SCCAjunkie

Hi,

Kinda new to this stuff, but you guys have really done a lot for me in the past year. I'm finally getting decent with this stuff, and have begun to transfer the knowledge over to my other love, turbo 2.2's.

My question is, how can I give back to the community? I'm considering doing tuning for local people to help get them started, but I'd also like to do something for the group that has already helped so many.

To just a few-
Glenn, you given me something to shoot for- highway mode, here I come!
Grumpy, you continue to amaze me. I love the no-nonsense approach.
Craig, where would we be without you?
Tim, I have no doubt that your info has saved me hours and hours of time tuning my VE tables.

Sorry to get sappy, but I think you guys don't realize exactly *how incredibly much* you've helped people.

So, how can I help?

Grumpy

You might try getting some folks together and doing a Source Code.

IMO, the whole DIY-EFI scene is getting stagnet. The info is out there for anyone who wants to read, on how to do about anything they want.

The next step is generating, the actual code in a form that can be modified. The means getting together some real sharpe folks and **just doing it**. It's gonna be alot of work, but it can be done.

Once you get to having Source Code, then the sky's the limit. You can get rid of all of the junk in the code. Stripe out the EGR, CCP, and use those outputs for any number of other options. Redo the IAC stuff, keep it for the choke function, but then get rid of the throttle follower stuff. Then use some of the later better idle strategies, there's just all kinds of stuff folks can do.

Making a group effort would make it so much easier.
Seems like to me, that all's that's really needed it getting it organized.

While you think your car might run good now, it can be made more tuner friendly and then you'll want to work on the details and see what perfect really is like.

BTW, glad to see your enjoying yourself.

anesthes

iTrader: (12)

Ideally it would be nice if someone or people could start working on code in C.

Assembler was great back in my Commodore days but after over 10 years of software development on Unix platforms, I have no interest in assembler.

-- Joe

Grim Reaper

Space is a major concern. No room for "overhead" that higher level languages tend to generate. Also, in private discussions I have had with Bruce and RBob, there are times you will find the code doing certain "bizzare things" that you wonder "why in the hell did GM do it that way when they could have done it more efficiently another way"...and the answer is "Sometimes it's an opportunity to improve the code and other times it is a necessary evil for timing purposes". Might be pretty tough to get the right "time" in some of those loops in anything except Assembler. So start learning Assembler - it probably won't work in C both in terms of "Bin Space" or "timing".

And Matt, welcome. Would be nice to see some work on the Turbo 4 stuff. With the high price of gas, I am seriously considering a Turbo 4 banger for my wife and our "commuting" vehicle.

Grumpy

Ideally world peace would be nice.

But, all the C3 and P4 stuff used in the 3rd gens is already in assembler. Reverse engineering the code and changing languages, just makes an already difficult item ridiculously complex.

What's needed is people willing to contribute, not complain about whatifs. Whatifs are what gets nothing done.

S10Wildside

I can't really contribute to this, but as a professional developer (a year out of college) I can see how the ECM could/should be written in C#, which is .NET. I'm just not smart enough at this point, but it makes perfect sense for the ECM to be object orientated. Even if I was fluent in C# and fully understood the existing ECM, I don't know anything about electrical engineering to actually put together some electronics and a chip to controll all this. I think you can buy chips with operating systems installed.

This doesn't help the cause, but this is what I've been thinking about lately. Actually I've been learning C# by creating classes such as, CoolantTempSensor(), Fan(), ThrottlePositionSensor(), etc.

RBob

iTrader: (1)

The overhead would kill ya'. Not enough processing time for all that !@#$^&*. Take any Windoze product and then tell me that it doesn't bring a Giga-Hz 64bit machine to its knees. Stock GM ECMs of the C3 class run at 1MHz, P4's at 2 MHz., at 8 bits.

Sorry to get worked up but I hear this stuff all the time. If you'd like to go C, C++, C#, then look into the MegaSquirt. Although it is a fuel only box. Just please don't bring up that you'd like to run on a PC.

Learning Motorola assembler is a piece of cake. It's not like you are being forced to learn Intel assembler . Heck, it would take more time converting all of the beautiful GM algorithm's to C? then to learn the asssembly. Oh wait, you would have learn Moto assembly first in order to understand what GM did in order to re-write it.

RBob.

anesthes

iTrader: (12)

Rbob,

I dont agree with you. The compiler is going to optimize the code fairly well, and in some instances better than you might even do in assembler. This argument is old, and very seldom do assembler
programmers actually write code that is better optimized than a compiler would have done.

The windows analogy doesn't fit. Windows development libraries are terribly redundant. Take a look at M68K Amiga stuff, and you'll get a better idea.

Yeah MC assembler is nicer than x86 stuff. x86 makes me sick, 6502, 68000 is great. The 68HC11 is very similar from what I've seen.

With that said, yes I agree reverse engineering would take plenty of time. But at the same time, I think if someone took the
time to API the driver interface, then writing the main loop and sub functions wouldn't be much extra effort.

The issue of loop timing however is valid, but thats because the GM guys didn't understand threading back in the early 80's.
Then again, looking at some of the code I don't think they understood 68HC11 either...

But whatever, I'll shut up now. An idea is just an idea until someone actually puts it to good use.

-- Joe

RBob

iTrader: (1)

>> Rbob,

>>I dont agree with you. The compiler is going to optimize the code fairly well, and in some instances better than you might even do in assembler. This argument is old, and very seldom do assembler programmers actually write code that is better optimized than a compiler would have done.

A complier can never create better assembler then a human.

>>The windows analogy doesn't fit. Windows development libraries are terribly redundant. Take a look at M68K Amiga stuff, and you'll get a better idea.

Having worked in CBM engineering for 5 yrs I know the Amiga stuff. Again no analogy can be made, the Amiga OS is multi-tasking with different requirements. It is also deeply ingrained with the video hardware.

>>Yeah MC assembler is nicer than x86 stuff. x86 makes me sick, 6502, 68000 is great. The 68HC11 is very similar from what I've seen.

Agreed.

>>With that said, yes I agree reverse engineering would take plenty of time. But at the same time, I think if someone took the
time to API the driver interface, then writing the main loop and sub functions wouldn't be much extra effort.

There is not only the human effort, but the overhead imposed on the hardware.

>>The issue of loop timing however is valid, but thats because the GM guys didn't understand threading back in the early 80's.
Then again, looking at some of the code I don't think they understood 68HC11 either...

I have trouble agreeing with you here. The GM engineers understood the 68HC11 instruction set very well. Remember that the C3's are not a 68HC11, they are a custom chip with a subset of the 68HC11 instruction set.

And what need for threading? It just isn't required. The code in these ECMS (as used in the f-body's) is well thought out, slim & trim. There are two minor loops that each run alternately 80 times a second. Then there are 16 major loops that run round-robin 10 times a second each. The GM engineering guys even included a test to check if any loops ran over. This is a true RTS.

Time critical code runs in the minor loops, non-time ctritical code runs in the major loops. The firing of the plugs (EST) and the firing of the injectors is independent of the software. To the point that I've run '747s to over 7,400 RPM without a problem.

>>But whatever, I'll shut up now. An idea is just an idea until someone actually puts it to good use.

And this is the gist of it all. I have seen a lot of hype and no action. It would be nice for someone to prove me wrong and run a C3 with code compiled in a high level language. With a minimum of the original functionality included.

RBob.

>>-- Joe

Grim Reaper

That's damn near blasphemy when you talk about GM ECMs.

anesthes

iTrader: (12)

RRob,

> A complier can never create better assembler then a human.

Like I said, this is the an old argument and I'm not really taking a side on it.


> I have trouble agreeing with you here. The GM engineers >understood the 68HC11 instruction set very well. Remember >that the C3's are not a 68HC11, they are a custom chip with a >subset of the 68HC11 instruction set.

See now here is my error, I didn't know _that_

I see the point with the loops thing. I've seen this type of timed loop fail in too many real-time applications but, I guess in an environment where the sensors are alwayws returning values, processing load is not an issue, Then its fine.


Glen,

> That's damn near blasphemy when you talk about GM ECMs.

I wasn't debating his experience with GM ECM's, I was dissagreeing from a software devepment point of view, which I think I'm qualified to do.

You have to also understand, my point about writing the code in C was from a developer preference point of view. I'm not going to get into that argument where "ok well write this in your language and tell me how fast it runs, how many lines it is". GM ECM is written in assembler and it allready works, I won't debate that. Its just if I'd *prefer* to be contributing in C..

Now, what I *am* willing to start doing is seeing what the hell the real differences are in $8D (7730) vs $58 (7749). Other than
timing, and fuel delivery based on load vs rpm, are there really
any tables that benefit on a repinned TPI type car.

Is $8D better code than $58 in areas other than >100KPA ?
Rather than running $58 and repinning, does it make sense to
rescale to 190KPA, and add some more >100KPA logic?

This is the type of stuff i'd like to get info on.


-- Joe

Grim Reaper

And I graduated in Computer Science in 1977 and worked in a variety of computer languages including various Assemblers and "high level languages". And RBob and Grumpy are also "old timers". All of us are speaking from "actual experience".

I will assume that you haven't really dived into the existing code because if you had, you'd understand some of the obtstacles we are talking about. There are MANY TIMES, when you see some bizarre algorythms where the GM Engineer did some weird "bit manipulation" all to do some form of a divsion. Your immediate reaction is "why in the hell didn't GM just do this"? But, you find that GM did this solely because they needed a little "cycle time" so they made the code use an "inefficent" method all to burn off cylcles.

I don't how you'd be able to emulate that if you used a higher level language to generate the object code. And, inspite what you may think of C, it WILL NOT create IDENTICAL object instructions as the Assembler - sometimes it may be as efficient (though you may wish it to be inefficient to burn off cycles) and other times C will generate extra object code you don't want.

My suggestion is to dive into the Assembler by documenting a "disassembled bin" (maybe just a block of code) and then think about the C code necessary to generate the indentical code...and you'll see that it can't. And unfortunately, you may have to.

Grim Reaper

I personally like the $8D, but that's because I've been working with it for the last couple of years (and still not as proficient as I would like to be).

Rescaling the MAP sensor to work as a 2 Bar or a 3 Bar is quite simple BUT you introduce two major problems. (I'll assume that a simple swap from a 1 Bar to a 2 or 3 bar is "linear" in it's corresponding voltage to kpa reading...which is not exactly true).

1) The tables relying on the MAP sensor loose "resolution". Instead of having a 5 kpa or 10 kpa increment in various tables relying on the MAP sensor, you now have 10 and 20 kpa increments (or 15 and 30 kpa if a 3 bar MAP sensor). This can be HUGE for fuel and spark trim in certain circumstances.

2) Most of the MAP tables have a "minimum" kpa with 20 kpa being the most common bottom boundary. Generally any thing below 30 kpa indicates deceleration and triggers DFCO. And, steady cruise in high altitudes can result in MAP readings below 40 Kpa. With a 2 BAR Map the minimum kpa is now 40 kpa and you now loose ALL of your DFCO and higher altitude cruise fuel settings. And a 3 bar MAP results in a 60 kpa minimum...you now don't have any fuel readings for most cruising.

To make the $8D work properly, you need to modify and create new MAP tables. 1) to increase the resolution you will loose and 2) to lower the bottom threshold so you can regain your DFCO and cruise.

RBob

iTrader: (1)

The biggest difference between these two maskIDs is in the fueling calculation. And not due to the 1 bar vs 2 bar, the $58 also runs as 1 bar with the flip of an option bit.

The $58 code uses the common BPC * VE * MAP * ~T * ~AFR fueling calc. The $8D code first calculates the airflow via ideal gas law and then calculates the required amount of fuel from that.

The $8D code is more refined and with additional functionality over the $58 code. Some of this is due to the use in a Corvette opposed to a Sunbird or truck, and also due to emissions requirements.

RBob.

anesthes

iTrader: (12)

Hi,

Thank you for those responses.

With all that said, Am I too assume your both in the opinion that

1) $8D is a better bin, and is much more refined for performance
AND economy but the map tables need to be redefined.

Now, what about importing the map tables, or the logic from $58?
If your resolution is from 30kpa tp 190, it would in theory be pretty much the exact same.

I guess I'd have know exactly where and how it calculates based on kpa, and so on.

-- Joe

need4speed

Yes it is linear. You just have to change the map calculations.

And it is not simple. You affect every table value and every value that is compared with the map calculations. It's not hard, just time consuming. The hardest part is hacking the ALDL programs to display correct map values. Just try tuning 2bar table with a program that displays a 1 bar map range




1) 20 KPA per scale on the ve table is fine. Make your lookup table
larger (more columns).
The MEFI3 ecm only has 13 columns for a 2 bar sensor. But is has three tables based on rpm range.

2)Not true.

I have already been though this with the 88051 Ecm. I have recalibrated it for a 2 bar map. My Ve table is a 30 -190 Kpa at 10 kpa per division. That includes all lookup tables,baro and offsets on both the event and time side. You just wind up with new formulas to base your math on.


Here is a clue. There are 2 major calculations for the map sensor readings. One is to get the map value range for use with airflow
calculations. It will become a 9-208kpa for a 2 bar map sensor.
The ecm always calc's the BPW from the airflow map value.


The second is scaling the map value for table lookup use.
This is used for all of the offsets.
It will become 0-255 for the 30-190kpa range. That is an "on center" value for the table cell. Entry point will be 1/2 of whatever your KPA per division is. In this case it's 5kpa.
Change the minimum entry value on your 3d table lookup to 0 and you now have an entry point of 30 kpa.

Other calc's include baro and vac.

You can also change the table to become a 17 column lookup.
4-point linear interpolation tables are the same regardless of the ecm. That code was not written by GM. they just used it.

Base your table inputs from different rpm values (12.5,25,33.25 rpm per bit) and you can increase the table rows by rpm ranges.
You can use the same algo. Just compare different rpms then point your index to whatever table location you need. Then
add or sub off your entry values as needed.

You can now have a low, medium and high rpm range VE tables that overlap.

All of Gm's code is modular. You can move subroutines around as required to make room for any changes needed. Assuming you have enough space in the Eprom. Otherwise something has to go.


I don't bother with any of the earlier ecm's. They are not fast enough. Any HC11F1 based ecm that runs on an 8mhz clock is fast enough for any motor. Not to mention there is plenty of free space on the eprom.


Need4speed

JohnL

This post illustrates both the advantages and the disadvantages of the internet as a tool for communication and organisation

SCCAjunkie waxes eloquent over the help he's received from this board and asks what he can contribute. When he apologises for being sappy, I know exactly what he means. This forum is a fantastic resource. We'd like to be part of exploring this really interesting stuff and do something useful too.

The most useful thing I've found in one place on this board is Tim's Prom Intro item. Lots of stuff from lots of work and posts is organised and located at one spot- newbies just get pointed there and, with some harder slog, wading through relevant and irrelevant posts, you can get the wonderful satisfaction of making your car run on "your own chip." And then you can pass this on to others and save the experts' time in responding to entry level questions.

RBob's code patching "sticky" appears to be a start in the direction of educating the ignorant like me, but "doing a Source Code" in Grumpy's words is several orders of magnitude in difficulty above using a chip editor and an eprom programmer.

And then this thread drifts off into a fascinating technical discussion about stuff I haven't even dreamt of, far off the point of "how newbies can help."

I can't believe that Grumpy says, "You might try getting some folks together and doing a Source Code." You might as well not have wasted the electrons on this, but you keep making similar comments on many posts, spoiling the value of your obvious technical expertise. You've confused a technical exercise with a managerial/organisational exercise. The man asks, "how can someone kinda new to this stuff help?" and you say, "just do it".

It would take most of us years to get up to your speed and we haven't got the time. There is no guide I've seen to "source code programming" even remotely like Tim's PromIntro. If I'm wrong, please correct me.

I think the answer to SCCAjunkie's question ATM is, "learn what you can from all the disparate sources of correct and incorrect info, try out some applications of what you think you've learned, and then post it for others' benefit." That is, continue the many individual useful efforts.

Expecting newbies to understand how to lead and manage a large cooperative volunteer technical effort is unbelievably naive. [I hope you enjoy the Grumpy writing style ]. If you want to make this go forward in an organised way, it needs both technical and managerial leadership, otherwise it's only going to go forward on the basis of some brilliant individual work.

Sorry for the long post.

John

Grim Reaper

That's part of the reason these post can be useful. Most people get into eprom burning because of "a need" (they modified their engine and cannot get it to run properly - possibly even after an aftermarket prom has been burnt for it). Few get into eprom burning as part of a "plan" to build their engine in the future.

Hearing some of the "possibilties" that you can do should you get into source code may be the catalyst you need to take that "next step". Simply going through the exercise of "documenting" your bin (even if a fully documented version of your bin exists) is a wonderful way to teach you Assembler and get your mind thinking of "the possibilities". Yes, it's time consuming. I've been doing mine for two years now as free time permits and I'm still not done. But whenever I find myself wishing the ECM would handle something or do something differently, I now can make it actually do it.

Trust me, the second greatest moment I had in eprom burning (after the first time I made a change to the bin, reburned it and it worked), was the time I actually modified my source code to do something GM never intended, had it assemble error free and it worked!

anesthes

iTrader: (12)

> I can't believe that Grumpy says,

Well. Grumpy doesn't owe any of us anything. His contributions have helped many. Besides, why do you think we call him Grumpy???

> RBob's code patching "sticky" appears to be a start in the
> direction of educating the ignorant like me, but "doing a Source > Code" in Grumpy's words is several orders of magnitude in
> difficulty above using a chip editor and an eprom programmer.

Sometimes people expect others to understand things. I've spent a good part of the last few years in Kernel space, so you can see where I was coming from in posts earlier. If you're not a coder, or a developer your prolly like "what the hell will I do with source code" Coding is for some people, for others editing registers with a hex editor, or a tuner is good enough.

A good example of editing code would be "making the shift-light useful", or changing timing when N20 kicks in because a driver
hit +5 volts.


Keep in mind, this isn't a development team. This is a bunch of guys who have hobby's. None of us get paid for this, and because their is soo many variables in operating environments (different cars!) its hard to have multiple people working on common goals.


-- Joe

need4speed

Isn't that the truth.
My main reason for getting into this years ago was not to make money on fuel injection.

It was not to SPEND any on fuel injection

Grumpy

Hey, well put,
too bad your so far out in left field you have little insight into the subject matter.

Feel free to do some actual research and look at things like Programming 101, 808, and the DIY-WB.

In Programming 101 *we* reverse engineered an ecm with a rough disassembly, an ecm bench and a few volunteers. And LOTS OF EXPERTS said that couldn't be done. Then we did 808, and AGAIN LOTS OF EXPERTS said Good Luck. Then *we* did the WB project, without reverse engineering a darned thing.
BTW, not one person directly involved with 101 or 808 was a colledge grad.. And only 2 had any colledge training in computer science. But 4 drove trucks with 747s, and I had my first gen ecm bench. 7 guys living throughout the USA, with just a *net* connection.
Ya, NAIVE, but it worked.
Hard to actually agrue with something that does in fact work.

So here I've documented 3 cases of couldn't be dones, it's just the limited vision folks that have no ambition, to actually do anything that slow things down.

Management Skills?, what a laugh, just takes the desire to do it. And someone with enough ambition, and forsight to get something actually done.

And while I did have a huge hand, I have actual Source Code for my car, in MAF and MAP configurations.

On another list private list we are doing a source code project.

So as far as all your comments go they are are just flat out wrong, or represent someone terribly misinformed.

To bad there are so many folks that don't want to even try. I just love hearing the same educated excuses people use for failing, and worse then that even trying.

Naive, right, hahahahahaha.
Naive got you here, if you actually think about it.

Ed Maher

Actually, you're just looking at things all wrong. When grumpy says "just do it" or offers suggestions, it is a purely technical excercise and idea. Doing some engineering isn't about sitting around talking and figuring out who is responsible, and when it needs to be done. THAT is management.

Technical 'work' is about saying "that would be a neat idea" and doing it. So what if you don't have the slightest clue how it needs to be done, that just means you have some learning to do on the way. But until you actually start doing something, YOU are the managment team, worried about factors besides the actual work.

Grim Reaper

There's a lot of wisdom in those three short words. The whole concept of the entire website is a bunch of guys who've "done it" and sharing their knowledge and experience with those who "want to do it".

The only "management" I need is to justify to the wife "why we need to do it". The "trailer towing package" excuse only goes so far. She keeps asking when I'm going to install the trailer hitch?

need4speed

I hope you guys are putting your efforts into something other than a 20 year old ecm.

Grumpy

And if not, why should you care?.
Your not contibuting to it.

If you look at the latest of the aftermarket ecms, they still haven't begun to match the oem systems, of the 80s. They do offer some slick eye candy but nothing more then that.

<other then closed loop WOT, which I personally feel isn't be done properly, ie, no redundancy, and risking an engine on one sensor, isn't my idea of good engineering>

If you were to do some actual research, you'd see how good the oem stuff really is.

Lets see my 148 was lasted used in 88.
With my Source Code, and data logging of the WB, using lockers gives me the same tunability and performance of the $2,000+ (and the WB option pushes the number closer to $3,000) aftermarkets.

So tell me *ONE* meaningful thing that I've missed.
The aftermarket has done well advertising, as advertising does, but eye candy, and silly claims not withstanding they have nothing new to offer.

If you read thru and experiment with some of GM's strategies, it becomes more and more apparent the genius in them. All you have to do is spend some time reading the various hacs around and it will become obvious, who really understands ecms.

Grumpy

So everyone here is just wasting their time working on cars that are about 20 years old.

At least some of us are making an effort to actually push the sport forward. Unlike some that just post as a means of posturing, and playing the *better then*, games.

need4speed

You really are dumb, aren't you.

I never said go aftermarket. I'm a big thorn in those guys side. I promote the use of factory ECM's.
I said check out some of the later ECM's. Or should everbody stick with old head and cam desings to?


Check them out. They have more features, are faster and can control an electronic trans. There is nothing quite like dialing in a trans to shift hard when you need it, but not bang the hell out of your car when it's not.


Yeah, I don't know the first thing about the algo in ECM's.
I think it's more like you only know 1 or 2 ecm's.

Grumpy

Reread your last paragraph, and when you actually catch up on the subject matter you'll be in a much better position to speak with some intelligence about the issues at hand.

For basic engine management, the ecms that have been in use since the mid 80s have been just fine, so that's amlost 20 years of evidence that they did operate fast enough. Once you get past the Windows mentality of faster meaning better, you can see that an engine only operates so fast. At 1Mhz an ecm is twiddling it thumbs (figuratively speaking) waiting for things to do, but if you want to run the math feel free to and then you'll have an idea of what your talking about.

And once Source Code is done, and folks understand that then they can progress to something else.

Have you looked at the history of ECMs and tuning?.
Feel absouletly free to read thru some archives and see where most all of this began. It all starts with baby steps.

But alias, you seem tied to the world of feature creep, and talking about things you have no conception of.

BTW, just maybe, I have checked out the later PCMs.

As just a point of information, in the GM world when the powertrain is controlled by a same module as the engine, that is called a PCM. PCM meaning Power train Control Module.
You might even try to use the proper terms.

SATURN5

Still laughing over the ignorance of this statement.

Over the course of several threads I have yet see you post anything that really contributes to them.

You seem more interested in bitching at Grumpy than backing up your statements.

What have you done to help others on this board?
You have been on thirdgen for 2 years, with less than 100 posts.
Which means you either, just lurk, or have nothing of importance to add...

My guess is the latter..

HighHopes85

https://www.thirdgen.org/techbb2/sho...highlight=Cell

Some of the best answers I have ever gotten, both on forums and through email, are from people with less than 250 posts.

Chuck!

How much different is the Motorolla Assembly from IBM mainframe?

SATURN5

Thank you for proving my point...

That is nothing compared to what Grumpy, Glenn, Trax, RBob, DOHC32v, Craig, and others have contributed.

Instead, its become a pissing contest....

need4speed

I give up. The hardware on the later ECM/PCM (I go by the internal code $F4, $F5 not the box) runs at 2mhz. The core
speed is 8mhz.

There is a little something called Delta/Rpm for update rate on the hardware. It can and does get skipped, missed on high hp
high rpm/sec engines. The hardware will remain in it's last state until it gets updated.

And one more note. The later MAF systems run the entire fueling routine in IRQ. So your choice of MAF or speed density is closer to real time. In fact it is so fast, there is a counter or cylinder ID that is used to skip the lookups below 2000 rpm. So yeah, I do know
how the timing is setup for sample/update rates.

Anyhow, it appears that you guys have it covered and don't
need any help.


Enjoy.

Grumpy

Was expected.

Helps appreciated, so when you actually going to offer some?.

So far up until this post all you've done is torpedo'd getting anything done.

You gonna take the lead and do a Source Code project, or just talk like so many others.

BTW, who cares if the ecm lags the engine by a cylinder event?
Or runs one more loop of code before the next update?

You can worry about that, and do nothing, or some what settle and actually get a Source Code up and working. Lets see, up and working vs sitting at the keyboard worring.

need4speed

Ok, I'll bite.

I use Ida Pro 4.3.0.740A for a disassembler,
and Cosmic Idea and Zap for writing and compiling.

I use a Unirom UR16-1M-45 for Eprom/Flash emulation, bench testing/debugging and "tune on the fly"

What ECM / routines are you interested in?
I'm not quite sure I understand what you mean by source code.
Assembly or C ?

If you are looking to identify routines and what they do,
take a look at this.

http://24.130.98.97:8080/31_IDLE_FUEL.ASM

This is what I wrote for the idle PI fueling code.
It is for the 16197427 ECM, subroutine L863D
The off-idle part was removed for demo purpose, but it's
in the same sub.

Is this what you are talking about?

You guys can contact me off list.

RBob

iTrader: (1)

Nothing to see here folks, lets move along. . . Really now, this stuff has been around for 20+ years. Although the band-aid you call 'AE clamp' may be nice to have, only if the AE can not be tuned correctly.

RBob.

JohnL

I love the smell of napalm in the mornin'

Need4Speed- are you taking your bat and ball home because Grumpy's been rude to you? What else did you expect- your only mistake was in using the same writing style. This is the internet, and none of the spilt blood is real. Even Grumpy may not be real, or he might really be a nice bloke. Please stay and provide newbies with advice- I enjoyed the post that got HighHopes going after a year of being stuck, and no-one else was answering his question. Did you post that to help someone who knew the answer already?

The question at the top of the thread is still, "how can I contribute". All the people who are warring above have helped me out either directly or indirectly, so it's obvious just by participating in this rather incendiary forum you can help.

When I was saying, rather provocatively that what was required to "do a source code" was a "managerial/organisational exercise", I wasn't implying that there was none on this board, only that it was naive to expect good managerial/organisational expertise from newbies. I still haven't seen anyone demonstrate otherwise.

Further, I don't think anyone could fail to be impressed by the organisational and technical impact of this board and other internet resources. This sort of cooperative/ community effort must be a fundamental threat to the competitive edge of major corporations. It's pretty to watch.

So Grumpy, you said, "Naive, right, hahahahahaha. Naive got you here, if you actually think about it." Close, but actually "Ignorance" got me here. "Naive" got me to build my own eprom programmer, leaving me to waste several weeks before I found someone who could fix the chip problem

It would be nice if Someone developed a "Source Code Intro" like Tim's masterpiece, and Someone organised to select an ECM to target and break up the task into bite-sized pieces. Does this sound like a job for a newbie? I don't think so. So, again, the answer to the question appears to be, "help out where you can individually".

John

Grumpy

It was **my** naivety that got things Public Domained enough to get the ball rolling.

Before Source code intro can be done the technic for generating source code needs to be figured out. So that the average guy can do it for his specific ecm.

I've mentioned it here, numerous times about getting together and *just dong it*. But, no one's taken the challenge and done it. yes, there is a heck of alot of editing that takes days to do that reguires only basic typing skills.

Grim Reaper

Not to be a bad guy. But if you had read through the "archives", the whole discussion of the BLM table and controlling the edge boundaries was discussed over two years ago by myself and others. In fact I get into discussing the advantages of modifying those limits to utilize more of the table.

It's just that people never want to read the old archives. The reason we have more than 250 posts, is we keep answering the same question all the time. If I only answered each question ONCE, my number of posts would be significantly less.

Grim Reaper

Why would I want to work/install and LS1 PCM in a 20 year old car? This IS a Thirdgen Site and caters to Thirdgen Cars.

Besides, if we "cracked" the LS1, some slimeball would just "rip off" all our work and start making money off everyone elses effort. No thank you.

I will work on an LS1 PCM when I get an LS1. As for the LT1 - it too is a 10 year old design and it isn't getting any youner either.

But most importantly, this is a Thirdgen site it's SUPPOSE to cater to Thirdgens. The fact we discuss anything else is actually outside the scope of the site.

HighHopes85

I wouldn't have minded a "told ya so" answer since I don't have much of an ego to deflate. I searched the archives 'here' and 'there'. I guess using the correct terms of BLM Cell just got me a lot of stuff that had to do with BLM values and not actually cells, and then the stuff that did deal with cells was actually "I changed them"...not "Here's how the hack can be broken down".

It was for a program that kinda went step by step through some of the calcs the ECM used. For me, for everyone. I could have fed garbage into the proggie and had it report false stuff. Wouldn't have helped anyone out. I'm not Einsteinian enough to get on the "Secret Decoder Ring Source Code Mailing List" so I had to post up here and beg for help. I got the help without kissing any ***. What a crying shame.

I should probably stay outta this. For me, PROM burning is a hobby. Opps, did I say "HOBBY" out loud??? YMMV

need4speed

So tell me, where exactly is the map math covered on this site.
And how the values are multiplied and scaled?

I did a search and could not come up with anything.

The info I gave him was not about how to find the cell limits, he knew that. His question was about some confusing numbers on the hack that did not jive with the Tunercat program.

The info I gave him was intended to show him how to look it up.
Not just follow simple branch instructions.


Need4speed

Grim Reaper

Whatever.

Grumpy

Try the links off of the www.tuncercat.com and there should be a site called the ecm guy's page, all the relavent math is there.

It's kind of hard to understand what it is your actually asking for.

If you have a guestion, ask it. Not every one stops by here every day and reads every new post.

Being vulgar, ain't no way to ask a guestion.

Grumpy

Look at the stickies at the top of the page.

Tunercat keeps the scalers hiden for some reason, so the editing is kind of weird.

TRAXION

... and finally we have a lock

Tim