What's new
Van's Air Force

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

Can Bus Length Concerns

mfleming

Well Known Member
Patron
RV-7, all Garmin avionics. Building my own avionics harness with the support of SteinAir.

This is an observation/question.

In all my reading in VAF about the Garmin Can bus maximum length, I see statements that one can't possibly exceed the maximum length (20M or 66ft) in a RV-7.

This interested me since when doing the basic wiring in my -7 I noticed the amount of wire from point A to point B was always much greater than expected. This of course is due to the serpentine routing.

So when I made a mock avionics harness and laid it out on a table I decided to measure the exact length.

61' and I don't have as many goodies as some.

So I'm here to tell future budding avionics installers like me....it's easy to exceed the Can Bus maximum distance of 66'.

Now for the question:
Is the distance really that important?

I can't imagine how the Garmin Can Bus system is installed in an RV-10 or larger aircraft than out RVs without busting this parameter.
 

Attachments

  • can-bus-mock-up.jpg
    can-bus-mock-up.jpg
    120.2 KB · Views: 305
My guess is that number is pretty conservative, don't have any first hand knowledge though.

I have not worked with can bus very much but in a past life I worked with a differential pair bus (like can bus) that was spec'd for a very short length like 3' that was usable up to 100' or more.
 
Someone asked how I got 61' on the harness. Here goes.

So in the post there is a photo of my mock can bus harness on the table.

That represents 45' and it starts jut behind the baggage compartment where a DB15 connects the aft portion of the harness already installed, it then runs up to behind the instrument panel across and then behind the sub panel, back down the firewall to the right wing DB15.

So with harness in the photo, the aft harness and the section in the right wing equals 61'

A lot is taken up with the LRUs and displays. If there is a 12" run to a LRU, that equals 24" because of the out and back. I measure 21.8' just for the LRUs out and back.
 
I do not know everything about CAN bus and Garmin may be able to give us the correct answer.

My understanding of the CAN bus length has to do with the data rate over longer distances. As the length gets longer, the data rate will get lower. Longer line means a lower amount of data transmitted reliability. Garmin may have picked a length that will allow a data rate that works with everything all the time. IF length gets longer, some devices may not work correctly or not work all the time. My understanding, see link above, "BIT rates up to 1 Mbit/s are possible at network lengths below 40 m". (131 feet)
 
Last edited:
I do not know everything about CAN bus and Garmin may be able to give us the correct answer.

My understanding of the CAN bus length has to do with the data rate over longer distances. As the length gets longer, the data rate will get lower. Longer line means a lower amount of data transmitted reliability. Garmin may have picked a length that will allow a data rate that works with everything all the time. IF length gets longer, some devices may not work correctly or not work all the time. My understanding, see link above, "BIT rates up to 1 Mbit/s are possible at network lengths below 40 m". (131 feet)

I've seen the reference to 40 meters before. Here's what Gamin say's in there G3X manual version _am.

"2.3.1.3.1 CAN Bus Architecture
The electrical architecture of the CAN bus takes the form of a linear “backbone” consisting of a single
twisted wire pair with an LRU connected (terminated) at each end (Figure 2-5). The overall length of the
CAN bus from end to end should be 20 meters (66 feet) or less. At each of the two extreme ends of the
CAN bus, a 120 Ω resistor is installed to “terminate” the bus. In the G3X system, termination resistors are
provided either within the LRUs themselves, or via termination adapters that plug into an LRUs CAN
connection."
 
If there is a 12" run to a LRU, that equals 24" because of the out and back. I measure 21.8' just for the LRUs out and back.

Just to clarify... There are no out-and-back "spokes" on a CANBUS. It is one continuous run, with very short "nodes" at each unit of perhaps 3 inches. And the units connected to it don't have to be in any particular order.

A 61-foot length seems very long to me. Is it possible you are going "out and back" too much?

For reference, I have a full G3X system in my RV-8, and the total CANBUS length is 39 feet. My CANBUS terminates at one end in the tail with the magnetometer, and out in the right wing on the other end at the roll servo. In between, it visits the firewall for my GEA 24, and of course about six different units in on or around the panel.
 
Just to clarify... There are no out-and-back "spokes" on a CANBUS. It is one continuous run, with very short "nodes" at each unit of perhaps 3 inches. And the units connected to it don't have to be in any particular order.

A 61-foot length seems very long to me. Is it possible you are going "out and back" too much?

For reference, I have a full G3X system in my RV-8, and the total CANBUS length is 39 feet. My CANBUS terminates at one end in the tail with the magnetometer, and out in the right wing on the other end at the roll servo. In between, it visits the firewall for my GEA 24, and of course about six different units in on or around the panel.

It sounds like your -8 and my -7 are wired the same.

So this is my interpretation of a CAN bus path length and how to measure it. If I'm in error, I'm eager to know.

I've included my avionics harness layout schematic and a sample from the current G3X manual. Notations have been added to the sample to explain my thinking.

-Michael
 

Attachments

  • CAN-bus-sample.jpg
    CAN-bus-sample.jpg
    257.9 KB · Views: 349
  • Avionics-Harness-Layout.jpg
    Avionics-Harness-Layout.jpg
    194 KB · Views: 355
I think if you have things like the GEA 24, GSU 25, GAD 29/27 mounted behind the panel that your "out & back" distances will be much shorter than you think.

I don't have any can bus devices in the wings on my 6A, just tail to panel but I figure I have maybe 30' of can bus wiring.
 
Hello Michael
There are several mixed messages on the subject of Garmin Can bus.
If you check the specs that Garmin gives for certified aircraft they have a max length of 85ft. This is probably because they have to use the more expensive can bus cable Part number CAN24 TST120 (CIT)

I had a similar problem on my RV-14 so i opted to use the more expensive can bus cable, which i Highley recommend! over the standard milspec.

One option is to shorten the can bus by installing a GMU22 magnetometer in place of the GMU11 ( the GMU22 is not on the can bus.)

Garmin's aviation training webinar, that had a section on can bus wiring.
If you send over your email i can send you the power point slides on this, if it helps. It has Some good installation tips & specs and also a piece on can bus testing and trouble shooting

Best Regards
Mark
 
I think if you have things like the GEA 24, GSU 25, GAD 29/27 mounted behind the panel that your "out & back" distances will be much shorter than you think.

I don't have any can bus devices in the wings on my 6A, just tail to panel but I figure I have maybe 30' of can bus wiring.

As I mentioned in post #3, the measured distance (not estimated) to the LRUs behind the panel eat up one 21' +.


GSU25-1 - distance is 9.25" = 18.5" total
GSU25-2 - distance is 13.25" = 26.5" total
GDU 460 - distance is 11.5" = 23" total
GAD 27 - distance is 11.5" = 23" total
GDU 470-1 - distance is 12.75" = 25.5" total
GI 260 - distance is 17.25" = 34.5" total
GAD 29 - distance is 8.5" = 17" total
GMC 507 - distance is 12" = 24" total
GDU 470-2 - distance is 12.25" = 24.5" total
GRT 200B - distance is 8" = 16" total
GEA 24 - distance is 11" = 22" total

Total for all branches behind the instrument panel and sub panel is 21' +
Thats just the branches, this does not count the distance between branches.

So my drawing looks a lot like the sample from the manual...unless someone can show me where I'm wrong, I think a lot of CAN bus structures are under counted for length.

Just a reminder the drawing of my CAN bus layout posted only depicts the main fuselage. The right wing and aft fuselage are not depicted.
 
Question

If the max recommended length is 66' and you are at 61'.........

Rest easy tonight and build on....correct?
 
Hello Michael
There are several mixed messages on the subject of Garmin Can bus.
If you check the specs that Garmin gives for certified aircraft they have a max length of 85ft. This is probably because they have to use the more expensive can bus cable Part number CAN24 TST120 (CIT)

I had a similar problem on my RV-14 so i opted to use the more expensive can bus cable, which i Highley recommend! over the standard milspec.

One option is to shorten the can bus by installing a GMU22 magnetometer in place of the GMU11 ( the GMU22 is not on the can bus.)

Garmin's aviation training webinar, that had a section on can bus wiring.
If you send over your email i can send you the power point slides on this, if it helps. It has Some good installation tips & specs and also a piece on can bus testing and trouble shooting

Best Regards
Mark

Thanks Mark, I just learned today about the higher grade CAN bus wiring. I'm using the 22-2 that SteinAir sent me. My total CAN bus length is just under the max specified in the G3X manual so I should be OK.

Your right about mixed messages, a SteinAir tech casually mentioned 55' max :eek:

My email is in the signature line...love to read the Power Point slides
 
If the max recommended length is 66' and you are at 61'.........

Rest easy tonight and build on....correct?

YES, exactly....I just was wondering if I was missing something. My measured distance seemed so much higher than others. I was also hoping someone would confirm my thoughts or correct them.
 
Mine is Similar

Mike will be about the same Mike. Mine starts at the magnetometer in the right wing to the right wing servo, then thru the panel Garmin Components, then hangs a 90 to the fuselage pitch servo where it all ends. I need to measure it but it will be somewhere between 40-60' as well.
 
Well I just had a long conversation with Garmin about my CAN bus layout. They confirmed my method of measurement is correct and the layout of my route is efficient and it is what it is...61'.

AND they answered my original question: Yes the distance matters.

So it's my advice for future Garmin installers, be very mindful of CAN bus length...very easy to bust..
 
Last edited:
CAN Bus Length

Hello Michael
There are several mixed messages on the subject of Garmin Can bus.
If you check the specs that Garmin gives for certified aircraft they have a max length of 85ft. This is probably because they have to use the more expensive can bus cable Part number CAN24 TST120 (CIT)

I had a similar problem on my RV-14 so i opted to use the more expensive can bus cable, which i Highley recommend! over the standard milspec.

One option is to shorten the can bus by installing a GMU22 magnetometer in place of the GMU11 ( the GMU22 is not on the can bus.)

Garmin's aviation training webinar, that had a section on can bus wiring.
If you send over your email i can send you the power point slides on this, if it helps. It has Some good installation tips & specs and also a piece on can bus testing and trouble shooting

Best Regards
Mark

As bus length increases, signal reflections become more of an issue, and signal delay between nodes increases. During the STC program, the need arose for a longer run than the originally specified 66 feet. We determined an increased bus length to 85 feet is acceptable if the CAN bus wire mentioned here is used.

Thanks,

Justin
 
I've included my avionics harness layout schematic and a sample from the current G3X manual. Notations have been added to the sample to explain my thinking.

Michael,

I think you'll find that your CANBUS runs will end up being somewhat shorter than what you've depicted in your first diagram. In many cases, there won't be a need to make those 12-inch "return runs" to a central point before moving laterally some distance, then back another 12" to the next unit. In many cases, the units will be right next to each other, and their CANBUS interconnects could be as short as 12" long TOTAL. I have a couple like that.
 
I've done a couple of RV10 panels where the length of the bus is way more than the RV7. When the GI260 first came out it had some issues that were causing CAN bus failures so on one of these 10's we looked very closely at bus behavior/faults, a comment from one of the lead guys at Garmin that it was one of the cleanest buses he'd seen. So unless you get really sloppy with the bus you should have no issues.

I can get the CAN24 TST120 cable at a reduced price if I buy 1000', would save about $1.50/ft.
If enough folks committed to most of it I would be willing to purchase and distribute.
 
Last edited:
The whole deal with a CAN bus is tolerance to noise, poor installation and other in-service problems. Garmin have to come up with a length, but its not pass or fail. A bus length of 67' doesn't mean your whole system will not work. If the installation work is not so good, low spec cable is used and your aircraft is electrically noisy there may be problems. Some planning is required to ensure the bus length stays at a reasonable length. There are steps that can be taken, such as using higher spec bus cable, if that is a better method for you than changing the cable routing.
 
Back
Top