What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

GMU22 intermittent "AHRS magnetic anomaly"

K

kaa

Well Known Member
Hi all,

I've been getting weird behavior from GMU22 lately, and would love any advice on how to debug this.

The setup is G3X Touch + GMU22 and a G5 (in a Lancair). At some point last year we started seeing frequent heading swings and heading divergence warning with G3X differing by 10-20 degrees from G5. The problem was likely in the G3X side of things, because sometimes we could see that the synthetic vision picture was significantly rotated compared to the outside.

We did the G3X magnetic disturbance test and it came back at ~8%. After that we did a calibration and heading disagreement warnings disappeared. Instead we started getting intermittent "AHRS magnetic anomaly" warnings.

After we upgraded G3X Touch to the latest software the warnings disappeared for a bit. They are now back. They are pretty rare, don't last long and don't seem to be correlated with anything in particular, so I'm at a loss for how to debug this. G5 and G3X headings seem to be the same (within 1-2 degrees).

Do they show up in any G3X logs? Is there a way to figure out what kind of anomaly the magnetometer senses?
 
Typically this would be caused by something nearby producing a strong magnetic field, e.g. an electrical device or ferrous hardware that has been magnetized.
 
Thanks. This makes sense, but why would it only cause intermittent problems? The only electrical device nearby is a strobe, and turning it on and off doesn't seem to have any effect. I assume anything magnetic nearby should produce permanent interference.
 
When you calibrate the magnetometer, you're teaching it what the earth's magnetic field plus the magnetic field created by your aircraft looks like. The earth's magnetic field is weak and hard to detect compared to a strong local source of magnetism, which is why you initially had problems. Then when you repeated the calibration process without fixing the problem, you produced a calibration that doesn't make sense, which is why it sometimes complains; that's what the "magnetic anomaly" message means. What you need to do is fix whatever has become magnetized or is otherwise generating a magnetic field, or move the GMU to a better location.
 
It's external, until it's not...

Are there errors on the RS-232/RS-485 datapath? I know there is some measure of fault tolerance in these protocols but a flaky ground reference, corrosion on one/more of the pins in the connector might cause an issue upstream.

Ground integrity is important, and difficult to achieve in "plastic" (no offense, I love the Lancair) airplane.
 
mburch said:
When you calibrate the magnetometer, you're teaching it what the earth's magnetic field plus the magnetic field created by your aircraft looks like. The earth's magnetic field is weak and hard to detect compared to a strong local source of magnetism, which is why you initially had problems. Then when you repeated the calibration process without fixing the problem, you produced a calibration that doesn't make sense, which is why it sometimes complains; that's what the "magnetic anomaly" message means. What you need to do is fix whatever has become magnetized or is otherwise generating a magnetic field, or move the GMU to a better location.
Click to expand...

Thanks. Any suggestions on how to figure out what it is? The installation is according to G3X Touch Manual, using Garmin bracket.

Now that I think about it, there is some evidence of static discharge on the wing leading edge nearby (common problem). I wonder if it could be messing things up.

Are there errors on the RS-232/RS-485 datapath? I know there is some measure of fault tolerance in these protocols but a flaky ground reference, corrosion on one/more of the pins in the connector might cause an issue upstream.
Click to expand...
I'd expect this to cause communication failure, not a magnetic anomaly error. Unless data is not checksummed, which would be strange?
 
kaa said:
Thanks. Any suggestions on how to figure out what it is? The installation is according to G3X Touch Manual, using Garmin bracket.
Click to expand...

Typically the advice is to start by using a handheld compass to look for magnetized items nearby. Ferrous hardware, cables, control surface weights, electrical devices, etc.

kaa said:
I'd expect this to cause communication failure, not a magnetic anomaly error.
Click to expand...

I agree that this does not sound like a problem related to the communication between the magnetometer and ADAHRS units.
 
My 10Friend had a similar problem and tracked it to seat belt hardware and the stainless cables. We degaussed them using a magnetic tape eraser, recalibrated and it did not come back.
 
Thanks everyone. I ordered a compass and will check things out this weekend.

In the meantime, just out of pure curiosity, I don't think I fully understand the engineering side of things. I get that the magnetometer perceives H_earth + H_airframe, and that the calibration essentially infers H_airframe to further subtract it from the reading to get H_earth. So if calibration succeeds it would mean that H_airframe is not strong enough to saturate the sensor. Why then would it intermittently cause problems in real operation? Does this imply that H_airframe changes with time? Or does the sensor have different noise characteristics far away from zero, so that large H_airframe causes lower SNR?
 
Magnetometer Interference

kaa said:
Thanks everyone. I ordered a compass and will check things out this weekend.

In the meantime, just out of pure curiosity, I don't think I fully understand the engineering side of things. I get that the magnetometer perceives H_earth + H_airframe, and that the calibration essentially infers H_airframe to further subtract it from the reading to get H_earth. So if calibration succeeds it would mean that H_airframe is not strong enough to saturate the sensor. Why then would it intermittently cause problems in real operation? Does this imply that H_airframe changes with time? Or does the sensor have different noise characteristics far away from zero, so that large H_airframe causes lower SNR?
Click to expand...

One example I can provide is an aircraft in which the Magnetometer was mounted about 2 inches above the rudder cables (not a Vans airframe). The rudder cables caused interference when the rudder was deflected to the left, triggering intermittent magnetic anomaly messages. Relocating the magnetometer ultimately resolved the issue.

Thanks,

Justin
 
g3xpert said:
One example I can provide is an aircraft in which the Magnetometer was mounted about 2 inches above the rudder cables (not a Vans airframe). The rudder cables caused interference when the rudder was deflected to the left, triggering intermittent magnetic anomaly messages. Relocating the magnetometer ultimately resolved the issue.

Thanks,

Justin
Click to expand...

As Justin stated, one of the items few people check during an interference check is full travel of the elevator. It caught us once after the fact in flight, which prompted me to check another aircraft in diagnosing an issue on the ground. Both were cable related. (Aircraft with elevator counterweights are more obvious to check when installing, but rudder cables can be overlooked)

pb
 
We did the full magnetometer interference test from the installation manual (except the gear swing). The magnetometer is also installed in a wingtip, pretty far from anything moving, except an aileron. Anyway, I'll check this weekend with a compass.
 
It's a common misconception that passing the magnetic interference test guarantees a clean magnetometer installation. All this test tells you is whether the magnetometer is located near a changing magnetic field, not whether the aircraft's magnetic field is overwhelming the earth's. The latter can typically be avoided through good installation planning, good maintenance practices such as avoiding magnetized tools, and in the worst case relocating the magnetometer.
 
Magnetic Interference Test

kaa said:
We did the full magnetometer interference test from the installation manual (except the gear swing). The magnetometer is also installed in a wingtip, pretty far from anything moving, except an aileron. Anyway, I'll check this weekend with a compass.
Click to expand...

We do our best to provide comprehensive installation guidance in the manuals, including lessons learned from troubleshooting previous aircraft, but unfortunately you cannot replicate every aspect of the in-flight environment, on the ground. For this reason, sometimes these problems cannot be detected during the on-ground magnetic interference test.

Some common (intermittent) culprits in the wing are lights that are locally grounded and high current carrying conductors (pitot heat power/ground).

There are other possibilities however. You will want to use Table 15-4, which is listed on page 15-3 of the G3X Touch Installation Manual to help guide you. As you move up the list from the bottom, it is increasingly important to include separation from each line item, with electric motors and relays at the top of the list, requiring as much separation as feasibly possible.

Thanks,

Justin
 
When you see intermittent problems with lights, do these problems correlate with turning lights on and off? We have strobe / nav lights nearby, but there seems to be no correlation to switching them on and off. Perhaps we have a bad contact somewhere, I'll check.
 
mburch said:
It's a common misconception that passing the magnetic interference test guarantees a clean magnetometer installation. All this test tells you is whether the magnetometer is located near a changing magnetic field, not whether the aircraft's magnetic field is overwhelming the earth's. The latter can typically be avoided through good installation planning, good maintenance practices such as avoiding magnetized tools, and in the worst case relocating the magnetometer.
Click to expand...

But if the aircraft's magnetic field overwhelms the earth's, how does calibration pass?
 
I finally had time to take another look at the installation, and it isn't as great as I thought - a bunch of ferrous hardware around the magnetometer. Here is a picture of the worst offender (could also explain why this was transient since this connector wasn't properly secured).

Thanks everyone for the explanation!
 

Attachments

  • magnetometer.jpg
    magnetometer.jpg
    299.3 KB · Views: 164
kaa said:
I finally had time to take another look at the installation, and it isn't as great as I thought - a bunch of ferrous hardware around the magnetometer. Here is a picture of the worst offender (could also explain why this was transient since this connector wasn't properly secured).

Thanks everyone for the explanation!
Click to expand...

Well - that's a test instrument not in everyones tool box.

Well done!!
 
What is that instrument called.
Thanks
Dave
Click to expand...

Gauss meter.

My Samsung cell phone has magnetic sensors in it and I presume most other phones do as well. I installed a compass app that uses them to drive a compass dial.

It also displays a total mag field strength number displaying usually about 40 u teslas.

I wonder, in lieu of that nice instrument mentioned above, if some common cell phones with a compass app may be useful to track down these illusive magnetic gremlins?

ron
 
You must log in or register to reply here.
Back
Top