What that says ^ . I am working on an ARAP-based bin for a friend's '89 IROC, 305TPI 5-speed, and I got pretty much everything done but the shift light just stays lit at all times.... I did a search but it didn't appear that anything came of the prior discussions about this.

Thanks,
Matt

------------------
1987 GTA L98 MD8
355, TFS Heads, LT4 Hot Cam
My GTA

The Minnesota F-body Club

 

Question, have you set the manual transmission bit?

 

Yes. That doesn't do anything with the OEM shift light operation. The threads that I read, said something about a new .tdf that would inlcude a table for tweaking the shift light function....

------------------
1987 GTA L98 MD8
355, TFS Heads, LT4 Hot Cam
My GTA

The Minnesota F-body Club

 

???? Well???? Anybody know what has happened in regards to this issue????

------------------
1987 GTA L98 MD8
355, TFS Heads, LT4 Hot Cam
My GTA

The Minnesota F-body Club

 

C'mon now!! Glenn? Grumpy?? Anyone?

 

I know thsi is old.

I tweaked the winbin ecu file so you can nail the light on, and off at certain rpms.

Limit is 6375, which is higher than anyone wants to shift anyway.

Mine's set to nail on at 5500, and off at 6375.. Figure by the time I grab the next gear, I'm at 5700 or so. Which is plenty for a boosted motor.

-- Joe

 

could you let me in on what the shift light offset addresses are for all of its features? im writing a bin editor for 6E and would like to make it as feature rich as possible on the first go.

 

I'm not sure what all of them are. I'm sure you can dig that information up. Some people have very well commented the $6E source.

What you need to consider, is using winbin/gmpro ECU files rather than writing an editor for just $6E.

Sure the interface to winbin is ugly.. But the ecu file's arn't a bad format. I have been working on and off on a editor written in C/C++ which should build on both Linux, and non POSIX OS's..
I'll incorperate winbin ECU file formats, rather than writing my own, or making it specific.

-- Joe

 

Winbin's ugly look and poor UE is a result of using the ECU's. (He must dynamically create his interface based on the ECU. This makes it difficult and limiting.

My editor will use separate DLL's for each type (6E, etc). Right now, as proof of concept, I'm only creating a 6E dll.

I could only find on explination of the shift light in 6E (below). It doesn't tell me all the info I need though (for instance, factors, axis labels, RPM breakdowns, etc). I can *almost* make sense of it, though. Perhaps someone could clear this up for m.

;==============================================
; E LIGHT PARAMS
;
;==============================================
LC2ED: FCB 120 ; 50C, (122F) MIN FOR LIGHT ENABLE

LC2EE: FCB 47 ; 47 MPH MIN FOR LIGHT ENABLE
LC2EF: FCB 0 ; 0 TPS MIN FOR LIGHT ON
LC2F0: FCB 0 ; 0 LV8 MIN FOR LIGHT ON

LC2F1: FCB 56 ; 1400 RPM THRES FOR LIGHT
LC2F2: FCB 208 ; 5200 RPM FORCE LIGHT ON

;----------------------------------------------
; E LITE ON TABLE
;
;
; %TPS vs RPM
;
; ARG = %TPS * (256/100)
;----------------------------------------------

;----------------------------------
LC2F3 FCB 13 ; 5% 800
LC2F4 FCB 46 ; 18% 1600
LC2F5 FCB 118 ; 46% 2400
LC2F6 FCB 192 ; 75% 3200
LC2F7 FCB 192 ; 75% 4000
LC2F8 FCB 255 ; 99% 4800
LC2F9 FCB 255 ; 99% 5600
;----------------------------------------------


LC2FA FCB 52 ; LIGHT OFF 625 RPM
LC2FB FCB 204 ; FORCE LIGHT OFF, 5100 RPM

;----------------------------------------------
; E LITE OFF TABLE
; TPS% vs RPM
; ARG = %TPS * (256/100)
;----------------------------------------------

;----------------------------------
LC2FC FCB 18 ; 7% 800
LC2FD FCB 51 ; 20% 1600
LC2FE FCB 123 ; 48% 2400
LC2FF FCB 197 ; 77% 3200
LC300 FCB 197 ; 77% 4000
LC301 FCB 255 ; 99% 4800
LC302 FCB 255 ; 99% 5600
;----------------------------------------------


;----------------------------------------------
; E LIGHT ON DELAY MODIFER
;
; ARG = sec * 10
;----------------------------------------------
;
LC303 FCB 2 ; E LIGHT MULT
;
; sec's %TPS
;----------------------------------
LC304 FCB 150 ; 15.0 0.0
LC305 FCB 150 ; 15.0 12.5
LC306 FCB 150 ; 15.0 25.0
LC307 FCB 150 ; 15.0 37.5
LC308 FCB 150 ; 15.0 50.0
;----------------------------------------------

 

What are you writing this in ?? I hope it will be fairly cross platform.

 

There are alternatives to the Winbin interface that still have .ecu file compatibility. Here is one that you might like :

GMPCM

It was originally created to be an .ecu file creation utility, but has since been extended to include full binary editing capabilities, with compare and graphical analysis features.
Since Winbin is limited to less than 300 address areas defined, it uses an internal file format capable of thousands of items, which makes it possible to "completely" define a 32k GM binary file. It includes a definition for 3.1 L V6 (VIN T), the $A1, that has over 800 address areas defined.

 

Cool beans.. I'll give it a try..

-- Joe

 

C++. it will be cross MS OS, but will not work on a mac. ;-)

I'm going to take a look at GMPCM and see what it looks like, Thanks!

 

So, what do you guys think of the GMPCM software?

It is designed to interface well with Microsoft Excel... you can save any table as a spreadsheet, manipulate the data with 3rd party optimizers, and load it back into the program.
Does anyone have Excel installed, that could check these features out, and give me your opinon?

 

Mechanic,

Perhaps a new thread is in order?

Downloaded and installed GMPCM. Install was rather quirky. Install was looking for several files that could not be found. Seemed to be an issue with short vs. long filenames, probably on my end as I still use a DOS ver. of PKzip. Got that fixed by renaming the files in question to proper long filenames. Install also had a problem with WININET.DLL, but seemed to finish properly.

GMPCM will open and run(?), I can open ABOUT, it will give ver. info and actually did an update of itself. When I try to open the sample pcm file it gives the message:

Could not open file
9 Subsript out of range
OK

Upon clicking OK the program closes. End of game. Any ideas? Hosed install? Thanks.

Regards,
Gary Anderson

 

The dos version of pkzip is probably not going to work well, the version of Winzip used is 8.1 if that helps. You can get the latest version of Winzip here:

http://download.com.com/3405-20-960683.html

After extracting with the latest version of Winzip, I would try reinstalling, and if it seems to have any problems writing a file, select "Ignore", and proceed. There are files that would be required for a non-internet explorer equiped system that are included, but files such as an older version of wininet.dll, may already be installed and are currently being used by your system, hence the failure to overwrite. The program will work fine with any version of wininet.dll that may already be installed, so ignoring the installation of this file is completely acceptable.

I should also mention that the software was designed to be used on a fairly high performance computer system (ie. pentium processor), and requires at least 16mb of memory to run in a Windows 95 environment, with about 24mb needed for a Windows 98 machine.
The fact that you were able to load the interface, but experienced a failure upon attempting to load a data file, could indicate that the error is associated with insufficient memory, but thats hard to say. Increasing your swap file size could possibly help here.

I really do appreciate this and any feedback from those that have successful (or unsucessful) experiences with this software, and would also like to encourage any suggestions as to improvements or ideas for desirable features that I could include in the future.

 

Mechanic,

Removed previous install of GMPCM. Downloaded Winzip and reinstalled GMPCM. Install went much smoother although still hiccupped on WININET.DLL. Started program and same as before. Did the update to ver. 1.0.66 and same result. Same error message as before. I have Win98 and 256megs RAM so I doubt memory is a problem. I'm clueless.

Regards,
Gary Anderson

 

Well, dont know why it would give that error, works fine on all the beta test machines... Its possible that you have a corrupted system file, but that is hard to say, as other than you being kind enough to offer feedback, I have had very little input to go by.

I did upload an update (1.0.67), which may be a marginal work around at most. Please let me know if this works for you.

At this point, unless I can verify that it even works ok for anyone at all, I am probably just going to scrap the project...

 

After updating to ver. 1.0.67 and selecting the sample pcm file it says:

"Can't open a database created with a previous version of your application."

After clicking OK 6 times it opens to a screen with tabs for Definion, Constants, Tables & Flags. There is no data in any of the boxes and any attempt to put anything in any box results in:

"Runtime error 91
Object variable or block variable not set."

Clicking OK then terminates all and closes program.

Regards,
Gary Anderson

 

Aw.. that explains it. What version of microsoft access do you have installed? I believe you may have a corrupted MS Access system file. The GMPCM software will always try to use the latest version of the Microsoft Access Jet Database Engine that it can find on your computer.
So, if you have a corrupted version of the jet engine, say MSJet40.dll, the software will attempt to use it, rather than the jet engine that I distribute - MSJet35.dll.
In this case, you may need to reinstall microsoft office so that it corrects the newer corrupted access jet engine file before the software will work as intended.

 

Mechanic,

I did indeed have Msjet40.dll and right next to it is Msjet35.dll. I renamed 40 to Msjet40.bak and rebooted. GMPCM acted the same way. I noticed a few more dll files with same names and versions 35 & 40 next to each other. Would any of these other dll files be causing this? Is renaming them a fix or is there more going on here?

Regards,
Gary Anderson

 

It is unlikely that simply renaming the files will do any good. The windows operating system stores information as to which version of the Jet engine it should use. There are really only two possible solutions to this problem :

1. Reinstall office 2000, so that the corrupted Jet 4.0 files are repaired.

2. Remove Microsoft Access 2000 using the Office 2000 setup utility.

IOW, if there are corrupted files, they need to either be fixed, or removed from the system. This error could also occur due to settings for Admin privileges in the Access 2000 environment having been changed from their default values, so it may be better to just remove Access 2000 and reinstall at a later date.

 

Mechanic,

I never had Access on my computer, just Excell and Word. I tried reinstalling with no change and I removed Excell and Word. Also no change. Thanks for your help. I'm done for the nite, it looks like the Packers need my help.

Regards,
Gary Anderson

 

Sorry that it did not work out for you. I do appoligize for any trouble that may have been experienced by anyone here.
Prior experience's reported by others have been quite positive, and the software has worked flawlessly on all of the beta test machines, (still does), so I am really at a loss at to what would have caused these issues at this time.

 

I was able to install it, update it, and open the sample file. The help screen says that I can import a .ecu file but there is no 'import' feature listed in any of the pull down menus.

 

The "Import" menu item should be located directly to the right of the "Edit" menu item at the very top of the application...
Here is a screenshot of what it should look like :

GMPCM Screenshot

 

Thanks, I get that screen when I load your sample file. Is there a way to load a $6E ecu file and use it?

 

Yes, you just import the $6E .ecu file by selecting "Import" from the menu at the top, and then "ECU File" from the list that pops down. Then you locate the .ecu file using the "Open" file dialog window that appears, and if the .ecu file uses a correct Winbin file format, it will load. (A status dialog confirming that the .ecu file was loaded successfully will appear.)

You can then simply import a $6E compatible binary file, (ARAP.bin for example), and you are set to go.

Once you have done this and have saved the .pcm file using a new file name (the "File" "Save As" menu commands), you will not need to repeat the above procedure.
A pcm file contains both the definition AND binary data...

You can make any desired changes to the definition areas in the pcm file, and then "Export" this modified definition information to a new .ecu file if you wish.

In fact, you can create a Winbin .ecu file entirely from scratch, which was actually one of the original functions of the software. It is quite similar to the Tunercat TDF editor, but much more advanced in that you can see how the definition relates to the binary data in real time.
The software was intended to be a VERY useful tool for hacking GM binary code, or for transfering information from a commented dissasembly into a more user friendly format. It's capabilities have since been extended to include all of the features of the Tunercat Tuner interface.

 

I did locate a problem with the software download package. It seems that the demo pcm file ($A1.pcm) had somehow corrupted possibly when the package was compressed. I have since fixed the problem with the download package, and also have the demo file available for separate download to save time. It is located here :

$A1.pcm

 

When I try to import a $6E or $32B ecu I get a corrupt ecu error message. I have the 1.0.69 version and the new $1A.pcm file.

 

Me too. (6Ev009.ecu)

 

The .ecu must have the correct Winbin formating in order to load.
A brief look with at the ecu file with a text editor such as notepad should reveal the improper formatting quite readily. If you have any questions on the Winbin .ecu file format, just ask them here and I can address them.

 

uh, the ecu has no troubles loading in winbin and correctly enumerates all specified fields.

 

Here is an example of a correctly formatted .ecu file. It is located on the DIY ftp site :

ftp://ftp.diy-efi.org/incoming/

It is called : 1227165 $32 BUA Version 5.ecu


The correct formatting is like this "

/* Item 41: Maximum Spark Advance */
{
/*startAddr =1E,
/*columns =1,
/*rows =1,
/*elementSize =1,
/*bitMask =0,
/*offset =0,
/*mulOrDivOrBit =1,
/*factor =2.844,
/*map_name =Maximum Spark Advance,
/*ylabel =Adv,
/*yaxis =,
/*xlabel =Degrees,
/*xaxis =,,
};

Note the blank line at the end. This is simply a requirement of the program at this time and is quite an easy thing to fix if it is incorrect. I think you will find that most .ecu files will load fine.
The fact that Winbin can still load the file is irrelevant, as this standard format is simply a requirement of this software...

Another thing that is important to note is that a "New" file should be opened first, and then "Import" and ecu file into the new "untitled" file, which can then be saved using a file name of your choice.

 

This is the key to getting it to work. I was not choosing new before importing. 2 out of three ecu files now load.

 

Yes, I probably should have mentioned this sooner... There would be no logical reason to use an existing pcm file for importing of a .ecu file, so I did not plan for that scenario. Just a programmer kind of thing. :-) At least I finally realized that it may be important to point this out.

 

you might consider revising your software to take these things into account. the 'standard' ecu format (created for PROMedit, a DOS based bin utility, and only subsequently inherited by winbin) does not specify that a line must be in between each new line (as they are called). This means it actually *is* relevant that it can successfully load into winbin.

Meanwhile, I appreciate your support. sorry for misleading you to believe i know nothing of the format. carry on.

Here is the actual format specification:

-----------------------------------------------------------------------------------

All lines start with "{" and end with "}"
All fields on a line are seperated by a ","

Line one:

filesize in hex eg 16 K is 4000
checksum address eg 6
checksum size eg 2
checksum end address eg 3fff
number of tables to follow eg 3

NOTE: one table per line
NOTE: addresses assume first address is ZERO (0)

Line 2 to the end: (each dash is a new item on the line)

- start address of table in hex
- number of columns in table in dec
- number of rows in table in dec
- number of bytes in each item in this table eg 1 (usually) for byte wide data in dec
- 0 for normal multi byte data editing, 80 if bitwise editing required
- offset is "b" in y = ax+b 0.0 in example
- 0 for multiply (eg factor MUL rom value), 1 for divide eg( factor DIV rom value)
- a conversion/scaling factor 0.125, 6.25 etc this is "a" in y=ax+b
- a name for the table eg "main Ignition" 40 char max
- a name for the units along the vertical scale eg "RPM" 20 char max
- a concatenation of the 5 char strings used to name the vertical scale eg " 400 600 800 1200 1600....." 100 char max
- a name for the units along the horiz scale eg "Map" 20 char max
- a concatenation of the 3 char strings used to name the vertical scale eg" 20 30 40..." 54 char max

 

I like the interface. Has some quirks but we'll give it more tries tonight.


-- Joe

 

Just thought I would add a note for those that were interested in the GMPCM software...
I have issued an update that addresses issues some may have encountered with Winbin .ecu file formatting, and also adds a new feature for multi binary burning that should work very nice with Craig's switching circuits. You can update the software quite easily, by selecting the "Update" option available in the Help menu.