> To combat the eventuality of a $20 cable and bootlegged
> software, they are
> placing a chip in the cable and the full-capability,
> supported software looks
> for the chip before it will work.
Actually, that chip is a fairly sophisticated microprocessor, with the
integrated hardware controllers required to talk to CAN bus (OBD-II) on one
side, and USB bus on the other. They are very different interfaces; and CAN
in particular is a very odd network protocol.
The ones sold to garage mechanics are also quite limited in functionality.
We just bought some more of the "good stuff" at work, and they were nearly
$2000 each just for the hardware and software library to communicate with
it.
> I guess it is simply too
> much to ask for
> all car manufacturers to use common or open source
> programming for their
> products. With no incentive, why would they?
There are in fact powerful disincentives not to. There is a great deal of
money spent on developing that software and it changes pretty much
constantly. Giving away source code that cost literally tens or even
hundreds of millions of dollars to develop and can benefit only one's
competitors just isn't going to happen.
But you are quite right, those diagnostic addresses are only the tip of the
iceberg for what is actually going on inside. I spend a good deal of my
professional life writing CAN bus code; and diagnostic addresses are just
something that gets stuck on afterwards, to help troubleshoot problems in
the field. That's pretty much the way it has to be, since a DA can only be
updated at 10 Hz or so, but practically any control loop (ignition, fuel,
ABS, etc) has to run much faster. We run up to 1 kHz in software, faster
loops are done in (custom) hardware.
-- Randall
_______________________________________________
Shop-talk@autox.team.net
Archive: http://www.team.net/archive
|