Fixing DJI's Compass Problem

[vrso2]

having same issue.

V1.1.1, in ATT does not hold altitude, in GPS no altitude hold as well and flights all over. Have green flashing, did Advanced calibration so many times, MOD 1554 and all values in good range. I replaced all 4 motors, and just ordered new compass.
Getting tired of it.

 

[dano7602]

I live in NY and am experiencing the issue. In GPS mode as soon as I let go of the controls when flying forward the P2 drifts right and the opposite when flying directly backwards. Would love to try the beta firmware to see if it solves the problem. I would be willing to record the iosd data using beta firmware to see if it is working in north eastern US which has high declination.

 

[xtonex]

having same issue.

V1.1.1, in ATT does not hold altitude, in GPS no altitude hold as well and flights all over. Have green flashing, did Advanced calibration so many times, MOD 1554 and all values in good range. I replaced all 4 motors, and just ordered new compass.
Getting tired of it.

This has nothing to do with whats being talked about in this thread...

 

[tom3holer]

For those that live in higher Dec areas that indeed are having big time compass problems check out the compass rotation thread(s).

Tom

 

[grahamgee]

I experienced this the last time I flew. I thought it may have been a calibration error, as I hadn't done it after a minor accident. I took it out for a test flight and there was a SLIGHT drift and it wouldn't hover in GPS.

I'm on the East coast of Australia, just north of Sydney.

 

[AnselA]

I experienced this the last time I flew. I thought it may have been a calibration error, as I hadn't done it after a minor accident. I took it out for a test flight and there was a SLIGHT drift and it wouldn't hover in GPS.

I'm on the East coast of Australia, just north of Sydney.

Was this with firmware 3.05? These "reports" are getting confusing...

 

[sar104]

I experienced this the last time I flew. I thought it may have been a calibration error, as I hadn't done it after a minor accident. I took it out for a test flight and there was a SLIGHT drift and it wouldn't hover in GPS.

I'm on the East coast of Australia, just north of Sydney.

Was this with firmware 3.05? These "reports" are getting confusing...

3.05 reports are over in this topic, and clearly marked as such.

 

[vrso2]

having same issue.

V1.1.1, in ATT does not hold altitude, in GPS no altitude hold as well and flights all over. Have green flashing, did Advanced calibration so many times, MOD 1554 and all values in good range. I replaced all 4 motors, and just ordered new compass.
Getting tired of it.

This has nothing to do with whats being talked about in this thread...

Oooops, sorry.

So what is going on with my Phantom than ?

 

[hamishue]

Same problem here in Vancouver, Canada. Can't wait for new firmware.
I drift to the left when I first start off in GPS and when I come to a stop it hooks left. Usually corrects itself after about 10 mins or at least gets better

 

[Erwan]

Any news when the firmware should be out ?!


Envoyé de mon iPhone à l'aide de Tapatalk

 

[varmint]

Any news when the firmware should be out ?!

Just a guess:

 

[Dadcat]

I think I've found out why it's taking so long for DJI to figure this out. They must have assigned the same guys that picked the list of No-Fly zones.

Here's the Del Bonita/Whetstone International Airport (Category B):

https://maps.google.com/?ll=48.997902,- ... 4&t=h&z=15

Can't see an airport in satellite view here? I'll give you a hint. It's the 4400 foot long grass strip, between the fields, perpendicular to highway 213. In 2009 there were 55 takeoffs and landings. That's more than 1 a week - clearly there's a need for a No-Fly zone in the same category as Portland International Airport.

Obviously they included every airport with "International" in it's name so there are whole bunch like this. In the U.S. and Canada, any airport where you can get customs/immigration service (usually by appointment) is an international airport. Scrolling the map on DJI's website http://www.dji.com/fly-safe/category-mc , it looks like they may have found all of them.

And just so you don't get to thinking this is off-topic, the magnetic declination at Del Bonita/Whetstone International is 13.5 degrees positive. This suggests another fun fact: The runway numbers there are 7 and 25. If the runway numbers were assigned today they would be 8 and 26. That's because the magnetic north pole has moved a lot since they were named and runway numbers are assigned based on their magnetic compass direction. In fact, there are many runways on the coasts of North America whose numbers are no longer correct, unless they were changed.

Magnetic declination here is about 15.72 degrees positve. When I started learning to fly, 50 years ago, declination was 6 degrees greater (almost 22). The earth's magnetic field is constantly changing, which is one reason to let the calibration function that's already in the Phantom be the method to correct for magnetic declination.

If the machine just saved that data, none of us would have ever noticed a problem.

 

[ianwood]

Borderline relevant at best. And hell freezing over? Cute. Everyone needs a not so gentle reminder that DJI is working on the problem. Ask yourself how many times Apple, Microsoft, Samsung, or other significant tech companies have paid any attention to user forums (especially unofficial ones) and then actually started working on a real fix creating ad hoc test groups to actively work the problem. I can count dozens of defects with products from all those companies where they just ignored it.

So you may have several reasons for bitching about DJI and I am not going to stop you but this thread and others related to it are not set up for griping. They are set up for figuring out ways to fix the problem, hence the title. Please stay on topic.

 

[mikeve]

Hi

I appreciate that DJI is working hard on releasing 3.05 to fix the compass problem on the Phantom 2s but we need DJI to be working to apply the same fix to 4.02 on the Phantom 1.

My Phantom 1.1.1 with fw 4.02 exhibits the very same characteristics. Off course straight flight and j-hooking that corrects after about 7 minutes. Leaving 3 minutes of good behaviour. I also get drifting with yaw.

 

[Dadcat]

So you may have several reasons for bitching about DJI and I am not going to stop you but this thread and others related to it are not set up for griping. They are set up for figuring out ways to fix the problem, hence the title. Please stay on topic.

I wasn't bitching about DJI. My opening was intended to be humorous. I was using that as a lead-in to point out some things that I haven't seen brought up before and that I think are perfectly relevant to an understanding of compass issues.

I don't see how it's "off-topic" to point out (again) what is likely the simplest and most direct solution to the problem we are all experiencing. The self calibration works really well. You, yourself, said the beta firmware doesn't remember the data between flights. No one has ever explained why it doesn't do that.

 

[Andy T]

So you may have several reasons for bitching about DJI and I am not going to stop you but this thread and others related to it are not set up for griping. They are set up for figuring out ways to fix the problem, hence the title. Please stay on topic.

I wasn't bitching about DJI. My opening was intended to be humorous. I was using that as a lead-in to point out some things that I haven't seen brought up before and that I think are perfectly relevant to an understanding of compass issues.

I don't see how it's "off-topic" to point out (again) what is likely the simplest and most direct solution to the problem we are all experiencing. The self calibration works really well. You, yourself, said the beta firmware doesn't remember the data between flights. No one has ever explained why it doesn't do that.

No worries. Ian sometimes forgets whose side he is on or he is on the DJI payroll. Your comment was perfectly fine

 

[Dadcat]

The other day I had an opportunity to fly a friend's week and a half old Phantom Vision+ along with my own. It was no surprise that his Phantom's compass error issues were identical to mine, J-Hook and all. What surprised me was that he hadn't really noticed it. The thing I find most troubling is when I'm trying to move in just a little bit closer to an object, it drifts out of the frame.

I was also able to compare speeds between mine on fw 3.00 and his on 3.04 on back to back flights. Compared to what I'm used to, I felt his flew like a pig. The 33% top speed reduction, slower max descent rate and a little bit slower response on the right stick with 3.04 was even more noticeable than I thought it would be. I'll live with those limits when the new firmware gets done, which I'm sure will be pretty soon.

On one run, soon after powering on, I flew mine about 1000 feet away at maximum speed and that hook when I let off the stick was really impressive. It went at least 100 yards left and embarked on what looked like about a 25 yard diameter toilet bowl circle. I'm ready for the update.

Ian, please pardon any off-topicness.

 

[Andy T]

What surprised me was that he hadn't really noticed it. The thing I find most troubling is when I'm trying to move in just a little bit closer to an object, it drifts out of the frame.

Yeah it took me a few days to notice the J and drift issues when I first started playing with my Phantom. It is a bit frustrating when someone in the same area as you says "mines fine". As for the drift when hovering, DJI specs say 2.5 meters horizontal tolerance. Mine is like 7 or 10 meters. It would take a pile of cash to shut my griping up after dealing with these clowns. They have been useless from the day I took delivery in any dealing I have had with them. Dealer antics have even been shady. ugh!

 

[robby666111]

On one run, soon after powering on, I flew mine about 1000 feet away at maximum speed and that hook when I let off the stick was really impressive. It went at least 100 yards left and embarked on what looked like about a 25 yard diameter toilet bowl circle. I'm ready for the update.

I've noticed the same. The longer the forward run, the bigger the hook.

 

[chinezbrun]

Hi, i noticed the same behavior. In my place the declination is not so high (Europe/Romania) but the symptoms are there. Drift to the left and TBE.