New to the EFI List; Reprogramming my 1982 ECM (long)

[Carboni,Noel]

Hello, group!

I'm new to this list but harbor a strong interest in do it yourself
automobile
work.  From the looks of things here (ignoring the flames and noise,
which I
assume are transient), you are a group that considers, as I do, the
modern
"Automotive Transportation System" as provided by the manufacturers
insufficiently permeable.

Here's a copy of a post I recently submitted to VetteNet to give you an
idea of
what I'm into.  I'll try to contribute what I can to this group, though
it
sounds like some of you are WAY beyond me technically.

--

A number of people, over time, have asked me about reprogramming my 1982
ECM.
Here's a description of what I did.  I plan to tidy this up, augment it
with
pictures, diagrams, and maybe some downloadable software, and publish it
on my
web site soon.

First, recall that my car is a 1982.  There have been no fewer than 4
completely
different Corvette ECMs since 1982 (I've lost track, there are probably
as many
variations as there are model years, maybe more).  

In my case, my 1982 ECM uses a calibration chip, which turns out to be a
2532A
PROM.

Back in the mid '80s I was moving ahead with my plan to improve my car's
performance, and upgraded to a 417 small block engine.  Though the stock
computer would run the new engine, there were real driveability problems
and I
confirmed my suspicions that it would be necessary to alter the car's
fuel
injection and timing calibrations to make it run right.  The 1982 uses a
MAP
(manifold absolute pressure) sensor, so you can imagine that a
completely
different engine, with better heads, more cubic inches, and a more
radical cam
profile, would have pretty different characteristics.

I started by asking around and going through performance magazines to
try to
find a company that would make me a custom chip.  I found a guy in
California
(whom I think may have been an ex-GM employee) that said that he would
make me
some custom chips for $$$ if I gave him my engine characteristics.  I
did and he
made me three different chips, in which he provided varying fuel
calibrations
(lean, nominal, and rich).  Being the curious type, and now knowing the
chip
type (he sent EPROMs), I loaded the data from all of the chips into my
PC using
a PROM programmer.  I also pulled out the stock chip and loaded that up.
I
dumped the data in hex, compared the files, and voila - the location of
the fuel
calibration tables, and the knowledge that bigger numbers mean more
fuel.

By the way, his custom chips didn't run my new engine even as well as my
original stock chip.

Looking a bit further into the hex files, and applying some left-brain
pattern
matching skills, I found several other tables, including one that looked
suspiciously like a set of advance curves.  I wrote some quick graphics
software
to help me visualize the curves and voila, there they were!  The stock
timing
curves were actually very interesting:  1.  They were very choppy,
indicating
the engineers probably hacked in the values during testing.  2.  The WOT
timing
is actually backed off between 4000 and 5000 RPM, probably as an early
attempt
at limiting RPM.

Now the job remained to determine the calibration of these tables.
Making some
guesses as to the parameters used to index into the tables (knowing what
sensors
the car has), I erased the three PROMs and starting with my stock chip I
replaced the fuel and timing tables with a variety of different data,
some of
which had some "glitches" built in to allow me to validate theories and
determine actual values (e.g., RPM, throttle position, etc).

Over time I determined the parameters, learned the calibrations, and was
able to
reprogram a chip to run my new engine quite well.  It turns out the fuel
table
appears to be indexed by MAP and RPM, and the timing table indexed by
throttle
position and RPM.

A transmission change from auto to manual muddied the calibration waters
a
little, and I never could get below 14.0 seconds nor break 100 MPH in
the
quarter mile.  I finally figured out that my stock (though ported)
intake
manifold was the restriction to going faster.

Not being happy with the physical performance limitations of my TBI
intake
manifold, I decided to eliminate the TBI system entirely and bolt on a
TPI
system (that story is in my web pages).  Now the GM ECM was only
managing the
spark timing -- the TPI computer was doing the fuel.  This necessitated
further
GM ECM chip changes, though, as the TPI was building WAY more cylinder
pressure
at low to mid RPM and I was seeing detonation between 1800 and 3000 RPM.

I fooled with the timing curves (which are now RADICALLY different from
stock)
and both eliminated the detonation at heavy throttle and improved
mileage at
light throttle to boot!  Today, I run 12.85 at 111 MPH in the quarter
mile, yet
get nearly 20 MPG on the highway.  Not bad, considering my car, when
stock, used
to run a 16 second quarter mile at less than 90 MPH!

Subsequent research has shown that even the 1984 TBI ECM system from GM
is
completely different from the 1982.

What's next for me?  Who knows... I may buy an Accel SuperRam system,
which
would swap out all of my existing computers and tie everything into
their DFI
(Digital Fuel Injection) system.  The SuperRam is supposed to flow VERY
well and
the DFI is touted as a good controller...  Hey, it's only money!

------------------------------------------------------------------------
                              ____________
                            //~~~~~~~~~~~~\\
                           // _____________\\
                     /~~~---~   |      |   ~---~~~\
                    /|    /______\____/______\    |\
                   ( | /      |          |      \ | )
    Real 'Vettes   | \============================/ |   Program in Hex
                   |\ \\[  ]   |corvette|   [  ]// /|   
                   | \/+========================+\/ |
                   |~==============================~|
                   |_____|                    |_____|
------------------------------------------------------------------------
       Noel Carboni, 1982 Collector Edition, 12.85 sec @ 111 MPH
                     http://home.att.net/~ncarboni
+----------------------------------------------------------------------+
| MS D112, Office D1U20            | EMail: Noel_Carboni at usa.racal.com |
| Racal-Datacom, Inc.              | Phone: (954) 846-6744             |
| Sunrise, FL  33323-2899          | Fax:   (954) 846-3822             |
| "We turn networking inside-out"  | WWW:   www.racal.com              |
+----------------------------------------------------------------------+