AHRS / Heading Indicator Gremlin

[rwtucker]

. . . any thoughts on diagnosis if the problem is happening in the air, at full power? Rather than electrical, could it be a metallic or disturbance of some sort?

I receive the Garmin Tech Bulletins sporadically and I recall situations in which certain weird combinations of flight plan edits, etc. can cause a loss of heading or other indications. Software bugs. Are you working with the software when this happens? I'm stumped. Let me think about it.

[h20wrks]

The majority of the time I loose the heading indicator is when I turn onto to the runway. Their is no metal structures near the plane when this happens. I almost think that something shifts in the plane when I turn and accelerate the plane that causes my temporary interruption. I am leaning towards a connector issue.

Jimmy

[Chris B]

This is small comfort, but given the high vibration environment (especially with a 4-banger! ), large temperature/humidity swings and plethora of electrical connections, I am surprised that we don't see more issues like this.

But troubleshooting intermittent electrical issues is a PITA. Sorry you have to go through this!

Chris

[ThomasD]

I had a similar issue, but less frequent. I spent weeks with various avionics shops, swapping units, re-calibrating, doing compass swings - even took it to the Diamond factory in Canada and THEY didn't find anything.
In the end (after a ferry flight) I took it to a UK avionics shop who found - in less than an hour - that it was a simple problem with bad crimps. The wires could just be pulled out of the crimp connector by hand. They re-did the crimps and no more problem.
FYI, between the magnetometer and the rack there are 4 pairs of connectors - this many I guess to make the wiring easier.
So I suggest you try the basics first - get someone to physically look at all the connectors from the magnetometer -> inbound to the rack: check for good/bad pins and give them all the wires a pull, and get them to buzz out to check impedance is low in all singles wires with no low impedance crossovers.
Likewise get them to 'wiggle' the wires while the buzz box is connected - this is actually how they found my problem.
Not popular with avionics shops as some of them are terrible to get to and takes a lot of contortion and/or scope to see, but that is what is needed..

[Antoine]

I think Thomas' recommendations are very wise.
Generally, it seems that we have to take a hard look at the wiring and connectors. Most if not all are gold plated, which is why I like Thomas' ideas.
Even Garmin told me that they would rather not sell me a replacement magnetometer (after my in-flight IMC failure) before an avionics shop has checked the cabling.
Regarding the wing root connector - no idea why they placed it there and did not provide slack.
My new A&P found this one after I asked him to dig up "all three of them" for me.
He showed number four to me on the electrical schematic and asked me if I really wanted the wing out. I said no and have not regretted it so far. I had enough misery getting to the one in the back AFTER he had located it for me....

Robert: was it you who pointed me to Stablilant 22 - I apologize for not mentioning the source, but I no longer remembered. Great stuff.

Arpad - sorry no specific ideas for your situation...

[h20wrks]

I guess you guys are motivating me to get a good weather day and do a little checking on my own. Unfortunately most of my experience with the shops have been with parts changers instead of with true mechanics. This afternoon I signed up with Savvy Aviators. It will be interesting to see how they address this and my starter issues.

Thanks,
Jimmy

[ThomasD]

I guess you guys are motivating me to get a good weather day and do a little checking on my own.

Actually, I would suggest not doing it yourself unless you are an avionics/electronics engineer. It needs someone with all the right wiring diagrams, knowledge of correct pull strengths on the crimps, impedance meters, someone who knows when connector pins look and OK and when marginal etc. - i.e. a good avionics shop that will properly check out the wiring.

[h20wrks]

I have already payed for 7 shop visits to resolve my issue with no luck, and although avionics is not my background my last electric accomplishment I was able to fix an electrical gremlin in my track hoe after the dealer valued on 3 attempts.

[ThomasD]

I sympathise - I went through 4 avionics shops then even the Diamond Canada factory (who had the plane for 3 weeks, did prop balancing, flew the airplane with an engineer in the back with a laptop connected to the avionics rack etc.) - none of whom found the wiring issues because none of them had properly investigated the wiring.
(I too am a BSc electronics engineer and have access to some of the equipment needed, but not all, nor the wiring diagrams nor the other info, and I couldn't find the issue as I could not find all the connectors.)

What happened with me with avionics shop number 6 (IAE at Cranfield in the UK) was I said about all the LRUs that had been changed and the response was "well it must be something else then - start with the wiring". This seemed sensible and logical: if changing the LRUs does not fix the issue then it MUST be something else. The G1000 should work reliably.

My experience is most of the avionics shops assume it is the LRU at fault - often fixing on past experience of faulty units / service advisories to this effect - and change these as a first step rather than the 'less glamorous' and hard work of buzzing out the wiring. Having had my previous experience, with any avionics issue I would insist that an avionics shop starts with the wiring and thoroughly check this first - and if they won't I would go elsewhere.

If you are in Europe I would recommend IAE - they found the issue within an hour ( !!! ) after weeks elsewhere with other avionics shops.

Sorry not to be able to help more; good luck with it.

[rwtucker]

Actually, I would suggest not doing it yourself unless you are an avionics/electronics engineer. It needs someone with all the right wiring diagrams, knowledge of correct pull strengths on the crimps, impedance meters, someone who knows when connector pins look and OK and when marginal etc. - i.e. a good avionics shop that will properly check out the wiring.

I understand Thomas' point but I have a different take. If you are competent in working with wiring and connectors, including using the right size crimping tool should that turn out to be the problem, there is no reason why you cannot perform these checks yourself -- assuming that you are so inclined. Anyone who is careful and methodical can check the soundness of the electrical paths, including those I outlined in my first post. One step should be to measure the DC resistance from the magnetometer connector sleeve, all the way to the connector sleeves (ground) that connect to the LRUs. As was suggested, wiggle things a little to see if the resistance changes. It could be loose crimps, loose connectors (I found several on my brand new Diamond), connectors that need to be cleaned up by disconnecting and reconnecting, etc. Who knows. Just disconnect and re-connect everything in the path that the diagrams and your good judgment tells you is appropriate. Given the experiences of other DAN members, the LRUs themselves would be near the top of my list. They report many G1000 fixes simply by removing and reinserting, with and without Stabilant.

One small caveat, DC Ohmmeters present a voltage across their terminals. The voltage at play is low, typically 3, 6 or 9 VDC. Since I do not know how the internals of the magnetometer function, I recommend disconnecting it as you do your tracing; leaving it connected might invalidate some of your readings anyway.

Good luck and let us know! Ours could be next.