What generally causes IMU heading error and Compass Error. Exit P-GPS Mode?

[mouseno4]

This morning I decided to get a nice shot of the sunrise from my driveway, since I am unable to leave with this crisis going on.

I placed my MPP on the ground, turned on my RC, opened the DJI Go 4 app, started the drone and waited for it to give me the all green to lift off.

Once it did, I lifted off and started going straight up, no lateral movement. 45m up the drone started ''drifting'' off in an easterly direction. I then receive multiple errors at a distance of 45m laterally from where I took off. "IMU heading error", "Compass Error. Exit P-GPS Mode" along with "Weak GPS Signal" despite the drone locating 12, then a few seconds later, 14 satellites. The signal bars were 'greyed' out.

I had no idea what was going on, so I manually took control and flew it back (through a car park no less!) whilst descending at the same time.

Reviewing the recorded flight on the map, I can see the drone seemingly teleporting the 45m away instantly.

Please note, I cannot provide any location information given that I took off from my home and not from a public place! If I were not at home, I would have no issues ... But I would rather not openly reveal where I live!

 

[sar104]

It was most likely caused by magnetic interference at the takeoff point but, without the flight log, that's going to remain a guess.

 

[mouseno4]

It was most likely caused by magnetic interference at the takeoff point but, without the flight log, that's going to remain a guess.

Is there a way to share that log without giving out location details?

 

[sar104]

Is there a way to share that log without giving out location details?

No.

 

[mouseno4]

No.

Looking at my data on AirData.com, I can't find the reason for the issues. Why did the drone behave like it did. I have never experienced such compass/GPS failure before nor this weird fly off in the past.

I am at a loss to explain it. If it is in fact what I suspect it to be - there must be some serious amounts of magnetic interference in my driveway. Or huge quantities of metal in the concrete of the driveway to cause this.

 

[sar104]

Looking at my data on AirData.com, I can't find the reason for the issues. Why did the drone behave like it did. I have never experienced such compass/GPS failure before nor this weird fly off in the past.

I am at a loss to explain it. If it is in fact what I suspect it to be - there must be some serious amounts of magnetic interference in my driveway. Or huge quantities of metal in the concrete of the driveway to cause this.

Or a really local magnetic field just where you started it up. Magnetic distortion due to rebar in concrete, for example, can be very short range - just a few inches. The definitive data will be in the logs, but not in the data you were looking at on the AirData site.

 

[mouseno4]

Or a really local magnetic field just where you started it up. Magnetic distortion due to rebar in concrete, for example, can be very short range - just a few inches. The definitive data will be in the logs, but not in the data you were looking at on the AirData site.

Hmm. There is a decent size police and emergency services building within say, 100m of my home and where I took off. But I have never heard of police, fire or SES broadcasting anything that disruptive.

 

[sar104]

Hmm. There is a decent size police and emergency services building within say, 100m of my home and where I took off. But I have never heard of police, fire or SES broadcasting anything that disruptive.

Well that was an interesting segue from local magnetic interference to public service radio transmission. I can assure you that was not the cause, by the way.

 

[mouseno4]

Well that was an interesting segue from local magnetic interference to public service radio transmission. I can assure you that was not the cause, by the way.

True. But when all other evidence doesn't make sense, the far-fetched must also be considered.

 

[sar104]

True. But when all other evidence doesn't make sense, the far-fetched must also be considered.

The evidence makes perfect sense - your description is a commonly encountered problem. It just can't be confirmed without actual data.

 

[mouseno4]

The evidence makes perfect sense - your description is a commonly encountered problem. It just can't be confirmed without actual data.

To you. It makes no sense to me.

 

[Meta4]

True. But when all other evidence doesn't make sense, the far-fetched must also be considered.

No need to look at extremely unlikely scenarios when you've already provided one that is much more likely.

I am at a loss to explain it. If it is in fact what I suspect it to be - there must be some serious amounts of magnetic interference in my driveway. Or huge quantities of metal in the concrete of the driveway to cause this.

It doesn't take "huge quantities of metal" to cause the issue.
It could be cause by something as simple as a bolt or nail if it's close enough to the compass.

But as you said that you launched from your driveway, here are some X-ray images of what reinforced concrete structures have inside them.
Can you see a possible source of magnetic interference there?

 

[bigswifty]

Looking at my data on AirData.com, I can't find the reason for the issues. Why did the drone behave like it did. I have never experienced such compass/GPS failure before nor this weird fly off in the past.

I am at a loss to explain it. If it is in fact what I suspect it to be - there must be some serious amounts of magnetic interference in my driveway. Or huge quantities of metal in the concrete of the driveway to cause this.

If it is the usual suburban Australian driveway with rebar acc to Australian Standards, then it has more than enough ferrous material to cause a compass error. I fly from home quite a bit, and one of my drones won't even go into "Green" Flight Mode if I'm within a few yards of my driveway, let alone on it.

 

[mouseno4]

If it is the usual suburban Australian driveway with rebar acc to Australian Standards, then it has more than enough ferrous material to cause a compass error. I fly from home quite a bit, and one of my drones won't even go into "Green" Flight Mode if I'm within a few yards of my driveway, let alone on it.

There is no knowing what is in this concrete. I took off from this same patch of concrete this evening, despite this morning's excitement. When I powered the drone on this time, I held it up about chest height during it's self calibration and start up. I then repeated the same flight I performed this morning on the second attempt.

First attempt, straight from the concrete produced the fly away. Second attempt after a compass calibration, I achieved the desired 120m straight up, took my shots and came down. No more issues.

This evening, I did the exact same thing. I went up to 120m, got my shots and came down. No issues. Only difference is that this time, I powered on the drone from above ground level. Maybe it solved the issue, maybe no. No way to really tell.

 

[sar104]

There is no knowing what is in this concrete. I took off from this same patch of concrete this evening, despite this morning's excitement. When I powered the drone on this time, I held it up about chest height during it's self calibration and start up. I then repeated the same flight I performed this morning on the second attempt.

First attempt, straight from the concrete produced the fly away. Second attempt after a compass calibration, I achieved the desired 120m straight up, took my shots and came down. No more issues.

This evening, I did the exact same thing. I went up to 120m, got my shots and came down. No issues. Only difference is that this time, I powered on the drone from above ground level. Maybe it solved the issue, maybe no. No way to really tell.

Yes - it's very easy to tell - it's all in the flight logs.

 

[Meta4]

There is no knowing what is in this concrete.

If your driveway is covered in deep cracks, there's no steel in it.
Otherwise it's loaded with the stuff.

I took off from this same patch of concrete this evening, despite this morning's excitement.

As I said .. it doesn't take much steel, as long as it's close to your compass.
Look again at the reo mesh in the images above.
Imagine an inch of concrete above the mesh and you place your drone.
You might place it with the compass in the middle of a square and be safe.
Or it might be just above a crossing point in the mesh or an overlap and zap goes your initialisation.

Don't launch from reinforced concrete, it's just asking for trouble.
Incidents caused by doing that are common and usually end up more serious than yours did.

 

[Dave Maine]

This conversation should be closed. Real data exists in the flight logs. There is no basis for silly speculation.

 

[slup]

...When I powered the drone on this time, I held it up about chest height during it's self calibration and start up. ...

That is a very good procedure ... continue doing that, the risk for a wrongly initiated IMU yaw is during this start up phase, keeping the AC well above ground minimize the risk for something unknown in the ground to mess the initialization up. Once everything is on & self calibrations is done you can place it on ground for take off. Just take off watches, rings & bracelets from the hand/arm you hold the AC with ...