G3X Canbus issue GSA28

[walker172]

Howdy Braintrust,

I got the first engine start completed on my RV-10 build and almost everything is working great, but I’m having a canbus issue that I think is isolated to my autopilot servos.

My electrical system was wired and designed by steinair and is a full G3X system. With everything plugged in, I get network errors on the canbus and my roll autopilot servo isn’t even detected by the system. My pitch servo and yaw damper servo both also show hardware failures. The status lights on the servos themselves are solid green for roll (not detected by system), and solid red for pitch and for yaw damper.

I originally thought that it must be the canbus wiring from the roll servo since it’s the one not being detected. Of note, my canbus termination points are the roll servo and the yaw damper servo. Interestingly, if I put one of those canbus terminator plugs on the harness at the roll servo, all of my LRUs show green with 0 network error. The only errors I get in this case are the hardware failed pitch and yaw damper servos. I checked all of the breakpoints to the servos for wiring, and they are all correct. No mismatched pins.

I’m at a bit of a loss. My servos are old (-00) old, but have never been used. I’m working to replace them under the new SB, but Garmin won’t do a direct RMA to Canada, so I’m looking for a dealer to facilitate.

I feel like it’s unlikely that a servo itself is causing canbus network issues, but with it showing no error with the toll servo capped, I also feel like the wiring must also be good. Anyone seen a canbus issue caused by faulty autopilot servos? Garmin said that would be exceedingly rare, as would having 3 bad servos all at once (even with them all being old).

What do you pros think?

 

[Richard Connell]

I had many Garmin related issues with my RV10 initially.
I’ve wired 4 trouble free canbus g3x systems in my life and I was very diligent about testing each node as I built it.
I had significant AHRS issues which turned out to be vibration related.
All the way Garmin said it was my canbus. I was dubious but that was about the extent of support I received. I actually reworked my canbus in its entirety which was quite a task. Made no difference…
I still had sporadic network issues. Also seemed to emanate from the roll servo outboard too. Terminated at the GMU11 in the wingtip.
After a series of inflight AFCS failures Garmin replaced all my servos.
Low and behold not a single network error since.
So… yeah Garmin may not believe you, but I do

 

[mfleming]

I'm not an expert and hopefully someone much more versed in this can help.

My system had a similar problem. Check that all your can bus pins are in the correct slot. A hi and lo pin swapped is all it takes. Start with the most accessible plugs and work from there. Even Steinair tech are only human

 

[walker172]

I had many Garmin related issues with my RV10 initially.
I’ve wired 4 trouble free canbus g3x systems in my life and I was very diligent about testing each node as I built it.
I had significant AHRS issues which turned out to be vibration related.
All the way Garmin said it was my canbus. I was dubious but that was about the extent of support I received. I actually reworked my canbus in its entirety which was quite a task. Made no difference…
I still had sporadic network issues. Also seemed to emanate from the roll servo outboard too. Terminated at the GMU11 in the wingtip.
After a series of inflight AFCS failures Garmin replaced all my servos.
Low and behold not a single network error since.
So… yeah Garmin may not believe you, but I do

This gives me some hope. I have a hard time wrapping my mind around 3 bad servos. But if that’s all it is, that would be a pretty easy fix.

 

[walker172]

I'm not an expert and hopefully someone much more versed in this can help.

My system had a similar problem. Check that all your can bus pins are in the correct slot. A hi and lo pin swapped is all it takes. Start with the most accessible plugs and work from there. Even Steinair tech are only human

I’ve gone through all the servo breakpoints and they’re good. I may have to dig into the other LRUs, even though they are all working error free.

 

[N804RV]

Howdy Braintrust,

I got the first engine start completed on my RV-10 build and almost everything is working great, but I’m having a canbus issue that I think is isolated to my autopilot servos.

My electrical system was wired and designed by steinair and is a full G3X system. With everything plugged in, I get network errors on the canbus and my roll autopilot servo isn’t even detected by the system. My pitch servo and yaw damper servo both also show hardware failures. The status lights on the servos themselves are solid green for roll (not detected by system), and solid red for pitch and for yaw damper.

I originally thought that it must be the canbus wiring from the roll servo since it’s the one not being detected. Of note, my canbus termination points are the roll servo and the yaw damper servo. Interestingly, if I put one of those canbus terminator plugs on the harness at the roll servo, all of my LRUs show green with 0 network error. The only errors I get in this case are the hardware failed pitch and yaw damper servos. I checked all of the breakpoints to the servos for wiring, and they are all correct. No mismatched pins.

I’m at a bit of a loss. My servos are old (-00) old, but have never been used. I’m working to replace them under the new SB, but Garmin won’t do a direct RMA to Canada, so I’m looking for a dealer to facilitate.

I feel like it’s unlikely that a servo itself is causing canbus network issues, but with it showing no error with the toll servo capped, I also feel like the wiring must also be good. Anyone seen a canbus issue caused by faulty autopilot servos? Garmin said that would be exceedingly rare, as would having 3 bad servos all at once (even with them all being old).

What do you pros think?

I wouldn't discount the faulty servos. "Solid red" indication means internal fault. The older GSA28s caused a lot of headaches for the certified world. I work part-time for a flight school that has endless problems with them.

But, also, go back check that the ID straps are set correctly on the servo connectors. For Pitch, it should be a 4" jumper between pins 5 & 8. And for Yaw damper, it should be a 4" jumper between pins 6 & 7.

The "strap" numbers don't match the pin numbers. Also, its easy to get the auto-trim servos and AP servos "strapping" mixed up.

 

[Bavafa]

Do you have two terminator or just one installed? The CAN bus need to be terminated at both ends and this can be accomplished either by wiring or the terminator plug.

 

[mburch]

Have you followed the troubleshooting steps in section 2.3.1.3.6 of the installation manual?

 

[walker172]

I wouldn't discount the faulty servos. "Solid red" indication means internal fault. The older GSA28s caused a lot of headaches for the certified world. I work part-time for a flight school that has endless problems with them.

But, also, go back check that the ID straps are set correctly on the servo connectors. For Pitch, it should be a 4" jumper between pins 5 & 8. And for Yaw damper, it should be a 4" jumper between pins 6 & 7.

The "strap" numbers don't match the pin numbers. Also, its easy to get the auto-trim servos and AP servos "strapping" mixed up.

I did look at the ID straps and they appeared correct as well. But I’ll double check them again. Thanks.

 

[walker172]

Do you have two terminator or just one installed? The CAN bus need to be terminated at both ends and this can be accomplished either by wiring or the terminator plug.

I tried a terminator plug on both ends and it also cleared my canbus issues. I currently have a terminator on the roll servo (one end of the bus) and the other end (yaw damper) plugged into the servo. This also provides no network error, but the servo itself is a red light. In the G3X config menu, the yaw damper servo calls out a hardware and power supply fault, and the pitch servo calls out a hardware/clutch failure. The roll servo had never registered on the canbus, but it has a steady green light.

My servos were purchased in 2019 during the build of the tailcone and wings and haven’t seen use other than initial install. Sounds like the older servos have a high fail rate. Mine are all (-00) lots. Maybe some new upgraded ones will do the trick.

 

[N804RV]

I tried a terminator plug on both ends and it also cleared my canbus issues. I currently have a terminator on the roll servo (one end of the bus) and the other end (yaw damper) plugged into the servo. This also provides no network error, but the servo itself is a red light. In the G3X config menu, the yaw damper servo calls out a hardware and power supply fault, and the pitch servo calls out a hardware/clutch failure. The roll servo had never registered on the canbus, but it has a steady green light.

My servos were purchased in 2019 during the build of the tailcone and wings and haven’t seen use other than initial install. Sounds like the older servos have a high fail rate. Mine are all (-00) lots. Maybe some new upgraded ones will do the trick.

I think @Bavafa has nailed it, based on your description. Your CAN bus must be constructed such that there are two terminators, one connecting each end of the high and low 'legs'. The servos are considered "nodes" that "see" the bus in parrallel with respect to the terminators, never in series.

Also, it sounds like with the 2 terminator installed, you still have 2 out of 3 servos showing solid red status lights. Those are internal faults for sure.

 

[Richard Connell]

Just to add to this…
The servos can be jumped to be a canbus end without a canbus terminator but you can’t do both.
So in all of your trouble shooting you haven’t ended up with 2lots of temination accidentally?
To mburchs point before have you ohmed a random node in the network to make sure you’re getting the 60Ω and not 120 or 40Ω?

 

[Traash]

CAN issues can indicate at one or both of the terminations but actually exist elsewhere on the bus. The technique of depowering all LRU's on the bus except for the PFD and one other at a time can help identify the problem LRU (Described in the attached doc). View additional CAN data on the Configuration/Diagnostic page by holding in both "NRST" & "MENU" buttons at the same time while powering on the PFD.

What type wire is being used? MIL-C-27500will work for small busses. (<60' and not too many? LRUS's) but can become overwhelmed. CAN specific wire, as called out by Garmin, transfers data much better since it is manufactured to maintain the desired 120ohm impedence. This can be sourced from ACS, Stein, and the manufacturer.

A good explanation of CAN bus architecture is in the attached video.
https://youtu.be/YBrU_eZM110?si=Qvl7cvfxxSRRMMeh

 

[lr172]

Howdy Braintrust,

I got the first engine start completed on my RV-10 build and almost everything is working great, but I’m having a canbus issue that I think is isolated to my autopilot servos.

My electrical system was wired and designed by steinair and is a full G3X system. With everything plugged in, I get network errors on the canbus and my roll autopilot servo isn’t even detected by the system. My pitch servo and yaw damper servo both also show hardware failures. The status lights on the servos themselves are solid green for roll (not detected by system), and solid red for pitch and for yaw damper.

I originally thought that it must be the canbus wiring from the roll servo since it’s the one not being detected. Of note, my canbus termination points are the roll servo and the yaw damper servo. Interestingly, if I put one of those canbus terminator plugs on the harness at the roll servo, all of my LRUs show green with 0 network error. The only errors I get in this case are the hardware failed pitch and yaw damper servos. I checked all of the breakpoints to the servos for wiring, and they are all correct. No mismatched pins.

I’m at a bit of a loss. My servos are old (-00) old, but have never been used. I’m working to replace them under the new SB, but Garmin won’t do a direct RMA to Canada, so I’m looking for a dealer to facilitate.

I feel like it’s unlikely that a servo itself is causing canbus network issues, but with it showing no error with the toll servo capped, I also feel like the wiring must also be good. Anyone seen a canbus issue caused by faulty autopilot servos? Garmin said that would be exceedingly rare, as would having 3 bad servos all at once (even with them all being old).

What do you pros think?

You need to pull the covers off the connectors on the termination point devices (roll and yaw servo). These servos have termination options internally and they are accessed with jumper wires on the DB connector. If those jumpers are in there AND you used the terminator plugs, you are going to have too much resistance and will get canbus errors. There shoiuld only be two terminating resistors and need to be at the correct devices, but can't double up on those resistors. It is either the jumper or the plug, not both. Are you positive that stein wired it for those two termination points? The fact that everything else works fine tells me that maybe there are terminating resistors activated via jumpers on other devices that you are not aware of. Most, but not all, devices on the bus have a jumper option for terminating resistors. Look at your wiring diagram to comfirm the end points, then check every db connector and confirm that only the two end point devices have the termination jumper installed.

Read the manual for canbus testing. It will tell you wha the resistance should be when probing the canbus (120, IIRC) as well as likely termination problems if it doesn't match.

Larry

 

[walker172]

I think @Bavafa has nailed it, based on your description. Your CAN bus must be constructed such that there are two terminators, one connecting each end of the high and low 'legs'. The servos are considered "nodes" that "see" the bus in parrallel with respect to the terminators, never in series.

Also, it sounds like with the 2 terminator installed, you still have 2 out of 3 servos showing solid red status lights. Those are internal faults for sure.

I may have explained it poorly. I have a terminator plug in the harness at the roll servo since it isn’t being detected anyways, and the yaw servo is plugged in (it’s the other end of my canbus) so it is terminated at both ends currently.

 

[walker172]

You need to pull the covers off the connectors on the termination point devices (roll and yaw servo). These servos have termination options internally and they are accessed with jumper wires on the DB connector. If those jumpers are in there AND you used the terminator plugs, you are going to have too much resistance and will get canbus errors. There shoiuld only be two terminating resistors and need to be at the correct devices, but can't double up on those resistors. It is either the jumper or the plug, not both. Are you positive that stein wired it for those two termination points? The fact that everything else works fine tells me that maybe there are terminating resistors activated via jumpers on other devices that you are not aware of. Most, but not all, devices on the bus have a jumper option for terminating resistors. Look at your wiring diagram to comfirm the end points, then check every db connector and confirm that only the two end point devices have the termination jumper installed.

Read the manual for canbus testing. It will tell you wha the resistance should be when probing the canbus (120, IIRC) as well as likely termination problems if it doesn't match.

Larry

I have pulled apart the connectors for all servos. Everything is pinned correctly and I only have jumpers on the roll servo (one end of the bus) and the yaw servo (other end of the bus).

 

[lr172]

I have pulled apart the connectors for all servos. Everything is pinned correctly and I only have jumpers on the roll servo (one end of the bus) and the yaw servo (other end of the bus).

With the power off and all connectors attached, back probe the canbus. How many ohms?

 

[walker172]

You need to pull the covers off the connectors on the termination point devices (roll and yaw servo). These servos have termination options internally and they are accessed with jumper wires on the DB connector. If those jumpers are in there AND you used the terminator plugs, you are going to have too much resistance and will get canbus errors. There shoiuld only be two terminating resistors and need to be at the correct devices, but can't double up on those resistors. It is either the jumper or the plug, not both. Are you positive that stein wired it for those two termination points? The fact that everything else works fine tells me that maybe there are terminating resistors activated via jumpers on other devices that you are not aware of. Most, but not all, devices on the bus have a jumper option for terminating resistors. Look at your wiring diagram to comfirm the end points, then check every db connector and confirm that only the two end point devices have the termination jumper installed.

Read the manual for canbus testing. It will tell you wha the resistance should be when probing the canbus (120, IIRC) as well as likely termination problems if it doesn't match.

Larry

It’s interesting that with the jumper and the plug on, that actually eliminates the canbus network errors.

 

[Dugaru]

Old servos

They caused a lot of headaches in the experimental world also. Both of mine went tango uniform, one after the other. A mechanic I know said he had seen them croak on four different experimental airplanes.

I wouldn't discount the faulty servos. "Solid red" indication means internal fault. The older GSA28s caused a lot of headaches for the certified world. I work part-time for a flight school that has endless problems with them.

 

[lr172]

It’s interesting that with the jumper and the plug on, that actually eliminates the canbus network errors.

That would imply that the jumper is not installed correctly or there is some other issue.

 

[walker172]

I think for sure the servos need to be replaced. I’ll start there and see. Probably too tough to try to find multiple issues at once.

 

[Dugaru]

I'd love to see the data

Garmin said that would be exceedingly rare, as would having 3 bad servos all at once (even with them all being old).

Just noticed this. Are they willing to define "exceedingly rare"? After my first servo died, I asked that they replace both, as I had by that time learned of the rash of servo failures. They declined, assuring me the second one would be fine. It wasn't. At least when they sent me the second replacement, they didn't make me pay in advance.

I've set a calendar reminder so I remember to get my third set of servos before the five-year deadline.

 

[walker172]

They did not define it, no. They also said it would be very rare that a faulty servo would cause a canbus issue. The reason I think it has to be the roll servo is that it has never communicated with the canbus and I have quadruple checked the wiring from the roll servo onwards. If I remove the roll servo and put a terminator plug on its harness, everything else shows zero network error, but I have two hardware failures on pitch and yaw.

To me, it’s most telling that roll has never communicated in the bus, despite me confirming the wiring multiple times.

 

[ronrapp]

I'm not sure anyone has mentioned it yet in this thread, but Garmin has a service bulletin out for the GSA28 servos.

It's more of a recall than anything else. Garmin refers to it as a "service life extension program", but the reality is that they'll replace your servos for free for up to 5 years after the publication date of the service bulletin.

See:

https://support.garmin.com/en-US/?faq=fDr3LoWvVh73y7YNV5WnjA
and
https://support.garmin.com/en-US/aviation/faq/2tVToAbyOO2EVxW6GmNTj6/

I haven't done mine yet, because a) they're working well, b) I may do some avionics work in the future which would be a good time to incorporate the Mod 1 units, and c) there's always a chance they may end up modifying the servos further if problems are still present. Five years is a long time.

Anyway, if you suspect the servos are at fault, given the recall on them, I'd be inclined to swap them out since that's probably something you'll wanna do anyway.

--Ron

Edit: I just re-read the original post and noticed you've got the -00 servos and Garmin won't RMA to Canada. That seems ridiculous. If the servos are bad, what does it matter where the airplane with the deficient hardware is located?

 

[walker172]

I'm not sure anyone has mentioned it yet in this thread, but Garmin has a service bulletin out for the GSA28 servos.

It's more of a recall than anything else. Garmin refers to it as a "service life extension program", but the reality is that they'll replace your servos for free for up to 5 years after the publication date of the service bulletin.

I haven't done mine yet, because a) they're working well, b) I may do some avionics work in the future which would be a good time to incorporate the Mod 1 units, and c) there's always a chance they may end up modifying the servos further if problems are still present. Five years is a long time.

Anyway, if you suspect the servos are at fault, given the recall on them, I'd be inclined to swap them out since that's probably something you'll wanna do anyway.

--Ron

That’s definitely the plan, especially given 2 of them are showing hardware faults. The issue in Canada is that Garmin won’t do a direct RMA. I have to work through a dealer. I’d bought these units through Aircraftspruce and their customer service has been less than stellar. I’ll contact some other dealers, but it’s hard for them to get excited about it given that they won’t make any money on the exchange.

 

[pilot2512]

I don't see where you have swapped the servos. Why not try swapping the roll servo with the pitch and see what happens. It would give you more information.

Jay

 

[walker172]

I don't see where you have swapped the servos. Why not try swapping the roll servo with the pitch and see what happens. It would give you more information.

Jay

Good idea. I’m going to replace them all anyways, but once I have them out, I’ll swap and see what happens.

 

[9GT]

I had issues with my canbus also. Both AP servos were replaced by Garmin. While in the process, I added a GHA15 and GMA 245 into the canbus mix. On power up I had a lot of canbus issues again and after two pin out verifications, took a week break. My new servos were at each end of the canbus as terminations. I started disconnecting things one at a time in configuration mode. When I pulled the GMA 245 from the tray, slowly steady green check marks started showing up. This time I took great care in inspecting the 245 canbus pins and found that the hi and low pins were reversed, even though I had inspected the pin outs twice already. All is good now.

 

[walker172]

I had issues with my canbus also. Both AP servos were replaced by Garmin. While in the process, I added a GHA15 and GMA 245 into the canbus mix. On power up I had a lot of canbus issues again and after two pin out verifications, took a week break. My new servos were at each end of the canbus as terminations. I started disconnecting things one at a time in configuration mode. When I pulled the GMA 245 from the tray, slowly steady green check marks started showing up. This time I took great care in inspecting the 245 canbus pins and found that the hi and low pins were reversed, even though I had inspected the pin outs twice already. All is good now.

I’ll check pins again. For me, the weirdest part is that my roll servo is greyed out in config mode. It has never been detected by the canbus.

 

[Mikejetstream]

G3x GSA 28

I too had 2 GSA 28 servos out of 3 have similar hardware faults, eventually was able to get a warranty exchange prior to the recall! Able to swop the serviceable one to check for canbus issues. The warranty exchange replacements were overhauled items and all three can be exchanged in the future. All three have subsequently worked without fault

 

[lr172]

I’ll check pins again. For me, the weirdest part is that my roll servo is greyed out in config mode. It has never been detected by the canbus.

and that makes sense. You plug it in and it starts creating canbus errors. Replace it with a terminating plug and all is good. That points to a problem in the servo. If the canbus wires were wrong, it should have still had errors with the terminating plug. It would be interesting to just disconnect the roll servo and see what happens. If it has no can bus errors with that connection open, that would possibly point to something different. SOP for canbus stuff is measuring resistance on the canbus wires, so still suggest you do that. Garmin isn't going to help too much without doing this test and providing the results. Pretty sure the install manual lays out the testing details.

 

[mburch]

SOP for canbus stuff is measuring resistance on the canbus wires, so still suggest you do that. [...] Pretty sure the install manual lays out the testing details.

Quoting myself from the first page of this thread:

Have you followed the troubleshooting steps in section 2.3.1.3.6 of the installation manual?

 

[walker172]

Just thought I’d post an update for those who were following. It turns out that my issues weren’t really related to autopilot servos other than the two that have hardware fails.

Examining my system information page, I discovered that my GMC507 autopilot controller was also not being detected. Taking apart the connector for it revealed that my canbus low pin was in the wrong spot. A quick swap later and it appeared, updated, and functioned as expected. But after that, my roll servo was still greyed out. After examining servo breakpoints for the 10th time I decided to start working backwards from my roll servo on the canbus. Next in line was my aux connector. Upon pulling it apart, it was also mis pinned. After fixing that, everything is now communicating as expected with no network errors.

After a servo swap, I should be good to go.

I’m relieved to have solved the mystery, but a little disappointed that the issues were there to begin with. It’s a SteinAir panel, and both wiring errors were in breakpoints that they put together. I trusted that they would be correct. My advice is if you have a stein panel not yet installed, take apart the connectors while everything is still on the bench and confirm the pins are correct. It would have been a whole lot easier then, compared to having them in the airplane.

 

[kiwipete]

Mike

It’s great to see someone close out a thread. Sounds like a frustrating time but well done finding the errors.

Regards Peter

 

[Traash]

Lesson learned...

As reported in my 11/4 post:

CAN issues can indicate at one or both of the terminations but actually exist elsewhere on the bus. The technique of depowering all LRU's on the bus except for the PFD and one other at a time can help identify the problem LRU. View additional CAN data on the Configuration/Diagnostic page by holding in both "NRST" & "MENU" buttons at the same time while powering on the PFD.

 

[LawrenceJames]

When we build panels, we build them fully populated with servos and all LRUs connected, configured and make sure green checkmarks all exist. There is no other way to build a panel especially if a vendor like Stein is selling the wire and the avionics and the panel. That should have been caught before you got it.

It isn't good enough to say people are human and make mistakes. QC is QC. You don't ship if it isn't tested and if anyone can connect and test a harness they can.