Mid Air "Compass Error, Exit P-GPS Mode"

[quid246]

In Phantom 3 Release note:
https://dl.djicdn.com/downloads/phantom_3/en/Phantom_3_Professional_Release_Note_en_20160613.pdf
Firmware update that touches the compass is this one:
--------------------------------------------------------------------------
2016.3.15 Overview： 1. All-in-One firmware version updated to：v1.7.0060 2. DJI GO app iOS version updated to: v2.7.0 3. DJI GO app Android version updated to: v2.7.1

What’s New： 1. Improved compass accuracy.
2. Fixed issue of image being corrupted in rare occasion.

Notes：  The firmware can be downgraded to v1.6.0040.
---------------------------------------------------------------------------

Can the latest firmware be downgraded, does anyone know?

Think you can only go back one version, that rules most of us out.

Now the release notes say they improve compass accuracy... was there any evidence that it was not accurate before this?

Funny that the compass "fix" came on March 15... same day P4 was released. Sounds fishy.

 

[zgembit]

Think you can only go back one version, that rules most of us out.

Now the release notes say they improve compass accuracy... was there any evidence that it was not accurate before this?

Funny that the compass "fix" came on March 15... same day P4 was released. Sounds fishy.

I have no direct evidence but in one thread on dji forum pages, regarding phantom 2 and compass error questions, dji representative states that compass has been made more sensitive for interference in one of the firmwares. I did link to that thread earlier in this one.

However, its totally fishy to not adress this in P3, makes you think that they want oss to step up to more expensive phantom. I have P4 and it has no such issues, for now...

 

[flyNfrank]

I have no direct evidence but in one thread on dji forum pages, regarding phantom 2 and compass error questions, dji representative states that compass has been made more sensitive for interference in one of the firmwares. I did link to that thread earlier in this one.

However, its totally fishy to not adress this in P3, makes you think that they want oss to step up to more expensive phantom. I have P4 and it has no such issues, for now...

I'm not sure what the exact details are, but P4's have also been experiencing some firmware related issues. You probably have seen some of this if you visit our P4 forum on this website.

 

[Meta4]

Now the release notes say they improve compass accuracy... was there any evidence that it was not accurate before this?

None at all ... DJI often dress up the words in firmware details to sound very positive.
It could be a language difference and/or a marketing thing.

I have no direct evidence but in one thread on dji forum pages, regarding phantom 2 and compass error questions, dji representative states that compass has been made more sensitive for interference in one of the firmwares.

However, its totally fishy to not adress this in P3, makes you think that they want us to step up to more expensive phantom. I have P4 and it has no such issues, for now...

This is irrelevant. DJI made some minor change to something compass firmware related back before the P3 was released.
So if it was important, it would have been included in P3 firmware right from the start.
And P3 Phantoms have been flying straight ever since they were released.
All this conspiracy stuff is nonsense and not helpful.
Almost everyone with a P3 does not have compass problems.
If you have ongoing compass issues, either:
You are doing something wrong ... or
There is something wrong with your Phantom
It's that simple.

 

[flyNfrank]

None at all ... DJI often dress up the words in firmware details to sound very positive.
It could be a language difference and/or a marketing thing.

This is irrelevant. DJI made some minor change to something compass firmware related back before the P3 was released.
So if it was important, it would have been included in P3 firmware right from the start.
And P3 Phantoms have been flying straight ever since they were released.
All this conspiracy stuff is nonsense and not helpful.
Almost everyone with a P3 does not have compass problems.
If you have ongoing compass issues, either:
You are doing something wrong ... or
There is something wrong with your Phantom
It's that simple.

Flight log data I have looked over with the most recent firmware has shown there is issues.

 

[zgembit]

None at all ... DJI often dress up the words in firmware details to sound very positive.
It could be a language difference and/or a marketing thing.

This is irrelevant. DJI made some minor change to something compass firmware related back before the P3 was released.
So if it was important, it would have been included in P3 firmware right from the start.
And P3 Phantoms have been flying straight ever since they were released.
All this conspiracy stuff is nonsense and not helpful.
Almost everyone with a P3 does not have compass problems.
If you have ongoing compass issues, either:
You are doing something wrong ... or
There is something wrong with your Phantom
It's that simple.

Well I can not see how you contribute helping to solve this issue? I think what you are writing is nonsense. If you dont have same problem or issue do not bother reading.

I have problem with P3P and vendor is not willing to even take it in for look because they already sent a couple of other P3-s with same problem to some head-service and they have no idea what it is. I spoke with them just Today!. They can only lock in your drone for weeks and you only can wait. That do NOT make me think everything is just fine with 2 weeks old drone that costs lots of money. So we search answer.

 

[Meta4]

Well I can not see how you contribute helping to solve this issue? I think what you are writing is nonsense. If you dont have same problem or issue do not bother reading. ... So we search answer.

I was attempting to be helpful by pointing you in the right direction and save you wasting your effort on unrelated and irrelevant conspiracy theories.

 

[Kamloops]

None at all ... DJI often dress up the words in firmware details to sound very positive.
It could be a language difference and/or a marketing thing.

This is irrelevant. DJI made some minor change to something compass firmware related back before the P3 was released.
So if it was important, it would have been included in P3 firmware right from the start.
And P3 Phantoms have been flying straight ever since they were released.
All this conspiracy stuff is nonsense and not helpful.
Almost everyone with a P3 does not have compass problems.
If you have ongoing compass issues, either:
You are doing something wrong ... or
There is something wrong with your Phantom
It's that simple.

There may be a bad batch of P3's or a bad firmware version, but for you assume that we are doing something wrong contributes to the conspiracy theory.

So far we DJI, (for the most part) the vendors, the 'knowledgeable' members here have the attitude of DENY, DENY, DENY any possibility of a bunch of owners having real issues, only that we all cannot calibrate a compass properly.

 

[Meta4]

There may be a bad batch of P3's or a bad firmware version, but for you assume that we are doing something wrong contributes to the conspiracy theory.

Read my post again
I did not assume anything.
I said ...
either:
You are doing something wrong ... or
There is something wrong with your Phantom
If you aren't doing something wrong .... then THERE IS SOMETHING WRONG WITH YOUR PHANTOM
And if there is something wrong with your Phantom, wasting time and effort of irrelevant conspiracy theories won't fix it.
Talk to DJI support if you want to address the issue.

 

[quid246]

I think the real litmus test would be to get a P3A/P (Amazon anyone?) that still has old firmware... calibrate and test it in a control location. If no errors, upgrade firmware, re-fly in the same area, notate if any errors or not, and then recalibrate the compass and notate if any errors or not.

If it errors, back to Amazon it goes. If it doesn't, back to Amazon it goes.

 

[whoreable]

My 2nd phantom advanced is flying perfectly, my first phantom however is a POS. Took her out in same spot as my new phantom and she had a compass error from liftoff which caused her to rock back and forth upon take-off and i almost thought it was going to crash as it rocked 3 feet to left and down.

2nd phantom is running firmware from amazon, 6.0.40

 

[zgembit]

I was attempting to be helpful by pointing you in the right direction and save you wasting your effort on unrelated and irrelevant conspiracy theories.

I cannot see that there is some full-blown conspiracy theory taking place in this thread.
Yes, I really get it that something is wrong with the drone The problem is that you aint getting much help from Dji or vendor where you bought it, so you are little bit on your own.

 

[quid246]

My 2nd phantom advanced is flying perfectly, my first phantom however is a POS. Took her out in same spot as my new phantom and she had a compass error from liftoff which caused her to rock back and forth upon take-off and i almost thought it was going to crash as it rocked 3 feet to left and down.

2nd phantom is running firmware from amazon, 6.0.40

Any indication that the two birds are from the more or less the same manufacture date?

 

[BudWalker]

My 2nd phantom advanced is flying perfectly, my first phantom however is a POS. Took her out in same spot as my new phantom and she had a compass error from liftoff which caused her to rock back and forth upon take-off and i almost thought it was going to crash as it rocked 3 feet to left and down.

2nd phantom is running firmware from amazon, 6.0.40

I'd like to take a look at that flight with your first phantom. From your description it sounds like something different than the recent compass errors that a lot of pilots have been experiencing. Would it be possible to retrieve the .DAT so that I could take a look? Go here to see how to retrieve it. The bottom half of that page contains advice on retrieving the right .DAT. It will be large and can't be posted here. You'll need to create a Dropbox link (or equivalent).

 

[AlexSP]

My oldest P3P is still running on older FW and my last compass cal was... I guess... 8 or 9 months ago, if not before that. I traveled many many times and far away sometimes (more than 100mi), and still never calibrated again. Zero compass issues on any of my DJI ACs so far. P3s have been flying straight since their inception indeed. So have the I1 and most other DJI ACs, in my experience. I agree that the only one capable to fix any compass issues (assuming everything is being done right by the pilot) is DJI.

That said, it's good practice to seek solution for a problem by isolating the possible causes. If the AC hasn't crashed, been mishandled, etc. then it may not be something physical with the compass or its internal systems. Next I'd look for what has changed since the problems started. A FW update seem to me like a serious potential, if not for that reason ("issue began after FW update" - I've noticed an increase but it could be coincidence or anecdotal, empirical) for the simple fact that these things rely heavily on FW for proper operation, and they're quite sensitive and precise.

 

[quid246]

Here's another data point... I remember reading people saying that the issues pop-up when the P3 faces west.

I took my bird up for quick flight, all was fine until I yawed to the West, Compass Error, it settled down after a few seconds and flipped back to P-GPS. Was able to do this several times in a row.

Back on the ground, I kept my P3A on and facing North... as I rotated to the West, Compass Error... it would settle as above flip back to P-GPS, but when I rotated back to North, once facing North it would give me an error then go back to P-GPS. Going to the East or South resulted in no error, only when rotating between North and West.

 

[BudWalker]

Here's another data point... I remember reading people saying that the issues pop-up when the P3 faces west.

I took my bird up for quick flight, all was fine until I yawed to the West, Compass Error, it settled down after a few seconds and flipped back to P-GPS. Was able to do this several times in a row.

Back on the ground, I kept my P3A on and facing North... as I rotated to the West, Compass Error... it would settle as above flip back to P-GPS, but when I rotated back to North, once facing North it would give me an error then go back to P-GPS. Going to the East or South resulted in no error, only when rotating between North and West.

Sounds kind of like what happened to this pilot

The red line is yaw; i.e. the P3's heading. MagYaw (green) is the heading derived from the magnetometers. Starting at 186 secs the P3 was rotated through 540 degrees. MagYaw should track Yaw but doesn't. In fact, the magYaw error depends on the actual heading (Yaw). When the heading is between -117 and 63 degrees the FC declares a COMPASS_ERROR_LARGE (shown by the red background)

 

[whoreable]

another perfect set of flights from the second phantom, did not even bring the first one today.

 

[quid246]

Sounds kind of like what happened to this pilot
View attachment 62554
The red line is yaw; i.e. the P3's heading. MagYaw (green) is the heading derived from the magnetometers. Starting at 186 secs the P3 was rotated through 540 degrees. MagYaw should track Yaw but doesn't. In fact, the magYaw error depends on the actual heading (Yaw). When the heading is between -117 and 63 degrees the FC declares a COMPASS_ERROR_LARGE (shown by the red background)

Thanks for the great tools.. here is something from a flight (I am guessing pre-flight due to the negative timeline). It's almost like a coding bug or something where the values are rolling over incorrectly...

And another flight (again wasn't looking for the compass error)...

 

[BudWalker]

Thanks for the great tools.. here is something from a flight where I encountered the compass error (looks like i was doing a pan around). It's almost like a coding bug or something where the values are rolling over incorrectly...

View attachment 62587

And another flight (again wasn't looking for the compass error)...

View attachment 62589

Yikes!! I keep mouse wheeling these images to try and zoom in on parts of it Any chance you could Dropbox (or equivalent) these .DATs?

Not sure what you mean by rolling over incorrectly.

In the first graph it doesn't appear that the Yaw/magYaw separation is dependent on the heading (Yaw). But this is before motorStart.

In the second graph it does appear that the Yaw/magYaw separation is dependent on the heading (Yaw).

You might want to look at gyro, accel, and magMod and see if there is any connection with the compass errors.