What's new
Van's Air Force

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

Garmin CAN BUS Shield Connection

mfleming

Well Known Member
Patron
OH BOY, I'm jumping in the deep end of the Garmin pool :eek:

All my connectors arrived from SteinAir to start building the harness for my all Garmin panel.

I'm having a hard time getting my head around the proper way to do a CAN BUS daisy chain connection.

There is a lot of conflicting information out there and even some of the Garmin info conflicts with itself. In fairness, some of the Garmin video info is older and the methods have changed over the last couple of years.

So what's the right way to do a daisy chain? The current Garmin manual schematic shows shield ground at each backshell connection.

dwg.jpg


OK, got it
icon14.gif


Now, how do we physically make that connection?
Garmin does not have a complete drawing of the connection but this is as close as I could find in the G3X manual. BUT it doesn't show both conductors from the CAN BUS wire...there should be 2 wires going to the backshell, (HI & LO).
Now in all my searching for an image of the daisy chain connection...none look like this :confused:

shell.jpg


Most look something like this.
But the long green ground coming out of the back of the solder sleeve looping back to the backshell seems to violate the max length for a shield ground. And somewhere in the manual it forbids the shield ground for coming out of the back of the solder sleeve and looping around.


loop.jpg



# 4 gives the min and max length for the shield ground. Now admittedly the Garmin drawing shows a flat braided shield wire not a insulated wire.

term.jpg


Whats a budding avionics tech to do ??
 
30-years ago making test cables for spacecraft, we would daisy change from shield to shield and only use one drain wire from one shield to backshell. The idea was always to try to keep minimum length but long enough to work with.

Make the wire long enough that it can be worked with but make it no longer than it needs to be.

Hope these comments help.
 
Here is a great series of video's that Garmin put out that will explain it nicely.

https://www.youtube.com/results?search_query=Garmin+G3X+Wiring+Fundamentals+Series
Unfortunately those videos are old and don?t conform to Garmin?s current thinking.
They loop the shield from cable to cable and only ground at the end.

30-years ago making test cables for spacecraft, we would daisy change from shield to shield and only use one drain wire from one shield to backshell. The idea was always to try to keep minimum length but long enough to work with.

Make the wire long enough that it can be worked with but make it no longer than it needs to be.

Hope these comments help.

Yes, thats the way Garmin did it up to about 2 years ago.Now they terminate the shield ground at each backshell like this:

dwg.jpg
 
Last edited:
You can check the Steinair website and view a couple of the many how to videos that they have provided on their website. Basically you are turning a 2 conductor shielded communication line into a thoroughfare. On an LRU (any unit connected on that daisy chain) that is not an end point in the CAN bus (communication bus), you will need 4 solder sleeves. One for each of the two wire leads (2 whites and 2 blu/whites) to marry those together, and one for each of the two wire CAN bus cables to marry the two shields together while also providing a tail to be connected the the shield ground on the d-sub connector body. Stein has a good video showing how to splice two leads (the 2 whites, or the 2 blu/whites) while providing the tail that you crimp on the d-sub pin or socket. For the shield grounds, there are a couple ways to do this, but they result will be the same thing. You are soldering the two shields together with solder sleeves and a jumper wire, or braid, and on one of the solder sleeves, also including another wire or braid that will have a crimped ring terminal that will connect to the backshell base with a #8 screw and lock washer. The screw and lock washer is not included in the backshell kit that you paid $50+ for, so get a supply of 1/8” #8 screws and split lock washers. The fourth picture in your post shows opening up a ‘window’ in the outer cable insulation to use as a solder sleeve access for the shield ground. That’s not the only accepted way of doing this, it’s just the newest way Garmin recommends. Stein shows another way that works equally well on his website.
After you have done a couple of these, it gets easier, and is really a very robust way of wiring sensitive avionics to prevent future problems with radiated interference on small signal wires.
Check out the many videos showing pictorially how this is done. You will need a sharp exacto blade, small snips, a good wire stripper, a d-sub crimper with standard and high density ferrules, a small screw driver, some heat shrink tubing (maybe) and heat shrink gun for solder sleeve melting, and some silicone fusion tape (very important). You could use plain electrical tape, but if you want it to last, get the fusion tape. You can buy it at Home Depot. Don’t forget to buy about 1000 4” uv resistant ty-wraps, also available at HD/Menards.
 
Last edited:
Take a look on YouTube. There is a series of videos that are pretty good at showing how to do these terminations.
 
You can check the Steinair website and view a couple of the many how to videos that they have provided on their website. Basically you are turning a 2 conductor shielded communication line into a thoroughfare. On an LRU (any unit connected on that daisy chain) that is not an end point in the CAN bus (communication bus), you will need 4 solder sleeves. One for each of the two wire leads (2 whites and 2 blu/whites) to marry those together, and one for each of the two wire CAN bus cables to marry the two shields together while also providing a tail to be connected the the shield ground on the d-sub connector body. Stein has a good video showing how to splice two leads (the 2 whites, or the 2 blu/whites) while providing the tail that you crimp on the d-sub pin or socket. For the shield grounds, there are a couple ways to do this, but they result will be the same thing. You are soldering the two shields together with solder sleeves and a jumper wire, or braid, and on one of the solder sleeves, also including another wire or braid that will have a crimped ring terminal that will connect to the backshell base with a #8 screw and lock washer. The screw and lock washer is not included in the backshell kit that you paid $50+ for, so get a supply of 1/8” #8 screws and split lock washers. The third picture in your post shows opening up a ‘window’ in the outer cable insulation to use as a solder sleeve access for the shield ground. That’s not the only accepted way of doing this, it’s just the newest way Garmin recommends. Stein shows another way that works equally well on his website.
After you have done a couple of these, it gets easier, and is really a very robust way of wiring sensitive avionics to prevent future problems with radiated interference on small signal wires.
Check out the many videos showing pictorially how this is done. You will need a sharp exacto blade, small snips, a good wire stripper, a d-sub crimper with standard and high density ferrules, a small screw driver, some heat shrink tubing (maybe) and heat shrink gun for solder sleeve melting, and some silicone fusion tape (very important). You could use plain electrical tape, but if you want it to last, get the fusion tape. You can buy it at Home Depot. Don’t forget to buy about 1000 4” uv resistant ty-wraps, also available at HD/Menards.

Yes I watched these videos.
Any photo or video that shows a jumper between two cable shields is old. Now all shields get run to the backshell.

My original question is how do we do the shield termination and still follow Garmin's length restrictions. A shield ground no more than 2.5". The ones that come out of the back of the solder sleeve and loop around to the backshell seem much more than 2.5"...

I'm hoping someone knows "why that's OK" or a photo of the Garmin way.
 
Hey Michael,

Mr Control from Sweden totally hits the nail on the head. The whole aim of the shield is to stop the wire it is protecting becoming an antenna and picking up stray signals that may affect the signal the inner wire is trying to transfer from A to B. Having the shield cover as much of the inner wires as possible is ideal, but not so much that you don't give yourself length to comfortably strip/crimp pins and place them in the respective D-sub positions. Ideally you place the solder sleeves just outside the strain relief clamp of the D-sub shell. The length of the ground wire that runs from the solder sleeve to the D-sub shell really doesn't matter. All it is doing is making sure the entirety of that shield is as close to the aircraft's ground voltage as possible, and that any induced current the shield may have picked up (because it was run past other high current wires) is bled out of the system. Whether this shield is 1" or 6" isn't really going to make a lick of difference. Just keep it long enough that you can comfortably shift things around if you have to change pin positions, but short enough that you're not wasting weight, money and making your wiring look messy. To answer your question, the reason length doesn't matter, is that the induced currents flowing along these wires are tiny, and when you run extremely low currents along a 22AWG wire capable of comparatively high current, the voltage loss will be minimal regardless of length. Also, don't worry about the whole braided thing for the grounding strap. The ground voltage doesn't give a **** whether its traveling through an insulated 22 AWG wire or a braided ground strap. You'll find that most of the solder sleeves will come with a pre stripped wire already in it and these are generally insulated 22AWG white with black stripe.
At the end of the day, there's a reason they have been using CAN bus in tractors, cars and even wind turbines for almost 30 years. It's super robust and can handle other electrical noise well. It's not going to give a **** really.
Good luck. I have total faith that you'll knock this one out of the ballpark.
Tom.
 
Well, no responses yet for the current Garmin method of daisy chaining the CAN BUS, so I followed the latest Garmin manual REV-AM released in April of 2020 and built a test article.

I used the exact dimensions shown in these current Garmin drawings.

I used figure 27-12 from page 27-4
dwg.jpg


And figure 25-8 from page 25-14.
shell.jpg


Here's the test article built to Garmin specs.
gm_dc.jpg


I used Amazon solder sleeves on the test article just for cost purposes.
 
Now imagine your picture but:
- with one yellow sleeve covering both shields and one grounding wire.
- and my can-hi and can-lo splices are so short that the sleeves for them are actually inside the shield sleeve.

End result is my latest daisy chain connection.

Having said that, you connection looks great and looks to be within the length limits :D

Very interesting. So you have the window like mine (Garmin method) on the cables and slide a yellow solder sleeve over both with a single ground wire coming out?
This would connect both shields and the backshell in one connection.

Something like this?

control.jpg


What brand solder sleeves do you use? I bought these Wirefy solder sleeves just to test out. Watch the video. I tested them and they work well.
 
Unfortunately those videos are old and don?t conform to Garmin?s current thinking.
They loop the shield from cable to cable and only ground at the end.



Yes, thats the way Garmin did it up to about 2 years ago.Now they terminate the shield ground at each backshell like this:

dwg.jpg

thanks so much for the post ... as a beginner, I learn every day from you guys every time I open such post.... my background is in forensics .. not engineering so please bear with meeee:).
my question, why run the CAN Bus as per your posted figure ... why not like the figure I am posting... in this case I would need 1 shielded wire for Can Hi and 1 shielded wire for Can Low and both shield are grounded to the shell .... repeat at every LRU and terminate the start and end. ... it is like running light bulbs in a series
 
Last edited:
From a practical standpoint the CAN is made using a two conductor shielded wire. This means you will have a single shield conductor for both high and low CAN conductors.

When you splice the high and low wires together at each LRU (leaving a pig tail to connect to the LRU connector), you will have the shield from both wires to connect. Simply add a pig tail with a ring terminal to ground to the connector backshell.

Your proposed scheme would use a single conductor shielded wire. One for high and one for low CAN. It looks to be more effort to fabricate, as well as less robust. It looks like it would work in theory, but I see no reason to reinvent the wheel. There are a lot of wires already without using single conductor where multi conductor is typically used.
 
From a practical standpoint the CAN is made using a two conductor shielded wire. This means you will have a single shield conductor for both high and low CAN conductors.

When you splice the high and low wires together at each LRU (leaving a pig tail to connect to the LRU connector), you will have the shield from both wires to connect. Simply add a pig tail with a ring terminal to ground to the connector backshell.

Your proposed scheme would use a single conductor shielded wire. One for high and one for low CAN. It looks to be more effort to fabricate, as well as less robust. It looks like it would work in theory, but I see no reason to reinvent the wheel. There are a lot of wires already without using single conductor where multi conductor is typically used.


thanks ... i see what you are saying :)
another question do I need Can terminaters if i only have 2 LRU
 
Last edited:
Back
Top