AC Delco ECU

[Paul Beam]

At 12:29 AM 6/26/96 +1000, you wrote:
>
>I'm told that it is based on a Motoralla 68HC11?? controller. Is this
>correct? Has any one got a memory map of the system and/or a good
>description of the connection interface? As a matter of fact, I would be
>interested in anything that amounts to more than is in the GM workshop
>manuals. From the way the unit can be reconfigured for such a variety of
>applications, I am lead to believe that the "entire" program is in the
>mem-cal or is there some basic boot code embedded in the controller chip?
>Has anyone ever done a full disassembly of one of the mem-cal units?
>

Bill,

About 1987, I had my hands on the complete source code for the DE ECU of the
Saturn automobile.  Unfortunately, I have forgotton most of it over the
years, but this is what I remember.

The processor is a 68HC11.  I wrote some code for the computer using a cross
assembler on a PC.  To the best of my knowledge, all of the program is
stored in the EPROM, not in the chip itself.  Granted, my work was on a
pre-production car, but I don't think they changed that in production.
There is some EEPROM in the unit that is used to store some data even when
the power is off, and there is some DRAM that has to be refreshed through
software -- which seems to me to be a real pain.

Most of the code is endless tables.  Telling a table from the program can be
a little difficult.  There is a real time kernel as part of the software.  A
timer generates an interrupt every millisecond or so, and a loop calls each
routine based upon a schedule.  I seem to recall that the interrupt vectors
are at the top of memory, so that would be a good place to start to find the
main entry points.  The watchdog timer is also used, so its interrupt vector
is important.

Someone with more recent experience may be able to add more.

Paul

R Paul Beam
Computer Systems Engineer
Johnson Bible College
7900 Johnson Drive
Knoxville, TN 37998
(423) 573-4517
(423) 579-2337 fax