Fixing DJI's Compass Problem

South Florida here. same problem.

I've replaced my compass, re calibrated, magnetized, etc. always spins left an extra 10-15 degrees when turning left after letting go of the stick. Also wants to curve left running forward at higher speeds.

P2, h3-2d

Glad it's not just me.
 
Wind_Shear said:
South Florida here. same problem.

I've replaced my compass, re calibrated, magnetized, etc. always spins left an extra 10-15 degrees when turning left after letting go of the stick. Also wants to curve left running forward at higher speeds.

P2, h3-2d

Glad it's not just me.

You may have a different problem. Do you get the symptoms as described in this video?

[youtube]http://www.youtube.com/watch?v=9BUilGeH74Q[/youtube]

If so, I would suggest recreating them as closely as possible:

1. Fly out straight and back when first powered up.
2. Fly for a few minutes and repeat the forward/backward test.
3. Fly again for a few more minutes and then repeat tests one last time.

In the last test, the hook should be gone pretty much completely.
 
Add me to the list...

Also got the problem

[youtube]http://youtu.be/BaFLZThsMok[/youtube]
 
Great Video Ian! You really demonstrated and explained the issue as well as show how it eventually corrects itself.
Unfortunately it starts all over again when you shut down the Phantom. Totally sucks when I am shooting a video project and go through multiple batteries and have to waste valuable flight time waiting for it to do the self corrections.
 
Hi Ian - what is the best way to check updates on this issue? I know it's posted in multiple threads so wondered which one, or how is best to watch for developments?
 
wow i'm here in NJ with P2. too cold to test. Will test as soon as weather clears up a little.
 
BigBadFun said:
Hi Ian - what is the best way to check updates on this issue? I know it's posted in multiple threads so wondered which one, or how is best to watch for developments?

Right now, I am simply collecting information and building a list of people who are impacted by the issue. Once I get a couple more people, that'll give me the 50+ that I want to present to DJI as part of making my case for them to fix it.

I'll keep everyone posted on updates here. Hopefully it gets the right attention from DJI without much effort but if not, I may enlist everyone here to help make some noise!
 
My Phantom 2 just seems to fly about 20 degrees to the left of straight ahead and it's been driving me nuts.
I'm so glad to find this thread. I'm in Dunedin NZ, where the magnetic declination is 25 degrees, even more than Jared in Christchurch, so your explanation sounds fair to me. I experience a toiletbowl effect upon hovering after takeoff too, usually fixed with a twitch of the sticks, but it's nasty and distressing at low altitudes. Hoping DJI comes up with a fix for this real quick, it has cost me one job already! Oh yes, and trying to correct for some wicked TBAs has cost me a couple of crashes and a Zenmuse gimbal. It's a wonderful thing, but these flaws make it difficult to work with.
 
You can add me to the list. 40 miles east of SF
Mine actually hovers the best its ever done since I moved the compass. It also fly's perfectly straight, however it drifts slightly left when I stop but not as bad as it used to.
 
I just ordered a new P2 with the H3-3D gimbal. I'm hopeful that it will miraculously fly straight as an arrow but am assuming that I will likely be adding my third Phantom to your growing list.

How is the list coming? Have your 50 yet? I'd like updates as soon as you hear anything. I'm thinking of reaching out to some contacts soon myself.
 
We are up to 52 people after only 8 days. The new firmware is confirmed to not have any influence on the problem one way or the other. So, it is a good time to reach out to DJI and make sure they are addressing this issue. I am going to start working the channels. If anyone has any high-level contacts at DJI or one of their main dealers, please PM me. I am open to ideas as to how best to get their attention. My only request is that we keep our efforts coordinated. I have a couple ideas already but I want to try the direct route and see if DJI will address it.
 
I have flown my P2 only 3 times, but it will not stay in one place when hovering. I don't know if it's something I'm doing wrong but it is moving.
 
Have you included a link to this thread on their facebook page?

I live in an area with -14º declination, but I haven't had any TBE. Jury is still out for me on J-hooking because of too many windy days skewing my results, but I'm not suggesting the problem doesn't exist.

I think you're probably on to something that DJI possibly haven't accounted for declination in the firmware, and that there is some arguing going on between the compass and the GPS, but I still stand by the claim that the "calibration dance" is a less accurate way of calibrating the compass. I'm sure it's only a Hall-effect sensor in there, but when you revolve like that, you're running the compass through many more planes of potential interference or WHATEVER it might be that could be throwing the thing off. When you rotate on an axis, if there is interference of any kind, at least every orientation is calibrated with respect to the same level of that interference.
 
This is a defect. Of that I have no question. I am pretty sure it is a signing issue where in one or more parts of the code, instead of applying the right declination offset, it is applying the opposite. This theory is supported by the fact that shimming by double the declination is the only way to get to fly straight right from start up. And shimming by double the declination will lose it's effect and start to over correct over time as result of the adaption. Furthermore, only the P2 has this problem. The P1 is not affected.

Wind has no influence either. I've done it in a cross-wind, quartering-wind, following wind, and no wind and it's all the same. IMU and GPS do a good job cancelling out wind effects. Wind may slow down the adaption process or it may speed it up. Hard to tell.

As for the compass dance, I've tried it both ways numerous times. I've even tried calibrating counter-clockwise both ways. No difference. I am pretty sure with 52 people reporting the same issue it's not because 52 people who happen to all be in areas of high declination are all doing the compass dance as opposed to spinning it in place. I won't discount that maybe the sensitivity to local magnetic fluctuation may be higher doing the dance but I am confident that it has nothing to do with this issue.

Also, declination related TBE only happens when the Phantom is in area with a declination of +/- 18 or higher. The primary symptom is hooking and the inability to fly straight which happens where the declination is +/-10 or higher.
 
Hawaii here. Just got my P2 yesterday, put the new 2.0 firmware on it and upgraded everything. Did the compass dance and first flight this morning and I also got the hook to the left. It wasn't as pronounced as in the video though. It did also boomerang when coming back and did get better at tracking straight after a few minutes. This is in Honolulu where our magnetic declination is about +9 degrees.
 
Ianwood, you clearly know far more than I but wouldn't even just the slightest bit of wind do this? I'm in NY and Mass and I don't know that I ever have any hope of my P2 flying straight just because of the wind. Also, my P2 yaws back and forth like a pendulum even with my hand off the stick!

I'll try to check mine this weekend if the weather cooperates, but I think I'd need literally zero wind to check this, otherwise there would be no chance of a decent test.

EDIT: Have you reached out to Tahoe Ed about this?
 
adanac said:
Ianwood, you clearly know far more than I but wouldn't even just the slightest bit of wind do this? I'm in NY and Mass and I don't know that I ever have any hope of my P2 flying straight just because of the wind. Also, my P2 yaws back and forth like a pendulum even with my hand off the stick!

I'll try to check mine this weekend if the weather cooperates, but I think I'd need literally zero wind to check this, otherwise there would be no chance of a decent test.

I know it wasn't the wind for me, because the hook was into the wind. If the breeze was causing a drift in my situation, the hook would have been to the opposite direction, with the wind. Also with full GPS lock and low wind, the Naza software should compensate for wind drift.
 
The Phantom does very well dealing with the wind generally. Once the Phantom spends several minutes learning to fly straight, I can put it into a 30mph crosswind and it will track perfectly straight. Wind may influence the issue but it is not dependent on the wind. It happens in zero wind just as much.

I've pinged Tahoe Ed. Waiting to hear back.
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,094
Messages
1,467,598
Members
104,980
Latest member
ozmtl