What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

G3X - Garmin Autopilot / Servo communication fault

Estevanb

Active Member
I am having a hard time to troubleshoot my autopilot / servos system and I was wondering if anyone in the forum has ever seen this error message in their GDU data log files:


Fail / Servo communication fault / 00000001 00000000


My system is unstable. Roll servo communicates when can-bus is not terminated properly and then, if I properly terminate the CAN Bus I loose all communications.

The error message appear either way and even if the wire harness at the wing root is totally disconnected.

Any opinions, advices? all welcome at this point...

Thanks!
 
The network architecture of the G3X Touch system is pretty simple -- 2 wires, daisy chained from device to device. It's also amazingly easy to mess it up - common mistakes are node lengths, missing/wrong strappings, missing termination, swapped Hi/Lo from device to device, etc. etc.

When you say "not terminated properly" and "properly terminated" what does that look like? ...Post Pictures if you have them...
 
The network architecture of the G3X Touch system is pretty simple -- 2 wires, daisy chained from device to device. It's also amazingly easy to mess it up - common mistakes are node lengths, missing/wrong strappings, missing termination, swapped Hi/Lo from device to device, etc. etc.

When you say "not terminated properly" and "properly terminated" what does that look like? ...Post Pictures if you have them...
by properly terminated I meant the jumpers on the servo connectors. may network has pitch servo in one end and roll servo on the other end. There was communication on going when the jumper on the roll servo was removed and network resistance was 120ohm (which is only 1 termination resistor - from pitch servo). When I install the jumper on roll servo, I read 60 ohms on any point of the network but the BUS crashes...
 
The network architecture of the G3X Touch system is pretty simple -- 2 wires, daisy chained from device to device. It's also amazingly easy to mess it up - common mistakes are node lengths, missing/wrong strappings, missing termination, swapped Hi/Lo from device to device, etc. etc.

When you say "not terminated properly" and "properly terminated" what does that look like? ...Post Pictures if you have them...
This is my CAN-BUS diagram.

I will re test all connections.
 

Attachments

  • RV-14 CAN BUS.pptx.zip
    86.5 KB · Views: 17
...I also forgot to mention, there are some D-sub connectors out there that have the incorrect pin numbers stamped into the plastic, verify pin numbering.

It's also very easy to be "off by one" when counting & trying to find the right pin/socket number, especially when you're upside down and your readers have fogged over and you've just swallowed one of the D-sub pins that you were holding between your teeth --- AMHIK :D
 
I would check and check again that the can bus has not become reversed somewhere. A communication fault must mean there is something wrong with the can bus as that is where all the communication happens. Take each D sub plug apart and ensure the correct pins have been used for Can Hi & Can Lo, and that you have not transposed the wires (connected hi & lo together) somewhere. Make sure the shield grounds have been terminated correctly at each point and are not shorting Can Hi or Can Lo. It is going to be tedious but you will find an error somewhere.
 
What is the exact message you are getting? There are a lot of simple things this could be.

Are you saying that the entire can bus network fails when you plug your servo in causing a complete failure, or just that the servos stop communicating properly and there is an error?

If the issue is localized to just the servos and the rest of the system is still talking, then check to make sure you have your pitch ID jumper in and that it is pinned out correctly. If you don't they have a conflict and it causes issues.

If you are having a complete can bus failure you will want to start trouble shooting your can bus.

Here is a quick can bus troubleshooting procedure that I train our staff to do.

1. Ensure that both termination points are plugged in. Test the resistance between can bus hi and can bus low at any LRU except the termination points.
  • If you read 60Ω then your can bus is properly terminated, move on to step 2
  • If you read 120Ω then your can bus is not terminated. This can be an unplugged termination LRU, a break in the line, or a missing termination block or jumper. Start by checking the termination jumpers. If those are in the correct pins then there is a break in the can bus up stream somewhere. If you have any break points in the system, start by checking those. From there you will need to check continuity LRU by LRU
  • If you read 40Ω or less then you have too many terminations. Check to see if you have any of the 9 pin termination block installed.
2. In your case, you seem to be having a localized issue with the servos. I would do the following: Once you have established that your can bus is properly terminated start by removing power from the servos. If the can bus starts working with the servos plugged in, but turned off, check to see if you have can bus hi and low flip-flopped at one end.

3. If pulling power from the servos doesn't change anything then we move onto a bit more involved troubleshooting. unplug every LRU on the can bus except your PFD and 2 termination points. If the system works (green check mark on those three units on the system info page) then you need to turn the system off and add one LRU back, repeating this cycle, until the can bus breaks - that unit is your issue and it is almost certainly a pinout error on the can bus wires.

4. If the can bus is not talking when brought down to its base components and you have confirmed that it is correctly terminated, then start by checking the can bus pinout at the PFD and both termination points. There is likely a pin out error at one of those units.

5. If none of the above works then the next most likely issue is a short to either can bus hi or low or a short between hi and low. This can be a bear to find. start by unplugging every can bus LRU and termination block. Once they are all unplugged check the Hi and Low against ground. If you detect continuity from either wire to ground there is a short to ground. If you have continuity from hi to low then they probably got soldered together at some point. With this type of failure it is usually faster to replace the entire can bus rather than to identify a specific point. ... Though it can be super frustrating to do this - its usually a last resort.

I'm sure that this is WAY more information than you need, but I wanted to type out a more thorough can bus trouble shooting guide for anyone who might stumble across it in the future.
 
Last edited:
Hi all! And thank you so much for all advice and tips shared! It helped my troubleshooting process and I was able to find and correct the issue. Long story short, the culprit was a poor connection on the wing root MOLEX connector.

As the wire harness exits the fuselage, it makes a turn and twists to get too the wing root connector. I found out that the tension on the wires was actually causing one of the can-bus pins to loose connection. I tried to shape the wire bundle in a way to alleviate any tensions on the connector and checked that the connector was all the way connected to the female receptacle and immediately all devices started showing free check marks and zero network failures.

On one hand I am very glad we solved the issue. On the other hand I wonder... how fragile and poor this connection system is. What if a simple bad connection on this connector happens during flight. It has a major impact on the entire Garmin suite. I willl probably be left with G5 and a radio only..

I am thinking now I need to make a anchoring point for the harness so that it doesn't vibrate to a point it will cause poor contact in the pins...

Thanks folks. Really appreciate the help troubleshooting!
 
Glad you got the problem solved. Not sure what model of RV you have, but unless it's an RV-12 where the wings are designed to be removed, I would consider replacing a known-suboptimal connector with more reliable crimped splices. If you ever have to remove the wings from any other RV besides a 12, it's because you are already involved in an ugly project that's not going to be quick or easy anyway.

Also remember that you have the option to connect backup RS-232 connections between a GDU and your GSU 25 and/or GEA 24, should you choose to do so.
 
Hi all! And thank you so much for all advice and tips shared! It helped my troubleshooting process and I was able to find and correct the issue. Long story short, the culprit was a poor connection on the wing root MOLEX connector.

As the wire harness exits the fuselage, it makes a turn and twists to get too the wing root connector. I found out that the tension on the wires was actually causing one of the can-bus pins to loose connection. I tried to shape the wire bundle in a way to alleviate any tensions on the connector and checked that the connector was all the way connected to the female receptacle and immediately all devices started showing free check marks and zero network failures.

On one hand I am very glad we solved the issue. On the other hand I wonder... how fragile and poor this connection system is. What if a simple bad connection on this connector happens during flight. It has a major impact on the entire Garmin suite. I willl probably be left with G5 and a radio only..

I am thinking now I need to make a anchoring point for the harness so that it doesn't vibrate to a point it will cause poor contact in the pins...

Thanks folks. Really appreciate the help troubleshooting!
I'm glad that your found your problem! If you don't already have them wired in, there are backup RS-232 connections for both the GSU25 ADAHRS and GEA24 EMS that give a direct line to the screen they are wired to in the event of a canbus failure.
 
The are wired. I noticed that my PFD doesn't loose the attitude indication nor engine data. I guess a failure in CANBUS will be a massive loss but the airplane will still be flyable. :)
 
Back
Top