the issue is across all firmware versions.

This has been my assumption. A lot of people said that it was fine in the beginning but then got worse. I am pretty sure most people like and you and me didn't notice it as we were to busy learning to fly. And only after we got comfortable flying did we notice the hook.

I still haven't used 2.0 or 3.0 yet but will be this weekend to do my testing for DJI.

 

[Bigbells]

Ditto. I think it may have gotten worse, but I went back and checked the video of my first flight with 1.08 and it is evident the j-hook issue is there. I didn't notice it right away because this was my very first time flying a quad, but after checking the I am convinced the issue is across all firmware versions.

This has been my assumption. A lot of people said that it was fine in the beginning but then got worse. I am pretty sure most people like and you and me didn't notice it as we were to busy learning to fly. And only after we got comfortable flying did we notice the hook.

I still haven't used 2.0 or 3.0 yet but will be this weekend to do my testing for DJI.

I will be VERY surprised if you don't find 2.0 or 3.0 far worse then 1.08 when it comes to hooking and swirling. For that reason I also am looking forward to learning the numbers/evidence you come up with one way or the other.

 

[jkramer5]

Fairfield california here, drifts left in forward flight and bad toilet bowl. Also poor altitude drift in a hover.

 

[obiwan_pierogi]

I turned the compass (clock wise as viewed from the top) = to the + 17 degree inclination.. Definitely worked way better, than what it is now-- having returned the setting back to flat against the leg

 

[obiwan_pierogi]

just out of curiosity, during calibration.. If you rotate the compass on the leg, it should be ok to calibrate when you turn the copter on the flat plane, as nothing has changed; however, when you turn it on edge, should you hold the copter, so it would be slightly rotated, so the compass's long edge would still sit parallell towards the ground?.. Ie: Instead of the two bottom prop arms being equidistant from the ground, one would be slightly closer

 

[John Shaw]

It would be helpful if those who rotate their compass on the leg would comment on performance latter in the flight and make a comparison. I believe the current thinking of many is that when you rotate the compass it helps initially but later in the flight the drifting starts. If with no rotation you drift to the left, with a rotated compass that appears to compensate for this initially, will later in the flight start drifting to the right. So it would be very helpful if compass rotation testers would watch performance later in the flight just as critically as at the beginning. (Not a part of the official DJI testers plan).

 

[Andy T]

I have the drifting problem as well so have been watching this thread and trying to figure out what the solution might be and I came across this interesting tidbit,

http://wiki.dji.com/en/index.php/Naza-M ... ode_Flight

It is not straight forward on the phantom to turn the gps. I opened mine up and discovered my GPS sensor was cracked.

 

[sar104]

I have the drifting problem as well so have been watching this thread and trying to figure out what the solution might be and I came across this interesting tidbit,

http://wiki.dji.com/en/index.php/Naza-M ... ode_Flight

It is not straight forward on the phantom to turn the gps. I opened mine up and discovered my GPS sensor was cracked.

The orientation of the GPS module is irrelevant in the Phantom because it does not contain the compass sensor.

 

[Andy T]

I have the drifting problem as well so have been watching this thread and trying to figure out what the solution might be and I came across this interesting tidbit,

http://wiki.dji.com/en/index.php/Naza-M ... ode_Flight

It is not straight forward on the phantom to turn the gps. I opened mine up and discovered my GPS sensor was cracked.

The orientation of the GPS module is irrelevant in the Phantom because it does not contain the compass sensor.

You have come to an interesting conclusion. The compass plugs directly into that thing and the document from dgi says it is relevant.

 

[sar104]

I have the drifting problem as well so have been watching this thread and trying to figure out what the solution might be and I came across this interesting tidbit,

http://wiki.dji.com/en/index.php/Naza-M ... ode_Flight

It is not straight forward on the phantom to turn the gps. I opened mine up and discovered my GPS sensor was cracked.

The orientation of the GPS module is irrelevant in the Phantom because it does not contain the compass sensor.

You have come to an interesting conclusion. The compass plugs directly into that thing and the document from dgi says it is relevant.

The compass plugs into it to send it compass data. Those data are unaffected by the orientation of the GPS module because the compass is unaware of the GPS module orientation. The standard (non-Phantom) Naza units integrate the compass into the GPS puck, and those units are therefore sensitive to orientation, as indicated in the document that you referenced.