16197429 Documentation Project

[RobertISaar]

initial purpose of this project is to determine just how capable of a unit it is for an entirely fresh codebase. as in, throw away 95% of all GM code(keeping useful subroutines) and make whatever you want of it. as such, this is mostly a hardware based project to see if the software is worth persuing. i do have a rough and sparse XDF built already though the checksum needs bypassed due to the complexity of it not yet being supported by tunerpro.

DIS can be done with any ECM, but yes, SFI on an 8 cylinder engine would obviously be possible.

[The Dark Side of Will]

The unit obviously has a lot of IO... what would you want to do with it?

One shortfall it has relative to more modern computers is that it only has one ignition output, which limits it to DIS or distributor and precludes coil per cylinder.

[RobertISaar]

depends how you utilize it..... could hook up more than one PCM and use what would be the injector outputs on a 7429 as coil driver outputs.

there is also the "simple" way of using CPC as waste-spark and just use the DIS ICM to trigger the coils.

[Caleditor]

Just wait until you get a few cals read out. It will make you rethink tuning them all together. I have seen several cals and maybe 2 of them would work on one def file. I know one of the calibrators that built them at GM

[RobertISaar]

i've got a couple...... i think 1 or 2 may have come from you actually. they appear to be split by year.

besides, i don't plan on really playing with the stock code at all.... if i do anything with them, full custom code.

[brian89gp]

How easy would it be to transfer the needed pieces of P4 code ($A1 and $2E which is documented and simpler) to the P6? It sounds like that might be what you are contemplating with of course all of the custom stuff? Other then addresses and I/O moving around, my poor memory seems to remember the P6 added a couple instruction but didn't necessairly remove any?

[RobertISaar]

easy? kind of.... i'd rather go full custom though. there are a lot of odd/too compromised things that happen that i'm tired of patching around in factory code.

addressing and I/O really is most of the rework needed.

P6 added XGDX and XGDY instructions off the top of my head.... which could save a few bytes in the code the couple of times they're needed. no instructions removed. searching for XGDX instructions is an easy way to tell if you're playing with a P6 vs a P4 actually, assuming you don't have the ECM in front of you. it seems that the cadillacs actually introduced them IIRC.

[RobertISaar]

semi-related since they can both have similar end-games:

grabbed a 96 LT1 BIN to see how the bankswitching is done.

total BIN size is 144KB.

seems that only one side is bank switched. one side, the actual BIN is E000(56KB) long, starts at $16000. that leaves 88KB for the other side(probably T-side if i had to guess based on how the 94-95 unit is setup).

0000-DFFF is a block, which leaves a block that is 8000(32KB) long.

looks fun to play with.

[Caleditor]

The 2 chips on the memcal are ECM and the other is TCM/TC

[The Dark Side of Will]

My Northstar .bins. There's a .txt in the .zip that says what everything is.

[RobertISaar]

interesting....

breakdown based on the BINs i've collected appears to be:
93
94 and 95(non-aurora)
95 aurora

the manual trans version's algorithm matches the 94-95 version with the exception of the blank bytes being set to FF as opposed to 00. those could be changed to match, but they're of no consequence. with those accounted for, the closest calibration i have is BKFX, which is 94 Seville/Eldorado with the LD8/3.11 gears.... total difference of 884 bytes, over half of which are just in the VE tables. haven't checked the trans side yet.

odd that the LO version would be selected for hooking a manual trans to instead of the HO.

[RobertISaar]

more interesting stuff: transmission side of BKFX and BKFZ are identical, with only the ID byte changing between them.

1,316 byte difference with manual compared to BKFX on the trans side.

[The Dark Side of Will]

My understanding is that the manual transmission calibration was used for dyno evaluation... I also got it from a guy who had a 275HP engine, who got it from a guy who worked at GM. I hope it helps you understand the code and hardware.

With 1316 bytes different, it's obviously not a simple switch that says "transmission enable".

interesting....
93
94 and 95(non-aurora)
95 aurora

odd that the LO version would be selected for hooking a manual trans to instead of the HO.

Which is which, out of curiosity? Do you have a listing of all the BCC's used with the Northstar versions?

[RobertISaar]

of the BINs i have:

BBSX: 93 E/K HO
BKFX: 94 E/K LO
BKFY: 94 E/K HO
BKFZ: 94 Deville LO
BMRY: 95 Aurora w/3.71 FDR
BMTU: 95 E/K HO

[WASyL]

of the BINs i have:

BBSX: 93 E/K HO
BKFX: 94 E/K LO
BKFY: 94 E/K HO
BKFZ: 94 Deville LO
BMRY: 95 Aurora w/3.71 FDR
BMTU: 95 E/K HO

can you share with that MRY: 95 Aurora w/3.71 FDR bin please ?

Are these 27c512 eproms are Texas Instruments ones or other brand/manufacturer ?

best regards