ODB1 ECU Project - Page 7 - Tuning Forum

Forum Post / Reply
You must log in before you can post or reply to messages.
Re: ODB1 ECU Project
Friday, August 22, 2008 2:48 PM
My understanding from reading up on GM is that single fire mode is where all injectors fire at once. This has its limitations of firing the injector both on an intake charge and and exhaust charge. The reason i found out they do this is at higher rpm or where more fuel is needed, the single fire can keep up.

However with what we're doing with larger injectors, and with the newer advent of Sequential fuel injection, I believe I can eliminate the single fire mode. We can't do a SFI mode on this pcm without finding 2 more injector driver ports, which I don't think this unit has. However i think we can probably setup the double fire tables to more closely act as a SUDO SFI, simply firing 2 injectors instead of 1 each.

This new discovery now makes me wonder if we shouldn't lean more towards building a custom os and modifying the pcm for sequential fire. I do realize it will mean we have to somehow integrate a cam sensor, but it could be worth the development work leading towards a more efficient and higher power engine.

Re: ODB1 ECU Project
Friday, August 22, 2008 2:51 PM
Ok so what I want to work on in the next few weeks is setting up a way to Easy VE tune like HPT has with this pcm.

Can anyone working with the current software pitch in any ideas.

The problem i've found so far is im not sure how to force this pcm to go open loop...
Re: ODB1 ECU Project
Tuesday, August 26, 2008 5:45 AM
C Smyth wrote:

Ok so what I want to work on in the next few weeks is setting up a way to Easy VE tune like HPT has with this pcm.

Can anyone working with the current software pitch in any ideas.

The problem i've found so far is im not sure how to force this pcm to go open loop...


What I did last year when I was borrowing a WBO2 was turn up the cold and warm closed loop timers in the constants menu. The cold/warm temp threshold is 185 degrees F and I set the cold closed loop timer to around 8900 seconds, so as long as the car was started under 185 degrees, it would take over 2 hours to enter closed loop. It worked just fine, the WBO2 verified. From that point, it would comand whatever was in the cold AFR vs. coolant temp. Im presed for time now, but ill explain what I did with the VE tables after work.





13.934 @ 97.82 ALL MOTOR
Re: ODB1 ECU Project
Thursday, August 28, 2008 4:53 PM
So, Im trying to come up with ways to burn programs fast, and not always have to pull my ECM,

Think geting This Bling!

With This More BLING

And maybe This Done with Bling

And maybe this to make it faster.. well?

So, it should work, right?

Chris




'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Thursday, August 28, 2008 6:38 PM
Thats the realy crappy way to do it. Do it this way...

All the BLING you need

You already have the lap top, download TunerPro RT and use the def. file you already have. Then you can make all the changes you want in real time as the car is running or driving down the road. No more chips untill your happy with the tune and ready to take the Ostrich out.





13.934 @ 97.82 ALL MOTOR
Re: ODB1 ECU Project
Thursday, August 28, 2008 6:47 PM
So... I guess i need to buy a lab top (Im using a desk top, but could use Brian's i guess).... get my burn done, and send it your way Brandon?


$79 VS $175, but either way i would need a socket booster.. and also have to register Tuner Pro...

hmm


Chris





'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Thursday, August 28, 2008 10:58 PM
Actually the odb1 test unit is this BEST BLING!

We did some work for moates.net a few years ago on his site, and so he hooked us up, and that autoprom is the best ever.

Datalog, real time emulation, man theres nothing like it, plus its nvram version is the fastest out there.
It also can be connected with bluetooth, making wires obsolete. I built a custom casing for my pcm and put the board for the autoprom and pcm in it so that it is completely waterproof.

I'm working on an odb1 project website, forums etc.. I'd still like to keep our info here on jbo, but build up some kind of how-to manual for the latest $36 and $f1 as well as some of the custom firmware i am working on.
Re: ODB1 ECU Project
Friday, August 29, 2008 7:19 AM
Think im going to get the Ostrich, as i already have Burn-1.....

this should work, right?

Chris




'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Friday, August 29, 2008 4:22 PM
Problem here?

How you going to read the PCM, do you already have a datacable and interface?

Do you want to simultaneously read pcm data and emulate?

If so you will want the autoprom, its the top of the line component.

I also have an ostrech and well, its a great product for straight emulation, but was designed for the guys who didn't need simultaneous datalogging/emulation.

Chris
Re: ODB1 ECU Project
Monday, September 01, 2008 2:01 PM
Straight emulation, once the tune is on, i would just burn the prom....

Right?


Chris


'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Monday, September 01, 2008 2:35 PM
How are you going to read for knock, timing, sensor data?
A wideband will only get you so far, you will still want to see what the narrow band sees, the blm's and any adjustments you will need to make will be based off of the VE tuning for hpt plus setting our pcm into open loop.

Re: ODB1 ECU Project
Monday, September 01, 2008 3:20 PM
wont i be able to see it through tuner pro......

my wide band is the narrow band.....has a narrow band out put.....

i can always hook up a speaker to the knock sencer to see if its hearing Knock.

For the most part i haven't played with timing, from what Brandon's tune had, just richened it up after the exhaust.

Chris




'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Friday, September 05, 2008 9:20 PM
C Smyth wrote:

How are you going to read for knock, timing, sensor data?
A wideband will only get you so far, you will still want to see what the narrow band sees, the blm's and any adjustments you will need to make will be based off of the VE tuning for hpt plus setting our pcm into open loop.


Either by finding a scan tool that works with the ALDL or playing around with the moats ALDL reader for 95 OBD1s.

I didn't have either in my case, so I couldn't look at BLMs. I did put the computer into open loop and drove around while I logged Map, RPM, WB02 and TPS for 45 mins on an LM-1. I used the output on the LM-1 to give the PCM a narrow band signal. Put that huge datalog into logworks, then setup a table of WB02 Vs. KPA Vs. RPM that uses the same dimentions of the low RPM VE table in cats. In excel, took the average AFR from the wide band table, and the commanded value in cats, divided the 2 to get a percent difference. Coppied that table, then multiplied the VE table in Cats by that percent table. After doing that a couple times, the general trend of the VE table was set and didn't change by much more than 1 or 2%. I still haven't gotten a scan tool to verify the BLMs, but I'm sure its close.



13.934 @ 97.82 ALL MOTOR
Re: ODB1 ECU Project
Saturday, September 06, 2008 12:27 AM
I have a scan tool that would work.. IF the code mask.. and maybe cheek sum(?) would jive with the OE ones..

as of now. the tool that talked to my OE prom, will not talk to my ECM...



Chris




'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Monday, March 09, 2009 11:01 PM
Ok, gonna drag this back up top for 2 reasons. One is to see if Mr. Smyth has made any progress either on the tuner pro definition file or any cool hacks on the 95 PCM and to contribute an ALDL definition file I made for Tunerpro RT that works on the 95 cavys that are OBD1.

95 jbody .ADS file

If you have a laptop, the moates extream ALDL cable with the 95 OBD1 plug or any generic 95 OBD1 logging cable that works with tunerpro RT this should work. What you have to do is connect to the ECM and then send the "Disable IPC" repeatedly till the Packet Error Counts stop going up. This will keep the dash data from flooding the ALDL with chatter and allow tuner pro to log the engine data properly. The only by-product is your dash wont work properly save for the speedo. Not a huge deal as TP will be logging the RPM and important stuff anyways. Protomec and I have tested it on his 95 Z24 already, so we know it can work. When my cars back together ill try it there also. If any other 95 OBD1 cars are willing to invest in the equipment and test it out, that would be awsome! Its fairly inexpensive, so look into it. Questions are welcome also.





13.934 @ 97.82 ALL MOTOR
Re: ODB1 ECU Project
Monday, March 09, 2009 11:41 PM
I have made tons of progress but am only about 75% through decrypting the bin. The hardest part is getting through all the manufacturer garbage to find useful subroutines. A majority of the code is emissions garbage put there to manage emissions or to make the code more "average user" friendly.

I'd like to post my files under the GPL license simply because while this work was extensive, it wasn't 100% me. I have to give credit to the tuner cat crew for figuring out the F1 mask which is a 36 mask exactly with a 1000 byte shift in code.

My work has come to a standstill as far as tuning goes due to my lack of a boosted ride. I had a daughter about a year ago and that put my boost plans on hold. HOWEVER, the reason I wish to post my research public is so that anyone with a boosted ride who doesn't mind dropping 400 into the right equipment can refine it farther.

I will sift through the ADS posted, as it may have very valuable information. My current part of the project is accurately predicting AFR from the narrow band o2 (as accurate as narrow band o2 can be) and hopefully crack the section of the code that reads the o2 sensor so we can modify or add a wide band to the ecu.

I've been looking at a few pieces of software for the DSM ecu's that allow simulations of how software changes will affect engine performance, and am looking at trying to build a piece of software that will let me simulate how the code of our pcm's works with hardware. My hope is to map out the hardware interface points so that a custom os can be built. The 95pcm has the capability of being used as a boosted pcm, as there are plenty of 5v inputs we can add sensors to, its just a matter of figuring out exactly where those inputs are located in address memory on the cpu.

I'm starting to get to the end of my knowledge base, as this is mostly a hobby for me. But if someone with Motorola cpu experience would like to chime in, I would be happy to provide what I have done so far. I can say that we are close to at least modifying extensively the GM mask software to do what we want it to do. I can already get rid of egr operation, disable emissions controls, shut off speed limiting devices, program in a studder box for boosted application, modify timing, ve, AFR tables, launch control. I did run into some minor issues building the studderbox/launch control, as you have to tap into the wiring for the auto trans on a manual so the pcm can sense the clutch is pressed in.

Sorry I havn't gotten much farther, kids are a lot of work.
Re: ODB1 ECU Project
Wednesday, June 03, 2009 4:23 PM
update??



Re: ODB1 ECU Project
Wednesday, June 03, 2009 5:18 PM
Mine is boosted now..... and the tune I got from Protomec.... did not work, it idled and ran EXTREAMLEY LEAN Unless I put the stock Map back in........
Even after riching it up........

Myself and Speedracer think the stock 95 Z24 is a 1.5 bar ECM... as it runs great with just PE turned up past 100 KPA (the tables in TunerPro go up to 130 IIRC)......

However I myself am at a stand still, I ran out of fuel.... I need a better pump.

SOOO....

After I mess around with it more, I will post up my findings.

Any newer XDF's CSmith?


Chris




'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Wednesday, June 03, 2009 5:47 PM
Progress on the project has slowed due to a new addition to the family and ensuing legal costs to be part of her life (16 month old daughter). However, I can say for sure that my last beta3 xdf and ads along with tunerpro and scannerpro work pretty good.

I'm not so sure on the map scaling just yet, you'll have to play around with it.

I'd be interested in seeing some VE tables and a list of current fuel mods. I have been looking into some coding projects to try to bring this project closer to the megasquirt project as far as tools go. For instance, VE auto calculator, timing calculator, etc.... but no luck so far.

I know i'm not that far off with the software but my Visual C++ coding is not that advanced.
There are some basic libraries out there to read the data coming across the ALDL that we can use to import data into a program, but i'm years away from actually making a useful tool.

Does anyone have a file repository they prefer, I am going to upload my recent xdf, adl, ads files so that the community can have a go at it.
Re: ODB1 ECU Project
Wednesday, June 03, 2009 5:57 PM
The trick to fooling the PCM in our cars for boost is swapping the map sensor to a 2 bar and then reconfiguring the XDF. Then just recalculate the values for the table in the .xdf so that instead of actually being 0-100KPA which is what my .xdf is supposed to be, it would actually be 0 - 200.
1 bar being atmospheric and 2 bar being 14psi.

The stock z24 ECM is a 1bar ecm, but the .xdf I have used extrapolated values from a different gm mask, hence the beta version and not release.
And for the VE tables the values are based upon efficiency percent, 100 being 100% efficient 0 being no efficiency. In a boosted application with our cars, you need to adjust your ve tables while under boost for higher than 100% efficiency because they are pumping in more air than sucking in.

Taetsch Z-24 wrote:


Myself and Speedracer think the stock 95 Z24 is a 1.5 bar ECM... as it runs great with just PE turned up past 100 KPA (the tables in TunerPro go up to 130 IIRC)......


Hope that helps, I'm still putting more work into this.

Special thanks goes out to all thoes who have helped catch me up to speed and get us this far.
Re: ODB1 ECU Project
Wednesday, June 03, 2009 8:07 PM
I know how a fake 2 bar workes, and the chip was set up for 2 bar, and had a 2 bar sensor......

BUT ran like CRAP. like 19.5 AFR......

had the same problem with the drag car and its 2 bar tune......


the only thing I can think of is IF it may be a 1.5.....


But IDN.... I really dont know.


Chris






'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08


Re: ODB1 ECU Project
Thursday, June 04, 2009 2:17 PM
Many apologies Chris i'm positive you and protomec know how to fake a 2 bar.

I'm currious as to why you ran so lean, in looking at the 95pcm code it should be faking the 2bar exactly as planned?

My only thought is that perhaps there is some undecyphered code in there from the odb 1.5 junk thats messing up the equation, but really thats a shot in the dark.

Have you tried playing around with my beta3 xdf and Base Pulse Width and Injector size constants? I've had some pretty good luck with them in riching and leaning out my cavi..

Now are you reading your AFR with a wideband (i assume so) or the ECU AFR calculated output in my adl file.

The calculated AFR is an experimental calculation, as its based on the rather inaccurate narrowband, but it should be accurate at 14.7.

I want to incorporate wideband directly into the code, and have been doing some research with previous gm code to see if I can locate the equation to do it, once I do I'll definately release a updated xdf.

Other than that, the only thing i can think of that would be leaning you out is fuel cut from high knock.

Lemme know how it turns out,

Chris
Re: ODB1 ECU Project
Thursday, June 04, 2009 5:26 PM
Yes, You gave me the $36Beta3XDF..... Works GREAT....
Ran 14.6's NA,....... (really bad wheel spin in 2nd and 3rd)

my Timing was set to 12-14*... ( all he\we really did was copy the '00 Supercharged tune)

I know my fuel pump is not up to the job, (stock 95),..... But with the 2 bar tune it ran lean all the time... with the one bar sensor, "fake" 2 bar timing, (stock timing up to 100KPA and then 12*) and messing with PE above 100 kpa (the tables go up to 130.... is that correct? ) ran GREAT Is faster then my GTO...... But would climb in to the 14 AFR's...... with the return line closed with a vice grip....

Once I get the new pump installed, and I tune until i get GOOD AFR's, Ill post\email you my .bin.

(all AFR#'s are from a Wide Band.)

Chris





'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Re: ODB1 ECU Project
Tuesday, June 09, 2009 5:54 PM
Interesting that PE would fix your AFR issues. The car only goes into PE when you slam the gas to the floor. I guess you could tune for WOT using PE but what about the VE tables for all other times?

I guess that kind of makes sense when your not in boost your stock VE tables would work really well, but under boost thoes VE tables are way off, and the PCM will have to correct but dumping massive amounts of fuel.

You can use a similar tuning principle to tune the VE tables as they use on the odb2 stuff on the org here.

You just have to find a way to trick the pcm to not go into closed loop, and then use the BLM values to tune.
I have some beta ads files for scannerpro that make graphing and logging a snap, i'll dig em up or you can make your own from the adl files.

Chris
Re: ODB1 ECU Project
Tuesday, June 09, 2009 7:17 PM
there is a PE enable TPS that I messed with....... around 30-40% if I remember...


I think there is one for KPA also.

But, like I said, I need to do the fuel pump this weekend for any real answers.


Chris




'02 Z-24 Supercharged
13.7 @102.45 MPH Third Place, 2007 GMSC Bash SOLD AS OF 01MAR08

Forum Post / Reply
You must log in before you can post or reply to messages.

 

Start New Topic Advanced Search