GFC-500 Disconnecting

[Pinecone]

So, my GFC-500 has started to disconnect. Red flashing AP annunciator, tones. Re-engage and all is fine.

It started a few weeks ago on a trip, it kicked off while holding, but was the end of the flight. The flight home had no issues.

A couple of days later did a 2.5 hour flight and it kicked off 3 times in the descent. A short stop and return flight. About 2 hours into the flight, it started kicking off in level flight. Re-engage, a couple of minutes later, it disconnects again.

Sent log files to my avionics shop. They see the disconnects, but don't see a reason, so sending the log files to Garmin. But I figured to try the online brain trust.

I attached the last flight log file. G3X Touch, G-5, GFC-500 (and other things).

Log File.xlsx

[toto]

For whatever it's worth, I had a loose inspection panel that was rubbing against a control rod and causing a GFC500 disconnect on every flight.  After tightening the inspection panel, it has never happened again.  The avionics shop looked at it and they were a little bit dubious about my theory on the cause, but it went from disconnecting at least once on every flight to never disconnecting on any flight.

[Rick Junkin]

@Pinecone I’m having the same issue. It disappeared initially after updating to G3X Touch V9.17, but it’s back again. Same kind of thing where it occurs at about the 2 hour point +/- and I’m able to re-engage, but this repeats multiple times after that point. I sent in log files about 6 months ago and was told through email that they added my data to an engineering ticket that was already open from other reports of the same issue. I haven’t heard anything more back.

I’ll be compiling failure data from my trip to KOSH to send to Garmin again.

Let us know if you hear anything new from Garmin, I’ll do the same.

Cheers,
Junkman

[PT20J]

One thing to check is the pitch/roll calibration of the GMC 507. If the GMC AHRS doesn't agree with the PFD, it causes a monitor fault disconnect. 

[Pinecone]

Thanks.

Yes, it seems it is at the 2 - 2.4 hour point.

I will see about having the 507 recalibrated.  But wouldn't that show in the log?

[PT20J]

4 hours ago, Pinecone said:

Thanks.

Yes, it seems it is at the 2 - 2.4 hour point.

I will see about having the 507 recalibrated.  But wouldn't that show in the log?

When mine did this, the Autopilot State column showed: Fail / Attitude comparator / 00002004 00000800

[MikeOH]

Sorry, I don't own a GFC500 (and, likely never will) so I do not have anything specific to check.  My comments are more of a general troubleshooting nature:

1) Does Garmin publish (in the manual, or elsewhere) a complete list of conditions that will cause the AP to trip off?

2) Is this failure temperature related?  Example: your flights start in the morning and a few hours later things are hotter.  This could be mechanical (loss of clearance), or electrical (component drift with temp change).

3) When the AP trips off is the plane well trimmed?  I.e., maybe it's very sensitive to even a slight out of trim condition.

4) Kind of related to 2 and 3.  With everything off, how smooth and easy to move are all the controls? Compared to other similar aircraft?

 

Good luck!

[dzeleski]

Make sure you aren't accidentally knocking the the trim switch if you have one. It will immediately disconnect the AP.

[chunko]

Garmin has a SB for the -11 servos they used to use for the GFC500 and they're replacing them with the -21. The -11 has some kind of problem where it intermittently disconnects. I'm not telling you to go to your avionics guys and tell them "hey the internet said replace the -11 with the -21!!!", I just had it happen a few times this year already.

[Pinecone]

I should have -21 servos.  This is a new install this spring.

Might be temp related.  But for which part?  The GFC-507 should not be handing a lot of power.

I will check the logs.  I called Garmin and sent the logs to them to look over also.

I will pay attention to the trim next time.  But I think it was a trimmed a bit nose down.

[LANCECASPER]

On 8/1/2024 at 2:22 PM, Pinecone said:

So, my GFC-500 has started to disconnect. Red flashing AP annunciator, tones. Re-engage and all is fine.

It started a few weeks ago on a trip, it kicked off while holding, but was the end of the flight. The flight home had no issues.

A couple of days later did a 2.5 hour flight and it kicked off 3 times in the descent. A short stop and return flight. About 2 hours into the flight, it started kicking off in level flight. Re-engage, a couple of minutes later, it disconnects again.

Sent log files to my avionics shop. They see the disconnects, but don't see a reason, so sending the log files to Garmin. But I figured to try the online brain trust.

I attached the last flight log file. G3X Touch, G-5, GFC-500 (and other things).

Log File.xlsx 4.33 MB · 6 downloads

Did the disconnects start before or after the gains were changed on the autopilot?

[PT20J]

Assert log shows the autopilot state is Fail / System not usable when it fails.  I have no idea what causes that. I think @Rick Junkin had that failure indication and Garmin was looking into it.

[Rick Junkin]

15 hours ago, LANCECASPER said:

Did the disconnects start before or after the gains were changed on the autopilot?

I experienced disconnects early on right after my avionics were installed in Sep 2023, with the -21 servos and the original gain settings that were producing pitch oscillations. I believe that was with G3X software v9.15. I have v9.17 installed now with the alternate gain settings from the M20 J/K GFC 500 gain addendum and the disconnects are still happening inconsistently at the 2 hour point, +/-. Once the disconnects start happening I can re-engage the AP briefly and then it disconnects again. But it isn't consistent and I've had some flights where the AP operated normally throughout. BTW the J/K alternate gain settings smoothed everything out nicely.

14 hours ago, PT20J said:

Assert log shows the autopilot state is Fail / System not usable when it fails.  I have no idea what causes that. I think @Rick Junkin had that failure indication and Garmin was looking into it.

Garmin hasn't offered any feedback other than to say they are aware and have added my data to their existing engineering problem report on the issue. The Garmin folks at OSH had no new or useful information for me.

The log file data doesn't give any clues as to what is precipitating the disconnects. Garmin isn't sharing anything they may be looking into. My next step is to delve into the rest of the data recorded elsewhere on the G3X to see if I can find anything in the diagnostic discretes. That will be an extremely tedious exercise because to my knowledge there is no way to extract that data from the unit and I'll have to look at it on the G3X screen. I'm going to start with exporting the entire data history from the G3X (I think I know how to do that) to see if that data is more granular than the log file data, and then set to gathering the rest of the diagnostic data from the G3X screens.

I'll let y'all know if I discover anything. It will be a week or two.

EDIT: Has anyone figured out who the new "Trek Lawler" is at Garmin? I haven't had any luck identifying a technically astute G3X person who's able and willing to talk openly about how to get some of these problems resolved.

Cheers,
Junkman

[PT20J]

3 hours ago, Rick Junkin said:

Has anyone figured out who the new "Trek Lawler" is at Garmin? I haven't had any luck identifying a technically astute G3X person who's able and willing to talk openly about how to get some of these problems resolved.

I doubt there is one. Trek told me that he was fielding about 150 emails a day. His primary job was to support the dealer network, plus his team was responsible for trade shows. He was trying to be helpful by showing a presence on the various user forums. That took on a life of its own and became time consuming and often not very productive because some members of the various forums mostly like to complain and rant. (check out the G3X thread on BeechTalk). I don't think anyone at Garmin has the time or inclination to get this involved with customers via the internet on a regular basis. They would much prefer that end users use aviation support or go through their dealer network.

The best support I have gotten for the G3X/GFC 500 has come from g3xpert because they have the most experience with this equipment.

[PT20J]

7 hours ago, Rick Junkin said:

I experienced disconnects early on right after my avionics were installed in Sep 2023, with the -21 servos and the original gain settings that were producing pitch oscillations. I believe that was with G3X software v9.15. I have v9.17 installed now with the alternate gain settings from the M20 J/K GFC 500 gain addendum and the disconnects are still happening inconsistently at the 2 hour point, +/-. Once the disconnects start happening I can re-engage the AP briefly and then it disconnects again. But it isn't consistent and I've had some flights where the AP operated normally throughout. BTW the J/K alternate gain settings smoothed everything out nicely.

Garmin hasn't offered any feedback other than to say they are aware and have added my data to their existing engineering problem report on the issue. The Garmin folks at OSH had no new or useful information for me.

The log file data doesn't give any clues as to what is precipitating the disconnects. Garmin isn't sharing anything they may be looking into. My next step is to delve into the rest of the data recorded elsewhere on the G3X to see if I can find anything in the diagnostic discretes. That will be an extremely tedious exercise because to my knowledge there is no way to extract that data from the unit and I'll have to look at it on the G3X screen. I'm going to start with exporting the entire data history from the G3X (I think I know how to do that) to see if that data is more granular than the log file data, and then set to gathering the rest of the diagnostic data from the G3X screens.

I'll let y'all know if I discover anything. It will be a week or two.

EDIT: Has anyone figured out who the new "Trek Lawler" is at Garmin? I haven't had any luck identifying a technically astute G3X person who's able and willing to talk openly about how to get some of these problems resolved.

Cheers,
Junkman

There are two G3X log files. If you turn on data logging in configuration mode and keep an SD card in the GDU, it will log a .csv file for each flight (actually each time the G3X is powered up) with 1 sec updates. The G3X also keeps a internal log that can be exported as a .csv file that appears to have essentially the same data (I have not compared every field) but at finer time resolution. This is a single very large file (it takes several minutes to download to an SD card) and is ordered by date/time, not flights. Here are instructions for accessing it: Garmin_G3X_How_To_Export_Flight_Data.pdf

In config mode there is a tile at the bottom right for Diagnostics and if you open this there is a section for Autopilot that will display a log of every error occurrence with a date-time stamp. I didn't see a way to export that, but you should be able to take a screen shot of it if G3X screen shots is enabled.

 

[Rick Junkin]

1 hour ago, PT20J said:

There are two G3X log files. If you turn on data logging in configuration mode and keep an SD card in the GDU, it will log a .csv file for each flight (actually each time the G3X is powered up) with 1 sec updates. The G3X also keeps a internal log that can be exported as a .csv file that appears to have essentially the same data (I have not compared every field) but at finer time resolution. This is a single very large file (it takes several minutes to download to an SD card) and is ordered by date/time, not flights. Here are instructions for accessing it: Garmin_G3X_How_To_Export_Flight_Data.pdf

In config mode there is a tile at the bottom right for Diagnostics and if you open this there is a section for Autopilot that will display a log of every error occurrence with a date-time stamp. I didn't see a way to export that, but you should be able to take a screen shot of it if G3X screen shots is enabled.

 

Thanks Skip, that's my plan. I have data logging turned on and upload each of my flight data from the SD card to Savvy for graphical analysis/review. I exported the complete internal log once just to see what was in it but didn't see ongoing value in doing that. However now I'll export it again so I have a single file with all of the data for parsing. I'll use the diagnostic fault log to narrow my effort.

I'm not optimistic that I will learn anything useful from this, but I'll give it a go. If the internal log doesn't contain additional data fields beyond what is in the flight log files I doubt I'll find anything that helps me identify the cause of the disconnects, but I have to look.

It sure would be nice to have access to the design documents that define the parameters that generate the AP fail status I'm seeing. That would at least show me where to concentrate my efforts.

Cheers,
Junkman

[Pinecone]

My disconnects started after the gain change, but there were no long flights before the change.

And it did not start right after.  

I am wondering, if it is an AHARS issue if cycling the AP circuit breaker would reinitialize the AHARS and start another 2 hours of good operation.

[PT20J]

3 hours ago, Pinecone said:

My disconnects started after the gain change, but there were no long flights before the change.

And it did not start right after.  

I am wondering, if it is an AHARS issue if cycling the AP circuit breaker would reinitialize the AHARS and start another 2 hours of good operation.

Possibly if it is a drift issue. But the failure was not logged as a monitor fault. 

What does the diagnostic log screen for the autopilot say?

[Rick Junkin]

4 hours ago, Pinecone said:

I am wondering, if it is an AHARS issue if cycling the AP circuit breaker would reinitialize the AHARS and start another 2 hours of good operation.

I’m thinking the same. I observed the ball on the G3X wandering left and right about half a ball in the climb while the G5 ball remained solidly centered. I recorded some video to send to Garmin and my avionics shop.

[PT20J]

20 minutes ago, Rick Junkin said:

I’m thinking the same. I observed the ball on the G3X wandering left and right about half a ball in the climb while the G5 ball remained solidly centered. I recorded some video to send to Garmin and my avionics shop.

Resetting the AP circuit breaker would only reset the monitor AHRS in the GMC 507. This sounds more like an issue with the GSU 25D.  You can look at the ADAHRS data in the log (column BQ I think). It should be XXXXXXXX N100 N100 where XXXXXXXX is a code. If the N numbers ever go to D it indicates a degraded status. The G5 can also log to an SD card and it might be interesting to compare G5 logs and G3X logs for the same event.

I asked g3xpert what Fail / System not usable means and they said that Engineering has told them that the G3X software should be upgraded to the latest version if that occurs.

EDIT: Trek once told me that since the ADAHRS uses GPS aiding, GPS interference can sometimes cause problems with the ADAHRS. A way to check is to look at GPS status in flight and see if you still have 3D DIFF NAV.

Edited August 5, 2024 by PT20J
Added comment about GPS interference.

[Rick Junkin]

46 minutes ago, PT20J said:

I asked g3xpert what Fail / System not usable means and they said that Engineering has told them that the G3X software should be upgraded to the latest version if that occurs.

Interesting. I’ve seen Fail / System not usable with every software load so far, including the current v9.17. Thanks for the interpretation of the ADAHRS codes. I downloaded the internal G3X log today and will have a look. I’ll also take a look at what I can get from the G5 and it’s SD card to compare.

[Pinecone]

2 hours ago, Rick Junkin said:

I’m thinking the same. I observed the ball on the G3X wandering left and right about half a ball in the climb while the G5 ball remained solidly centered. I recorded some video to send to Garmin and my avionics shop.

Hm, not so good, hard to reset just the G3X in flight.

[Pinecone]

2 hours ago, PT20J said:

Resetting the AP circuit breaker would only reset the monitor AHRS in the GMC 507. This sounds more like an issue with the GSU 25D.  You can look at the ADAHRS data in the log (column BQ I think). It should be XXXXXXXX N100 N100 where XXXXXXXX is a code. If the N numbers ever go to D it indicates a degraded status. The G5 can also log to an SD card and it might be interesting to compare G5 logs and G3X logs for the same event.

I asked g3xpert what Fail / System not usable means and they said that Engineering has told them that the G3X software should be upgraded to the latest version if that occurs.

EDIT: Trek once told me that since the ADAHRS uses GPS aiding, GPS interference can sometimes cause problems with the ADAHRS. A way to check is to look at GPS status in flight and see if you still have 3D DIFF NAV.

OK, I see a column BU that is SFD AHRS Status.   It is a number (080187FF) and then two 3 characters that start with N or D

For about 8 seconds it is NXX D99.  XX ranged from 34 - 58, increasing each step.

The systems stabilized at N100 N100 at about 30 seconds.  Stays there for the rest of the flight.   Occasionally the first number would dip to 99 or 98.

I found the AP fails and looked at the pitch and pitch trim torques, but they are not out of line.   

I did see that at some times, the PTrim torque was as high as 22.  And that it would fly for a bit with the Pitch Torque as high as 8 without trimming.

[PT20J]

1 hour ago, Rick Junkin said:

nteresting. I’ve seen Fail / System not usable with every software load so far, including the current v9.17.

g3xpert mainly deals with the non-certified customers and the latest experimental version is 9.41 so perhaps this is something already addressed that hasn't made it to a certified release yet. 

[PT20J]

13 minutes ago, Pinecone said:

OK, I see a column BU that is SFD AHRS Status.   It is a number (080187FF) and then two 3 characters that start with N or D

For about 8 seconds it is NXX D99.  XX ranged from 34 - 58, increasing each step.

The systems stabilized at N100 N100 at about 30 seconds.  Stays there for the rest of the flight.   Occasionally the first number would dip to 99 or 98.

I found the AP fails and looked at the pitch and pitch trim torques, but they are not out of line.   

I did see that at some times, the PTrim torque was as high as 22.  And that it would fly for a bit with the Pitch Torque as high as 8 without trimming.

Did the D99 happen coincident with the AP Fail? If so, perhaps the GSU 25D is the problem rather than the autopilot. Without understanding the design, it's not possible to know, but maybe the "system" referred to in the failure message is the ADAHRS and not the autopilot. I could see the autopilot perhaps wanting to disengage if the ADAHRS was degrading but perhaps not degraded enough to cause a MISCOMP with the standby. Or maybe it is something external that affects both ADAHRS.

Were the disconnects in different locations to rule out GPS interference in some location?