Old 486 Board for ECU? Why?

[Jim Tyler]

> From: Frederic Breitwieser <frederic at xephic.dynip.com>
> Subject: Re: Old 486 Board for ECU? Why?
> 

> If you build a PC based system, if the
> motherboard blows for whatever reason, and your ISA card is what does
> some of the work, toss system board, buy new one for $79.95, and reboot
> :)

Sorry, makes no sense to me when you can buy any GM ecm for 50.00
at any junkyard or a reman for 100 and change.  

If you are doing this for academic reasons, fine, great!
If this is for a dyno that makes some sense.

If you are going to put this in a car it will be an expensive kludge.
You are going to build a  one-off ISA card, with all the hardware to 
properly condition the power, do all the control functions, a watchdog,etc.
custom enclosure and harness, then 100's of hours of code?????  
 
> The more customized you make it, in general, the more of a pain in the
> ass it is to replace things when they break.....  My custom card however, will be
> a nightmare to fix, repair, or replace.  One off's are very expensive.
> I must have spent close to 100 hours soldering surface mount parts by
> hand.  I would not recommend it.  

Are you going to mass produce these things? Doubt it. That is the only way 
this project could ever be cost effective.

> >What does make sense is a general
> > purpose programmer for the most common and powerful GM ecm.  Sort of like
> > the SY/Ty editor but for something more common than the 749 ecm.
> 
> Because Java is interpreted, and inherantly slower than machine code.
> Unfortunately, because anything that works cross platform is a good
> thing. Scale up and down as necessary.
> 
That really doesn't matter much, in fact it is great so you can use a MAC or
whatever box you like to run the app.   The target GM ECU would be running
HC11 of course.

 
> My card supports 32 injectors, two TBI sensors, one MAP or MAP sensor,
> eight O2 sensors, a crank and a cam sensor (electromotive stuff, not OEM
> stuff), and a variety of other stuff.  Its generic enough that in
> theory, you could have 4 injectors per cylinder (or not) of different
> fuels, use OEM injectors with small flow on huge-ass turbocharged
> engines (which is what I'm doing), and of course, a variety of inputs as
> well.  I haven't written the software yet, so don't hold your breath.
> FOr me, this is nothing more than an expensive idea that uses a spare
> 486 system board.

Your card means little to the rest of the world, even if you make the 
design free and give away the circuit boards.   It will still be a 
kludge running on a cheap motherboard.   

To make this concept real world, the final version would have to be 
on an industrial controller, and I think any cost 
benefit is long gone by then.

Jim
----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org