Quote: Originally posted by ben73 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???? What kind of files does datamaster output? It may be easier to do the adjustments in a spreadsheet program like excel after you do the scan. Or maybe datamaster can do the conversion for you. A lot of the parameters are sent at the binary numbers and converted to human readable by the scan tool.

I was hoping I could do the same with my ecm but the MAT works differently tehn a MAP or a TPS voltage type input. It probably has a resistor from a +5 volt source to the MAT input pin. Changing resistance of the CTS/MAT changes the voltage at the pin. Youd need to do some soldering to actually have it sense a 0-5 volt input.

That would be a good idea worth checking into. Not that it would make a HUGE difference but having the ability to use a thermocouple in place of a resistor style cts would make turbo stuff a lot quicker to react. That all depends on how often the ecm looks up the cts which might negate the faster acting thermocouple. Just an idea to chew on .
Or if you're already in the c3 dim, just solder a few of the unused a/d chip pins to the some unused pins in the connector.

Typically, yes, if you go to Ludis's site and look at the ecm wiring for the 730s, 749s, 8253s there's several unused inputs, also some unused outputs.

Nope, it's not as easy as you might think.
You need a diagram of the pcm your using, then figure out the hardware reference in the code, and then write a patch similiar in execution to what the 749 WB patch does.

I am now speaking (been away from home for 2 weeks). Count me in. I have a vested interest as I have been working with the $0D mask for nearly 2 years now. Hopefully I can contribute to this project. Glad to finally see the interest in the PCM's on the rise.

I wired up the unused pins over teh weekend. Havnt had a chance to try it out yet, though. Was hoping to try it and teh car out this week but I had a bit of a problem with the rear. Broke a damn tap off in the yoke when I was repairing the threads.

The CTS seems to fall into the non-critical catigory. The lookup may only be done a couple of times a second. However, the lookup simply could be moved to one of the main fuel or spark routines.

Quick question for you guys.
What if any advantage does the 8625 have over the 7060 pcm?
I ask because I know the 7747 to 7060 swap has been documented before & gives you the 8192 baud data and transmission control capabilities. I'm considering a 7060 swap myself but would consider the 8625 if there are additional advantages.

If you're gonna do the 7747 to 8625 swap consider the 16197427. It is the latest and greatest. It superceded the 8625, 16156930, 16196395. This PCM can be used with several Mask, $0D, $0E, $31, $63, $8C, or the $E6. The $0D is probably the most widely used and best understood of the above mentioned Mask for the 16197427 PCM.

89350STRO

I think you will find that the 7060 is made to control a 4L80e and the ones we are talking about control a 4L60e.

Any of the PCM's (8625, 16156930, 16196395, 16197427) will control the 4L80E or the 4L60E or for that matter the 700R4 (see Ben73's application above), all you have to do is change the mask being used. The $0E is one of the mask for the 4L80E there may be others.

89350STRO

Well, time for an update...
I have been racing and driving around with this PCM for some time now and I must say that it is LIGHT YEARS ahead of the older 7747 and 8746 ECM's.

The good news today is that I just managed to get the PCM to control the 700r4 TCC!!! It took a fair bit of experimentation, but it is now working... So the swap is now complete and fully functional ;-)

Care to share that TCC knowledge? btw, awesome work.

Sure, basically, you connect the TCC to the 3-2 shift solenoid output and enable the 'shift light' function in the bin, along with the autotrans and 700r4 flags, and also tell it that there is no DRAC module..
It is pin E2.
Its not difficult, just requires a combination of elusive changes to make it all happen ;-)

Ben,

At what point did you utter the magic words "Presto, Chango"?

Good work!

I finally started working on a good $0D hack for this damn thing so I can do some coding on it. It looks like a real PITA.

One question, though, is this a dual MPU PCM? I heard, and possibly see, that it has two MPUs to help with the control of the trans.

Dunno about that.. I can pop the cover off and look if its going to be visible/obvious??

I use the $E6 code, but its almost identical to the $0D. I have a good hack and .xdf if you want to use them?

Hmmm. I dont know if it would be obvious but Id assume youd see two large surface mount chips? Ill have to get one for myself.

As far as the hack goes, every little itty-bitty thing has to be commented and understood if your making changes. Real pain in the butt. Im building on top of the $0D hack and adding further comments to the swaths of code that arnt commented.

It is a single processor PCM. See pic.

I guess all those rumers are just that, rumers.

Its apparent that the engineers are really out of control when theyre programming the more complex PCMs Cant even fathom what the newest pcms look like.

1000 text lines in to the hack (out of nearly 50,000 ) and theyre already using dual 256 byte coolant temp tables, each with their max resolution located at a different area, to form a high resolution coolant term. They even base the filter coeff for the slow filtered tps off of coolant temp. Looks like there are a whole slew of factors for just about everything imaginable. And, I havnt even gotten out of the power-on/reset vector yet. I guess this is GMs take on 'Its not more then you need, just more then your used to.' I think a quote from Scotty best summarizes how I feel: 'The more complicated you make the drain, the easier it is to stop up the plumbing.'

I know this helps with drivability, but damn, it looks like you could spend eons tuning all the little this' and thats. One could just leave all this stuff alone, but then all of it turns into little gremlins that make the computer do unexpected things.

dim,

I think Ben73's experience has been that he can leave most of this stuff alone and have it run better than a highly tweaked 7747. I'm hoping to do this shortly. Have the 8625, harness adapter, and K/S mod done.

Theres definatly no doubt that itll run better then a stock 7747. The big question for me is, how much of that will have to be touched?

Having worked with the 7747 for the last 3 years, the 8625 feels like your drinking out of a firehose. It would have been nice to stick with the 7747 albeit with 8192 baud from the Speedreader and the increased resolution available through RBob's Ultimate code. But I'm not sure of the timelines on that project, and I need somenthing that allows me to extract better idle, part throttle and WOT performance than what's available through the stock 7747 code.

Dominic,
Once you get going with the later PCM, I'm sure you will agree it is easier than working with the 7747, especially with the Ostrich and proper datalogging capability via Datamaster. Because you can see EVERYTHING thats going on via Datamaster, it removes all the guesswork and experimentation/trial & error stuff you do with the 7747. ie, you can see what it is trying to do, target AFR's, actual SA, target idle, IAC targets etc etc..
Imagine how much easier all that 7747 idle crap would be if you could see what the ecm was doing! How many times have you made a change with the 7747, like 'Cold Comp. Spark Adv." and started the car back up and wondered if it is now actually doing anything different??? Now you just type a new number into Tunerpro, the 'O' updates it in the blink of an eye, and then you instantly see the change reflected in Datamaster. When you have come from a background of stopping the car to remove and burn chips, you will be far more productive this way.. No more guessing/wondering..

"Now you just type a new number into Tunerpro, the 'O' updates it in the blink of an eye, and then you instantly see the change reflected in Datamaster. When you have come from a background of stopping the car to remove and burn chips, you will be far more productive this way.. No more guessing/wondering.."



Hurray!!!

Interested in doing the swap, but have a few questions...

1) what is the B7 mod and (why not or can I) use the single wire knock sensor that works with the 7747?

2) what sort of O2 sensor came stock on the truck that ECM came out of (#wires?)

3) My truck currently has a TH350, eventually going to swap that for a T56. I don't *really* care about the transmission as it only has to last for another 500 miles, tops, but would running in manual mode destroy the auto, or would running in 700r4 mode set a code?

4) Having only done some tuning on the 7747 I'm not really well versed in "masks", what is the stock mask on the 8625 and what others is it capable of running?

5) what is the stock EPROM in the 8625 and has anyone used the ostrich with the 8625 yet?

1. ok from memory, the early (7747) type knock sensor is 100kohm and the later type is 3.9k.. if you don't shunt it with the resistor, it throws a code and is not happy.. with the resistor in place, and the ESC module bypassed, all is well.

2. dunno, but it doesn't matter.. if you have a 1 wire and its located close enough to the head to be hot enough to work properly, no prob.. if you use a heated one now, thats cool too....

3.. with a th350 or a manual trans, just tick the flags that tell the pcm the engine has a manual trans bolted behind it, and it won't care about any trans/tcc issues..

4. stock is $E6, but i am certain it will run $0D and one other that escapes me ATM..

5. stock eprom is a memcal mounted 27c512. i have an ostrich in mine and cannot rave about it enough! it is truely A-W-E-S-O-M-E and so is Craigs service..

I dont have an emulator and I already could datalog at higher rates via a subroutine in the ecm so the only thing Im really after is the extra power, hardware capabilities, and the cavernous eprom. All of which this appears to have

Ok, so the ostrich should just fit in by removing the 27c512 from the board and soldering in a chip socket, right? Or do I need a different adapter from craig (I already have a G2 for my '7747)

Gotta agree on craigs service, I've been extremely happy with everything I've purchased from him and gotten it fairly quick too.

What? No emulator! Wow, whoda thunk it. You're right about the cavernous Eprom. There is a ton of unused memory space, enough for all kinds of cool patches and as Jprevost mentioned, all the code is on the Eprom which opens up a whole new world of code possibilities.

I will have one at some point. Just havnt gotten around to getting one yet. Havnt even gotten enough time to patch the WB-O2 into one of the A/D inputs yet. Too much to do, too little time

Thats one thing I forgot about, that cursed on-board rom. I have RBobs romless board as well as the extra 4K of addressable ROM space mod on the C3. Only gives 16K of total eprom space, though, but with the C3, thatll hold more code then the proc. can execute.

EDIT: Ah, now I remember why I never got the emulator. I havnt had any power in my interior for the longest time so there was no cigarette lighter to power stuff and the laptop would only go for a 1/2 hour on the shot battery. By the time I got done datalogging it would crap out and Id return home to quickly recharge and update the chip. I have power again so Ill be able to use all the extras.

i desoldered the eprom from the memcal and put a socket in its place..
or you can buy a G1 from craig.

I just picked one up off ebay for $34 shipped. Will let eveyone know how the swap goes when I get it (should be here later this week)

I might be moving to the dark side soon... no not TPI, but this p4 pcm . I've got a hopeful link to a 94 4L60e with only 9k miles and upgarded clutchs. If I get it, I'm going $0D like I had been wanting to do for YEARS. Check out this tread I posted in 01; http://www.thirdgen.org/techbb2/show...threadid=38951
As proof that I've wanted to do this. I just wasn't in any position at the time (being a daily driver) to fiddle around with the wiring. Now that's it's been tried and true, I can dive in. I hope to make some patchs for this ecm. One thing I really want is a switch for performance mode shifting. You can do it with the 94-95 f and b-body lt1 pcm's with some wiring so I bet it could be done with this. The code for controling the trans is probably very similar. That's the ultimate control. Having done a few tunes on the electronics trans I'm hooked. I'll save the ultimate TBI for another vehicle, hopefully f-body.

the pcm is cool, but its also much more complicated then the tbi ecms on the code side. I could work on the C3 in my sleep, but this P4 is about 4x as complicated with 4x the ammount of code. It could take me close to a year or more to complete the hack

Yeah, 64KB, even with a quarter of it blank, is a lot of code. How much is code and how much is just calibration tables? I've looked at the hack but only briefly, I haven't disassembled it because I'm still a noob with this stuff. I have a hard time playing darts and counting in my head!

From memory it has 2 sets of shift tables already, 'cold' and 'normal'. This would be the cheats way to achieve what you want.. Set the 'cold' threshold at a temp just above your ambient temp, so it starts up in 'normal' mode, then have a switch that switches a resistor in place of the trans temp sensor that makes it 'cold' and program the 'cold' tables as your 'performance' tables..
I'm really liking the recent interest in these PCM's, the more guys that are running them, the more patches and tools we will have.. Some cool patches would be a WBo2 input to put into the ALDL and also cooling fan control IMO..

I'd like to do a $0D vs $E6 comparison. I run $E6 and have put a heap of work into a really nice .xdf, and am therefore reluctant to change.. I have looked thru the $0D hac and it seems essentially the same.. Anyone know of any significant differences?

ben73,

Lord don't I wish RBob and JPrevost would gaze upon this PCM and enlighten the rest of us.

Ben,

However, you are the trailblazer.

The calibration section spans just over 11K on the prom.

I'm still struggling with a part of the TCC stuff and rapidly learning to read assembly.. Some of the real guru's getting involved would really speed up progress!

I'm really liking it too. Been waiting over 2 years. Ben, I worked out cooling fan control months ago as I mentioned to you, just haven't had time to test it what with the new wife and all. Anyway here it is and I will be testing it this weekend. I am using the CPI Manifold Tuning Valve/A C Clutch control output and modifying the valve control routine slightly to get it to work with temp instead of RPM or TPS depending on whether or not you want hysteresis. It is enabled here:

From the $0D hack:
L400E FCB $00 ; MODE WD, AFR 4 0000 0000
; b7 1 = not used
; b6 1 = not used
; b5 1 = LATCH ERR 45
; b4 1 = USE L496F WITH ASYNC FUEL DELIVERY (LOW BPW OFSET vs BPW)
;
; b3 1 = VAR TUNE CNT'L
; b2 1 = SHIFT LIGHT ENABLE
; b1 1 = USE ALT CMAP Vs. MAP LD FOR FUEL CUR HYST PAIR
; b0 1 = USE ALT CMAP Vs. MAP LD & AD MAP FOR BLM ENABLE
;---------------------------------
L400F FCB $18 ; MODE WD, DIG I/O 0001 1000
; b7 1 = MAN, (0 = TCC)
; b6 1 = TCC (Non Elect xmish)
; b5 1 = not used
; b4 1 = CONV OVER HEAT
;
; b3 1 = KNK RETARD
; b2 1 = A/C CLUTCH CNT'L, 0 = VAR TUNE
; b1 1 = not used
; b0 1 = 1 DO RPM/MPH LMT, (GOV'R OPT)
;---------------------------------

Turn on bit 3 at L400E, turn off bit 2 at L400F. This enables the output for Manifold Tuning Valve control. Here is the cal table:

;---------------------------------------------
; CPI MANIFOLD TUNING CONTROL CALIB
;
;---------------------------------------------
ORG $412B ;
L412B FCB 0 ; 0 LO RPM, LOWER HYST
L412C FCB 0 ; 0 LO RPM, UPPER HYST
;
L412D FCB 0 ; 0 HI RPM, UPPER HYST
L412E FCB 0 ; 0 HI RPM, LOWER HYST
;
L412F FCB 0 ; 0 % TPS, LOWER HYST
L4130 FCB 0 ; 0 % TPS, UPPER HYST
;
L4131 FCB 0 ; -40 c COOL, CUT OFF
;---------------------------------------------

Here is the routine:

**********************************************
* CPI MANIFOLD TUNING CONTROL CALIB
*
**********************************************
D399: LD399 LDX #$400E ; MODE WD, AFR 4
D39C: BRSET 0,X,#$08,LD3A2 ; BR IF b3, 1 = VAR TUNE CNT'L
; .. else
D3A0: BRA LD3C6 ; EXIT

D3A2: LD3A2 LDX #$412B ; INDEX CPI QUAL'S LIST
;
D3A5: BRSET L0052,#$08,LD3B0 ; BR IF b3, VARIABLE TUNING CNT'L

D3A9: LDAA L00A2 ; COOLANT VALUE
D3AB: CMPA 6,X ; -40 C COOL CUT OFF
D3AD: BCS LD3C6 ; BR IF COOL LT THRESH

;
; CK RPM QUAL'S
;
D3AF: INX ;
D3B0: LD3B0 LDAA L0062 ; CURRENT RPM/25
D3B2: CMPA 0,X ; 0 LO RPM LOWER/UPPER HYST
D3B4: BLS LD3C6 ; EXIT IF LESS THAN OR SAME

D3B6: CMPA 2,X ; 0 HI RPM LOWER/UPPER HYST
D3B8: BHI LD3C6 ; EXIT IF HIGHER

;
; CK TPS QUAL'S
;

D3BA: LDAA L01FC ; CURRENT % TPS
D3BD: CMPA 4,X ; 0 % TPS LOWER/UPPER HYST
D3BF: BLS LD3C6 ; EXIT IF LESS THAN OR SAME

D3C1: BSET L0052,#$08 ; SET b3, VARIABLE TUNING CNT'L

D3C4: BRA LD3C9 ; EXIT VIA RTS


D3C6: LD3C6 BCLR L0052,#$08 ; CLR b3, VARIABLE TUNING CNT'L FLAG

D3C9: LD3C9 RTS
;----------------------------------------------

If you want hysteresis, simply replace the current RPM lookup at D3B0 with the current temp term (L00A2) and set the cal table with the upper and lower entry and exit temp values instead of RPM. If you don't want hysteresis, replace the current TPS term at D3BA with the current temp term and set the cal table with entry and exit temp values instead of TPS. This routine and output could also easily control a Nitrous solenoid. I have more cool stuff in the works as well like using the Linear EGR valve position input to log WB02. It is a real A/D input. Enjoy

Awesome! I am spending some time tomorrow nutting out the TCC stuff, so i'll check that out then also!

Yeah, I picked up a TPI system last week. Going forward with MPFI control. Should be cool to have a Prom based OBDI TBI PCM controling a TPI truck. The people on the truck boards say it can't be done. We'll see who has the last laugh.

Ahh, all the more reason to try! Be sure to thank them for the inspiration
I plan to try it with a converted LT1 intake on the 406 i have started building...

If you do get it to work well with a TPI system, I'd definitely be interested...I want to put an LT4 intake/vortec heads on my truck and if I could avoid switching ECM's in the process that would be awesome..

What MPU did this thing get?

Looking at teh pic above, from the # of pins and the location of the pins for the crystal it looks like the closest match in the PDF from Moto is the MC68HC11F1. It, and the PCM, have 1K of RAM, which is nice.

Nother question,

Does the ign. module send out a seperate signal when the engine is cranking?

On the surface of it, it looks like the cranking fuel routine is interrupt driven by a cranking signal (tied into the IRQ pin) and runs each time there is a DRP.

HaulnA$$,

Have you actually gone through and gotten a handle on how everything works in the code? Im going through it but itll be at least a couple of months before I have anything to show for my efforts.

Are you guys familiar enough with CVS or a similar code versioning setup, that it would be worthwhile to set one up? I'm considering hitting the boneyard today to snag a couple of these to play with and hopefully contribute to this effort.

If you're not familiar, CVS is nice because two people can be working on the same file at the same time, and when they both try to check their changes back into the server, it will attempt to "merge" the seperate changes back into the file, so everyone can always have the latest copy by syncing what they have against the server, and not everyone trying to reinvent each others wheel. Also, you can tag your changes with comments, so the other can surf the comments to see what is in each new version.

Idea?

Teeleton

Eldebrock has been selling a port manifold set-up that uses the the 7060 for some time now.

There was a posting on another board about grounding some pin on the memcal, and then what cal changes were need to run in a normal batch fire method.

The above is for the $85 mask.....