[maus92]

I installed the latest version of system software, v4.20 from 4.1. After updating both displays, I go into config mode and check to see if all the various components / LRUs have had their updates applied as well (if applicable.) Since I normally keep the autopilot servos off, a fun thing I like to do is watch the servos get updated when I turn them on. The roll servo updated on the first try, however the pitch servo failed on several attempts. I recycled power, went into config mode and observed another failed attempt, but on the second try, it worked. I flew after that, and everything worked properly, AFA I could tell.

The question is what could be causing this, and is it anything to be concerned about? Network error is 0%. The servo is second to last on the CANBUS, and the last device (GPS 20A) on the bus works fine.

 

[dweyant]

I installed the latest version of system software, v4.20 from 4.1. After updating both displays, I go into config mode and check to see if all the various components / LRUs have had their updates applied as well (if applicable.) Since I normally keep the autopilot servos off, a fun thing I like to do is watch the servos get updated when I turn them on. The roll servo updated on the first try, however the pitch servo failed on several attempts. I recycled power, went into config mode and observed another failed attempt, but on the second try, it worked. I flew after that, and everything worked properly, AFA I could tell.

The question is what could be causing this, and is it anything to be concerned about? Network error is 0%. The servo is second to last on the CANBUS, and the last device (GPS 20A) on the bus works fine.

I had the same issue. Took several tries, but finally updated and seems to be working just fine.

-Dan

 

[bkthomps]

I have issues too, I've resolved them by leaving one screen on at a time while upgrading, usually my AHRS 1 that has the errors and issues upgrading

I have discussed at length with Garmin but never came to an understanding of the "why"

 

[Brantel]

Have not experienced anything like this before but may this time who knows... I always ensure all LRU's are powered up and running normal before installing new firmware. Not sure if that has anything to do with it or not. I will do the same for 4.20 and report back.

 

[g3xpert]

We have multiple indications that Brian T. has a marginal installation that needs to be improved (including difficult code loads), but what Dan and Charley describe is likely nothing to be concerned about. More on Brian's issue later.

Most communication on the CAN bus is periodic - data which is transmitted over-and-over many times a second with slight changes (like airspeed, altitude, and pitch/roll).

Having a data packet delivery failure of even one of these hundreds of packets which are transmitted each second is quite rare with a good installation, but even if it does occur, the system is tolerant of this and devices like the displays and autopilot servos can tolerate the loss of this periodic data for a fraction of a second.

Code loading, however, is quite different. If there is some issue which causes even one packet to be dropped during the code load process, the process fails and the system will automatically re-try until the process is complete. This is what Charley and Dan saw happening. Again, probably nothing to be concerned about, but it could be a sign that improvements could be made if desired.

Brian has had a persistent issue with code loading, and as he describes, has generally found it necessary to power down some or all devices which are not being code loaded to "clean up" his CAN bus communications so the code load process will complete without error. We may have some good news related to this.

CAN bus devices spread around the airplane rely on a common ground bus to reference signals on the wires. While we don't recommend anyone rely on airframe structure ground for their avionics, since Brian has a fiberglass Lancair, it is even more important that every device has a high quality connection to a common ground point.

We have G3X and G3X Touch systems installed in lots of Lancairs, but we recently ran into a second instance of a Lancair that was having issues like Brian - only quite a bit worse. The entire CAN bus was replaced (actually < 10 ft in this particular aircraft - quite short), and that didn't help at all.

Next, we discovered that powering down the GAD 29 solved the CAN communication issue - rock solid. The dealer assisting in this installation wanted a new GAD 29 sent immediately, so we did, but since we hardly ever see a device failure, we asked that the grounding to this device be carefully checked and measured. We were assured it was fine.

The replacement GAD 29 was received and installed. This had no affect on the issue (as expected), so we dispatched a service engineer to help out.

The problem turned out to be a poor ground connection to the GAD 29. Once that was corrected, everything was rock solid.

So, Brian, please check the grounding to all of your CAN devices and let's see if we can get your aircraft straightened out!

Thanks,
Steve

 

[dweyant]

Code loading, however, is quite different. If there is some issue which causes even one packet to be dropped during the code load process, the process fails and the system will automatically re-try until the process is complete. This is what Charley and Dan saw happening. Again, probably nothing to be concerned about, but it could be a sign that improvements could be made if desired.

Thanks,
Steve

Steve,

Thanks for the well written explanation.

In use everything has been rock solid since we got a few early issues resolved. I've put about 150 hours on my 9A in the last year, and it has been great!

However, I am interested in some details on what you think I might be able to improve? I'm going to be finishing up my first condition inspection this weekend (since the weather looks crappy), so if there is anything I can/should look at while the plane is open this would be a good time.

Thanks,

-Dan

 

[g3xpert]

Steve,

Thanks for the well written explanation.

In use everything has been rock solid since we got a few early issues resolved. I've put about 150 hours on my 9A in the last year, and it has been great!

However, I am interested in some details on what you think I might be able to improve? I'm going to be finishing up my first condition inspection this weekend (since the weather looks crappy), so if there is anything I can/should look at while the plane is open this would be a good time.

Thanks,

-Dan

Hi Dan,

Glad to hear things are going great with your RV-9A.

We sent you an email with some information related to this discussion.

Thanks,
Steve

 

[Brantel]

Upgraded to 4.20 today and also got the update error on the pitch servo. Everything else updated normally. I just sat and watched and as soon as everything else updated, the pitch servo tried again and also updated normally. Have never noticed this before but then again I was not looking for it.

Bottom line is that it does not seem to be anything to worry about and it appears that there is plenty of error checking and fault tolerance built into the firmware update process to prevent bricked units.