We have had various threads on CAN but no detail. I think all avionics applications are proprietary i.e. undocumented but perhaps not encrypted. Certainly that is largely the way vehicle CAN BUS applications (the major CAN market by far) have gone.
AIUI, nearly all cars are CAN 2.0 and only a few are CAN FD. I would thus expect avionics to use CAN 2.0.
Can anyone please post examples of CAN applications in avionics?
Garmin G3X uses CAN between the GDU460/470 screens, GSU25 ADAHRS, GMU11 magnetometer, GEA24 engine interface unit, G5 standby display, GM 507 autopilot controller, GSA28 AP servos and the GAD29 ARINC 429 converter.
I’ve not seen the protocol published anywhere so no whether you can tap into it is unknown.
There is also ARINC 825, but I could not verify if it is implemented by any GA avionics manufacturers.
Never heard of it, so looked it up.
Looks like for avionics but based on CAN.
Current Standard 2018-09-25 
The link to what looks like the document is dead (on any browser I try) but it looks like it was designed by a committee so it’s going to be very complex 
At some point, not soon, I will be looking for an aircraft with the fluxgate connection, to connect to with a CAN bus analyser. I am doing a project for the uncertified market.
The ROTAX is engines (fuel injected) have an ECU that is communicating through a CANBus interface to the engine monitor system of the different EFIS that you can install. The ROTAX protocol seems to be based on canaerospace
Austro Engine FADEC uses CAN for Diagnostic Interface and sending engine data to G1000.
Cessna uses CIES fuel qty senders with can bus for NAVIII (G1000) equipped planes.
All GIA 63/63W/64 (The G1000 COM/NAV/GPS/AP box) have 2 can bus ports for such purposes
Not sure if it’s relevant, but I mention it because it might be hard to discover, but even encrypted CANBUS doesn’t rule out third party stuff. I run open source autopilot software in my car and that uses encryption on the bus. I had to go through a process of extracting the keys. Quite what the point of using encryption if the vehicle gives up the key is I don’t really know, but point is, such things are possible.
Quite what the point of using encryption if the vehicle gives up the key
Presumably they tried to make it not completely trivial but used some silly scheme? Otherwise, hard crypto is trivial – AES256? I use that, and TinyAES, 1k of binary from C, 100kbytes/sec… Better sw version, 4k, is 800kbytes/sec.
Between CAN modules, any encryption will prob90 involve stored (fixed) keys. One can attack that but it is way too much work.