TBI guys, its time to update!
 

I have just swapped my '7747 for a '8625 pcm the results are amazing!!
The advances over the 7747 include 8192 baud data, full size VE and spark tables, 50% TPS AE table, much easier to use idle control and control over a 4l60e if desired!
The list goes on....

The 8192 baud data is an awesome step forward. I have been using datamaster and have already used its feedback to make my idle better than it ever was before. Having data on SA, BPW, injector DC and a whole host of other variables makes tuning soooooo much easier. It has removed a lot of the guesswork.
I have attached a couple of datamaster screen shots so you can see what you are missing.


http://users.bigpond.net.au/ben_p/co...datamaster.jpg
The main datamaster screen showing a startup at idle from cold to closed loop. The graph can show any 4 variables at any time and you can zoom into any section for a closer look.




http://users.bigpond.net.au/ben_p/corvette/dm/blm.jpg
This is the blm histogram overlay from the 1st chip i started it up on, needs some work huh ;-)


http://users.bigpond.net.au/ben_p/corvette/dm/bits.jpg
These are all the status flags. As you review the logged data these change at the appropriate time!


I know the TPI guys are probably looking as saying 'whats the big deal', but this is all new to anyone who is still running a 7747 or 8746 etc....

That's pretty cool. How easy was the swap over and what did that pcm come stock in?

It is cool!
It came in '93 light duty trucks w/4l60e, #16168625, running $E6 mask.

The 94/95 one would probably be easier to find, and its almost identical as far as i can tell, it runs the $0D mask, but i forget it's part no...

Ben.

The service numbers for three of the other 94/95 PCM's are 16197427, 16196395, and 16156930. Any of these can run several different Mask, 0D, 0E, 31, E6, 8C.


Ben73,
Do you have a pin to pin cross over chart for the swap??

Are you using it in a 700R4 or 4L60E tranny application??

What is your application??

Sorry for all the questions, just very exited about this!! This is something I have been working on for some time, and it's good that someone is doing it as well.

Thanks
89350STRO

I was intending on doing this as soon as I finished my current projects. Doesnt this series of PCMs have dual processors?

I didn't draw up a 'cross over chart'. I made an 'adaptor harness' using pinout drawings from my original 1226026 and the new pcm. It was simple enough to do without having to draw it out.
Current application is a 383, afr195's, roller cam, fed by a 4bbl holley 900cfm pro-jection tb.
Currently have a 700r4 near death, and will be converting to 4l60e soon.
Ben.

Anyone interested in developing a HAC for this? I'd like to learn how to do a disassembly on a real project.

There is a hac out there already, thats how I was able to create a .ecu for tunerpro. I think i found the hac at diy-efi.org
Ben.

What are (if any) the differences in the way the 3 are wired as I'm going to swap to a 4L80E when time permits and use one of those with $31 code for the trans control. I've been searching and researching and have not come to a conclusion on the wiring between the 3.

Thanks

Ben73,

Is the 1226026 the same pinout as the 1227747?

If possible could you post or email me your pin to pin comparison?

Did you have to do anything special to get the 700R4 TCC to work?

I have a 1227747 to 8625/7427/6395 pin to pin comparison and I just wanted to check it against yours since you have completed the swap.

eric305TPI,

It is my understanding they are all direct replacements, the service number 16196395 was superceded by 16197427.

89350STRO

Cool, also if you don't mind could you E-mail me the 1227747 to 8625/7427/6395 pinout if it is accurate.

89350STRO,

I'd appreciate the 7747 to 8625 Pin to Pin layout as well. Been talking to Ben about his swap and it seems like a better alternative than the 7749 swap.

Yeah, this sounds like a pretty solid swap. The only thing that would keep me from swapping over is that I have a Prominator hooked into my 8746 now.

Here Ya go,

I have attached a 16197427 to 1227747 pin to pin chart. This is untested.

Ben73,

Could you verify the pinout is the same as you used?

89350STRO

edit

well I tried to attach a file??!! I'll email it those interested.

Eric, Dominic you have mail.

Shes all yours:D

I was thinking of possibly doing it but I dont have much free time. I might work on it, though.

BTW, where are the pinouts for this ecm listed?

EDIT:
NVM. Found them

I was thinking, how 'bout we publicly hac it? Like on that WIKI site there could be a page where people post commented routines from the code. Start off at the power on vector and then work from there and go through it just as the ecm would. Not quite sure what the best way to organize it would be, though. There would also have to be a standard for how things would be posted and commented.

When I get home tommorow Im definatly going to take a stock bin from one of the download sites and disassemble it to get a fresh start.

I have a $0D mask, BCC BJYL, that I have succesfully dis and reassembled and it compares correctly. I have started to hac and have had a little success. I would be willing to share what I have so far, or, if you just want the BIN, let me know. This is a bone stock untouched bin file that I personnaly read.

I think doing it on the wiki site is a great idea. Anyone Else?

Count me in, I know several other people I have been talking to here that would be interested. But I'll let them speak for themselves.

First I think it would be a good idea to latch onto a common BCC. Such as the BJYL, YES, of course I'm biased.

89350STRO

I can't really help with a 7747- 8625 pinout as I made a 6026-8625 harness.

I just had the 2 pinout drawings in front of me and worked it out as i went, its that simple...

The other thing you must do is remove/bypass the ESC module by jumpering the sensor direct to the ecm. Also needed is a 3.9k resistor from ground to the KS input, or a later type KS.

The other issue is the VSS. All i did here was change the 'pulses per revolution of tailshaft' parameter in the EPROM from 40 to 2 and it works perfectly. Diacom matches what my speedometer says.

If someone wants me to draw up a 7747 - 8625 conversion pinout listing I can do that.

Here is the link to the $E6 (16168625) hac.

I have created a good .ecu file for tunerpro, but i just need to add the transmission stuff before I make it public
Ben.

Ill take a copy of the bin if you wouldnt mind. E-mail is dimented24x7@aol.com

Ben,

Send me the 6026->8625 mapping and I'll convert it to
7747->8625. I have the 6026 -> 7747 mapping from my conversion. Since I did mine prior to the availability of the Xfire to 7747 adapters, my 7747ECM is hardwired to the original harness.

Are the pins in the connectors on the 7747 and 8625 the same? I looked on this site and it looks like they may be. If they are then it should be possible to simply remove the pins from the 7747 connectors and insert them into correct locations in the 8625 connectors. Is it that simple?

The pins are identical I can verify that as I have de-pinned both the 7747 and the 8625 connectors.

Dominic,
When you get the 6026 to 7747 to 8625 pinout completed shoot me a copy, I want to compare.

89350STRO

pin out attached

It only takes this dumb ol country boy three tries to figure he was attaching a file in excess of 102500 bytes. (always read the directions last!!!)

The "Not Avail" pins must be made up in a transmission wiring harness, If you plan on using the 4L60E tranny.

The Mod B-7 is the modification Ben is talking about that must be done to the ESC module and replacement of the Knock Sensor or modify the circuit.

89350STRO

Maybe?
Even if the pins fit, the 'keys' that prevent you plugging it in the wrong way are different.
The later PCM has the 'keys' on the connector in different places, so the connectors won't plug into the later PCM without filing the keys off your old 7747 connector.

I'll sit down and compare that to what i did later. One extra thing I did was connect a couple of the trans wires so the PCM knows when the trans is in pk/n or one of the forward gears.
Thats just a temporary measure, as i'll be fitting a 4l60e before long.

BTW, I haven't experimented much with it yet, but in the hac i found flags to set the PCM for 4l60 (700r4) OR 4l60e, OR manual trans.

I completed my tunerpro .ecu file for $6E last night. It now has all the engine and trans parameters that tunercat has, and a whole lot more..
Its ready for distribution.... ;-)

I was just thinking about the pins. I know the connectors need changing, especially on the one that is smaller on the 7747. Sounds like the pins can be removed from the 7747 connectors and inserted into the new connectors.

I'm planning on doing a swappable so I can swap a 7747 back in if needed especially during the tuning portion.

LOL! Just like me!. I dont read 'em untill I break it, then I read the instructions. Just like not asking for directions when your lost:D.

Thanks for that pin-out sheet. Ill have to hang onto that.

On a side note:

I disassembled that bin, looked at a pre-existing hac, and then familiarized myself with the 68HC11 a bit. Then I commented a couple hundred lines or so of the disassembled bin to get a taste for it. All I can say is its way more then Im looking for. Its actually a REAL computer instead of a bunch of transisters tossed into a metal box like my TBI ecm. For me that would be like driving a finishing nail with a sledge hammer. It looks like a really sweet peice, but I think Ill stick with my transisterized carbeurator:D.

You were able to comment a couple hundred lines in an evening!!?? thats twice more than I have done in a couple of weeks!! Please, can you share what you did, are you still interested in the project?

89350STRO

Im not going to stop working on it if thats what your thinking. I probably jsut wont do any serious source coding on it since its so damn complicated compared to my tbi ecm. Keep in mind that Ive hac'd and done lots of assembly on my 8063 so Ive already gone up some of the learning curve. The only things ive done so far are some of the startup and subroutines and I think know what theyre doing, just not what theyre there to do. I also looked in the 'pink book' or whatever its called and started to learn what the registers do. Its much more complicated then my tbi computer. In that theres nothing more then a pinch of ram, a few outputs, and a couple of accumulators. Cant get much easier then that:) This thing looks like it can even write to an EPROM, giving the possibility of flashing teh ecm directly via the UART or something. Id think Id need a few more letters after my name to really understand and work on it:D.

I found that a good way of making progress is to look for accesses to the special registers, like the SPI port, to give you an idea of what is happening.

Another thing I like to do is follow JSRs and BSRs to get as far down as possible... it's like pulling on a thread, everything starts to unravel. :)

Thats teh same thing I do almost. I follow the code as it would be executed by the ecm/pcm.

What dimented24x7(x365) is saying here is important. The more complicated the code the more difficulty there will be in tuning it. With any Mask/ECM one can only tune it if they understand how the tables/parameters operate.

That is exactly how I got so involved with the code. In order to enact valid changes I needed to know how the tables/parameters affected the fueling and timing.

Without knowing that its like throwing darts into the wind, on a dark, stormy night ;)

RBob.

So that's why I can't hit the Bull’s-eye!! Not enough light and too much movement caused by outside sources?

Seriously, How do I find such info as the "special" registers, memory maps, etc. I have looked, searched, prodded, read, and read some more for info on the 68HC11. Here is what I have uncovered so far and believe to be correct;

Please feel free to correct any mistakes and add any comments.

Moved to: http://wasabi.dynu.com:8080/wiki/ind...8HC11_%240DHac

89350STRO

uhmm, fully commented $0D file has been out for ages, check http://www.moates.net/fileman/ look at source code and hacks

Its a good start, but its still very incomplete.

Yes a good start but very incomplete, agreed.
Thats why I decided to attempt commenting, a dis, with the assistance of others hopefully. Keep in mind I'm new and still on the real steep end of the learning curve.

It's just like everyone has said, if you don't know how the PCM's innards, (love that country slang), work then it makes it diffcult to understand the tunning, and if you don't understand the code for yourself, then you can't know the PCM. Just my opinion.

I have done a chip to chip comparison of the 16168625 and the 16197427, the chips are identical, the board layout is the same except for a few caps, resistors, inductors being left off/added on the newer PCM, looks like a few traces may be moved around and eliminated as well.

Does anyone know how to add a "new" page to the wiki page? I would like to start adding my findings there if possible.

89350STRO

Take the base address of the site like this:

http://wasabi.dynu.com:8080/wiki/index.php/

and add the name of your page on the end:

http://wasabi.dynu.com:8080/wiki/index.php/The0DHac

This link will work because it will automatically start a new page for you.

Feel free to call it whatever you want.

Yeah, the whole HUGE code that I don't understand is a bit intimidating. It definatly has potential of being a good ecm if some guys get together and put together a nice commented hack for it. The ability to have all code on chip makes it nice too. This way you could run just about any code you want with some code work, no hardware work needed.
I find the c3's in stock code form to be more confusing and harder to tune than a p4. The adder tables to save size really aren't all that intuitive. Same with the open loop AFR tables, 2 2D arrays instead of a 3D table.

Hmmm, so with the Ultimate-TBI code I should change to a 3d desired AFR table from the two separate tables? Only needing to do a AFR vs. coolant lookup 5 time a second to a 3d table 80 times a second adds to the overhead, that's all. Can-do otherwise. . .

RBob.

RBob,

Was that english you just spoke???

AlexJH,

Thanks, I have created a page at the following location for any and all things related the $0D BJYL hac.

http://wasabi.dynu.com:8080/wiki/ind...8HC11_%240DHac

Please join in and help out. I will post what I have which isn't much and not sure if its correct, but hopefully others will jump in and correct my mistakes.

89350STRO

I changed to a 3D table in place of the 2D stuff for the open loop AFR. Same info that the 2D ones had, but its definatly more pleasing to the eye.

Sweet... Are tehre any schematics for this pcm?

None that I have found, any volunteers to trace it out??

89350STRO

Hey guys I just was looking on ebay and found this '6395 for sale, thought you guys that wanted to look at one might be able to pick it up cheap.

Here's a question for you guys who are working thru the code, its beyond my knowledge....

I have pinned an extra wire into my '8625 and now have a 'spare' input working at the MAT sensor input. I have a pot connected and I can see its working in datamaster. It has no effect on the spark or VE, as MAT compensations are turned off in the .bin at $400B, bit 6.

I want to use this input for my 0-5v WBo2 output. It is working now, but I'd like to be able to change the scaling on the MAT to match the WB, so I don't have to remember that 12.4-12.8 AFR is 48.5-44.8 degC on the MAT indicator.

I'm guessing that the scaling is handled by the scanning software (datamaster) and the only way to achieve this would be to create my own aldl definition using tunerpro.
Is this correct??

If this is correct, is it possible to alter some code in the pcm to scale the output so that datamaster reads 125degC (or degF, whatever) for 12.5 AFR and 147 for 14.7, 130 for 13.0 etc etc...

Ben.

I don't think it's necessary at all, I'm just pointed out the hassle of having to do the math either by pencil, in your head, or with a program like excel.
If you want I could offer up my software for the ultimate TBI code to make it more user friendly. It'll do all the combination charting, smoothing, self-tuning, blah blah blah, it goes on :) .

The temp inputs, generally have two different pull up resistors, they use, for better temp resolution, I'd suggest you use a spare MAP or TPS input. Not an EE just repeating what I've read.

Thanks Grumpy, but how do you mean 'spare' tps or map inputs? Are there more than one of each?

I imagine that there must be an a-d lookup table somewhere that converts the non-linear MAT voltage to 0-255 raw counts, and this is what ends up in the aldl stream.. I was hoping to manipulate this a-d conversion to match the curve of my wb, giving 147degC for 14.7 etc...

Not that easy????