[Gmecm] Back to square one

Sat May 13 04:16:01 UTC 2006

The main VE tables cover up to 1 BAR.  In 2 BAR mode, higher pressures 
use a fuel multiplier against the 100 kPa entry for the appropriate 
rpm.  Spark tables cover the range from 30 kPa to 200 kPa.  I haven't 
looked through the code to see if flipping the 1 BAR bit forces the 
code to use the 20 to 100 kPa entries, or if it rescales the table for 
100 kPa max with 2x the resolution.  If it's the latter, then the 
values in the timing tables would be incorrect for the MAP range they 
apply to.

What I can say with certainty is that flipping the bit to 1 BAR and 
connecting to a 1 BAR sensor produced poor idle and driveability in my 
car.  I had no desire to chase the reason, I simply switched to a 2 BAR 
sensor.  I valued the time spent more than the potential for education.

Wasn't making fun of you on the code modification.  I hope it didn't 
read that way.  That's just the way I talk sometimes.

There used to be a turbo Sunbird calibration in the .bin library.   I'm 
not sure where they disappeared to.  Are you using a manual or 
automatic transmission?  If you want to send your mystery sunbird.bin 
calibration I can probably identify it for you.  There are several 
factory calibrations for the 'Birds.

You can load the 4 cylinder calibration directly onto the memcal.  
There's no need to play with the 6 cylinder version.

Yes, the information you linked to on Thirdgen.org looks like what I 
was thinking of.  Ludis Langens made a nice equivalency chart for 7727 
<-> 7730 pin functions.  Sorry, I don't have a direct link.  I just use 
Google.

The AYBN hac which is floating around was likely authored by the same 
person who wrote the 8D hack.  It's not complete, nor is it completely 
accurate.  But it is a decent starting point.  It's acceptable for use 
with the Sunbird code, although certain minor functions differ between 
the two applications.

Zaphod

-----Original Message-----
From: Robin Handley <Robin at FuryWorld.fsnet.co.uk>
To: gmecm at diy-efi.org
Sent: Thu, 11 May 2006 10:31:43 +0100
Subject: Re: [Gmecm] Back to square one

   > $58 for the 7749 will work in an NA application.  It's going to 
expect the
2 BAR MAP sensor.  To use the 1 BAR sensor you'll need to re-scale some 
of
the tables.

Doesn't %KAFOPT3% bit 5 tell the code whether a 1 or 2 bar MAP sensor is
connected? Does this not cause everything to be scaled automatically?

> The turbo Sunbird was the first to use the $58 code.  The Syclone and
Typhoon were second.  It's better to start with a calibration for a 4
cylinder car.

Agreed. So, could I take the AYBN commented hack for a 6cyl 4.3 Turbo 
Sy/Ty
and overwrite the lookup values with those from a genuine 4cyl Sunbird 
.BIN?
Are they a direct map?

> The $58 calibration does not have tables which extend to 7000 rpm,
although it does provide a means to adjust fuel and spark after rpm 
exceeds
the maximum tables entries.  It probably won't be an issue.  At that 
engine
speed you're not usually looking for high resolution tuning.

Agreed.

> Other people have used the 7749 calibration in a 7727 / 7730.  I 
believe
there's a page on the thirdgen.org site which details the correct 
pinout,
but I might be wrong.

This is where I found my pinout conversion info:
http://www.thirdgen.org/techboard/diy-prom/213539-58-frequently-asked-que
sti
ons.html

Is this the one you meant? I guess it just needs mapping from 730 to 727
pinout for my ECM. I was just looking for confirmation that I hadn't 
missed
anything subtle, since the above (and other) discussions just talk 
about $58
code on a 730 (not a 727).

> The turbo Sunbird calibration is 16k large.  It will fit on a 32k 
eprom
and work.  You will need to move it to the "top" of the chip.  Look for
instructions on "bin stacking" for one method to do this.  The memcal 
must
match the calibration in the number of cylinders it's designed for or 
else
there will be a "cylinder select" error set.

No problems there. I've now looked at the AYBN hack and its associated 
.BIN.
The hack is 16k and the .BIN has the first 16k padded with $FF - so 
already
in the right format for my 32k EPROM.

> To make the "special secret super duper code modification" you'll 
probably
need to learn the $58 code.  There may not be an equivalent function in 
the
$58 as you worked with in the $8D.

The 'not actually a quote from my Email' code modification allows OTF 
map
changes from a laptop, so I'm pretty keen to be able to reproduce it! 
:-) I
just wondered if there was a Sunbird specific hack that I've missed - 
that
might make code changes easier. If not, am I right in thinking that 
AYBN is
the best/most commented $58 hack to use? The comments with the AYBN 
code are
a LOT sparser than ANHT :-( although the description of the lookup 
tables
seems better - and the Sunbird document that I've seen is a massive 
help.
:-)

BTW: Is the Sunbird code/bin unique? Does it have a 4 letter name, like
ANHT?

Thanks,

Robin

> -----Original Message-----
> From: Robin Handley <Robin at FuryWorld.fsnet.co.uk>
> To: gmecm at diy-efi.org
> Sent: Wed, 10 May 2006 10:04:42 +0100
> Subject: [Gmecm] Back to square one
>
>
> After a couple of years of researching and playing with GMECM, I 
thought I
> was getting my head around it but, having started thinking about 
changing
> the code in my 16149396 (1227727 with SRAM) from $8D/ANHT to something
> possibly better hacked and easier for me to understand all the
arrays/tables
> required in applying it to a 2 litre 4 cylinder naturally aspirated 
engine
> that will rev to 7000+RPM, I am now confused and need help/advice.
>
> I've read quite a lot recently about Turbo Sunbird code and it being
> identical to Sy/Ty code. What I've read tells me that both originally 
ran
on
> 1227749 hardware, that they are part of the $58 family, and that the 
EPROM
> size fitted to this hardware was 16k, whereas 1227727 & 1227730 
hardware
> came with 32k EPROMs fitted (what I have). Also, that it is possible 
to
run
> $58 code on a 1227730/1227727 if a few ECM pins are wired 
differently. I
> don't have or want IAC or EGR or any transmission connections, so I 
think
> the only one that might matter here is the wire to connect to my
injectors.
> Have I got the above correct? :-)
>
> So, if I am to do the switch from $8D/ANHT to $58/Sunbird code, and 
get
back
> up to where I am now, I need a .BIN (I have something called 
Sunbird.bin,
> but no idea what 4 letter designation this relates to) and I need a
> commented hack - for ALDL monitoring, reference, and code modification
(I've
> done a nifty change to ANHT that I really need to recreate with any 
new
> code). A .ECU would be nice, but I am capable and happy to generate 
one
from
> the hack.
>
> I've looked at the .BINs and hacks on www.moates.net (which seems to 
be a
> copy of the www.diy-efi.org equivalent, but with hacks) and found 4
1227749
> $58 4.3 Turbo .BINs (1576BBZA58, 1618BBZB58, 4175AYBN58, BBZB1580 - 
all
16k)
> and one $58 hack (AYBN - which is configured for 6cyl but mentions 
Turbo
> Sunbird in the header, and the accompanying .BIN is 32k !!! - This
confused
> me, can anyone explain?)
>
> Could somebody with a good overview of GMECM put me straight, please. 
:-)
Is
> the Turbo Sunbird code the best start point for a 2.0 4cyl NA engine? 
Is
it
> the exact same code as the 6cyl 4.3 Turbos, but with different lookup
> values? Will Sunbird code run on 16149396/1227727 hardware (I'm in 
the UK,
> so getting new ECM hardware is not trivial) with a dizzy reluctor 
trigger
> feeding an 8 pin GM HEI? (I don't have/want EGR, IAC, or any 
transmission
> related stuff - just something that will run my NA engine's ignition 
and
> injection - after tuning (obviously) - but with clear visibility about
what
> needs to be set, and what it actually does.) BTW: I have a WBO2 for
tuning,
> and a MAP sensor with the following ID: 460 7236 (which I believe to 
be a
1
> bar sensor, but this number does not match that on a web site of GM
> sensors!)
>
> If so, please advise which .BIN should I start with, and which hack? 
And
can
> I reuse my MAP sensor (I believe the $58 code has an option for a 1 
bar
> sensor)?
>
> Help! (Sorry for all the questions.)
>
> Robin

_______________________________________________
Gmecm mailing list
Gmecm at diy-efi.org
http://lists.diy-efi.org/mailman/listinfo/gmecm

___________________________________________________
Try the New Netscape Mail Today!
Virtually Spam-Free | More Storage | Import Your Contact List
http://mail.netscape.com