Re: Bench tests needed...

[Shannen Durphey <shannen@grolen.com>]

Ludis Langens wrote:
> 
> Shannen Durphey wrote:
> >
> > I'd like to add
> > about 80 degrees offset to the timing calcs, and it looked like a
> > simple matter of adding it in right before the maximum advance check
> > is done.
> >
> > Ok, is this done in hardware or software?
> 
> The problem is in the software.  It computes advance as a 16 bit value,
> and then it tosses the upper eight bits when it goes to convert from an
> angle to a time interval.
Oh, geez, there it is right in front of me. : )  

This has puzzled me for a long time:
LD3B5:  ldx     #0x0037
D3B8    ldaa    *SAC
D3BA    jsr     8X16mult
D3BD    std     *L0044

What is this?  SAC looks like the 8 bit spark angle, but what
significance is 55 decimal?
> 
> > > Is this something to do with the DIS/dizzy conversion?
> > Yes.
> 
> Someone ought to look at the differences between real GM DIS and dizzy
> applications.  A good choice might be to compare software between the
> 1227748 and 1228321 ECMs.  The former is DIS, the latter is dizzy I
> think.  These ECMs have the same hardware except for the limp home
> calibration.  They also use (small) 27128 EPROMs - so the programs can't
> be too complex.
I have a Q4 749 disassembly that I compare to the dizzy software. 
It's difficult for me to tell what's happening in either.  I have
found some small items:  The syty code calculates dwell based on
ranges of time between reference pulses, and adds for low voltage and
during acceleration.  DIS dwell is ref/2.  

I had dropped this project for a while.  Let's see what happens this
go 'round.
Shannen

> 
> --
> Ludis Langens                               ludis (at) cruzers (dot) com
----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo@lists.diy-efi.org