Iron duke (pontiac151) MPFI build

[windsorcastl]

Hello everyone,

This is my first home brew project so have mercy on me,

I have a wonderful iron puke from 1988, its in a Pontiac Fiero, and I am also young and poor (hence why i own a fiero). I know about its performance issues, the block is weak, main bolts yada yada yada, and its lack of after market. I AM aware. The engine currently runs on an ECU#: 1227748

Anyways, I'm poor so I'm using what i have to create a MPFI set up. I built a manifold from scratch (its tight and leakless)
I'm using a computer from a 95 Impala LT1, ECU#: 16188051.

On this engine block there is a DIS system.
There are 2 connectors, a two pin and a 6 pin.
a.) The two pin has
-positive
-Ground

b.) The 6 pin has:
1.)According to the Michel Manual's diagram
-EST
-Bypass
-Return
-Igntion Reference High
-Ground
-Tach

2.)According to an online diagram i found
-Ground
-Reference
-Bypass
-EST
-6X
-Tach

So "6X" must be the "bypass" wire. This is where a 5 volt source is applied to the module to switch timing from the module to the computer correct?

From the computer I will be using there is:
-Low resolution signal
-Distributor reference Low
-Ignition control
-High Resolution signal

I have read that the low res signal is a 4 pulse/crank signal, the high res is a 360 pulse/crank. But doesn't that leave redundancy for the reference signal?
How do i hook up the wires to be compatible with the computer?
Is it possible to change the firing order in the computer?
I also read about a guy soldering the terminals inside the module to bypass the "igniters", what this means i haven't the foggy-est but is it possible to alter the module to make it compatible with the computer?

The firing order for the iron duke is 1-4 and 2-3, a 2 event engine. The Lt1 is 1-8, 4-3, 6-5, 7-2, which i believe is a 4 event engine. So because of this i hooked up the injectors as cylinders (iron duke#-LT1#) 1-1, 4-8, 2-6, 3-5.

Last questions, will the VATS system effect me getting the engine to run, or does it cut out after it starts? can I program out the VATS system?
Do i need to apply 12 volts to the "fuel enable signal" at the computer, or is this an outgoing signal?


Any help at all will be appreciated,
Aaron Newbill

[steveo]

hey there, nothing wrong with an iron duke, but you picked a currently impossible to use ECM.

my personal suggestion is to sell it online for a hundred bucks and use the money to buy an ecm that will actually work for you, might be able to afford a cheap chip burner to boot


I have read that the low res signal is a 4 pulse/crank signal, the high res is a 360 pulse/crank. But doesn't that leave redundancy for the reference signal?

they are both cam trigger signals from an optispark distributor. both are optical signals, so digital.

your stock sensor is a hall effect sensor, i think? even if it is optical/digital/whatever, you'd at least need to fabricate a new encoder wheel or something.

they aren't redundant really, the 4 pulse per crank signal is keyed not just to cam revolution overall, but it has a varied pulse length to determine TDC and current firing order, and the 360 pulse signal is used on top of that for more precise timing within strokes.

this is definitely different from your current sensor

there is a lot of per-cylinder sequential stuff going on in the LT1 ecm, i really doubt this is going to work well on a 4 banger unless you do some serious code hacking? i have never heard of this ecm being used on anything but an 8 cylinder engine successfully.


How do i hook up the wires to be compatible with the computer?

you don't. the signals are just too different.

you really need an optispark, or something approximating it, or the ecm wont function correctly and this whole project will flunk. everything in the ecm's ignition and fuel calculations relies on optispark signal in some way. at the very least you'd have to build an optispark simulator that takes your current signal and imitates an opti. i'm not sure if this is a project you'd be willing to take on. hacking an opti and building a direct cam drive for it would be the easiest way.


Last questions, will the VATS system effect me getting the engine to run, or does it cut out after it starts? can I program out the VATS system?

the least of your worries, but disabling the VATS flag in your bin will solve that problem completely. with the flag enabled, it needs a fuel enable signal from a VATS module before the injector pulse routine works, with the flag disabled, it doesn't care anymore.

[steveo]

the outer wheel is the high res pulse, the 8 slots a bit further in are the low res pulse.

http://shbox.com/1/optical.jpg

[1project2many]

Hello,

Sounds like you have a neat project in mind. Are you stuck on using the 8051? It's not compatible with any ignition system except the Optispark. There are other ecm's which are better matched to the engine and ignition you currently have.

Do you have a six pin connector with your DIS system? The Bypass wire will be tan with a black stripe. If you don't have a connector, or if your connector has all the same color wires, it's time to locate an accurate diagram. Do not mix and match diagrams as guessing can kill the module.

According to a training manual I have the 2.5l module wiring shows the Tach wire furthest from the two wire connector. Wires are Tach, unlabeled, EST, Bypass, Reference, and Ref Low or Ground. This is repeated in two different diagrams.

The ecm does not care what the firing order is. It does nothing to control which cylinder comes next. The Optispark handles the job of synchronizing spark in the LT1 with 8051 while the DIS module handles the chore on the 2.5. There is no need to change the firing order.

VATS will prevent starting the engine if it believes the vehicle is being stolen. The VATS signal is not steady voltage. It must be the correct frequency or it will be misinterpreted by the ecm.

If you have access to a 1227727 ecm that will do a fine job of controlling your 2.5 with pfi.

Edit: Looks like Steveo posted while I was confirming DIs wiring. :)

[windsorcastl]

First of thanks for all the help sofar,




Sounds like you have a neat project in mind. Are you stuck on using the 8051? It's not compatible with any ignition system except the Optispark. There are other ecm's which are better matched to the engine and ignition you currently have.



I might be able to afford a different pre obd2 ecm. I've looked into the 1227749, which came on the 2.2 and the 2.0 turbo in the sunfire, but its batch fuel injected and I'm looking for multi port. I'm not sure if "port" extends to multi port, but i am willing to compromise depending on whats out there.

What I'm looking for in an ECM is
- individual fuel calculations for each cylinder as well as spark
- the ability to handle 2 bar pressures
- spark advance of more than 115 ms (or hundredths of a degree, or whatever number they are using in tuner pro for the 1227748)

It would be nice if it included
-support for a mass airflow sensor as well as vacuum/pressure sensor
-knock sensor support
-air injection support
-at least 32 cell tables for tps, map, and temperature calculations (the 1227748 has 4 cell tables....)
I'm almost hard-lined on having a maf and map sensor (for future turbo application), but the other 3 are just icing on the cake.

Does that 1227727/30 offer these things? I see that they have used it in v6 and v8 applications, is it easy to reform the spark timing? I know the 1227749 is batch fuel, so its low on my list but i will compromise for it if i need to because they only cost 30 dollars on ebay.

[1project2many]

but its batch fuel injected and I'm looking for multi port
"Batch" means more than one injector is triggered together. "Multi-port," at least in GM, means on injector is placed at or near each intake port. The 7749 operates on "batch-fired Multi-port injected" engines.


What I'm looking for in an ECM is
- individual fuel calculations for each cylinder as well as spark
- the ability to handle 2 bar pressures
- spark advance of more than 115 ms (or hundredths of a degree, or whatever number they are using in tuner pro for the 1227748)


This is a very steep list. The 8051 you referenced previously does not meet all of these specifications. Actually, I am interested ecm's that do meet this spec. How would the ecm get feedback to determine an individual cylinder fuel calculation needs adjustment?

The 7727/730 is a very capable ecm for the engines most people build. It does not match the specifications you have posted. It will read a pressure signal that ranges from 0-5V and the voltage can be interpreted as 2 BAR. Maximum rate of spark increase is limited in code and for most engines, the spark resolution in the tables is greater than what can be accurately delivered at the cylinder (reference this graph of crank speed variation on a four cylinder engine. (http://www.epi-eng.com/images/Engine/ET-X-4%20Cyl.gif))

Engine control will be achieved through proper selection of a combination of hardware and software. You can use a 1227749 or a 1227730 ecm with a four cylinder "chip" or memcal along with a program written for the four cylinder Turbo Sunbird to control your PFI injectors with forced induction.

[StandUpToFight]

Its my understanding that GM didn't use full sequential on most ECU's until 94-95, and then the hardware still could only support it up to 4000RPM, and batch fire takes over. I would look for a Quad 4 or 2.2L ECM from that time frame, and make sure an XDF may be available, unless you want to build your own disassembly.

[StandUpToFight]

The GM DIS module is a digital distributor. Its feed the crank speed signal (with a reference pulse - so 7X) directly, When running in bypass mode it runs on a fixed timing, this is for cranking. When engine speed raises high enough the ECM will cut the bypass signal and the module will take the advance signal from the ECM to change the spark timing. It does all the math internally for coil igniting based on the advance signal from the ECM, it cannot do individual cylinder timing or dwell. You would need to run different hardware to control the coils independently from the ECU. At that point you would really want to look at an OBDII or one of the later MEFI ECUs to control coils individually.

[windsorcastl]

"Batch" means more than one injector is triggered together. "Multi-port," at least in GM, means on injector is placed at or near each intake port. The 7749 operates on "batch-fired Multi-port injected" engines.



This is a very steep list. The 8051 you referenced previously does not meet all of these specifications. Actually, I am interested ecm's that do meet this spec. How would the ecm get feedback to determine an individual cylinder fuel calculation needs adjustment?

The 7727/730 is a very capable ecm for the engines most people build. It does not match the specifications you have posted. It will read a pressure signal that ranges from 0-5V and the voltage can be interpreted as 2 BAR. Maximum rate of spark increase is limited in code and for most engines, the spark resolution in the tables is greater than what can be accurately delivered at the cylinder (reference this graph of crank speed variation on a four cylinder engine. (http://www.epi-eng.com/images/Engine/ET-X-4%20Cyl.gif))

Engine control will be achieved through proper selection of a combination of hardware and software. You can use a 1227749 or a 1227730 ecm with a four cylinder "chip" or memcal along with a program written for the four cylinder Turbo Sunbird to control your PFI injectors with forced induction.

Standup mentioned sequential, that's what i meant. And sequential is what gm calls MPFI correct? Does the 1227727/30 do sequential injection? The reason why i want sequential is for gas mileage purposes because this car is usually my DD. I'm sure I can compromise on pressures for MPFI.

So then I may have to have a wide band MAP sensor for pressures above 1bar(1atm=1bar=105Pa)? Or are you saying that the compiler interprets the XDF's differently based on the application? (ie: the 1227749 came on the Cyclone v6, turbo sunbird, and regular 2.2 N/A). I've been inside many XDF's on a text editor and can not figure out how this works (the reprogramming), it seems that the XDF's only address table size and memory storage locations, is it possible that the interpretation of all the sensors is found in the Bin files? if so how do i edit those? I'm totally willing to learn everything it takes to do this, this is going to be my future, I really just don't know where to start on the programming side.
Do I Need:
-circuit diagrams?
-some other sort of compiler or code interpreter for binary files?
-A basic education in assembly?

Aaron

[1project2many]

MPFI only means that the injector is at the intake port. Early systems were TBI so MPI or MPFI was a "Big Deal." Today's systems with MPFI are sequential at low RPM but switch to batch mode as rpm increases. This strategy allows improved idle and low speed emissions but also allows small injectors.

I'm not sure what you are looking for in terms of knowledge. Code we work with in an ecm is already compiled. "Tuning" involves changing variables used by the code. "Reprogramming" means transferring a complete calibration consisting of code and variables, generally with altered variables, to the pcm or chip.

Yes, the interpretation of the MAP sensor values represented in the tuning software is nothing more than a different scale for a given voltage range. You can change it yourself in Tunerpro. The big difference with code is what functions are available in a forced induction program vs an NA version. Turbo Sunbird program is set up with boost reduction on knock, desired boost vs throttle angle and rpm, extra spark tables for boost, and some other features to help dial in a calibration used with boost. You can understand the software and tuning tools without spending a lot of time learning the details in the same way you can enjoy a good meal without understanding exactly how taste and digestion work.

[The Stickman]

OK I am kind of an Iron Duke Fiero guy. There were actually lots of aftermarket stuff. There was the Holley big bore TBI, Holley TBI intake and an Edelbrock TBI intake. I had one. Plus atleast two different Offy intakes one of which was a dual plane. None of this takes into effect all of the Super Duty stuff that will work on a stock Duke. There was a 2 barrel intake, 4 barrel intake both with fuel injection bungs cast in. There were ITB intakes. And there was even a dual quad intake. I had a 2 barrel intake that utilized a Porsche/VW/Audi mechanical Bosche FI system. Cool stuff.

[Xnke]

Ok, so we'll take this one point at a time, to try and clarify-I think the OP is starting to get confused-or at least confuse the terminology.

MPFI = more than 1 point of fuel injection, generally meaning 1 injector per intake port. Can be sequential or batch fired

Sequential injection = Injectors are timed specifically for the cylinder on the firing stroke-this does not mean you're trying to shoot fuel directly into the cylinder past the valve! The valve is not open long enough for that at higher RPMs...which is when most manufacturers just go back to batch-fire anyway.

Batch injection = Injectors are grouped in banks or all together, and fired multiple times (usually) per engine cycle, keeping at least a portion of the manifold and port walls wet with fuel at all times, and some fuel atomized into a vapor state. Batch injection can be throttle-body injection, multi-port injection, but not direct injection.

Sequential injection won't improve fuel economy in any significant way once you're above idle speed-it does reduce emissions at low speeds, and offers a smoother idle, but on every engine I've played with it on, no measurable change in fuel economy was found. The idle behavior is much improved, though!

You will get a bigger return on fuel economy in just tuning-most of the GM tunes I've looked at/cars I've measured run fairly rich through the operating range, and significantly richer than needed in power enrichment mode.

Sounds like you're planning a turbo 2.5L, but you're worried about fuel economy...Turbochargers don't help with fuel economy. The "turbo fuel economy" argument usually comes up when comparing a 200HP turbo 1.5L vs a 200HP n/a 3.0l engine...the turbo 1.5 will produce a better fuel economy due to having a wider range of power delivery available, meaning you can "turn down" the turbo 1.5L more than you can the N/A 3.0L. When producing 200HP, the turbo engine will take approximately 5 to 10% more fuel to do so than the N/A engine. The Brake Specific Fuel Consumption, or BSFT, is generally 5.0 for a N/A engine, 5.5 to 5.8 for the same engine turbocharged, and 5.5 to 6.0 for the same engine, but positive-displacement supercharged. This means, for fixed displacement and fixed power output, fuel consumption will be highest for the supercharged engine, slightly better for the turbocharged engine, and best for the N/A engine.

Your best solution is to find a 7749 or 7730 ECM and wiring harness for a 4 cylinder engine-pretty much any 4 cylinder that the ECM was used on would be fine.

The XDF is a file used by a program called TunerproRT to interpret the BIN file for your ECM. The ADX file is the datastream interpretation file, so you can watch the live data coming in. The BIN file is the actual calibration and operating system for your ECM, and you would use Tunerpro (and the appropriate XDF) to view a human-readable interpretation.

You need:

Tunerpro RT-a computer program
A laptop-to run the program
An ALDL interface-to log data from the ECM
Some way to write the program to the EPROM, be it an emulator or an EPROM burner

For TunerproRT to work with your ECM, you need the appropriate XDF, so tunerpro knows how to interpret the BIN file for your ECM, and the ADX, so tunerpro can listen to the data the ECM pumps out on the ALDL line, and so Tunerpro can send commands back to the ECM.

As for circuit diagrams, make sure you get the wiring diagram for not only the fiero body, but the fiero ECM harness, as well as the wiring diagram for both the body and ECM harness of the vehicle you get the computer and engine wiring harness from. You will need to do some rewiring to merge the two together, and you need to make sure you keep track of what size fuses feed what lines in the ECM harness, as well as what fuses your fiero can provide. When I swapped my S10 engine wiring harness, I had to add 8 fuses to the engine harness, as the S10 truck harness did not have them, but the 94 camaro harness did have them.

The 7749 and 7730 computers seem to have the widest support base, so it would be a good idea to start with one of those.

[windsorcastl]

So i swapped over a 1227749 ECM. It runs and drives (barely), and paired it up with that short runner manifold I was building. Now to fix the bugs.

I am dealing with a quad driver error and low vacuum from MAP signal error and am forced to use the 2.0 turbo XDFs for tuning (because I see no Q4 options). I see a checker for 1bar/2bar but I'm more concerned right now about getting that DIS off of start crank timing. I notice that the 4 main wires are the same with either of the 3 engines (original Tech 4, 87-89 Quad 4, and 2.0 Turbo for the respective computers 1227748, 1227749).

-EST
-Bypass
-Return
-Ignition Reference High

even though the 2.0 is HEI, the Quad 4 is DIS like the Tech 4 and I've heard the signals are the same.
The only main difference between the Quad 4 DIS and the Tech 4 DIS is the presence of a crankshaft sensor, but I cant figure out how to disable that.

Is the computer actually sending and receiving from the DIS and I'm just not noticing?
Is the CS sensor the only problem I'm Having?
Is it possible to HEX edit it out?

Aaron

[1project2many]

I can give you answers but you won't like them.


Is the computer actually sending and receiving from the DIS and I'm just not noticing?
1) Possibly, but testing is required to know for sure.


Is the CS sensor the only problem I'm Having?
Unknown. Testing is the only way to know for certain, and since you're the person at the car only you can do this.


Is it possible to HEX edit it out?
You're referring to the crank sensor? No, it cannot be hex edited out.

No for some related, and possibly helpful, information.

There are three values you need to change when using a distributor based 749 calibration on a DIS equipped vehicle. The reference angle, the max advance, and the min advance (or max retard).

Use this thread for reference:
http://www.gearhead-efi.com/Fuel-Injection/showthread.php?2135-58-timing-with-DIS

[dave w]

I'm late to the party on this thread. This thread reminded me of a similar project (started by the late eaglemark) http://www.gearhead-efi.com/Fuel-Injection/showthread.php?1997-Ideas-for-4-cylinder-MPFI-Toyota-22RE-with-GM-ECM It's a long thread. Basically an external trigger wheel (which I designed) was used. See attached pics.

So let's fast forward a few years ... I would suggest the "What If" the 4 cylinder MPFI system was leveraged of the OBD2 Oldsmobile Quad 4? Basically I would design a dual trigger distributor ( cam and crank triggers relocated to the distributor), similar to the OBD2 dual trigger I make for a SBC using the '0411 PCM. See attached pic.

dave w

[windsorcastl]

I'm late to the party on this thread. This thread reminded me of a similar project (started by the late eaglemark) http://www.gearhead-efi.com/Fuel-Injection/showthread.php?1997-Ideas-for-4-cylinder-MPFI-Toyota-22RE-with-GM-ECM It's a long thread. Basically an external trigger wheel (which I designed) was used. See attached pics.

So let's fast forward a few years ... I would suggest the "What If" the 4 cylinder MPFI system was leveraged of the OBD2 Oldsmobile Quad 4? Basically I would design a dual trigger distributor ( cam and crank triggers relocated to the distributor), similar to the OBD2 dual trigger I make for a SBC using the '0411 PCM. See attached pic.

dave w


So if i were to get the tuner cat quad 4 ecu file I would still have to design a new sensor wheel? (Not that it probably makes any difference, but between 87 and 91 they used this computer which is OBD 1 on the Olds Q4)

[windsorcastl]

1Project2Many,

Naw testing is what I'm interested in! It looks like there's still hope after reading 6 shooters thread.

Now questions about starting tuning. The old computer, 1227748, had a M27C128A PROM and you could flash it with UV and I would write new with the GQ writer. The new set up, 1227749, has a huge inline processor bundle with a blue cap for the knock sensor, that doesn't fit the chip burner.

What are my options for tuning?
Can I tune through the knock bundle with a custom board?
Can I buy a different prom that fits?
Should I just invest in an Ostrich?
etc...

Aaron

[1project2many]

What are my options for tuning?
Tunercat and Tunerpro are the two software tools that I use.


Can I tune through the knock bundle with a custom board?
Inline processor bundle? You mean the socket containing three chips and the knock filter? There are ways to make an adapter so the eprom can be connected to the burner, or you can remove the eprom and install a socket. The 7749 eprom is usually a 27C256 although some have had 27C512's. I'd have to look at the calibration size to see if it fits on a 27128.


Can I buy a different prom that fits?
Yes, there are several. Eproms and eeproms both. Moates site used to sell eeproms or flash chips that are compatible.


Should I just invest in an Ostrich?
An emulator makes tuning much, much easier.

BTW, the DIS Quad 4 uses the same 7 notch trigger wheel as the DIS 2.5 and the DIS 2.2.

[windsorcastl]

Tunercat and Tunerpro are the two software tools that I use.


Inline processor bundle? You mean the socket containing three chips and the knock filter? There are ways to make an adapter so the eprom can be connected to the burner, or you can remove the eprom and install a socket. The 7749 eprom is usually a 27C256 although some have had 27C512's. I'd have to look at the calibration size to see if it fits on a 27128.


Yes, there are several. Eproms and eeproms both. Moates site used to sell eeproms or flash chips that are compatible.


An emulator makes tuning much, much easier.

BTW, the DIS Quad 4 uses the same 7 notch trigger wheel as the DIS 2.5 and the DIS 2.2.

Thank you so much for answering the trigger wheel dilemma. There has been so much misinformation about that out there. Anyways I contacted Moates , they have a board that bypasses the memcal (the G1), for a very reasonable 35 dollars and paired it with a 512 so I can throw code $59 (www.code59.org) on there, and no soldering involved.

So until that shows up and I can play around with it, my only remaining question is are there any other hardware changes I need to make to be able to handle the $59 code? Do I need a wide-band Oxygen sensor?

[1project2many]

I've never used $59 but I believe it will work with stock Sunbird / Syclone sensors. Changing to 1 BAR is a Big Deal. Many of the tables and several of the scalars will need to be adjusted to reflect 1 BAR values. I do not believe I have a complete list of the changes. I might use a 2 BAR sensor to make the process easier.

[windsorcastl]

I've never used $59 but I believe it will work with stock Sunbird / Syclone sensors. Changing to 1 BAR is a Big Deal. Many of the tables and several of the scalars will need to be adjusted to reflect 1 BAR values. I do not believe I have a complete list of the changes. I might use a 2 BAR sensor to make the process easier.

Already preemptively purchased one, but the "sunbird only $58" and the $59 have 'flags' for 1 bar/2bar, and $59 has an additional flag for 3 bar and 3.2 bar :). Thank you for all the help 1P2M! I will be updating soon with pictures and details for everyone.

Aaron

[1project2many]

That flag is one piece of the puzzle. There are many others. Look for posts and pages by Robin Handley as he used $58 with a 1 BAR sensor. You'll probably need to use cached Google pages and the Internet Archive's Wayback Machine to view them.