kyanox

Ok guys,

So i've been working on the 16196285 ecu decoding, and have come to an impass, more related to my not being able to understand asm code more than anything else.

Thing is, i've identified the hardware locations that store aldl data and so from there have been able to reverese some subroutines.

The thing i need your help with is, i've found a version of the $0D mask for the 16196395 ecu, and I'm wondering if they're hardware compatible with the 16196285 ecu or if i have to buy a 16196395 and convert it to the jbody connector.

The $36 and $F1 masks are what run the 16196285 as far as i can tell, and aren't commented very well, but the $0D is highly commented thanks to you guys here.

If someone could answer this simple question, i think i should be able to go from there. OR if someone is willing to work with the asm code, it would not only be greatly appreciated but I believe some of us over at Jbody would be willing to start a virtual collection plate to pay someone to do it.

HaulnA$$

I have been working with the $0D, $0E, and $31 masks and assosciated hardware for a few years. I don't believe the 16196285 is hardware compatible although it is possible. A PCM bench is the only way to find out. Your best bet is probably to convert PCM's. Be careful if you run an auto tranny though as there are some subtle differences between '94 and '95 code and hardware. HTH

dimented24x7

iTrader: (2)

What where some of the differences? The different TCC's where one that I can think of. Are there others? Since I cant use an auto even if I wanted to I havnt looked at that end of the PCM yet.

HaulnA$$

That's it TCC. In '94, the 4L60E used a conventional on/off type TCC solenoid, same as the 700R4 but in '95 a new valve body was added with an additional PWM TCC solenoid to smooth out the TCC operation. Some of the '94 wiring harnesses are wired for this, others aren't. Between model years, the $0D code for the 4L60E did not change, only the calibration. The changes on the 4L80E side, which I am in the process of researching, prompted a change in code from $0E to $31. HTH

kyanox

I run a manual tranny, but we're looking at a full solution for Jbodies, and while this isn't really an 3rd gen fbody thing, in a way it does use the ECM.

In the ecm swap to 0d, i've designed a harness, considering gm used relatively similar sensors between the years, and espeically the 16196395 used in an LN2 for the chevy S10. Our kit will consist of mostly the ecm, and converter harness as I believe the engine sensors are identical.

Long story short, the 2.2L jbody stuff also works on the 2.3L stuff which is what the 4cyl tuner's and racers are using.

Let me know if anyone has any info that could lead to a working 0D mask on 16196285 as this would save a lot of time in ecm conversion. Even if it means reprogramming hardware interupts to work, at least then the code would be commented.

dimented24x7

iTrader: (2)

Id imagine that youd probably have to go through both and match up the hardware, or at least see where comprable hardware is used. If you know what hardware is in use in the j-bod ecm/pcm, then adapting it would come down to revising the $0D code to use those outputs instead. There is alot of hardware in the custom HC11F1 they are using. It also has alot of RAM, and inputs/outputs (A/D, on/off, PWM, high frequency inputs, close to 1k RAM etc), so that may be some inputs/outputs or RAM thats not available in the PCM you wish to load it on. You might still be able to use the $0D mask, but some features might have to be disabled if the hardware isnt compatible or doesnt exist.

As far as the code goes, if you have the latest $0D hac that Ive been working on (in 'TBI guys, its time to update!') then that should give you an idea of what has to be done to make it compilable. Its rather tedious, but worth it in the end.

As far as the kit goes, I assume you will be selling it? Im ok with this, and will provide input, provided the code remains open source so we can all continue to learn.

kyanox

Yea the code will remain open source. The chevy S10 LN2 pickup has commented source code for it's memcal. The memcal is 0D, meaning your latest code would probably benefit our tuning files. Ideally what i'm looking at is buying a 16196395 ecm from the local junk yard, and doing first a harness to plug it in, then plugging it in, loading up the code and seeing if it will fire our test engine.

If it does fire, i should be able to connect with tuner pro rt and monitor to see it's operating ok. I also need to do a cross reference of LN2 parts off the 94 S10 vs the 95 + LN2, but i think they are close. If nothing else, there might be a few sensor swaps or recoding of the memcal to make it work. But like i said, the code must remain open source.

if your willing to help with the 16196285 code, that would be great. We have many users who are using this ecm with mantapart memcals who would rather tune with moates hardware.

dimented24x7

iTrader: (2)

A test bench would be the best way to test it. You can subject the PCM to whatever conditions you want and see how it works. The mid 90's truck PCMs where sort of unique so I wouldnt be surprised if it required some mods to the code to get it to work on the j-body PCM.

dimented24x7

iTrader: (2)

Re-re-edit: Actually, if you could, can you post what you have for the code in the j-body PCM? That may answer some questions right there just looking at how its structured.

kyanox

Sure, i'll do better than that. I'll post the src file.

The S10 LN2 pcm is setup for an LN2, i've checked the actual ln2 that is in the S10 truck and it's virtually identical to the cavalier ln2 physically except mounted front to back for rear wheel drive.

The pcm differences are auto trany routines, but i've not got that far yet, i'm manual tranny and we're looking at first a working port to manual trans then auto.

The hardware addresses for the aldl routines are in the source, its really just that i lack the asm knowledge. If you've dove into any serious pcm work, you could probably finish the routines and check whats there. There's also 2d table lookup, 3d table lookup, some FDB constants, but just to make sure see what you can do. If you get anything or find anything let me know, I'm not opposed to pouring through code you've already been through to figure something new. Also, any changes you make I will apply to our test engine and see if we can't figure it out.

I should have a test bench up soon.

dimented24x7

iTrader: (2)

The $36 is definatly different from the $0D, $0E codes. Its longer, the hardware is mapped to different address range, etc. All that would make adapting the $0D code difficult. On the other hand, though, in some ways, its similar. GM seems to have a peticular style, format, and structure in their codes, so things flow a certain way, even when comparing the early ECMs to the PCMs. They also like to share routines between different masks. I can see similarities just skimming through it.

I would say you have two options:

1) Adapt the PCMs running the $0D, $0E, and $31 codes to run in place of the 6285. This would be easier, but it would also require replacing the PCM. These where also only produced for a limited number of years, so finding them in large numbers could be difficult.

2) Use the current available info from the $0D and $0E hacks floating around to reverse the $36. This is definatly the harder option, but with the progress made with the other PCMs, you definatly should be able to make some good headway and be able to map out the $36, and $F1 masks. It took me a good 6-9 months to get the $0D to the point that it is currently. That was with little information about what the code actually did, so it involved alot of head scratching. That info's now available, though, so you could probably be able to go into your code, and find the similarities and figure out, at least in part, what does what.

If you've dove into any serious pcm work, you could probably finish the routines and check whats there.

I dont mind helping, but diving into another PCM right after the $0D project would probably make me go . I like working with them but even I have my limits Ive also spent way too much free time lately working with bits and bytes rather then nuts and bolts, so Id like to actually finish up the car and drive it before gas costs $100 dollars a gallon.

kyanox

Its understandable. I've just had a hard time decoding routines. The aldl hardware addresses are in that source, its just a pain to try to figure out the src when im not comfortable with assembly yet.

I'm actually adapting an 0D pcm from a 94 chevy S10 into my setup, and if its successful, i can pick these pcm's up for $50, as the local parts yard has agreed to search for them and sell them to me in lots of 100.

I agree it would be easy to swap in a 0D pcm, but even more i think it would be better to decrypt the 16196285 pcm, though the jbodies are either swapping in a pcm to begin with or going with hptuners.

kyanox

Where the ECM lists the aldl addresses in source, is that something in can change?

Meaning, can i change the code where aldl addresses are defined to display a different variable.

I was thinking to map the unmaped addresses i could change the aldl addresses i'm not using at the time, or simply pick 10 or so and change them to map out how each subroutine functions.

The thing that holds me back is how the data is processed, for instance if i find a subroutine that has to do with IAC but the only info i have is iac motor, there are also about 9 different ram addresses that are currently in use. What im thinking is, changing the aldl table to pull the ram data to the aldl table and display it in the aldl output. I realize i have to relabel my .ald file for the corresponding outputs, but this would help me effectively study each routine with an output i can store to xls in rt tuner pro.

What do you guys think?

83 Crossfire TA

iTrader: (2)

I hope you guys don’t mind me asking some general “point me in the right direction” questions going in a similar general direction, but I’ve been looking for info on the ’95 J-body, 2.3L ecm for a while now with what seems like no real concrete info going anywhere (I’ve seen a few mentions of the J-body OBD1 ecu project around ’06 but no info).

Trying to confirm, the 16196285 is used in both the 2.2 and 2.3 in ’95 (wife’s car, can’t just take it apart…)? Are they both $36 masks? Is there a scan tool out there that works on them (I really need to figure out what this thing thinks it’s doing)? Is there an .ecu, .xdf or tuner cat definition file that works on the ’95 2.3/quad4 engine?