[Gmecm] RE: More '730/$8D tuning questions...

Mon Apr 23 23:19:51 UTC 2007

Hank,
The learning curve is still near-vertical for me, so I expect to make plenty
of mistakes.  By all means, please point them out!  As long as I can sort
out the mistake, I can keep going, and the learning curve gets a little
flatter.  Feel free to fire away with anything that looks like I might have
screwed up, all I ask is for an explanation of why.  I'm a civil engineer
with a water resources background, so dirt and water are more my thing than
digital devices.  I do a lot of hydrologic and hydraulic modeling, so at
least I'm familiar with complex inter-related systems whose functions are
described with variables based upon approximations and known physical
constants.  I'm also pretty good at putting an analog tune on a car, and
have been tinkering for better than 20 years now.  Sheesh, I'm getting old.

Here are the details of what I'm using for chip reading / tuning / burning /
emulating:

Read/burn/emulate hardware is Craig Moates' APU-1.  With an adapter, it can
read GM OEM memcals, but can't erase or write to them.  Instead, it writes
to one of two types of loose chips (AT29C256 & AM29F040) that are attached
to the ECM with a an adapter that replaces only the chip part of the memcal
and leaves the resistor pack directly connected.  The emulate function
attaches to the memcal chip adapter by a ribbon cable from the APU-1.  The 1
Mbit NVRAM in the APU-1 replaces the chip - no EPROM swapping necessary
while emulating - and can be edited/replaced on the fly by connecting to a
laptop.  I can leave the APU-1 in place without the laptop, and the engine
will run with the calibration that's stored in the NVRAM.  FWIW, the laptop
is a 700mHz Toshiba running WinXP.

Read / burn / emulate / tune software is Mark Mansur's TunerProRT.  I'm sure
everyone knows what it can do, but in a nutshell: I select an appropriate
32KB .bin, choose the appropriate code mask ($8D so far), edit tables &
constants as necessary, and write the .bin to the chip on the APU-1.  I
really don't change anything but VE and SA values at this point.

The base calibration I have is what was provided to me by the guys at
www.AffordableFuelInjection.com, based upon info I gave them about my
engine/vehicle.  They burned the custom .bin on an OEM memcal for a '730
ECM.  This .bin runs the engine fine at idle, but is pig-rich, and is
virtually undrivable.  It is also the base file for my retuning attempts.

With all this in mind, I'll try to answer your questions below:

1,2) I'm about 99.44% sure the EPROM type I'm using is compatible with the
727/730 ECM.

3) I'm ASSuming that the TunerProRT software is placing the code at the
correct offset in the chip memory, as I have run the engine on both burned
chips and the OEM memcal.  I don't know what you mean by "absolute
addressing".

4) My create & burn details are listed above, but checksum & verify checksum
are two operations I'm not familiar with.  Again, I ASSume that this is
being handled appropriately by the software.

So there you go.  I'll try to answer any other questions as best I can.

Clair

----- Original Message ----- 
From: <dozierhc at aol.com>
To: <gmecm at diy-efi.org>
Sent: Monday, April 23, 2007 12:24 PM
Subject: [Gmecm] RE: More '730/$8D tuning questions...

> Clair...
>
> I guess we now need to start looking at details of your equipment, and how
calibration is transferred to the EPROMs.
>
> 1) What EPROM type are you trying to use (eg: 27C256-25, etc?)  Maybe
EPROM is too slow?
> 2) If using a 1227730/1227727 series of ECMs, are you using a 32K (27256)
EPROM, or something else.
> 3) If #2 is other than 27256 type EPROM, are you placing the code at
correct offset in chip memory, as well as absolute addressing for SW?
> 4) Maybe details on how you create, checksum, verify checksum, and burn a
EPROM can shed light on this, as well as the equipment used.
>
> Not trying to point out any mistakes here in the above questions, but
something is going on that should be easy to fix when thoroughly understood.
>
> Hank
>
> *****************************************
> Message: 1
> Date: Sun, 22 Apr 2007 20:50:31 -0500
> From: "Clair Davis" <clair.davis at charter.net>
> Subject: Re: [Gmecm] More '730/$8D tuning questions...
> To: <gmecm at diy-efi.org>
> Message-ID: <003101c78549$c7013380$03000004 at davis>
> Content-Type: text/plain;   charset="iso-8859-1"
>
> Well, I'm at a loss.  I've tried to copy over the necessary bits of my
> previous tune attempts (injector size, displacement, etc) to both AXXC and
> AXXD, supposedly known-good 5-speed .bin's, and in every case, they drop
in
> to limp-home mode almost instantly.  The only code I can retrieve is 12,
> ECM, so something is very seriously wrong.  If I'm switching .bin's on the
> fly, I can upload my old last attempt, and all gets better.  SES/CEL goes
> out, mixture leans out to stoich +/-, and MAP jumps up in to the mid 40's.
> I've been through almost all of the various items you can edit through
> TunerProRT, and anything obviously strange is evading me at this point.  I
> think about all I can do is go through the various tables that appear
> related to an auto trans in my .bin and copy in the parts of the 5-speed
> .bins I have (AXXC/AXXD/.  I don't know what that will really do for
> drivability, but at least it makes it a bit more "right".
>
> Now for the question.  My gut feeling is that I need to seriously adjust
my
> Accel and Power Enrich functions to make this thing more drivable in the
> near term.  I believe this is the case because when the car is in warm-up
> mode, throttle response and transition is pretty dang good, but AF is in
the
> 11-12 range.  I think that's fat enough to cover most enrichment sins
until
> it goes to closed loop.  Where can I find a good description of the
> variables in these functions?  I need to be able to change the variables
> that make the greatest contribution to the adjustment factor.  So far, the
> XDF file for TPRT hasn't really given me what I want, and the book I got
> from Customefi's has only discussed the BPW this way.
>
> I'm eventually going to need the same formula info for the spark advance,
> but I think PE/AE is the big problem at the moment.
>
> Thanks again,
>
> Clair
> ________________________________________________________________________
> AOL now offers free email to everyone.  Find out more about what's free
from AOL at AOL.com.
> _______________________________________________
> Gmecm mailing list
> Gmecm at diy-efi.org
> Subscribe: http://lists.diy-efi.org/mailman/listinfo/gmecm
> Main WWW page: http://www.diy-efi.org/gmecm
>