Reply-To: "ark"
From: "ark"
Newsgroups: comp.arch.embedded,sci.electronics.design
References: <3DCDB1A5.34E114E3@bellatlantic.net> <3DCFBAF5.94BFFDC9@xympatico.ca>
Subject: Re: CAN Protocol and SAE J1850, ISO9141 and J1939
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
Message-ID:
NNTP-Posting-Host: 24.91.110.210
NNTP-Posting-Date: Mon, 11 Nov 2002 20:00:27 GMT
Organization: AT&T Broadband
Date: Mon, 11 Nov 2002 20:00:30 GMT
"Joseph Legris" wrote in message
news:3DCFBAF5.94BFFDC9@xympatico.ca...
>
> Does anybody know which vehicles implement CAN, or how this can be
> discovered? Also, what is the relationship between ISO9141 and CAN, if
> any? How about between SAE J1939 and CAN?
Basically, the car OEM wants you. - To go to their dealership. So, the
protocols are wildly incompatible; both on physical level and maybe all the
way up the protocol stack.
CAN (2 wires) has been a de facto standard in European cars. High-end cars,
if I am not mistaken, are slowly migrating to a multimedia-grade network -
for entertainment. Controllers are still on CAN.
GM used to do multiplexed UART communications on 2 wires (similar to RS422
but of course, different).
Then they switched to J1850 VPW (1 wire) with 3 byte header (AKA SDL Class
2); Pontiac and Saturn were very late in this changeover; if I am not
mistaken, Opel and Saab are CAN. AFAIK, GM is switching now to a 1-wire
flavor of CAN (AKA GM LAN).
Ford used its own 9141 _and_ ISO9141 on different models (2 wire UART);
switched to J1850 PWM (2 wires) (AKA SCP), _including_ some European models.
Chrysler used 1-wire J1850 VPW (1-byte header) but I don't know the details.
If I am not mistaken, J1939 is a _diagnostic_ protocol describing the
content of a message (=packet). In theory, it can work over any low level
protocol; in practice, it is, I think, bound to CAN.
- Ark