Question for Padgett, ie. MAF

[2seater]

Does anyone know if the MAF tables can be expanded to read past 170 gm/sec? My turbo project max's the tables pretty easily. Is this even important? There is a need to get a spark retard problem under control but otherwise it is a real rush.

[Guest Greg Ross]

Thrilled to hear you've got-a-runner. I would have to think you're going to thoroughly enjoy the new personality you've created.

Spark retard-come to think of it I may well have the same condition, hadn't even thought of it in the context of the output constraints of the stock ECM. Was just assuming the peak I have the MSD Dis-4 set for was phasing in. "Hello" Yes of course.

Arrgh! That's more rpm then I really need to be turning and the tires would be spinning and....

[padgett]

Anything can be done. Be difficult to change the number of tables but the range of each would be possible. What did you have in mind ? Are you leaning out at the top end ? (definately a no-no). Have you considered water injection ?

Knock retard is interesting - you can hit the engine with a wrench and trigger it. Obviously a turbo is going to need premium fuel and still may have an issue though is generally low speed-high load that causes though engine temp is a factor.

[padgett]

Does anyone know how far the L-67 tables go ? Noticed that Tuner Cat has the L67 descriptions that even use the same computer ($5B4) but haven't invested in the package. Note the computer change in 1994-95, must be when GM doubled the PROM size.

Really wish the time was available to really get into this but am home very rarely these days and the honeydos are piling up.

[2seater]

I am certainly no computer wizz so please understand I may use terms incorrectly. The older turbo cars, such as the GN's apparently had MAF tables that extended to 255 gm/sec, and I make the assumption this reasonably matched the capability of the engine/turbo package. A well tuned one with small modifications can exceed the tables, so the aftermarket chip makers would dial in parameters that would take over when the stock system was exceeded. Their chips were plentiful and cheap, so much so that there were makers that had chip holders with switches to access up to ten or twelve different chips to match the type of fuel being run, track conditions etc.. Essentially select your performance level. Again I make the assumption that at some point, probably at the limit of the MAF table, I need some sort of modification made that would reduce timing to a selected value. Fueling is of course another issue, but I suspect timing is the main culprit right now. I have tried it with the high timing chip (Fastchips), and I saw up to 15 degrees of retard. I switched to the stock chip and saw 10 degrees retard with one spike to 13. Still too much. I always run 93 octane. There are lots of variables of course. Air distribution in our manifold would favor #1 & #2 being lean and #5 & #6 rich, all of which will affect the timing required plus the tendency to knock. O2's never dropped below .88, which is somewhat rich, but that can be caused by lean misfire as well. I have a lot of tuning to do, and I appreciate any input. I belatedly contacted Rinda about a Diacom (no reply so far), and there is no doubt recorded sensor readings will be required to properly tune and plot a course of action. The main thrust of my original question is if there is any room in the chip for additional programming, or failing that, modify the existing program for this application. One last thing, he high timing chip seems to spool up quicker and low speed response is better. <img src="http://www.aaca.org/ubbthreads/images/graemlins/confused.gif" alt="" />

[Guest tempest68]

Comment on gas...

If you have a Sunoco near you, they have 94 octane also. And the one here in York PA also sells 100 octane "Race Fuel" (for about $4/gal!).

[Guest imported_barra]

Re High octane gasoline;

If you have lost your catalytic converter, check into local airports. They have 100 low lead aviation gasoline for less than $4.00/gallon. It is a blue color and has stabilizers in it to keep it from turning to varnish if it sits in your tank too long ( a problem with small aircraft in wintery climates when it is too dangerous to fly for months (sic) at a time ).

[padgett]

Most chip manufacturers just juggle the tables. Simple data mods. In this case we need to extend the table which requires reprogramming. The Turbo GNs used a much simpler program than the Reatta and memory that is half the size. This is why I suggested basing the MAF programming on the 92-93 L-67 supercharged 3800 which used the same computer/PROM. Comparing the difference between a L-67 and a NA 91 should reveal the programming differences much quicker than trying to diassemble one.

The problem as mentioned before is the serial stream between the Reatta's ECM and the BCM which controls the dash display and other things. The data stream must be correct or the dash goes wonkey.

So what you need is a MAF subroutine and advance tables from a 92-93 L-67 inserted in the PROM program for your car. Considerably more complicated & needs all the help possible.

[2seater]

As usual, you have cut right to the heart of the question/problem. Substituting more appropriate programming on the existing chip is probably the best solution and the most elegant. Now the big question, how can this be accomplished? There seems to be precious little information available, and even less help. As the only known expert in this field, can you help with this? I hate to even ask, but any guidance would be much appreciated.

[sahein]

The problem with reprogramming is that the source code is not available.

Whithout the source code to logically see how it functions, it is next to impossible to try to totaly decipher the object code. Trial and error strides can be accomplished for small segments of the code with alot of time and effort.

If all of the specifications were known it might be better to start from scratch and write your own new code.

[padgett]

That is not entirely true though it helps to have a few decades of experience in reverse-engineering of programs at the machine code level. The biggest difficulty is that the decompilers available for the Motorola 68xx (not 68xxx) series of processors are somewhat crude in comparison to WinTel programs like SOURCER and ADA.

That said there is a lot of data on the GM ECMs available online both free (DIY-FI/gmecm) and commercial (Tuner Cat) including some maps to the various variables. For instance if you know the location of the MPH value in memory then is is simple to identify diect accesses to that location which can give you the calling subroutines.

Some like the fan settings or desired idle speed can be "adjusted" just by changing the reference data byte or table once the location is known.

Similarly, knowing the locations and bit entries of both the current and history error codes can help find the soutines that set/reset and read them (this is how I told the ECM to ignore the transmission and EGR error codes on BB).

So, yes it does require some skill in reading assembly code particularly without the mnemonics. It also takes trial and error and instrumentation to determine what the real effects are (am presently trying to find the idle step up routine and conditions). This is difficult when constantly on the road.

Lacking a "mobile test bed" other than my '88 which I see occasionally or the factory programming manual (or just the memory map with mnemonics) things are a bit difficult but doable for the ECM since there is quite a bit of data available. Unfortunately this is not the case for the BCM. I have dumps and disassemblies but no information at all and expect it would take a concentrated effort for a month or two make sense out of, a month or two I just do not have.

However just "fixing" the MAF and advance tables to accomodate a boosted engine should not be two difficult if a dump of the factory 92-93 L-67 (think I may have one - anyone know the letter codes ?) ECM PROM and a memory map like the one from TC was available. And if I could find the time (often have some on weekends when away from home but then away from equipment as well).

What we need are

1) A 20-something CS or CT grad knowlegable in assembly language, interested in retuning cars, and who has a lot of spare time (I avoid the "H" word).

2) Someone in GM who can find the programming information or memory maps, or mnemonics/locations for 88-94 "P4" 3800 ECMs and Reatta BCMs

[sahein]

I have a 93 L67 prom from an Oldsmobile 98.

[2seater]

Both Sahein and Padgett lost me after about the first sentence. Is the L67 the supercharged one? I checked in on this post to see if it would be worthwhile to try to purchase a PROM for the suggested engine, but if already available, that would be wonderful. Believe me, if I could find someone that was conversant in what you are telling me, I would certainly talk with them. Fastchips doesn't seem interested and so far I have come up empty, but I will keep looking.

[padgett]

OK here it is: LN3 is the GM order code for the normally aspirated 3800 88-90, L27 is the 1991 NA, L67 is for the Supercharged 92-93 3800. I suspect the easiest conversion would be to take the 92-93 Supercharged program and the 16141470 ECM (computer used for both L27 and L67) and use that to run the engine since the tables etc would be correct.

Now for the dash in a Reatta to work properly, the data stream of the supercharged program would need to be changed to match the data stream expected by the BCM so that the dash functions would work properly. I *think* this would be the easiest way to get what you want but would really need to compare the wiring diagram of the computers.

What this means is that to put a supercharged or turbocharged engine in an '89 Reatta you need to use the Supercharged engine program and replace some of the subroutines from the L-67 program with the equivalent from the '89 LN3.

The replacement is not hard, just a matter of having about $200 of special equipment & takes about an hour (50 minutes to erase a PROM, ten minutes to program it). Knowing *what* to replace is the hard part.

The good news is that the two computers belong to the same family - physical size is the same, they are designed for the same basic engine, the PROM/Memcal size is the same, and while some wiring changes might be necessary, they would be simple.

Now I have located what apprears to be a prom dump from an L-67 (PROM code BADD3657) but do not have the memory map (am looking). Sahein: what is the code on yours ? Having more than one may make things easier

Question: doesn't the L-67 engine have a different EGR system than the LN3 ? This means that if you drop in the whole L67 engine, you need the later EGR program but it you just put a late (turbo)supercharger onto an early engine then you need the early routine.

In 91 GM also added the routines for the transmission to the ECM and later years also added the ABS but think that did not happen until 94-95.

Now I know we started out talking about a turbocharger but from an ECM standpoint, the program does not care how the boost occurs, just that it does. The lags are different but airflow is airflow (MAF) and it does not matter where it comes from as long as the range is right.

[Guest Greg Ross]

Padgett,

Let me know what FSM make and model year you'd like to have for the Wiring Diagram comparisons, which/ any other sections copied etc. and I'll look after getting it to you. Have a inside-the-Dealer source who'll loan the book.

And did Chuck from CA get in touch with you?

Greg

[padgett]

91 Reatta and 92-93 L-67 (Park Ave Ultra ?) ECM 16141470 wiring pinouts - should be two or three pages each. I have the pinout for the others. Also the PROM letter codes for the L-67 (do not seem to be in E-book - think BADD is one)

Other than that while there must be a service manual for the ECM and BCM, I have never seen one and doubt that they were released to dealers.

Chuck ?

[sahein]

The prom I have from the 93 Olds 98 L67 is BDND7419. It looks as if the PCM may plug in but the ECM's pinnouts are significantly different between the Reatta's ECM and the Olds PCM. Cruise Control and E-trans were integrated into the ECM, hence the name change to PCM (Powertrain Control Module). My guess is that the change occurred for 1991. The EGR on the L67 is the same as the Reatta until it changed for 1994. Also went to bigger(larger hole)throttle body in 1994. Superchargers are different for 91-93, 94-95, 96-? The differences are that the 91-95 will fit the the same SC manifold but the 94-95 SC has a larger inlet to accomodate the larger throttle body. 96-? is a different setup with the injectors in the heads rather than in the intake manifold. 91-95 SC's are 62 cubic inch displacement (m62) and 96-? are 90 cu/in (m90). I believe that all years of L67 use the same engine block.

[2seater]

Oh my god, you guys are way ahead of me <img src="http://www.aaca.org/ubbthreads/images/graemlins/shocked.gif" alt="" /> I had contacted Tuner Cat, thanks to your info. and they essentially said the same thing, same computer family but the PROM does not interchange. Possibly change the newer ECM to run the older engine? They weren't sure.

[Guest Greg Ross]

Hal,

Heard of Motorvation?

Listed in the last two Summet Racing catalogs. Contact(1-800-230-3030)

Summit sales for the referal and a Motorvation tech rep. will contact you? They're quoting $389.95 for custom chips for 85-93 GM Vehicles.

[padgett]

Any chance you could send a .bin of the Olds Prom (or the Prom, will be home for a few days next weekend).

TC gave you the short answer. From what I see the 92-93 PROM is physically/electrically the same as the Reatta (a 27C256 32 kbyte EAROM). With the series II they went to OBD (the On Bard Diagnostic, an industry standard) and a 27C512 64k byte memory). This was because the ECM became the PCM (Powertrain Control Module) and combined engine, transmission, ABS, and BCM (and probably a few others as well) controllers in one box. Wonderful as the Reatta is, having all of those separate black (silver) boxes is really a pain and a maintenance nightmeare.

However, if we stick to the original 92-93 series, mapping the boosted advance, mixture, and airflow tables onto a Reatta should not be very difficult.

(Less than a day including documentation for a GM engineer with access to the source code and ICDs but a whole bunch more difficult without).

Going to the larger throttle body (and a much larger air filter/inlet path) *probably* required recalibrating the MAF tables (could have built the bias into the MAF which would have made the programming much easier), certainly flow would exceed 170 near the shift points. For sure the 4T60 would be questionable in that configuration.

Must admit that is a project I would enjoy doing but the problem is that my employer expects other things/places and my family expects the paycheck.

Will say that to do it properly would require a facility (well my garage is big enough after application of a skip loader) and a test bed. Fact is that a lot here is guesswork. There is a ton of data out there but much is rong and telling which is often difficult.

For example I have complete disassemblies of a number of PROMs and while I know where many of the variables and some of the subroutines are, I am far from having a complete understanding particularly of the calibration routines.

The good news is that if a subroutine works GM tended to use it forever - see routines in code that did not change for over a decade of different cars and models - the bad news is that no one has ever pulled it all together. The only complete information is buried inside GM. So we make SWAGs and flip bits and hope for the best. Sometimes it even works.

[sahein]

It may take a while to get another .bin created because I have to use a freinds' equipment.

I have created a preliminary program to control the 4T60E in the Reatta.

CAREFULL OF WORD WRAP! IF IT DIDN'T PASTE HERE PROPERLY!

'{$STAMP BS2}

'{$PORT COM1}

' -----[ Title ]-----------------------------------------------------------

' Author.... Robin Sickels 15Apr2003 - 17Apr2003

' -----[ Program Description ]---------------------------------------------

' Purpose... To Control GM 4T60E Transmission on 88-89 Buick Reatta

' -----[ Variables ]-------------------------------------------------------

gear VAR Nib 'variable of program selected gear output

vss VAR Byte 'count number of vss pulses input

tps VAR Byte 'variable of tps input

sw2 VAR Bit '2nd gear switch

sw3 VAR Bit '3rd gear switch

sw4 VAR Bit '4th gear switch

vss=0 'to initialize vss value or to simulate input vehicle speed pulses

gear=0 'to initialize gear value before loop starts value range 1-4

tps=0 'to initialize tps value or to simulate input value range 0-100 (0-5vdc)

debug "reset occurred "

' -------------------------------------------------------------------------

Main:

Read_vss:

count 9,500,vss 'count the number of pulses on pin 9 in .5 second

debug ? vss ,cr ,cr 'display value of vss

Read_tps: 'read tps (throttle position) potentiometer value

high 7 'discharge capacitor

pause 1 'give time to fully discharge

rctime 7,1,tps 'read charge time and store tps variable

debug ? tps 'display value of tps

'use tps value to determine which set of vss shift limits are used

if tps <22 then tpslow

if tps <48 then tpsmid

if tps <65 then tpshigh

if tps >=65 then tpsmax

debug "tps error "

tpslow: 'the low acceleration speed constraints

debug "acceleration = low "

if vss < 7 then gear1

if vss < 12 then gear2

if vss < 22 then gear3

if vss >=22 then gear4

debug "vss error "

tpsmid: 'the medium acceleration speed constraints

debug "acceleration = mild "

if vss <12 then gear1

if vss <17 then gear2

if vss <25 then gear3

if vss >=25 then gear4

debug "vss error "

tpshigh: 'the high acceleration speed constraints

debug "acceleration = high "

if vss <16 then gear1

if vss <21 then gear2

if vss <32 then gear3

if vss >=32 then gear4

debug "vss error "

tpsmax: 'the max acceleration speed constraints

debug "acceleration = max "

if vss <19 then gear1

if vss <31 then gear2

if vss <42 then gear3

if vss >=42 then gear4

debug "vss error "

gear1:

gear=1

low 0: low 1 'make pin 0 (solenoid low and pin 1 (solenoid a) low

low 10: low 11: low 12 'indicate not 2nd 3rd or 4th to ecm

debug ?gear

Goto main

gear2:

gear=2

low 0: high 1 'make pin 0 (solenoid low and pin 1 (solenoid a) high

high 10: low 11: low 12 'indicate 2nd gear to ecm

debug ?gear

Goto main

gear3:

gear=3

high 0: high 1 'make pin 0 (solenoid high and pin 1 (solenoid a) high

high 10: high 11: low 12 'indicate 3rd gear to ecm

debug ?gear

Goto main

gear4:

gear=4

high 0: low 1 'make pin 0 (solenoid high and pin 1 (solenoid a) low

high 10: high 11: high 12 'indicate 4th gear to ecm

debug ?gear

Goto main

[padgett]

Control the 4T60E using what hardware ? Why not read the TPS directly with an A/D rather than timing a cap charge ? What is the VSS scale (1:2.55) ? Youth wants to know <img src="http://www.aaca.org/ubbthreads/images/graemlins/laugh.gif" alt="" />

BTW I suspect that all necessary information may be found at

Delphi Energy and Engine Management Systems

1200 N. Dort Highway

Flint, Michigan, 48556

[Guest Greg Ross]

[color:"#666666"] BTW I suspect that all necessary information may be found at

Delphi Energy and Engine Management Systems

1200 N. Dort Highway

Flint, Michigan, 48556

Barney, do we know anyone who works at DE&EMS? Do we know anyone who knows anyone per chance? <img src="http://www.aaca.org/ubbthreads/images/graemlins/confused.gif" alt="" />

[2seater]

I appreciate the feedback and conjecture but I feel somewhat like I am on the Tower of Babel, the language is out of my realm. Relative to the L67 ECM, if I follow the thinking correctly, either the tables in the chip are expanded to allow a wider(higher than 170 gm/sec) reading, or the MAF is somehow different and the output covers a greater flow rate in the same output band? Padgett is correct that the 170 gm/sec is exceeded, but somewhat before the shift point. I calculate I need something in the 190-200 gm/sec to cover the entire potential air flow. I know there are MAF translators made to allow larger MAF's to be fitted and still allow the ECM to read them. I would imagine the reverse might also be possible although I do not understand the principles of how it is done. If I understand the MAF output, it is a variable frequency rather than voltage output, or maybe a combination of the two. If a translator was available, would it then be possible to tweak or posibly use the existing tables? This would be a 12%-18% compression, which my conjecture would seem to indicate would lack as fine a control as the original, but perhaps useable? If I am on the planet Ozone, I would not be offended if you told me. <img src="http://www.aaca.org/ubbthreads/images/graemlins/crazy.gif" alt="" />

[padgett]

Thread is getting a bit long but will add one more. MAF tables are frequency related but are non-linear - cannot simply mutiply by a fudge factor. GM slit into 8 tables for the Reatta. Highest is 120-170 gm/sec so extending to 220 would double the range.

Am corresponding with a gentleman who extended his range to 255 gm/sec and hope to find out how soon. BTW he also went to #28 injectors are yours stock ? (think #24).

[2seater]

Yeah, this is a long thread, and perhaps it should be taken private, but there have been a lot of views of the thread, so perhaps there is some interest? As for the injectors, they are the stock 19# Bosch 901's. The stock injectors for a GN are 29.5#, which I have been told, (John Spina @ Casper's Electronics) are too large for my stock ECM to control well. I do have a set of GN injectors, plus another set from a Ford 351 which are rated @ 22 lb/hr. All of thes injectors are Bosch high impedance types, which would be operable by the ECM. BSFC calculations indicate the stockers are good for 225 hp @ 100% duty cycle, which can be extended to approx. 250 hp if the pressure is raised 8-10 psi. Much research needs to be done yet as to how large an injector may be needed. At this point the O2 volts are staying in the stock .9 range, but this can be caused by a misfire or knock retard as well, so that has to be brought under control to make a valid evaluation of how the engine is working internally. From what I have seen from Accel and MSD injector math, the 22lb injectors, (15% larger) "should" work with the stock computer calibrations. I will have to be careful with the throttle in the mean time. Thanks again for the support.

[sahein]

I intend to use a BS2 (Parallax Basic Stamp 2).

The A/D would be a more elaborate and I believe unnecessary addition.

VSS is 4000 pulses per mile. I sample for a half second.