Fixing DJI's Compass Problem

[WL Production]

I'm flying a F450 made up of left over Phantom 1 parts (don't ask).
My FC is a Naza lite, but I'm using a Phantom GPS+compass.
I'm obviously getting magnetic declination issues (TBE is pretty bad).
No firmware update for the lite...
Any suggestions?
Thanks!

 

[apadua]

I just crashed my 3.06 phantom 2 Vision Plus today... will need replacement camera/gimbal. Expensive crash.

The issue was that i travelled... and got TBE again at the new location. Keep that in mind!

 

[AnselA]

I just crashed my 3.06 phantom 2 Vision Plus today... will need replacement camera/gimbal. Expensive crash.

The issue was that i travelled... and got TBE again at the new location. Keep that in mind!

Did you recalibrate the compass at the new location?

 

[John Shaw]

I just crashed my 3.06 phantom 2 Vision Plus today... will need replacement camera/gimbal. Expensive crash.

The issue was that i travelled... and got TBE again at the new location. Keep that in mind!

I think this is consistent with our best guess at what DJI did to fix the declination problem. They may have done other things but we did see that it went through a brief learning process. On the first runs I made with the second 3.05 beta I thought nothing had changed but quickly it got better and worked well ever after - even when I loaded 3.06 it carried over. But inherent with the learning feature is if you move to a significantly different declination it has to learn the new declination so I would expect to see TBE etc for the first few runs.
So your caution is valid.

 

[apadua]

Did you recalibrate the compass at the new location?

I did not. Failure on my part. It didn't ask for it and I forgot to do it. This could also be fixed in software... if the GPS sees it's far from previous flight, ask for calibration, i suppose.

 

[AnselA]

Did you recalibrate the compass at the new location?

I did not. Failure on my part. It didn't ask for it and I forgot to do it. This could also be fixed in software... if the GPS sees it's far from previous flight, ask for calibration, i suppose.

How "GPS" would know your previous location(s)?

 

[apadua]

The ideal solution would be for the drone to keep in flash my last home point coordinates and if it's more than a specified distance, force a calibration. If this isn't doable in the drone's software, it could also be done on my phone, although that would have a risk that if i chose to use a different device (ipad vs iphone, say), it wouldn't have the info available. And that's assuming the compass calibration was in fact a factor in this scenario.

 

[rrmccabe]

Problem in my eyes is it depends where you are in the world. Where I live I can drive hundreds of miles without issue. But other parts of world you can't go far. That said since they know where you are they should be able to calculate that.

Regardless its a good idea no matter how far you travel.

 

[HarryT]

Problem in my eyes is it depends where you are in the world. Where I live I can drive hundreds of miles without issue. But other parts of world you can't go far. That said since they know where you are they should be able to calculate that.

Regardless its a good idea no matter how far you travel.

Yes, magnetic declination is virtually zero (less than one degree, at least) in the UK, so compass recalibration is simply not necessary here unless you mess it up with exposure to strong magnetic fields.

 

[ianwood]

Yes, magnetic declination is virtually zero (less than one degree, at least) in the UK, so compass recalibration is simply not necessary here unless you mess it up with exposure to strong magnetic fields.

Calibration does nothing with respect to declination. Areas of low declination should follow the same compass calibration policies as anywhere else.

One caveat: it could be that in 3.06, the compass calibration will wipe the adaptions held in memory which would make sense.

 

[N017RW]

Problem in my eyes is it depends where you are in the world. Where I live I can drive hundreds of miles without issue. But other parts of world you can't go far. That said since they know where you are they should be able to calculate that.

Regardless its a good idea no matter how far you travel.

Yes, magnetic declination is virtually zero (less than one degree, at least) in the UK, so compass recalibration is simply not necessary here unless you mess it up with exposure to strong magnetic fields.

I agree. I recently traveled about 600 miles NW of my home but at a route almost exactly along the -6 deg declination in Southeast USA.

Upon arrival and flight testing, there was no need to run compass compensation. Same with my return home.

 

[apadua]

I live in -20 declination, and my trip was to -22.5. It was enough to hit a tree where I usually wouldn't have, during take-off and hover. Now i need the frickin' camera ribbon cable . yay, my next 100 page thread.

 

[ianwood]

I agree. I recently traveled about 600 miles NW of my home but at a route almost exactly along the -6 deg declination in Southeast USA.

Upon arrival and flight testing, there was no need to run compass compensation. Same with my return home.

Again, calibration has nothing to do with declination. It has to do with deviation. At 600 miles, you should definitely calibrate. Localized magnetic variances (i.e. deviation) can be no different or very different even if declination doesn't change.

 

[N017RW]

I agree. I recently traveled about 600 miles NW of my home but at a route almost exactly along the -6 deg declination in Southeast USA.

Upon arrival and flight testing, there was no need to run compass compensation. Same with my return home.

Again, calibration has nothing to do with declination. It has to do with deviation. At 600 miles, you should definitely calibrate. Localized magnetic variances (i.e. deviation) can be no different or very different even if declination doesn't change.

I appreciate your input Ian.
I use the flight characteristics to determine the need for compensation unless I have changed something on/in the a/c itself..
In my case there was no need so then the deviation is similar or below a/the threshold of variation in which it is necessary.

 

[ianwood]

Be careful, I've seen many videos where compass issues aren't obvious until well into the flight.

 

[N017RW]

Thanks again.
Spent a week in the Carolina mountains.

As a result,GPS was nearly unusable so I spent most of the time in Atti.

 

[John Shaw]

I am slowly trying to dig out the facts but I am getting suspicious we are barking up the wrong tree...
I know there are a number of pilots in our midst...
All aircraft have magnetic compasses, everything from a Cessna 150 to a Boeing 747...
How many of you calibrate your compass each day you are going to fly??? NONE
How many of you calibrate your compass whenever you fly to some destination even far away??? NONE
How many 0f you calibrate your compass after you roll the plane out of a metal hanger??? NONE
In general aircraft compasses are re-calibrated when the equipment on the airplane and therefore the vehicles magnetic field changes.
With my Phantom I calibrate occasionally, but not to often and I haven't seen any problems...
Some how I think we are barking up the wrong tree.
Comments Please.

 

[N017RW]

I am slowly trying to dig out the facts but I am getting suspicious we are barking up the wrong tree...
I know there are a number of pilots in our midst...
All aircraft have magnetic compasses, everything from a Cessna 150 to a Boeing 747...
How many of you calibrate your compass each day you are going to fly??? NONE
How many of you calibrate your compass whenever you fly to some destination even far away??? NONE
How many 0f you calibrate your compass after you roll the plane out of a metal hanger??? NONE
In general aircraft compasses are re-calibrated when the equipment on the airplane and therefore the vehicles magnetic field changes.
With my Phantom I calibrate occasionally, but not to often and I haven't seen any problems...
Some how I think we are barking up the wrong tree.
Comments Please.

If you like to dig into things like theses magnetometers then review the manufacturer's data sheets for them.
While I have not ID'd the specific one in the Phantom's, reviewing several indicates similar device properties.

You'll find compass calibration is done at time of manufacture and compensation is to nullify the 'hard and soft iron' distortions which occur due to nearby components and features of the device it's embedded or integrated into. This is the purpose of the 'dance''. Those nearby items rotate with the module allowing the distortions to be characterized on both X & Y axes and thus compensated for post procedure.

There are, IMO, a lot of superstitious rituals practiced by some because of coincidental this or thats. Some to the point of seemingly making flying more work that fun (for me).

Not here to argue or debate, my mind is made up.

But I've said before: TEHO & "if it makes you happy...".

 

[ianwood]

All aircraft cockpits I've been in (which granted is limited to a handful of Cessnas and Pipers and one G4) there is a magnetic deviation placard. These reflect the deviation caused by the plane and it's equipment and need to be adjusted from time to time, mostly when equipment is changed. Same with boats.

And don't airports provide local deviation information in the form of a compass rose?

It is expected the pilot should be able to calculate his/her actual magnetic heading based off of the deviation from the airplane and the deviation published at the local field. Those two things are the same as calibrating the compass on the Phantom.

 

[John Shaw]

All aircraft cockpits I've been in (which granted is limited to a handful of Cessnas and Pipers and one G4) there is a magnetic deviation placard. These reflect the deviation caused by the plane and it's equipment and need to be adjusted from time to time, mostly when equipment is changed. Same with boats.

And don't airports provide local deviation information in the form of a compass rose?

It is expected the pilot should be able to calculate his/her actual magnetic heading based off of the deviation from the airplane and the deviation published at the local field. Those two things are the same as calibrating the compass on the Phantom.

I at least partially agree... and thinking about the rest.
Your certainly right about the placard. (The compass has compensation adjustments which isn't a pilot task.) The placard covers what the adjustments can't compensate for. The compass is "swung" referencing magnetic north, not true north. The pilot does use the placard every flight as you describe.

I am not aware of the pilot using the airport compass rose other than for general awareness of where north is, but... your essentially right that the magnetic declination (variation) is on every aircraft navigation chart and the pilots use this on every nav calculation.

But you, more than anyone, knows that the compass calibration did not and could not take care of magnetic declination. Remember (are you sitting down) Operation Drunken Sparrow. We know that 3.06 doesn't learn it from the dance. It has to go through a brief learning process with TBE, JHook, etc (even earlier FW did this) and now it remembers what it learned and on subsequent flights it takes advantage of that - thanks to Ian and the beta testers. We have seen at least one report (apadua earlier) that if you move to a new location, do the dance, and fly you again get TBE, JHook etc until it learns this new location.

I haven't reached a final conclusion but the more I thought about it the more I thought we don't fully understand what the compass dance can do and what it can not do. Still looking for factual answers.