Actually I just got to thinking on this problem. I think we have been looking at it backwards. Leave the MAF sensor alone and on the TOS input. Switch the TOS input to the TIS input. Eliminate the DRAC and wire the TOS speed sensor directly to the TIS speed sensor inputs. Eliminating the DRAC in my Van is no big deal, I already have the housing from JTR that gives me both a cable driven speed sensor (behind the mechanical driven speedo head) and the factory style one in the tailshaft housing.

The TIS read-ins are there, it just needs to be activated in the code. It will still see shaft RPM, but how much of the code would need to be changed is the mystery to me. I would think that a simple routine would allow one to go from RPM to MPH by simply knowing the Gear Ratio and Tire Diameter. A simple equation could turn the signal from RPM to MPH. Then the calculated MPH could be used by the transmission code in place of the value calculated currently by the coding for the 2002 PPM TOS Input.

Haulin- We pray for a successful operation and a speedy recovery.

 

THANKS.
Im new to it and trying to learn.

 

Hauln, I have heard about the winter gas vs. summer gas mixes but haven't put 2&2 together, didn't think about it this way. Thanks for the direction and definately good luck on the surgery. If it's an endoscopic surgery hopefully they are using one of my cameras. HA!

Good luck and Gods speed on the recovery.

Brian

 

Hmm... Didnt think of that. Good idea. The TIS input and TOS input are the same type of input. Assuming itll trigger properly, its just a matter of replacing one input with another in the code.

 

Lookt at the pink book and it works MANNY THANKS.
I needed to set rpm and tps adresses to temp to get it work(fan of with no tps).
This is a garage test,my fan is blowing the 40A fuse when it starts.
Its a 2 step from volvo and need 32A when its run but a lot more when its starts (dont have the box that steering the fan).

 

Any splicing of the fan wires has in my years, resulted in blown fuses when using the factory rated fuse.

Redoing and heavily inspecting the wiring and even sometimes using a fuse rated 5a extra has solved it.

I just make the harnesses in one piece now.

Them things eat juice when they turn on.

later
Jeremy

 

Dimented's MAF TBI programming on my G20s 350 TBI, ENJOY.

http://www.youtube.com/watch?v=7nVQePIYzhI

http://www.youtube.com/watch?v=c4CY3PvklvM

 

I'm alive! Thanks to all for the prayers and well wishes in my surgery. All went well and I am on the road to recovery. I "broke out" on Fri. and spent the weekend resting and now that I am up and around, I need to keep the boredom to a minimum. This helps.

Yes, there is a lean cruise mode. Some calibrations use it, some don't but it is easily enabled. By enabling the "Open Loop AFR Enable" flag (L400D, bit 0) and ensuring that the min temp setting constant at L48C7 is reasonable, the PCM will use the "Open Loop AFR vs Coolant Temp vs Vacuum" table in closed loop. This table is always used in open loop but by enabling the above parameters, these AFR's will be targeted in closed loop. I have verified this by datalogging desired AFR in place of Batt Volts. BLM's will be affected in closed loop when the desired AFR is above or below stoich. When tuning fuel, this should be taken into consideration to ensure VE table accuracy. When I tune open loop fuel, I disable EGR (and block it off to prevent any valve leakage) as well as CCP and Cat converter overheat protection then bypass the Cat (I run with a gutted one). This cuts the fueling modifiers down to AE and PE. I then set the table values in the operating temp range to something like 13:1 across the board and datalog desired AFR vs WBO2 AFR and adjust the VE accordingly. This ensures that the VE tables represent actual VE and not some fudged value. When I move to closed loop, I disable lean cruise and tune the O2 thresholds for BLM's of 128 across the board knowing that my VE tables are accurate. This is especially helpful when the factory O2 sensor has moved from the factory location. Once thresholds are tuned, I move on to proportional gains to get the O2 voltage swings where they need to be. Once all this is set, I enable and tune lean cruise. Oh yeah, I forgot AE. I tune AE in open loop with the WBO2 as well as WOT AFR.

As for the Barometer, it is a calculated value that compensates for changes in altitude or quick drastic changes in weather. Say for example, you started up a mountain road at the base and continued up 10,000 feet to the top of the mountain without stopping. If there were no Baro compensation, by the time you got to the top of the mountain, you would be running considerably rich. HTH

O.K., I fell better already!

 

I'm not sure what you mean by AC square wave. The TOS input does have a 4.7k pullup and is ideally suited for the open collector output of the MAF but the MAF is definetly not compatible with the reluctor input of the TIS sensor. You may be able to bypass some of the TIS input circuitry since at some point it must be converted to a 0 to 5V square wave in order to be read by the CPU. When I unpack my PCM bench from the move, I will take a look into it. Remind me as I tend to forget. I have ADHD. Seriously. I am a walking medical experiment. HTH

 

HaulinA$$,

What I had mentioned is running the TOS sensor straight into the TIS sensor circuitry in the PCM, eliminating the DRAC on my non-electronic cluster van. My 4L60E is already setup for the mechanical speedo retaining the stock VSS on it thanks to Jags That Run. My other VSS is behind the speedo head. My thought was use the TIS sensor circuity to read TOS RPM and calculate the MPH needed for the 4L60E transmission code based off of RPM/Tire Size/Final Gearing. Once the ECM knows the RPM of the transmissions output shaft, some simple constants and mathematics will give you MPH. The calculated MPH can then be used anywhere the standard MPH input is used. As you can tell, I am still stuck on trying to get my 4L60E back in AND keep the MAF setup. Both items provide an equal positive driveability change and I want them both at the same time.

 

I understand what you were suggesting as well as what you are trying to accompish, keeping the MAF on the TOS PCM input and moving the TOS sensor output directly to the TIS sensor PCM input and do away with the DRAC. My contention is this, doing away with the DRAC is possible but would require major code changes as the DRAC does more than just supply the PCM with a 40 PPR signal for the tranny code and drive the speedo. It also supplies the PCM, with a seperate 2000 PPM DRAC output for the engine code as well as the speedo. It would be far easier to parallel the TOS sensor output to the PCM TIS input as well as the DRAC and keep 2000 PPM DRAC signal to the PCM for the engine code. This would not require any major modifications to the code at all except switching the VSS input from the TOS input to the TIS input which is easy to do. The main engine and tranny code would not have to be touched at all. Moving the MAF to the TIS input would accomplish the same end result. JMHO

 

My G20 still has the 2000PPM VSS behind the Speedometer in the instrument cluster, in fact it is operating now with the 700r4 to control the engine and the 700r4s converter. My 4L60E has the $300.00 (I paid $75.00 on Ebay) tailshaft housing from Jags That Run that allows both the cable driven speedo AND the transmission mounted VSS.

Doesn't the PCM currently read the TIS input as RPM for slippage calculations?

Moving the MAF to the TIS input would require internal PCM modifications, would it not?

 

I know you have the 2000 PPM vss in the dash, I was thinking about others who may wish to do this setup in the future that have electronic trannys and dashes.

The PCM only reads the TIS input for slippage calcs on the $0E and $31 4L80E masks. The $0D mask doesn't even have a TIS sensor. It uses the engine RPM for this and fudges in the converter slippage as a factor IIRC although I could have it confused with the 4L80E masks. It has been a while since I looked at that part of it. $0D only detects major tranny slippage.

Moving the MAF to the TIS input does not require PCM hardware modifications. They could be done externally with a simple signal converter, however, having said that, a simple PCM hardware mod (like bypassing some circuitry with a single jumper wire) may be the easiest solution. I just don't know yet until I dig into it. HTH

 

Ok, that is understandable. Sounds like you are going to want to try it sometime.

The only thing I would worry about by using the converter and the TIS input is loss of resolution, depending on where the trigger points are.

 

As long as the input is triggered reliably, which should'nt be hard, resolution is not an issue as the MAF is a frequency controlled device and will trigger the input at the same frequency no matter where the trigger points are. There may be a slight propagation delay but that is not an issue either as the timing of the frequency input is not critical. The only item bieng measured is the frequency of the MAF output. HTH

 

Like haulin said, the torque converter slippage is inferred from the engine RPMs and TOS RPMs with the 4L60-E as far as I can see in the code. The PCM reverse calculates the TIS from the TOS and current gear and compares that to the engine speed to estimate torque converter slippage. Accurate so long as your frictions are still in good shape.
----------
I think the code mods could be done reasonably. It would be a matter of saving the calc'd MPH term from the trans calcs instead of the 2000PPM calcs when an E-trans is checked off.

 

That is sitll more work then paralleling the VSS to the TIS and keeping the DRAC, and I'm all about sitcking to the KISS principle. JMHO

EDIT: Oh yeah, happy thanksgiving to all you guys. I know I have a lot to be thankful for. One thing is all the joy I get from the dialogue on this forum, another is a healthy outlook and for the latter, I thank the man upstairs.

 

I just came across and read all of this thread. Lots of good stuff here. This ECM sounds promising for running MPFI with Peak&Hold injectors and a 4L80E.

Will it run 8 P&H injectors???.......is it like the 749 with two injector drivers or like the 730 with one injector driver?

I am guessing all it takes for MPFI is to change the CYL select CAL value (ground it) and change some PROM values. Change the fuel injector P&H hold resistors for 8 P&H injectors or ground the INJSENSE line for 8 saturated injectors. Just like how the 730 and 749 works for saturated and P&H. Since the INJSENSE isn't brought out to the ECM connector it will need an internal jumper wire to ground.

I may have to pick up one of these ECMs to play with.

EDIT: It sounds like the 427 is the best/latest ECM to get, is it?
Anyone know where there is a list of masks/BCC and which trans. was used and which is the latest code that was released?

 

The latest code is $63 that was used in Export trucks with the Vortec TBI engine and 4L60E.

 

junkcltr,

I have an 8625 for sale if you're interested.

 

It has two injector drivers but like the 749, it cannot drive 8 port P & H injectors. To drive 8 port P & H injectors or 4 TBI P & H injectors, either the driver must be modified or an external driver box used. I successfully drove a Holley 4BBL TBI with a quad P & H driver box with the '7427. The same could be done to drive 8 port P & H injectors. I have not tried the 8 injector mod that the '749 guys do. Maybe I'll try it sometime. As for masks and applications, see TunerCats site. The EFITune BCC site is gone so BCC lookup will have to be done some other way. HTH

 

Dominic,
Thanks for the offer but I am picking one up tomorrow that has the connectors on it (I need them) and a 5.7 MEMCAL.

Hauln,
That helps a lot. The fact that it has two injector controllers and two injector drivers/transistors makes it worth giving it a try. The small amount of unused I/O might become a problem.

The was I see it I need outputs for FAN1, FAN2, WATER_INJ_PWM, N2O_PWM, WASTEGATE_CNTRL1, WASTEGATE_CNTRL2.
Then I need inputs for EGT, MAT1, MAT2, Exh_gas_press, oil temp, WBO2

Once I get it in hand I can figure out what kind of ckts are on the unused pins. I like the 730 because it has all the extra I/O that I need, but the 4L80E ckts on the 427 sounds good.

This thread said that the 427 has 1K of RAM. That is another plus for the 427 over the 730 in terms of adding new code.

I guess I will have to dig around for BCC stuff to figure out with MEMCAL is a 5.7, 5.0, 4.3. The code doesn't bother me because I can reprogram. The knock and fuel NETRES stuff is the bad part. The EFITUNE site is down and so is the DIY-EFI FTP site. I guess I am hosed for now until I figure out something. Anyone know if the TunerCat BCC program is worth buying??

Thanks

 

BJYM, BJYN, and BJYL are all 350 BCC. BJYN is in Fullsize vans. BJYM is Tahoe/Yukon. BJYL is from a 1500 pickup.

 

if your using an e-trans and emmissions, then theres not much available output wise. If no e-trans or emmissions, then there are tons of unused outputs. One nice thing about the outputs is that theyre almost all PWM'd. Variable DC with 32-4096Hz switching speeds.

 

I picked up two PCMs. One has a BKMN (or BKNM) BCC. The other is a BKBJ BCC. I think one if from a 5.7 and one from a 4.3. I am going to probe the MEMCAL CYL later to see what they are setup for.

The app. would be an e-trans w/o emissions. This would be mainly to learn how the e-trans works and to get better MPG in the long run (TH350 right now). I took the cover off one of the 427 PCMs and it looked like the two injector drivers have giant heat sinks on them. The thing is loaded with PWM drivers. I noticed the PCB is marked"P5" but the oscillator is the 8.3xx MHz which is the same as the 730 and other older ECMs. I might start playing with the PCM on the bench for the I/O stuff this weekend if I can get the IC pipes done on the current project.

EDIT: The other good thing is that the SST27SF512 is a drop in replacement for the PROM from the looks of the pin-out that TunerCat shows. Now I just have to do the hard part and figure out the BCC and NETRES stuff. New projects are fun. Thanks to all that pioneered this PCM. If it wasn't for this thread I would have probably made some hardware for the 730 to control the E-trans. The 427 having 1K of memory is much better than the 730 ECM.

 

Thats one sucky thing is that the PCM still runs at 2.1 MHz. Its amazing it can run all that code. Theres very little spare processing time, though. Every so often, the loops overrun when there are alot of interrupts.

 

After looking at it more it does look like it runs at the 8.3x/4 MHz clock speed. If it does have the I/O I need then some code will probably have to be removed for the new CL and Boost code.

I compared the chips side by side with a 730 ECM. It appears to only have one A/D chip and one A/D MUX. It should be enough for what I need if the inputs are wired to the connector. I started checking some of them. If the outputs are all wired then it will have enough outputs for my app. It appears that all the output drivers are populated.

The injector driver ckt is way different than the 730 injector driver ckt. I need to trace it out and make a schematic to see how it works. Might need an O'scope on it to see how it functions before changing it for V8 MPFI.

The two PROMs I ended up with are BNKM (5.7 liter 4L80E trans) and the other is BNBJ (4.3 liter). The BNKM unit was $50 with the connectors and 1.5' of wire. Seems like a cheap easy way to go with an E-trans if the I/O is there for my other stuff. At worst, I will learn how the E-trans is controlled and can add it to the 730 AUJP code.

 

Keep in mind the TBI drivers are PEAK and HOLD type vs. Saturated and need modification to drive 4 regular TPI style injectors.

 

I hear you on the injectors. After looking at what is there for chips I started tracing out the injector drivers/FETs and controllers. I have sat. inj. that I will run in it at first but will change to PFI P&H inj. later on.

Both inj sets will require mods to the inj ckt due to them being both different than the TBI inj. I think the interesting part was that the ckts are tied together through Rs & Cs. It could be I lost track somewhere, but it seems like one ckt feeds into the other. I need to draw it out to be certain. I will need to figure it out enough to run 8 sat. inj. and 8 P&H PFI injs. I might have some time tomorrow night to start tracing it out. This will be a slow going project. Probably all winter because I also need to come up to speed on the code side of things. I have to check if my assembler will handle code that big......I had to make some mods for the large AUJP file last year so it wouldn't crash.

First I need to make sure it has the I/O I need. If it doesn't have all that I need it probably isn't worth it and I am better off looking at the E-code and sticking it in the 730 code. Learning a new ECM is a good project though. I like how they did the inj. ckt different. It makes you think a little.

 

From my testing, it seems that there is injector current sensing as is typical in all P & H drivers but it also seems that there is injector voltage sensing as well. It is difficult to tell as the injector driver is a black box but this seems to be the case. Changing the sense resistors does change injector current so switching to .05 Ohm resistors and changing the FETs to handle the current (as well as the PCB traces), like the '749 guys do, should work but as I said before, I have never tried it as I have never had the need to run 8 port P & H injectors. I'll eventually get around to testing it but I like the idea of an external driver box. Bypassing the sense resistors for 8 saturated injectors works fine. BTDT.

Getting the code to disassemble and reassemble cleanly is not as easy as some of the other code, like $42 or $8D, but it can be done. It took me a while to get a clean, correct disassembly but I eventually got it after playing with it. You could always do as dimented24x7 did and start with the hack and make it assemble. The only problem is that the assembly is not "clean". It works but will not compare to the original. HTH

 

The trick to getting a clean assembly is to compare the final assembled code to its source bin using hexworkshop or something similar. I did this for the hack I made to find all the errors. There will always be some differences. For instance, they where using external addressing for a direct address when updating the shift solenoid states and also there where a few minor errors in the original code that needed fixing. Also, the empty space in my source code is FF while 00 in the bin. Going through them I was able to determine that the code and bin produced the same output, less the empty space and original errors.

As for the external addressing, I wonder if that has any effect when that addressing mode is used to address a direct RAM address? I guess it probably didnt as the trans still worked.
----------
The ASM11.EXE that comes with miniIDE works great and is lightning fast as its a single pass assembler.

 

When assembling from the hack, yes, there will always be some differences since there are "holes" in the disassembly. To finally determine that I had a clean disassembly/reassembly I used the compare feature in TunerPro. It works great. I am able to disassemble the .bin, reassemble the disassembly, convert it to .s19, then convert it to .bin and get an exact match to the original. This is how I was able to make the code changes for e-fan control, N2O control etc. You must remember, at the time I did this, the hack was incomplete and had many errors. There was no other way to do it. Since you completed commenting the hack to the point you have, I see the errors in the original code but it doesn't matter since I get what I need. I guess my point is this, not having a known complete hack, if you start with a disassembly of a flawed .bin that works then reassemble it successfully to its original state, at least you know that you have a disassembly that can then be modofied and reassembled correctly. I have been playing with TunerCats disassembler so I can get a clean disassembly that is like the original hack, then I can paste the "holes" into your hack and get a commented assembly that exactly matches the original. I guess I am a little **** when it comes to comparing files.

 

I know what you mean. I started off my first project with the $4D (8063). Read the embedded ROM on it and started disassembling from scratch with the stock bin.

As far as comparing goes, hex workshop is my personal favorite. Points out and labels all the types of errors it encounters, and automatically resyncs within the bins when there is a frameshift or some other error. There was some frame shift, which is what your seeing, but it doesnt effect the contents of the code. I verified this by cross checking the binary machine code in each bin with the actual hack. There are no holes code and calibration wise, other then the ALDL diagnostics code and units table that was omitted. Theyre not essential unless one wants to have mode 4 diagnostics enabled and only appear in some of the cals as far as I can tell.

This is just my opinion, but Im only mainly concerned with the code and calibrations being bang on. Thats where I get very ****.

 

I can relate but seeing as I had no concrete frame of reference, i.e. a correctly commented complete hack, I had no other choice but to do it the way I did. What can I say, it worked for me. As for Hex Workshop, I am an old 6800 machine language guy so I still do some things caveman style.

 

Mine was $99 for the 299 ECM used in the 1992 fullsize van that I pulled my TBI setup from. I dug deep enough to find the VE tables, the spark table, the TCC, and a few other things.

 

Yeah, definatly, especially if you just have straight assembly with no comments.

Ive been substituting alot of beer and TV time for working on the dis. to get it complete as possible so there is a road map for the OBD-II stuff. Id like to move onto the vortec PCM so I can get used to *groan* working on a 32 bit MPU in assembly. That should be lots of fun... The instruction set aone is 600+ pages. I dont even know if there is a disassembler out there thatll work for it.
----------
Thats how I started as well. Matching up tables to find things in the ECM.

 

I have never tried the MiniIDE assembler. I may later on because it looks like a Linux ASM11 is available. I noticed the $0D hac has labels that are different than the Baldwin assembler labels. I have always used my own compiled version of A Baldwin's as11. Open source is a great thing.

I have probed the 427 enough to see that it may make it in terms of the I/O I want. I need to look at the ckts on pins B6 and B11 more. B6 has some 3 pin diodes or transistors on it. Almost looks like a CTS type ckt for swicthing the pull up or something. B11 is somehow connected to B12 (Fuel Pump input). I need to apply power to figure out what two of the components connecting it are.

The fuel ckt is setup much like the 730. The controller IC has changed but the external pin devices are similiar. The short sense res. looks like it will work fine for the sat. inj. The P&H will require a resistor change. One thing I was thinking about while checking the resistor value........the 749 people change to .05 ohms to get the peak current correct. What about the 1 amp hold current? That seems kind of small for 4 PFI injectors. Especially when only one TBI inj gets 1 Amp for the Hold current. I guess I need to check the LM1949 datasheet before doing anything for the P&H stuff.

You two were talking about MAF and using the TOS output because it is an open collector input. Also mentioned was the TIS input. It appears that one TIS is tied to ground. The other goes to a 14 pin IC through Rs & Cs. The chip looks like it is set to switch (trigger) at VCC/2. It seems like it would work with a square wave also. That is guessing from the external Rs & Cs connected to it. In the coming weeks I should have time to apply power to the ECM and see how that input works.

Has anyone looked at the "Brake" or "A/C clutch" inputs for the MAF? It looks like they go to the Timer chip and not the UC A/D pins or A/D MUX pins. Almost like they use a timer input. Not sure though becuase they could just be on/off inputs of the Timer chip.

EDIT: I guess using the brake input would be bad because the PCM has full control of the TCC. It isn't wired like the 700R4 TCC. One other thing I just thought of......I hope once I start reading the $0D/$0E code that I don't forget or confuse it with all of the AUJP stuff.

EDIT AGAIN: It appears the hold current will be fine for the 8 PFI injs. The ckt should reduce the current to 2 Amps for 4 PFI injectors.

 

I looked at the unused inputs in more detail. The single ended inputs that have the Rs & Cs to make it to the A/D converter all seem to have pull-up resistors on them. The MAT input (B9) does also. They look to run around 3K ohms.

So what does that mean for things like WBO2 sensors. Not much if the output resistance of the WBO2 is small. If the output resistor is small then the voltage will not change much. For the DIY WBO2 I changed the original output resistor from 1K ohm to 100 ohms. The 100 will not have a problem but the 1K would have been a big problem.

One interesting input is the B16 pin (EGR pintle pos.). It has a series RC pair to the MEMCAL to the Knock module. I need to power it up to see how it affects that input. Also, some inputs are configured like a typical ECT/CTS so that the input resistor can be switched. I need to figure out if the ECT and unused input that has switchable resistors are independent or if they are switched together. Switched together would be a pain to deal with unless the sensors had similiar resistances when switched.

The differential pair of pins F4 and F5 do not have any Rs and Cs installed are require a fair amount of parts to use them. It has the same form as the TIS signals.

So far the inputs that I want is more lacking than the outputs that I want.

 

I think I finally have the I/O traced enough to apply power and check out more things. It looks like pin E12 (A/C signal in) or F8 (unused) would be good candidates for the Frequency MAF. They have pull-ups and go to the Timer chip. I don't plan on using a MAF, but it might work for you guys so that the TOS pin can be used again.

It turns out I am hard up for outputs more than inputs. I thought the 4WD was an output, but it is an input. There seems to be enough inputs for what I need. I may have to use up the EGR and/or Evap Can later on or maybe do a different dual fan control so that one output can be used for something else.

I looked into the NETRES change for the 8 PFI mode. The grounds and VCC on the netres are the same as the 730 and it looks like the cyl sel is in the same place. I need to power it up to verify the change.

Pin B16 does not connect to the MEMCAL like I thought earlier. They just use the same ground/power plane.

 

One thing about the MAF is that it switches at a very high frequency. The I/Os can actually switch that fast from my experiments, but in the code, one needs a low overhead method to read in the MAF pulses. The TOS has a pulse counter and timer capture tied to it, so it functions just as the MAF input would on a later PCM. The MAF can also just plug right into it w/o any external hardware. All you have to do in the code is read in the pulse counts and captured timer value at 40 or 80 times a second rather then every time a pulse is recieved (3-12 kHz).

As for outputs, dont forget pin E11 (PWM TCC/non-CC TCC output). I finally found this in the code at $306A and got it to work.

Also, if possible, could you trace pin E10 (TCC output)? I cant for the life of me get it to work on the programming side. Ive managed to get just about all the other I/Os to work except for that one. Is it a simple on/off or is it PWM'd?

 

I traced the TOS to a chip with Rs & Cs on it. It appears to be a signal converter chip. It also has a trace tied to the FailSafe chip. I could not find a trace to the Ucontroller or the Parrallel Timer chip. I still need to figure out which chip it goes to.

The 730 Parallel Timer chip has counter/capture regs. That is how the ref pulses and VSS are done. I am guessing the 427 is the same since it seems to have kind of the same layout. I think the 427 TOS signal ends up at the Timer chip but the HC11F1 probably has timer/capture in it which the 730 ECM doesn't have. The 730 Parallel chip could handle the high freq. MAF. From memory the counter runs at 64 KHz.

Pin E10 or E11 is targeted for the Water Injection. It depends on the trans. I end up with. Pin E3 (CPI/AC Clutch) gets FAN_1, pin E7 gets FAN_2, pin E1 (EGR) gets Wastegate control, pin F16 (EVAP CAN) gets N2O control. That is in order of priority.

The inputs are still not determined for which pin will get what. The switchable resistors need to be looked into more. The odd pin B11 thing connected to the fuel pump voltage via a diode and cap needs to be explored. I am still missing one input for the 3Bar MAP. Since MAT_2 isn't necessary after initial tuning I will substitute that with exhaust back pressure off and on.

I traced Pin E10. It appears to be layed out like pin E11. The Parallel Timer chip seems to drive a custom 14 pin "gate" IC for both E10 and E11. The custom IC then drives the QUADCHIP outputs. So there is a complete path for E10 from the Ucontroller to the Parallel chip to a 'gate' chip to the QUADCHIP to the ECM blue connector. It could be PWM or simple on/off. The 730 Parallel chip has both and I am guessing the 427 is the same. The only way to figure out the type is to look through the code. A simple on/off will be a 1 bit field for a reg in the Parallel chip address. PWM will be a 16 bit register to set the freq and duty cycle.

I took a quick look at the code. It has the form of the 730 code with real time interrupts and subroutines. Oh yeah, the 427 has less outputs than the 730 ECM. The 427 has two QUAD drivers and the 730 has three QUAD drivers. The 730 ECM has way more inputs for the A/D. The 427 has more inputs for freq. varying signals. I *really* hope the 427 does have 1K of RAM because that is the *biggest* reason for me using this ECM.

I looked at the changes needed for the MPFI to get the thing to do double fire batch (squirt every 4th DRP). The change done for the 730 V6 MEMCAL to V8 MEMCAL applies. The injector ckt does need changes to run either 8 P&H or 8 sat. injectors. Both MUST be done to get it to run V8 MPFI properly.

I think I need a new battery for my multi-meter. Time to take a break from it....tiny solder dots get to your eyes after a while.

 

WOW, I just looked over the $0D (strange to type after using $8D for so long) and the code flows just like the $8D. That is good for me. It is exactly what I am used to looking at.

One big thing jumps right out. That is the Parallel Timer chip I talked about. Register wise it is just like the 730 ECM with the external pin out different. You said you were having a hard time getting pin E11 to work. I bet it is a bit in double register 0x3FFC (that is 0x3FFC:3FFD). That register controls the 1 bit fields internal to the chip and some of the outputs. The SES light is one example of this. Try playing with the bits for that register or look at the code to see where a bit is set in the transmission sections. It could be one of the PWM regs but I am guessing that you tried all of them.

One other thing to do is take a look at the Ludis 730/749 schematic page 2 chip U2. That is the Parallel Timer chip. You will see that the SES and F3OUT do not have a PWM address on the chip. The 0x3FFC:3FFD register controls them with a single bit.

One other interesting thing is that the Parallal Timer gets initialized to 0xB91A in the 427 PFI mode. The AUJP inits it to 0xFB1A. The lower bit change is due to the AJUP messing with FAN 2. I am not sure what this is in the 427 yet. The higher bit change I can't figure out......AUJP leaves it set to 1 always. $0D leaves it as 0 always. The rest of the bits are the same. That is good because some bits control the EST and fuel mode to the Fuel Management Devices (FMDs).

EDIT: I forgot to mention that the MAF code you added uses the same input register that the $8D AUJP uses for the VSS (reg in Parallel Timer chip). The 427 pin E12 and F8 I mentioned also go to the Parallel Timer chip so I guess you could read those timer registers for the MAF input.

 

Pin E11 is very inconsistant. I got it to work once multiple times (appears to be a standard PWM'd output w/ freq. and DC) but it never worked again and I cant figure it out. Even when I select the option to use the non-CC TCC code, which uses that output, it STILL doesnt work. Im totally baffled... and I have yet to get pin E10 to do anything. The only plus side is that now I now how the PWM'd TCC output works and confirmed what address it is in the code. I almost wonder if the hardware is hosed in my PCM.

 

When the E10 and E11 signals leave the timer chip they look to be gated with other stuff. From memory something from the FailSafe chip. It could be something from the Ucontroller. That could be keeping the signal from making it to the QUAD Driver. I will look at the 14 pin IC and see where the other signals go. I know one of them was going to another 14 pin Custom IC and signals feed into that one also.

 

One thing that is definatly true about the TCC is that its one of the hardest outputs to use. I eneded up moving the TCC to another output for Fast355 just to get the TCC to work. Hopefully the code will reveal more about how to use the outputs.

 

It has around 1K of RAM, but alot of it is in use by the engine, trans, and stack and free ram is hard to come by unless you ditch some stuff.. Ive always been a little leery of digging into the stack for free RAM, especially given how much stuff can get thrown onto it with all the interrupts. If you oust the e-trans code, youll get ALOT of ram and I/O's, but if your going to use an e-trans, then your sort of SOL.

 

If I took my notes right and remember right my TCC is on pin E2 now which was originally the 3-2 downshift sequencing solenoid or the shift light.

Dimented -
As far as the other outputs that you have, what pins are they on? I have not had time to play around with the different outputs to see what is going on. Such as are you using the EGR output for one of the cooling fans? If so that would explain my initial trouble of the EGR valve opening only when the A/C was turned on. I assume the unfinished Wideband coding that I found in your hack will be using the Linear EGR Position input(0-5 volts).

 

The wideband coding should work to save the wideband AFR x 10. I know it works with a 0-5V signal, but I still have to verify that the pintle position input will actually accept a 0-5V signal without mods. Im not completly sure about any of the I/O's until Ive had a chance to try them.

I am using the EGR output for a cool fan. The computer operates the fan based on the vehicles speed, use of A/C, and temp thresholds. Other then that and the 3-2 sol. output for the non-CC TCC code, I havnt really taken any of the other outputs over.

 

Linear EGR pos. works fine on the bench, BTW. Just need to calibrate the WB output table to your WB unit, add the address into the ALDL xmission list, and youll have the WB AFR in your datalog. Pin B16 for the WB O2 5 volt signal. The code you see is complete. All it does is reads in the voltage, looks up the AFR, and saves it to the RAM.

 

I looked at pins E10 and E11 because they seem to have a higher priority right now. They are gated with the FailSafe chip and..........you guessed it.....the BRAKE signal on E13. It seems to fit that they did this in hardware instead of code. In terms of the TCC problems and you thinking the ECM was bad, I don't think this is the case. The BRAKE signal has a pullup to VCC. I am guessing at one point you had it tied to ground and the TCC worked. Not sure though.....the ECM looks to have BRAKE and A/C inputs pulled to a VCC. I need to power up the ECM to see what the VCC is. It looks to be +12V.

The schematic shows conflicting data with the Pinlist on diy-efi. Diy-efi pinout says Brake is 0V when the brake is applied, the schematic says Battery voltage. Since the ECM has the BRAKE pulled to voltage, I am betting that if you have E13 tied to ground then the TCC E10 or E11 will not operate. Anyway, it is the BRAKE signal that is causing the problem and changing the polarity of what you have now should make E10 and E11 work properly.

I looked at the MAF input section of the code. You have a piece that does not update the MAF HZ if the delta of MAF pulse counts is zero. What if the counter wraps and is negative (not zero)? Or does the clock counter and pulse counter clear when read? That would mean the code is never excersized. If the other case with the neg. result the MAF value would be wrong for a few ms.
Not that I want to use a MAF I just look over the code that is out there. The way you wrote it is very clean and readable.