Call for Cal Docs

[1project2many]

Cal docs that made it out of GM engineering years ago have been saved, copied, stashed, and regarded as precious metal. But for the most part they haven't been released to the public. So this is a call to get some of them released. They refer to systems that in some cases are approaching 25 years old and don't have much to generate in the way of wealth for the holder. They're not going to enable people to undercut a business with huge profit, nor will they cause amatuer tuners to stop using the tools currently available. What they will do is enable or assist in cleaning up and completing disassemblies that have been floating around in various forms for years.

If you have one of these documents or if you know anyone who has one please take a moment to consider releasing it. There's an honest effort going on right now and much of the work is just covering ground that has already been covered. Make the effort to improve the hobby for all. If you're a member of a forum that may have members with access to this info, consider posting a similar request there. The more of this info we get into the public the more we can do overall.

[Playtoy_18]

Throw that out on gmecm list also maybe

[1project2many]

I know we have a couple of people here who are members there. Maybe someone could forward the request?

[RobertISaar]

i don't mind doing it, i just have to remember how to start a new topic, it's been a while.

i should probably have a useful fairly early 3800 document up soon, from that, a lot of the newer stuff can get a rough(but mostly complete) map out as well.

[Playtoy_18]

Not sure if any of this helps from random searching,but here it is.
random interesting GM stuff regarding pcm's,mostly obdii.

http://www.mathworks.com/automotive/...s/michaels.pdf
http://www.obdclearinghouse.com/inde...t_file&id=1357
http://www.obdclearinghouse.com/inde...t_file&id=1522

I thought this was interesting,as you can get more bits of info that might help.
Link below shows a model of their process for building engine calibrationsit looks like.
nice screenshots of whatever they use?
http://www.mathworks.com/industries/...ngs/berger.pdf

A techlink post regarding aftermarket calibrations.
http://www.archivedsites.com/techlin...-coverage.html

It's got math on it?
http://www.mathworks.com/automotive/...coldengine.pdf

according to this
http://www.slideshare.net/pmaloney1/...ration-general
they use this which is for generating optimal engine calibration?
http://www.mathworks.com/automotive/...coldengine.pdf

[1project2many]

That info above looks very interesting. I wish I still had access to fully functioning versions of matlab to try this out.

This is the calibration parameter summary for $58, the turbo Sunbird and Syclone code. http://www.gearhead-efi.com/caldocs/...58_summary.pdf

This is the caldoc. http://www.gearhead-efi.com/caldocs/...$58_caldoc.pdf
The caldoc contains primarily the same info as the parameter summary. It also has some engineering notes as well as the memory locations / variables used by the code.

Use those docs with this file (can be found in various places on the 'net) http://www.gearhead-efi.com/caldocs/Turbo_P4_Doc.pdf and you can work out quite a bit of the code.

[1project2many]

Robert Isaar has contributed this paper for the 1228253 \ $5B calibration. Excellent description of software operation includes alternate data communications modes for BCM / CRT / Digital HVAC equipped vehicles, mode 4 operation, and general software function.

27 Mb file: http://www.gearhead-efi.com/caldocs/...Cal%20Info.pdf

Thanks, Robert!

[RobertISaar]

something important to note about the calibration section, i'm going through and making an XDF now, there is a 2 byte shift around the ~$200 range(in the middle of the diagnostics cals, oddly enough), causes weird values if you don't catch it, i haven't seen anymore like that yet though.

[1project2many]

Not something we typically work with, but still might come in handy for someone.

This file is for a MEFI2 controller. The MEFI controllers were designed for marine use but these days are used in all sorts of vehicles from Harleys to hotrods. The MEFI2 was programmed using a flash tool and I'm told the procedure was similar to the early flash GM units. The early MEFIs used no VE tables. Instead, you program injector on time vs MAP and RPM which means it would work with a 2 BAR MAP sensor. It may support CNP ignition and it uses the same min/max/ref values as an automotive cal so DIS is an option. O2 and VSS can be enabled as well as reduced power mode for high CTS or low oil pressure. It supports a scantool so a definition file can be made for TP if desired. If you have a chance to use one it makes a great underhood, standalone controller for an EFI retrofit.

[1project2many]

This is a little light on info but it's still handy to have around if you're working on a disassembly. It's a calibration map for an LT5 calibration. The calibration valuesand ecm variables are listed along with GM's names for the variables and tables.

[daleulan]

according to this
http://www.slideshare.net/pmaloney1/...ration-general
they use this which is for generating optimal engine calibration?
http://www.mathworks.com/automotive/...coldengine.pdf

This is a very slick way to generate calibrations. There is another very similar way that is used which is called 'non-design of experiments'. You can use the same Matlab tool with a few scriptfiles to extract information from essentially random driving and generate calibration tables from that. I have done it to get misfire threshold data from a few days of on-road driving. What I did was basically drive for a few days while logging crankshaft acceleration and jerk values along with engine speed, load, and transmission gear and a misfire generator counter. I would then repeat the test with a misfire generator set to maybe 5% misfire. If you extracted the data using this tool and a small script to get the actual misfire data out (synchronized with the misfire generator) you end up with four surface plots - a reference surface and a misfire surface for each of jerk and acceleration. Yet another small Matlab script to generate the threshold surfaces, export into INCA (the ETAS calibration tool) and voila, misfire detection acceleration and jerk values calibrated. A humanoid still needs to look at the data to make sure it makes sense and there are no cusps in it. Also any data gathering that is limited by atmospheric conditions needs manual correction or extrapolation. I was just playing around with CAGE and thought that maybe it could process existing logged data. It managed to do it so I thought I would try it. Later I talked with Peter Maloney and he gave me a few hints on making 'non-design of experiments' work even better than my initial playing-around could do.