What's new
Van's Air Force

The definitive Van's Aircraft support community! Buying, building or flying an RV? Join our exclusive family of mentors and enthusiasts!

Struggling to complete G3X magnetometer calibration test

Prescott Flyer

Prescott Flyer

Active Member
I had to put in a replacement GSU 73 and have completed the AHRS alignment and Magnetometer Interference tests. But the Magnetometer Calibration test repeatedly fails. I believe this is for two reasons. Either the area I'm attempting to perform the test is not free from magnetic interference, or I am doing something wrong during the test. Here's a rundown of what I've done so far.

Aircraft: RV-9A
Equipment: G3X (GSU 73, GMU 44, GDU 370 (2), GDU 375)

Attempt #1 (last week)
  • Location: On my ramp (tarmac) about 30 feet away from a row of Port-a-Port hangars, and about 20-25 feet away from a couple of tied-down Cessnas
  • Wind: Breezy, occasionally gusting over 15mph
  • Engine running
  • Started the test pointed North.
    • Immediate failure was reported due to the aircraft not being still
  • I waited a few moments then restarted the test
    • Test procedure would not progress beyond the initial instruction to hold position

Attempt #2 (last week)
  • Location: Compass Rose
  • Wind: Breezy, occasionally gusting over 15mph
  • Engine running
  • Started the test pointed North
    • The test procedure would not progress beyond the initial instruction to hold position. I waited for 5 minutes, but I never got the instruction to turn the aircraft.

Attempt #3 (last night)
  • Location: On my ramp (tarmac) about 30 feet away from a row of Port-a-Port hangars, and about 20-25 feet away from a couple of tied-down Cessnas
  • Wind: calm
  • Engine OFF
  • With the help of an assistant in the cockpit, we started the test and it proceeded as expected, alternating between holding and turning the aircraft, roughly every 30 degrees.
    • Test calibration repeatedly failed over many (6+) attempts with "Calibration Error".
When performing the test with the engine off my assistant noticed the following:
  • Often where we were instructed to hold it didn't align closely with the next expected 30-degree position (usually we'd be past the expected magnetic heading)
  • We tried to slow the rate of turn as the dots "counted down" to the next expected hold position. But if I slowed the turn down too much, the dots would reset and start the countdown over. It makes me think I'm doing something wrong with how I'm turning the aircraft.
Maybe my presence at the front of the aircraft is having some influence? I emptied my pockets of anything metallic. I tried the test holding onto the prop at the hold position points, as well as keeping a few feet away from the nose of the aircraft until it was time to continue the turn. The prop is composite so it shouldn't be conducting any static from my person, right?

We plan to repeat the test with the engine off again but over the Compass Rose this evening to rule out magnetic interference that could be causing the failure on the ramp.

Any advice on things to check or do differently next time is appreciated!

Thanks!
 
IMHO you should not attempt this with only 30’ clearance from a steel hangar. Remember the earth’s magnetic field is incredibly weak. Also, do you have any headsets close to the sensor? (If you have a wet compass, watch what happens if you bring a headset close - the compass will swing way off. The headsets have small magnets in them). Since you cannot avoid having some small magnetic fields in the plane (because you have currents flowing) you should do the tests as described - engine running, avionics and other electrical stuff ‘on’. Double check that all high current wires are routed away from the sensor. Ordinary shielding (metal braid) will not help. However, a ‘twisted pair’ can greatly reduce any magnetic fields generated, as long as the pair really flows all the ‘hot’ current and all the ‘ground’ current, e.g., there is no ground path thru the airframe.
 
BobTurner said:
IMHO you should not attempt this with only 30’ clearance from a steel hangar. Remember the earth’s magnetic field is incredibly weak. Also, do you have any headsets close to the sensor? (If you have a wet compass, watch what happens if you bring a headset close - the compass will swing way off. The headsets have small magnets in them). Since you cannot avoid having some small magnetic fields in the plane (because you have currents flowing) you should do the tests as described - engine running, avionics and other electrical stuff ‘on’. Double check that all high current wires are routed away from the sensor. Ordinary shielding (metal braid) will not help. However, a ‘twisted pair’ can greatly reduce any magnetic fields generated, as long as the pair really flows all the ‘hot’ current and all the ‘ground’ current, e.g., there is no ground path thru the airframe.
Click to expand...
When you are attempting the test with the engine off, are you pulling the aircraft in a circle or pivoting on a wheel?

I think the test is designed to go in a circle…
 
Prescott Flyer said:
I had to put in a replacement GSU 73 and have completed the AHRS alignment and Magnetometer Interference tests. But the Magnetometer Calibration test repeatedly fails. I believe this is for two reasons. Either the area I'm attempting to perform the test is not free from magnetic interference, or I am doing something wrong during the test. Here's a rundown of what I've done so far.

Aircraft: RV-9A
Equipment: G3X (GSU 73, GMU 44, GDU 370 (2), GDU 375)

Attempt #1 (last week)
  • Location: On my ramp (tarmac) about 30 feet away from a row of Port-a-Port hangars, and about 20-25 feet away from a couple of tied-down Cessnas
  • Wind: Breezy, occasionally gusting over 15mph
  • Engine running
  • Started the test pointed North.
    • Immediate failure was reported due to the aircraft not being still
  • I waited a few moments then restarted the test
    • Test procedure would not progress beyond the initial instruction to hold position

Attempt #2 (last week)
  • Location: Compass Rose
  • Wind: Breezy, occasionally gusting over 15mph
  • Engine running
  • Started the test pointed North
    • The test procedure would not progress beyond the initial instruction to hold position. I waited for 5 minutes, but I never got the instruction to turn the aircraft.

Attempt #3 (last night)
  • Location: On my ramp (tarmac) about 30 feet away from a row of Port-a-Port hangars, and about 20-25 feet away from a couple of tied-down Cessnas
  • Wind: calm
  • Engine OFF
  • With the help of an assistant in the cockpit, we started the test and it proceeded as expected, alternating between holding and turning the aircraft, roughly every 30 degrees.
    • Test calibration repeatedly failed over many (6+) attempts with "Calibration Error".
When performing the test with the engine off my assistant noticed the following:
  • Often where we were instructed to hold it didn't align closely with the next expected 30-degree position (usually we'd be past the expected magnetic heading)
  • We tried to slow the rate of turn as the dots "counted down" to the next expected hold position. But if I slowed the turn down too much, the dots would reset and start the countdown over. It makes me think I'm doing something wrong with how I'm turning the aircraft.
Maybe my presence at the front of the aircraft is having some influence? I emptied my pockets of anything metallic. I tried the test holding onto the prop at the hold position points, as well as keeping a few feet away from the nose of the aircraft until it was time to continue the turn. The prop is composite so it shouldn't be conducting any static from my person, right?

We plan to repeat the test with the engine off again but over the Compass Rose this evening to rule out magnetic interference that could be causing the failure on the ramp.

Any advice on things to check or do differently next time is appreciated!

Thanks!
Click to expand...
I had a similar issue with a new GMU 22 magnetometer. Couldn't even get it to calibrate on a compass rose on a calm day. I finally gave up and got a refurbished GMU22 (original was out of warranty, even though it was unused) from Garmin and it calibrated on the first try.
 
BobTurner said:
IMHO you should not attempt this with only 30’ clearance from a steel hangar. Remember the earth’s magnetic field is incredibly weak. Also, do you have any headsets close to the sensor? (If you have a wet compass, watch what happens if you bring a headset close - the compass will swing way off. The headsets have small magnets in them). Since you cannot avoid having some small magnetic fields in the plane (because you have currents flowing) you should do the tests as described - engine running, avionics and other electrical stuff ‘on’. Double check that all high current wires are routed away from the sensor. Ordinary shielding (metal braid) will not help. However, a ‘twisted pair’ can greatly reduce any magnetic fields generated, as long as the pair really flows all the ‘hot’ current and all the ‘ground’ current, e.g., there is no ground path thru the airframe.
Click to expand...
The GMU44 was working correctly with the previous GSU 73 installed in 2013. The only thing that has changed is the installation of a "newly overhauled" GSU 73 Garmin sent me. I was getting the correct heading indications on the GDU 37x displays before I ran the AHRS alignment test. After completing the ARHS alignment test the Magenetomer Calibration test must be completed.

I'm the 2nd owner of the aircraft so I can't speak to specifics of how everything was wired up, other than the panel was designed and wire harnesses done by Stein Air. At this point, I don't have reason to suspect a wiring issue with the GMU44.
 
I once struggled trying to figure out magnetometer calibration issues on a plane...
Turned out the magnetometer was perpendicular to the aft fuselage side... in other words about 10 degrees off of correct fore/aft alignment.
Took a while to notice that!
To paraphrase Sherlock Holmes... if I explain it to you, you'll say that's obvious!
 
So many of the "usual suspects" have been mentioned, but..: Your own body won't affect the calibration. If the asphalt apron you mention is overlaid over concrete, there may be rebar in the concrete. The steel buildings sound close (30') -- I've only calibrated farther than that, don't know if 30' is too close. Diameter of your turn should not make a difference (unless it takes you close to some EM field). You might contact [email protected] -- I had repeated failures of an accelerometer, they concluded (correctly) it was the instrument & replaced it (successfully). Good luck!
 
Quick update...I thought I was going to have better luck completing the calibration pulling the plane in a circle (as opposed to pulling it around in a tight pivot by the nose wheel). But we had no luck on our ramp. Fired the engine up to taxi over to the compass rose and once again my G3X lost AHRS, ADC, and EIS after powering up the master avionics and PFD1 and PFD2. Sat in the plane for 2 hours in the hangar the following day and noted Network Error Rates zooming to 100% when powering up the MFD and PFD2, with PFD1 in config mode. Talked to Stein and he believes that since I just replaced the GSU 73, the issue is likely in the CAN bus wiring, or one of my 13-year old GDUs has a problem. So back to troubleshooting. Luckily I captured screenshots and videos of the anomalies I observed while trying different combinations of turning screens on and off. I documented it all and emailed the gory details to G3Xpert.

I'll post another update in a few weeks when I know more.
 
Hate to tell you this, but the replacement GSU 73 I got a few months back lasted exactly 2.3 flight hours before it went belly up over Klamath Falls as I was headed to do a DAR assignment in Oregon. Fortunately I had installed a GSU25 to give me AoA a couepl fo years ago, so I still had a good AHRS - just lost engine data. With help from a friend, we fairly quickly determined that indeed the “new” GSU was dead, and G3Xpert had a new one on the way to me in less than 24 hours. Bottom lien is that if you bought a “refurbished” one (the new ones are like $7K, so I suspect you got a refurb), they are not perfect.
 
Yes I got a “newly overhauled” unit in exchange. This one hasn’t even been installed for an hour of ground time. I need to see if I can conclusively rule out the CAN bus and GDUs. If I can’t I’ll ask for another replacement and try that.
 
Here's a quick status update...
  • I've done more "wiggle" testing of the wires connected to the GSU and GDU harnesses. We could not correlate any movement of wires with inducing failures or network error rates on the CAN bus.
  • Disassembled the GDU connectors and confirmed all wires were seated and pins were "tight"
  • CAN bus resistance measured 60 ohms (normal).
  • Consulted a second time with SteinAir and Garmin. I sent them the video I took demonstrating the problem. The Garmin engineer we spoke with thinks, frankly, in all likelihood, that the replacement GSU 73 I received probably failed shortly after I installed it. So today, I put it in the mail and am awaiting a 2nd replacement unit.
  • I also had a local Garmin tech that maintains avionics for Embry-Riddle's fleet take a look at my issue. He also could not determine the cause on his first visit. In his experience with G1000s, on more than one occasion replacing the config module solved a similar problem.
So the next steps will be to install the 2nd replacement GSU 73 and see if that fixes things. If not, I'll consider installing a new config module. If that doesn't work, we'll set some hours aside to see if we can find a short in the wiring.

This is not how I had hoped to spend my summer. Grounded for 2 months now :(
 
I had similar calibration problems for dozens of attempts. Apparently these units are extremely sensitive to pointing exactly N before you start the calibration. I solved by lining up 2 whiskey compasses on a steel bar 18 inches apart with spinner and rudder. Voila!
Also, you have to tow it in a 30 ft diameter circle
Greg Beckner
RV 14A built flown sold
RV 14 in storage waiting on Lycoming
RV8 about 70%
 
Having swung many GMU's I only rarely have problems.
As far as pointing north, well I go out to the spot where there used to be a compass rose and point it where I think north should be, no precision here but close enough.

I have found tail draggers (GMU angle), windy conditions and rough running engines can sometimes cause issues because the airplane is bouncing around.
The one thing I think has largest effect is either magnetized structures, or magnetic interference caused by buildings or rebar.
An inexpensive handheld compass is your friend for locating disturbances.
 
Walt said:
I have found tail draggers (GMU angle), windy conditions and rough running engines can sometimes cause issues because the airplane is bouncing around.
The one thing I think has largest effect is either magnetized structures, or magnetic interference caused by buildings or rebar.
Click to expand...
Walt speaks the truth, these are the most important points.

It's typically fine to do it with the engine off, if necessary... easier with two people participating, and if you have a slider canopy, don't forget to close it! I also like to use a rope instead of a steel towbar, or just push on the airframe.
 
Prescott Flyer said:
Hmmm well mine was open when we attempted the calibration test. I’ll try it closed next time. Do you have experience with the test failing while it is back vs passing when closed?
Click to expand...
Well, think about the amount of steel that moves closer to the magnetometer when you have the canopy open, then think about how weak the earth's magnetic field is...
 
Yes it makes perfect sense the steel frame being closer to the GMU could be the source of interference, and I’ll be sure to try it when I get my replacement GSU back.
 
Prescott Flyer said:
Hmmm well mine was open when we attempted the calibration test. I’ll try it closed next time. Do you have experience with the test failing while it is back vs passing when closed?
Click to expand...
Different scenario, no doubt...mine is a Dynon magnetometer (part of the ADAHRS) mounted right behind the baggage compartment bulkhead, but it was uncalibrateable with the slider canopy open.
 
Well I am thrilled to report that after installing a 2nd replacement GSU 73 my airplane is back flying and after about 5 hours so far so good.

I was able to calibrate the magnetometer on my ramp about 35 feet from my port-a-port hanger and 35 feet from two parked Cessnas. I did close the canopy this time but I don’t know if that made a difference or if the previous replacement GSU was to blame. Either way, I pulled it around in a circle and it passed on the first try.

So both Paul Dye and myself received a faulty replacement GSU 73 and had to get a 2nd one. What are the odds?
 
You must log in or register to reply here.
Back
Top