Hi guys,

Its been a while since I've done much with my car. Its a 350 '92 Z28 with AFR 190s 58cc, CC 212/218 112 lsa cam, 1 3/4 SLP headers, Zeitronix WBO2, and stock pistons and bottom end. I'm thinking about getting a procharger D-1SC.

Will the 58cc heads put the compression to high?

I'm looking at 36# Blue Top injectors and one of the 10 or so 3-bar MAP sensors listed at code59.org. Any suggestions for which MAP sensor to use or are they all the same?

The ECM is a 730. I've burned chips in the past and have flash chips, a burner, Ostrich emulator, tuner cat that I used for $8d bins, and tuner pro rt.

If I get a procharger, I'd try using the $59 code. It supports wideband O2 and a 3-bar MAP (I'm not planning on running anywhere near that much boost) - very attractive features. Plus they have an autotune script that may be useful.

I believe the wiring changes I need for $59 are:
F6 --> F4 since I have a 700r4
WB --> F14 for $59 input (which will equate to E14 --> F14)

Nothing would be connected to the stock O2's E14 pin. Do I still use the E15 ground pin?

Which $59 bin is the one to start with for a blown 350? Anyone have one they've modified for their 350 that they'd share?

The $59 code is supposed to come from source, not modified assembly, correct? Is that source public? I didn't see it on the code59.org site.

Anything I've over looked?

Thanks!
Joe

 

As far as GM map sensor, it must be a 3 bar and I would suggest a GM one as they seem to be more reliable. Not sure which list you saw but I suspect is just different part #'s for the same sensor but different manufactures.

For a base bin since you have 36# injectors I would start with the 33# (aka stock injector) bins we have posted. The -1 -2 etc is for different fuel levels.

Load one up, if lean/rich try another see how it goes. You will need to just swap it to 8 cylinders from 6, and select your WB o2 from the flags in tunerpro. Make sure only 1 WB flag is choosen.

Yes $59 is based on source code which is then compiled. Lots of guys spent a lot of hardwork making it compile, and the new changes so it isn't public. The bins and supporting info is all there for free.

Pin F14 is the wb o2 input. If in the future you want to run closed loop, code59 can do that directly from the WB input, no need to simulate a NB input. There is a table where you can run closed loop for any AFR you desire also. I wouldn't suggest closed loop until your tune is dialed in really well though.

 

Thanks for the info ty1295.

Here's the 3-bar MAP sensors listed on code59.org:

GM Style 3 Bar Pressure Sensors
Manufacturer Part Number

GM (old)* 16040749
AC Delco (old)* 213-101
GM (new)** 12223861
AC Delco (new)** 213-1562
MSD** 2313
MSD (compact)** 23131
Borg Warner** EC1678
NAPA CRB219481
Electromotive Engine Controls 71130
Holley 538-23


You suggest running open loop until the tune is close? The computer can't try to correct for a bad tune when in open loop, so isn't it safer to be in closed loop when testing a new tune?

 

ty1295 covered pretty much everything. I will say I'd be a little nervous...alright...alot nervous on 58cc combustion chambers and boost. Do you know what your CR is right now? I know on my old 350 my CR was 9.75:1 with 67cc heads. nd wouldn't think of giving it anything over 10lbs with 93 octane. My new combo is 72cc heads with a 8.7:1 CR.

 

Yes, those sensors are all just different brands. Should all function the same.

By no means throw a chip in there and go give it hell, take things easy. Hard to hurt a car if your careful with idle, then slowly work yourself up and into boost. Take your time, be careful and pay attention to what it is telling you.

If you "try" to run closed loop while your tuning, you and the computer will be fighting each other. "No my turn, you didn't tell me you did that" are going to be words you and the ecm will exchange. Like trying to shoot a moving target.

You may be confused as a lot of other ecms, guys will tune for a BLM of 128. That really is not even possible way of tuning $58. We have 2 BLM's 1 for open throttle, 1 for closed throttle.

You want to know what the ecm is doing from the tune directly, not with some unknown "adjustment" thrown in. The autotune won't work properly either if your in CL.

Hope that helps.

 

Hmmm, that's not what I was told by Paul, one of the site administrators at Code59.org. He said, "Well we already have fully commented $59 Source. Don (1981TTA on thirdgen) gave us fully commented and documented $58 code with wideband fueling already built in and we started from there."

There are not really many changes to $58 code other than for WBO2. As I remember, there are two small changes for the 2 to 3 bar conversion, and two for the "V8 EGR fix," that's it. And, these, as far as I can tell, were not developed over there but rather came from other sources and added to $59 code.

Reverse engineering a .bin into assembleable source code is not that difficult. I did it for both $58 and $59 in a weekend while vacuuming 64+ gallons of flood water from my home, and playing with grandkids. My $59 source code is very sparsely commented for now but following it is not terribly difficult if you are familiar with assembly language as every data point has a meaningful label (for example, MALF43_55_Mask instead of L8037). In my $59 source code I've identified the changes in code between $58 and $59, all additions and deletions, and an obvious error where someone along the line mixed up decimal with hex numbers. I have not included in it, the changes to the standard $58 tables, which there are many; although I do have that in a separate document somewhere. Anyway, L98 Z28, PM me if you would like my $58 and $59 code.

You might think twice about switching to a 3 bar MAP. Bigger doesn't mean better. If you will not be producing more than 15 psi boost then you will lose precision. If you are going to produce more than 15 psi then 3 bar is the way to go, else you are fueling blindly above 2 bars. Let's use an oil gauge analogy: Suppose your maximum oil pressure will be 50 psi, your reading the pressure will be more precise if you are reading a 60 psi gauge then if you are reading a 120 psi gauge, especially if reading the lower end of the scale.

As for your heads, what were the CCs of your stock heads? My '87 came from the factory with 58cc chambers and a 9.8:1 compression ratio. 9.8 is kind of high for supercharging on pump gas so water/alcohol injection is a good idea. It provides several benefits and is one of the reasons I wanted source code to work with so I could more easily add control for it as well as other goodies.

 

This thread is getting hijacked some.......

Jeff(ty1295) is also an administrator on code59. Jeff said exactly the same thing I did. Don worked on the code, documented it and go it to compile. I am sure Don had help from others to do this. Don then gave us the code and asked us not to give it out but ok'd us to make changes to it. Don is also a member of Code59. Turbodig has then proceeded to add a bunch of hacks and patches we had for 58 into 59. Turbodig has been making code patches for 58 for years. He help with the inital 3 Bar 58 patch and made all the wideband patches for it along with the wastegate table changes to be MPH based. So he rolled them into 59. All of Dons wideband sensor code and tables for 59 came from Turbodig. (if you search thirdgen you will see where I gave 1981TTA the tables and patches)

There are alot more underhood changes than you can see. Turbodig rewrote the 3D Table code to handle larger 3D Tables. There is a bug in the stock 3D Table code on large 3D Table. He correct it so we now have a 16x32 fuel table in 59. Not the 9x9 that 58 had. We are trying to put together something like Super AUJP project but based off the 58 code. So I hope you support it.

Here are some other changes that were done besides the closed loop wideband fueling and sensor selection.

Already in 59
Expanded Chip to 32K
3 Bar Required - At this time, we may make a 2 bar version.
ALDL Code Fix - Faster connection to ALDL and also makes it so the Autoprom will work in Autoprom mode.
Wide Band Sensor Selection
Closed Loop Wide Band Fueling
16x32 F29_AFR Table - Closed Loop AFR Table
16x32 VE Table F29x covering (400-6400 RPMs, and 0 - 300 kPa (30psi)) - Open Loop Fueling (WOT) - Tables F29 and F30 are no longer used for VE fueling.
F70_MPH - Change Boost control from TPS% to MPH based.
F71_MPH - Change Wastegate Control from TPS% to MPH based. Helps to reduce boost spiking in 3rd/4th gear.
Updated ALDL Outputs
- Actual AFR
- Desired AFR
- AE/DE/PE/DFCO Flags - Not in 58
- $3f map variable (Real 3 Bar Map Variable, used by F29x, F29_AFR)
- $3e Filtered Map
F29x Table Only - No longer use F29c (Closed Throttle Table) at idle
Expanded Boost Multipler F77x Table - Covers a wider range from 84 - 292 kPa
Updated BARO LOGIC - Fixed Bugs in the Stock Code Baro Logic so it actually works - Final change still in testing.
V8 EGR Fix

Upcoming Features
Knock gauge output - Already running on my Truck using a Moates o-Meter to show my how many Degrees of KR you are getting. Matches output in ALDL. http://www.code59.org/forum/topic.asp?TOPIC_ID=632 page 3 has the Circuit Diagram to drive the Moates Gauge - well any gauge that is programmable from 0 - 22.5 and there is a table to tune the output for the gauge so it matches the aldl. This code is not in the V18 release. It will be coming in a future release.
Alky/NOS pump/fuel control - drive an external circuit to control Alky pump - Low Fuel pressure cutoff nitrous, revert fueling/timing, low Alky revert to differnet fuel/timing, tuneable table for Alky Fuleing to drive pump, etc.
Fuel Pressure Input into ECM and output into ALDL
EGT Input into ECM and output into ALDL
RPM/32 - 8000 RPM support
MAT correction for WG init DC
Learning Boost Control - Based off Lotus boost control into 59. Basically Boost BLMs for wastegate control
Spoolup Mode - Faster Spool up from a stop
Kick Down Mode - Faster Spool up from a Roll when you get on it
Security Mode
2 stage rev limiter
LeanCruise Mode - AKA Highway mode for Improved MPG

So we have a lot of things in the works, might take some time before they come out.

This is somewhat of a misnomer. Yes you will be losing some resolution but not as much as you think. If you do the math the difference in the ECM's MAP value between a 1-bar and 3-bar sensor is 0.4 kPa/bit vs. 1.17 kPa/bit. I'd be interested to see the person who's tune is so well defined that they comprehend ~0.7 kPa.... :-)

Anyway back to Joe's original question, I would start with the stock 33bin. You will need to make sure you Change the cylinder select for a V8, you might need to change your timing table some to match your CAM Specs but that chip should be a good place to start from.

 

Blown87IROC if you choose to not use code59, great. Nobody has a chain around your ankle to use it. If you don't like how we have done the details again don't use it. I am fairly certain though our project has released more information and in a cleaner format than any other. Not saying is better or worse, everybody is smart and can make their own choice.

Glad you disassembled it. Your my hero.

Now what point would we have to be telling everybody false information, and claim we did stuff we didn't? Now some may try to profit off "false" claims. hmm, we give all this away. Can't be that.

We like to walk around with big ego's is the only other thing I can think of. And if you actually knew us, that isn't it.

So we have no reason to claim stuff we have not done.

But to each his own, have a GREAT day.

 

I said there were very few "code" changes and mentioned there are many changes to the stock tables but didn't mention anything about the added extended tables (none of these tables have to do with code changes) but did mention that most changes were actually developed elsewhere, which was confrimed above. Sure, there was a lot of work done but all of it originated with General Motors.

I didn't simply disassemble it. I edited the disassembly into fully relocateable and assembleable source code (which you incorrectly term as compiling).

What???? What false information?

L98 Z28, here's a RapidShare link to $58 (AYBN) and $59 source code.
http://rapidshare.com/files/10569568...eCode.zip.html

 

Woah. Things are getting a bit too out of hand, and from what I can tell, over nothing, really.

For anyone wondering the lineage of code59, it's based on a disassembly by Don Dibble, originally. There are comments in there that I'm fairly sure came from other sources- there have been various fragments of routines floating around the internet since the late 80s/early 90s. If you look at the source commentary, you'll find lots of references/credits to other folks.

From day one of the project, it was an underlying principle that all references would be credited where credit was due. (If it was possible to determine the source) If we've missed somebody, please, let us know.

Everyone directly involved in the development realizes that we're standing on the shoulders of giants. If it weren't for Ludis' schematics, ECMguy's disassemblies, Rbob's disassembly and contributions, Bruce, and the kind souls who trusted a few folks with some good inside information, we wouldn't be here. None of the stuff that people take for granted in terms of tuning tools, equipment, and know-how would be here if it weren't for the work that these guys put out there.

It was also intended, from day one, that the source would be released to the public. Up to now, that hasn't happened- not because we changed our mind, but because we wanted to get the commentary more complete, and we were also waiting for a stable release. (We're still technically in beta versions)

We have, and will continue to give access to source for people who are interested in helping with the effort, with the understanding that all the work will be freely released. We do currently ask that people with the source keep it close to the vest until we're ready to release. That's pretty much the case with lots of alpha/beta grade products, since there's no way to insure that they won't hurt something until they've had a chance to be tested. I haven't yet denied access to anyone who requested it; who was qualified and was interested in helping the project.

We will release the code when we get to a good release point, which we hope to be soon. I'm currently the only developer that's really active on the project, and I'll admit to not being as active as I like.

We have many, many planned upgrades (as per the feature list), as well as a few things that are floating around in my head that aren't on the list. (Mainly because I haven't worked out whether or not they're feasible)

I'm quite proud of what we've accomplished so far, albeit quite small. The code does what we needed it to, and we managed to resolve a few shortcomings of the code as released be GM. And, we've learned a lot.

Hopefully this can clear up some of the understandings. Please let me know
if it doesn't. The last thing we intend to create is hard feelings among the community that we're doing this all for.


Dig
code59.org development

 

I'm certain Blown87IROC, that had you made your intentions and knowledge of assembly known or even offered to assist in the development of the project, these guys would have had no problem providing the source to you. As you know, many people who USE the bins do not want or desire the source. There are not many willing to participate in the actual code work so distributing it out to anyone who asks can do nothing but cause problems for those who only need to make parameter changes.
I have met Don and discussed code sharing if it would benifit each others cause. He offered to share with Z69' and myself (S_AUJP V4) once they got the bugs worked out. I don't run boosted apps but possibly in the future the two could be combined to run N/A or boosted with one set of code. I haven't taken them up on it because my knowledge of boosted performance is not good enough (for me) to go after such an task.
We always look for people who are willing to help with the thought process.
From the sound of it, you could be a great help in adding code features and confirming operation in the $59 if you are willing.
Just my $0.02

 

I did that, both here and there, and have been ignored for the most part. Over there, trying to understand what I was seeing in the code I'd disassembled and convey my level of competence, I asked about several issues that didn't jive between the code and the xdf file, as well as what I'm fairly sure was a simple editing error in the source code that has gone unnoticed in the machine code. No biggies, though. No one besides me has answered my questions yet. When the above issue of source code came up I intimated that I was not a beggar but was certainly not offered any. I also offered a simple suggestion in circuitry for water/alky injection that is easily calibrated for any application, which TurboDig thought was a good idea.

Over here, to convey my level of competence, I found a not-too-old thread asking how to add a high-temp engine shutoff that no one provided a solution for, which I wrote some very simple code for it that should work but still cover the cautionary concerns raised. Then, in another thread, rather than beg for commented source code, I asked if anyone had a map of RAM usage to further my own commenting effort--no offers for even that much information.

I was actually a member here many many years ago (but forgot my login and password, and no longer have the old email address) when Grumpy (God rest his soul) was just starting to disassemble code and only had a rudementary knowledge of assembly language. Having written a checksum program, I had already been burning my own chips for a years thanks to some help with table addresses from the Meaney brothers (former GM engineers who founded DFI, later bought by Accel), and had several years experience reverse engineering so-called "secure" microcontrolled systems. When I offered my assistance Grumpy scoffed at me--oh well--I found other things to do with my time. Perhaps people shouldn't be judged by their quantity of posts.

... or would know what to do with it BUT there are the few do and they shouldn't (IMO) have to somehow join some secret clique, especially when the vast majority of code was produced by GM, and most of the added hacks come from public info several years old. By keeping code secret that's even only in beta certainly doesn't assist others who are intersted in developing and might assist, and doesn't seem in the spirit of the GNU principle.

My short range goal is to make the NVRAM module work with $58. I think I may have found a slight hardware difference between 730s and 749s that can be easily overcome but if that doesn't pan out I'll likely just port over $8D ALDL code. My problem now is that I'm not connecting with the 2-transistor or RS-233 adapters I've made. My test bench appears to be working (I can flash codes, etc), my serial port works with an old 2400b modem, and my scope is seeing serial data on my cables in both directions. I've got to get communicating before I can do any testing.

 

Seems like a good time for me to jump in here....

First, let me say Turbodig nailed every point exactly. I'm just going to try to add some historical context to some of the comments.... I did one of the original disassemblies of $58 and used existing disassemblies for 7730/7747 ECMs to help guide me through the "common" portions of the code. (Things that show up virtually identically like table lookups, lag filters, etc.) I used the resulting disassembly to implement a number of changes that had either been previously patched at the HEX level (V8 EGR, WBO2 voltage logging) or needed a bigger effort like moving from the original 16k chip to the 32k chip. The link here pretty much outlines what was done (with a LOT of help from a number of people posting in this thread!): https://www.thirdgen.org/forums/diy-...band-code.html

This was definitely a personal effort applied to a much larger public effort that had taken place. For another example, here's a link where JP86SS and I are discussing ECM I/O using Ludis' schematics to figure everything out : https://www.thirdgen.org/forums/diy-...tml?highlight=

There are a number of other examples throughout the years of messages on this board. You can see the encouragement, constructive comments and help everyone was very willing to offer throughout this project. The history is there if you want to look around a little. I can't imagine anyone was trying to mislead you on the origins of the project.

As mentioned previously, $59 was "born" of sorts with the changes you've identified (and were also spelled out in the first link I've posted above). Effectively, this was version "1" of $59. LOTS of things have been added since that original version.

I understand assigning emotion to internet posts can lead to dangerous assumptions and bad feelings. And, I obviously have a personal bias towards the work done on $59. But, I don't think I'm far off base saying I'm getting a very condescending feeling from these quotes. There was (and is) much more to $59 than simply disassembling/reassembling. I *did* sit down for hours trying to add comments to the assembly so it would be easier for others to understand. Further, it was among the first efforts (like the s_aujp work) to get the software in a state where it was well enough commented and partitioned that more than one person could understand the logic flow *and* make software changes at the assembly level (as compared to HEX editing). Further, you've seen the list of functions/improvements/modifications that have been implemented since that first version. Lots more than a simple disassemble/reassemble job. Many of these changes are effectively developed from scratch as compared to simply changing a single address. I apologize if I'm taking your statements out of context. But, it really seems (to me) as if you're saying "there wasn't much work done on the project since you've achieved the same results in a weekend". Once again, if I'm misreading what you typed, I apologize. From your posts, I get the impression you probably have the knowledge and skills such that you could (or can?) execute a project like Code $59. It just seems like a cheap shot minimizing the efforts of people who have given their time and effort to help others.

Joe, sorry for taking your thread a little further into the ditch....

 

L98Z28, I hope you've gotten off the ground with what you're trying to accomplish. Im still gigglin like a little kid because I was able to get my project fired up after a 4 year build using the $59 code. I can't even imagine the time and effort these guys put into this (although this post highlights some of it). Use any of those 3bars IMO, I personally used a MSD 3 bar I ordered from Summit.
BE CAREFUL with the boost until you have 100kpa and below tuned and YES, run in open loop untill you have everything good there.

 

This is sorta what I'm talking about, WRT to misunderstanding. There are only 2 people on our team that were really qualified to answer your question- Don and Myself. Don isn't around as much as he has been in the past, so that pretty much leaves me to answer the questions.

I was in Texas from the 28th (When you posted) til April 4th. I replied to your disassembly question (at least one of them) on the 7th. (Yesterday)

code59 isn't my only project, unfortunately... there are many. I've got a motor to build, a flowbench to finish, DIS ignition to fab and work out, etc.
Not to mention, a 3yr old who likes to watch cartoons with Dad.

I would think, that given your recent flood experience, having other priorities would be understandable.

We DO appreciate it when things are pointed out that are wrong/may be wrong. There's a lot of ground to cover, and the more eyes looking at it,the better.

I did reply to one of your posts, and I'm still researching the other one.

Once again, I think this is rooted in misunderstanding.

Bruce (Grumpy) scoffed at everybody at one time or another. That was just his way.

The number of people that have the kind of information you're talking about I can count on fingers and toes. And, for the most part, they don't watch the boards waiting for a chance to reply with good information. Thankfully, some still do. Also, a lot of them are like me, and have their fingers in too many pots, and don't have a lot of free time on our hands.

There has also been a real ugly past of commercial ventures or other posers taking the info/work that many have spent years on, and profiting. This, understandably, leaves people a bit reluctant to share. As I said earlier, we have given quite a few people access to the real source, but we knew the people involved, and that their intentions were good. I don't think this is an unreasonable thing to ask.

If you're using Monte's tcl code modified for the $58 stream, you'll likely have problems. The appear to be bugs in the Mode 2 routine, and I'm still trying to work out where they're happening. (via a logic analyzer)

Also, $59 has a patch in place to fix a bug in the ALDL receive routine. We also made it available in a binary form for $58 users as well, as it addresses connection issues with many scan tools.

Needless to say, this is something that we're also interested in getting working. We do work off a "schedule" of sorts, in that we're trying to finish previous list items before we move on to others, so it'll likely be a while before I can get to this. Hopefully, you'll have some luck in your efforts.

Dig
code59.org

 

1981TTA,

I’m sorry you feel demeaned but you’ve taken me out of context. ty1295 said,

Quote:

Yes $59 is based on source code which is then compiled. Lots of guys spent a lot of hardwork making it compile, and the new changes so it isn't public. The bins and supporting info is all there for free.

“Making it compile” doesn’t take that much work, period, but that fact in no way demeans the effort it takes to produce commented source code. I pointed that out because, as Paul said, Code59 began with fully commented $58 code that already had the WBO2 in it, and if you take out the factory code and WBO2 code you are left with very few lines of code, code derived mostly from public hacks so I felt ty1295 was misleading. I made no comparison between your source code and mine; in fact I said mine was sparsely commented. Over yonder I mentioned that the speed of which I was able to produce fully relocateable, assembleable source code was in a great part due to the efforts of folks like RBob who had a hand in creating the xdf files.

Sorry if I come off condescending here but “moving from the original 16k chip to the 32k chip” takes almost no effort. That is the essence of relocateable code, which is produced by eliminating all hard coded addresses and replacing them with labels, as are my posted source files. By taking my posted $58 code and changing the ‘org C000’ to ‘org 8000,’ and adding ‘ fill 0,16384’ after the last line of code (FF7D), you will have a working 32k $58. Then, you can add in any code you want, where ever you want (as long as appropriate, i.e. not in a table or in the middle of some unrelated subroutine, etc), then adjust the number of fill bytes to account for the added code.

I never said I didn’t like Code59 or that I didn’t think it was a good compilation. The reason I mentioned the 3 bar issue was not only due to the loss of resolution but, from reading archives, just changing the MAP multiplier in two places doesn’t produce, as I recall, a linear output. This deficiency is overcome by kludging the VE tables to try to make up for the problem. Why should someone’s tune suffer any because of the addition of unneeded code for the application? (Of course, this could be easily overcome by adding two code references to a new 2bar/3bar flag bit that allows easy selection, and a change back to more stock VE tables.)

Quote:

Originally Posted by 1981TTA

There are a number of other examples throughout the years of messages on this board. You can see the encouragement, constructive comments and help everyone was very willing to offer throughout this project. The history is there if you want to look around a little. I can't imagine anyone was trying to mislead you on the origins of the project.

Unfortunately, from my reading here, it appears much of the encouragement, willingness, and sharing has subsided here since Grumpy’s passing.

It’s obvious there is a communication problem here that is surely my fault. Since life is short and I have other things to work on, I will leave y’all in peace and wish y’all the best!

 

[QUOTE=ty1295;3708827]Now what point would we have to be telling everybody false information, and claim we did stuff we didn't? Now some may try to profit off "false" claims. hmm, we give all this away. Can't be that.

We like to walk around with big ego's is the only other thing I can think of. And if you actually knew us, that isn't it.QUOTE]

My hat goes off to TY1295, Turbodig and the unnamed others for new and old code that have given everyone another choice besides just the $58. I for 1 was having way too much troubles with the Moates Autoprom and $58 .

 

I have a lot of reading to do tonight...

As for compression, stock is listed as 9.7:1 with 64cc iron heads. I'm using 58cc aluminum heads with a FelPro 1010 gasket (my notes say is 0.039" compressed, 4.166" bore). My old calculations say 10:1 compression - I'm not a guru and this could be wrong. Guess I wasn't planning on getting a blower back in the day when I built it.

Do I need to add heads to the parts list? Ouch!

Joe

 

Joe, I would say 10:1 is too high personally...a D1 is going to put you in the 8-10 PSI range I think with its stock diameter pulley. You'd be tied to 93 octance even with a 3 core intercooler.

TALK to procharger before you buy. Their extremely knowledgable and alot of their employees run prochargers.