[Efi332] Where can I find info about 4 layer board

[BowTieVette at aol.com]

Sounds like you have a good software background to work from Darrin. One  
obvious question though.. if you have direct experience with the GMWPX computer,  
why not salvage one of these and use your knowledge of the interfaces to 
adapt  it for your own use?  At one point, there were some folks trying to hack  
this computer in order to insert their own code. 
 
Many of the diy listers are really software guys seeking stable  hardware on 
which to experiment with.  Hacking OEM stuff without knowledge  of the 
hardware interfaces is probably more work that just designing  from scratch.
 
 
In a message dated 4/17/2006 10:13:20 P.M. Eastern Standard Time,  
mrgrimes1 at yahoo.com writes:

Thanks for the  intro.
 
I, like many of you,  wrote drive train control software for over 12 years 
starting in 1988 for a  little company call Airsensors, Inc.  That company was 
purchased by  another and so on, then starting around 1996 we were building our 
 own 332 based controller for use in the Heavy Duty/Industrial automotive  
market.  Two years later we gave up on rolling our own and worked a deal  with 
Delphi, through our vendor for GM  relationship, to purchase the 1997 
vette/truck ECU, the GMWPX 68332  controller.  It was flashed  only with a minimal boot 
loader which we used to load our out boot loader in  ram and then re-flashed 
it with our code.  Our RTOS was of our own  making using a 2ms interrupt to 
schedule other 4, 8, 16, 32, and 128ms  interrupts that the application code 
used to do work in, like interface the  QSPI or other peripheral devices.   It 
was written in C++ and we used this language?s conventions to  abstract out each 
physical device into a BSP that allowed application layer  scheduled 
interrupts, when fired, to run till complete all the while accessing  the desired 
peripheral device through its C++ defined interface.  We were calculating speed 
density A/F,  spark dwell and advance parameters in crank angle domain from TPU 
inputs, for  TFD outputs, drive by wire throttle control, jerk and knock 
detection every  4ms.  The TFD was another onboard  peripheral asic that had a 
simple mask made up of logic devices whose output  was used to trigger injector 
firing and coil charging and firing. Its  input was registers, similar to a TPU 
that the application code would write to  via the BSP device interface.
 
Any way, I still have  my SDS compiler, and PandE BDM software, and a whole 
bunch of C/C++ source  code for the 332.  I also have some accurate speed 
density code as well  as many PID algorithms.  The speed  density code needed to 
have the temperature of the air at the intake valve,  when air flow was above 
sonic at the throttle body, calibrated accurately  before in was dead on.  When 
sonic  air flow at the throttle body was seen, a different approach was  used.
 
So why did I come  calling to you message board?  Well, all aftermarket fuel 
Injection  systems, including the HC12 based one mentioned quite often in this 
board, are  limited in their control accuracy.  Once you go TPU, you will 
never go  back! 
 
If any one reads this  post and has any hardware you are willing to sell, 
please email  me.
 
Darrin



 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.diy-efi.org/pipermail/efi332/attachments/20060418/81bce25e/attachment.html