ODB1 ECU Project - Page 5 - Tuning Forum

Forum Post / Reply
You must log in before you can post or reply to messages.
Re: ODB1 ECU Project
Saturday, August 26, 2006 4:50 PM
Quote:

the file for tunercat works AWESOME but has no funcionality with emulation or real time tuning/monitoring.


The definition file doesn't have anything to do with emulation. It's just a list, a translation if you will, which allows you the human to easily and quickly change the calibration variables in a meaningful way. Tunercat sells TunercatRT, the RT stands for Real Time, which interfaces with Moates' Ostrich and Intronics Romulator. You can download Tunercat RT from his web page. If the problem has to do with your emulation hardware and the size of the $36 mask, TunerproRT will not fix that.

Quote:

Im stuck though, in tunercat they list switches and flags as hex offset, in tuner pro its listed by the bit number.

The easiest way to identify a bit is to id the byte location, then the bit. TC lists flags as, for example, Location $01DF, mask $10. The easiest way by far to work with this is by using the Windows Calculator in Scientific mode. To get scientific mode, select view, then scientific. Once you've done that , select hex, type in the number (10 in this case) then select bin(ary). Motorola Land is different from Intel World in a few small but important ways. In Motorola Land, bits are counted from the RH side starting with bit 0. So $10 converts to b10000 and the bit in question is at location 4.

HTH

-->Slow

Re: ODB1 ECU Project
Saturday, August 26, 2006 5:07 PM
if you wanna see the difference, look up big-endian and little-endian. just had a class on a new motorola/freescale proc that uses munged little endian. the data appears to be little endian from the CPU side, but the byte order is big endian.

slow, wouldn't the location be in 3? if you're using the 0-7 naming structure?


Desert Tuners

“When you come across a big kettle of crazy, it’s best not to stir it.”


Re: ODB1 ECU Project
Saturday, August 26, 2006 5:25 PM
Thats just it slow, the Tunercat RT product says on the $36 file that it isnt compatible with emulation.

The reason for tuner pro is exactly that, and Mark is also a good buddy of mine. My plan is to sell and promote both software packages, but ultimately anyone wanting to use them will have to choose.

In the TDF editor of tuner cat, it lists a Hex Mask, for when you change a flag i think the software sets or unsets the mask, in tuner pro it doesnt rather it says which bit mask to use the 0 - 7. So what i need is a map as you will of what the hex mask does in tuner pro for the 0 - 7 scheme.
Re: ODB1 ECU Project
Sunday, August 27, 2006 2:13 AM
Quote:

slow, wouldn't the location be in 3? if you're using the 0-7 naming structure?


Not for the Moto semi.

  Binary  0 0 0 1 0 0 0 0
Location 7 6 5 4 3 2 1 0



Quote:

the Tunercat RT product says on the $36 file that it isnt compatible with emulation.

Ahh. And Mark's fairly certain that his software will work? Interesting.

Quote:

In the TDF editor of tuner cat, it lists a Hex Mask, for when you change a flag i think the software sets or unsets the mask, in tuner pro it doesnt rather it says which bit mask to use the 0 - 7. So what i need is a map as you will of what the hex mask does in tuner pro for the 0 - 7 scheme.


Lol. I gave you the "map." Use the Windows calculator. You're just overthinking the problem. The "mask" that is used in TC is just the hex number represented when a bit (or bits) of a byte are set to "1". "Mask" is just a programmer's shortcut of saying "So what bit was that again?" And why do we do this? To avoid exactly the issue that's cropped up here between FReQ Z and me.

FReQ Z says "what bit?"
I say "Bit 3"
He says "From right or left"
I say "from right."
He says "start from 1?"
I say "start from 0."

You see what a PITA this is ???

Much easier is :
"What mask?"
"$10."

Now when you convert to binary you can see exactly where the bit is.

I'm surprised that Mark doesn't have a program to read the TC def and write a Tunerpro XDF from it.

-->Slow
Re: ODB1 ECU Project
Sunday, September 03, 2006 10:57 AM
So i havnt quite figured out how to do the bitmask to bit number conversion yet.

But something in need right now is, what general parameters do i need to change in order to add a supercharger to a 2.2 2.3, am i just upping the VE or adjusting anything having to do with map.

So far the only supercharge maps i have are modifications from the 8D v8 mask, and im not sure i can even rely on them.
Re: ODB1 ECU Project
Sunday, September 03, 2006 2:09 PM
Quote:

So i havnt quite figured out how to do the bitmask to bit number conversion yet.


What part is giving you a problem?

Quote:

But something in need right now is, what general parameters do i need to change in order to add a supercharger to a 2.2 2.3, am i just upping the VE or adjusting anything having to do with map.

Well, this is the point of tuning! Finding out what variables to adjust to get the best performance is part of the learning process. Here are the general rules to follow:

1) As boost increases, spark advance decreases.
2) As air temperature increases, spark advance decreases.
3) As boost increases, VE increases, fuel needs increase.
4) As air temperature increases, fuel need decreases.

Do not use the $8D tables. I have posted spark tables from my turbocharged 94-96 spec engine to JBO in the past. You can compare those to the stock 95 tables to get an idea how spark needs can change with boost. Remember that each engine family is a little different and that the 2.2 MAP may be completely inappropriate for the 2.3 engine. I believe I've also posted VE tables from my car up to 1 BAR. If you're using a 1 BAR MAP sensor, it's almost pointless to use anything but stock tables anyway, since the ecm will not be able to affect spark or timing under boost.

-->Slow
Re: ODB1 ECU Project
Wednesday, October 25, 2006 8:04 AM
this may b off topic but does the obd1 ecu in our cars use the map for fuel or the tps (wonderign cause i can get a deal on a apexi super afc 2 and it uses the map for fuel corections)



Re: ODB1 ECU Project
Thursday, October 26, 2006 3:35 PM
Sorry i been gone so long, thanks for the advice on the boost fuel/map tables.

As for sk843143's question, sure you can do an apexi SAFC 2, i have 2 buddies using SAFC2 on cavaliers, and they work okay for fine tuning, but your best deal on the odb1 is to grab a copy of tuner cat and get the tuning file, or email me for pricing on an autoprom and a tuner pro rt setup, both i've gotten to work successfully, but the tunerpro has the added advantage to run realtime emulation and tuning, as well as reading the sensor data and with some advanced programming, the ability to program maps based on aldl data.
Re: ODB1 ECU Project
Saturday, December 09, 2006 1:35 PM
The project is not dead, i realize it may seem so but is very much alive.

Im working on a set of tdf/ads files for tunerpro that will be ready in the next few weeks. For thoes of you keeping an eye on this project it has been a great ride.
I will post links to files when completed, perhaps now we can start tuning the odb1 just as easily as the odb2.
Re: ODB1 ECU Project
Friday, January 05, 2007 7:07 PM
so if i go with the $36 definition file and tunerRT from tunercat with the APU1 from moates, i wont be able to tune in real time with the emulation?
also i want to datalog, is there anything available (ALDL definition file)?
thanx



Re: ODB1 ECU Project
Saturday, January 06, 2007 1:28 AM
Magik Just hang on for a few more weeks, I'm in beta 3 of the TDF/ADL setup for tunerpro rt, tunercat currently doesnt offer simultanious tuning/datalogging features that I'm aware of.

You can datalog with apu1, but there's a trick, you have to force the pcm to lock up the chat cycle using the ign key, i havn't yet figured out which part of the code removes the chat cycles, but if one of the other programming guru's wants to that would actually be a large contribution to the project.

What our pcm's do is they talk to a BCM and it throws off tunerpro's aldl. i've tried getting winaldl and some of the other datalogging software that can graph blm etc.... to read the pcm, but i can only get it to work when i trick the pcm into dealership diag mode.

I think this has to do with 8192 data mode that can be somehow tricked into disabling the chat from the Pcm, but our Pcm's are chattier than the lt1 pcm.

In beta 3, all the tabular data is correct, but my scales are off. I'm adjusting them so they are perfect. After all i wouldnt want to release a beta file and hear that a Jbodier blew up his car.

For thoes of you working with me on the project, drop me a line and i'll send you beta 3, especially if you already have the tunercat with $36 file, as you can probably fix the tables quicker. Another thing, im working with the tunerpro software to autocorrect the blm tables. Once i get it configured correctly we should be able to essentially tune on the fly with a laptop doing all the work.

Chris

Re: ODB1 ECU Project
Thursday, January 11, 2007 8:36 PM
man, i cant wait.....
let me know when you have some update, thanx!




Re: ODB1 ECU Project
Saturday, March 31, 2007 1:53 PM
http://www.j-body.org/forums/read.php?f=40&i=133983&t=133983
Hey guys check this thead, it may be the solution to our injection tuning problems.

Right now we have the ability to tune the 16196285 with the files I have. Now they are only beta versions, so i would also buy the tuner cat version and verify scales etc..... Use at your own risk. Reason i post this, and bring this up. Simple, the thread posted above has information on how to top feed convert your ln2 without spacer plate.

This means cheaper topfeed injectors....

Post to come with beta files as soon as i get them uploaded to the server.
Re: ODB1 ECU Project
Sunday, April 01, 2007 6:47 PM
As promised, here is beta 3 for testing...... By downloading the file, you assume full responsibility for what is done with the file, you are free to edit or distribute the file under the GNU license agreement. It may not be sold, only given or transfered.

With that said have fun.....

http://files.cyberstopweb.com/36xdf.rar
Re: ODB1 ECU Project
Monday, May 21, 2007 9:37 AM
Anyone have any luck yet with the tuning files?

I have a ? for anyone who codes out there. I need to eliminate the o2 threshold subroutine because its not needed on test pipes, and i cant seem to locate it? Anyone wish to help.
Re: ODB1 ECU Project
Monday, May 21, 2007 7:56 PM
I have looked at them. Haven't used it, but have looked at it.

Constants seem to work well.
I don't use the flags so I haven't looked.

Tables seem to be accurate for the key items most people would change.


I've been using TC for chips and changes and I am pleased with it. I hope to use TP for datalogging at some time in the future, but I haven't figured out how yet and I don't have a Comm box to even try anyway.

So I guess I'm saying I don't have a valid opinion, but I have looked at it.

sig not found
Re: ODB1 ECU Project
Tuesday, May 22, 2007 4:05 AM
Quote:

I need to eliminate the o2 threshold subroutine because its not needed on test pipes, and i cant seem to locate it? Anyone wish to help.


What do you mean by "O2 threshold subroutine"? There is more than one threshold related to the O2 sensor. Are you trying to eliminate closed loop operation?

-->Slow
Re: ODB1 ECU Project
Tuesday, May 22, 2007 11:13 AM
Sorry slow, i musta been sleeping when i wrote that. You know how the pcm on our cars includes code to protect a catalytic converter? Where the afr is managed in a wave when at stoich. I've noticed that on the ms we got, we removed the ripple from the afr and the car ran better, this one is used for racing so it doesnt have a cat on it.

What i need to do is remove the ripple effect so thte pcm will just adjust to the set prescribed afr that is commanded and hold it unstead of alternating rich/lean and adding up the o2 crosscounts.

Chris
Re: ODB1 ECU Project
Tuesday, May 22, 2007 4:32 PM
The ripple you're talking about isn't really for cat protection, it's done because the O2 sensor is only accurate at stoich. Even without converter overtemp protection, the ecm toggles from rich to lean and back. It's easier to keep the AFR "averaged" at 14.7 than to try and keep it exactly 14.7 using a narrow band O2. If you disable closed loop the ecm won't toggle back and forth, plain and simple. Many race cars switch to open loop to make tuning easier. If you're dead set on using closed loop and tightening up the closed loop AFR values then you're looking for the O2 toggle voltage, the O2 rich / lean threshold, or the O2 switching voltage or AFR. All of these refer to the same variable. I don't have any files here at work to look through but if you still can't find anything I'll look at my notes later. FWIW it's easier just to switch to open loop.

-->Slow
Re: ODB1 ECU Project
Tuesday, May 22, 2007 9:16 PM
I'm 90% through the code, i wonder if i should just eliminate the nb o2 and put in a wb02 with code modifications.
Re: ODB1 ECU Project
Wednesday, May 23, 2007 4:20 AM
WB isn't a bad choice. You'll want to change the way the target AFR is reached. Since the WB can be used for all desired AFR's you'll effectively be in closed loop most of the time with a varying desired AFR rather than the ~14.7:1 fixed value it uses now. Short term, switch to open loop. Then when you've got new code ready, install the WB.

-->Slow

Re: ODB1 ECU Project
Friday, August 31, 2007 7:28 AM
Bump.......

For thoes of you who are new to the org we have working definition files for tunerpro rt on the 16196285 pcm. The mask is $36, but if anyone would like to modify for the 2.3l, the $36 definition is exactly 1000 bytes shorter than the $f1.
Re: ODB1 ECU Project
Friday, March 14, 2008 11:58 AM
BUMP: ODB1 project goes well. Plus we have a software addition.

The author of tunerpro released a beta on scanner pro, which lets us datalog and dyno program.

Anyway, i'm looking for somewhere to host the project files, and i'll post a link shortly.

For thoes of you who havn't kept up, the 16196285 pcm is the 95 pcm used in the jbodies. It uses the $36 and $F1 mask GM programs. I'd be happy to offer anyone help who wants to use this pcm, it's not only a very responsive and nice pcm to tune with, but changes can be made on the fly with moates.net hardware.

I'm pretty close to being able to develop a custom os, and this pcm can be picked up at a junk yard for about 50 bucks.
Re: ODB1 ECU Project
Friday, March 14, 2008 12:28 PM
thats great news man....
now send everything you have for that pcm to me
thanx





Re: ODB1 ECU Project
Wednesday, March 19, 2008 7:34 AM
C Smyth wrote:BUMP: ODB1 project goes well. Plus we have a software addition.

The author of tunerpro released a beta on scanner pro, which lets us datalog and dyno program.

Anyway, i'm looking for somewhere to host the project files, and i'll post a link shortly.

For thoes of you who havn't kept up, the 16196285 pcm is the 95 pcm used in the jbodies. It uses the $36 and $F1 mask GM programs. I'd be happy to offer anyone help who wants to use this pcm, it's not only a very responsive and nice pcm to tune with, but changes can be made on the fly with moates.net hardware.

I'm pretty close to being able to develop a custom os, and this pcm can be picked up at a junk yard for about 50 bucks.


I have not pursued this much lately. When I last did anything with it, I was trying to work out datalogging, to which I have been totally unsuccessful to date.

If you need hosting, I can host them on my Miller-performance site or JBOM's. That would be no issue.

I am not too concerned about a custom OS myself, but I would willing to give it a try.

I do have concerns about the $36 Tunercat files and their total compatibility with the 2.3 file. Some items do not work as they should. Also, the 2.3 and 2.2 files are different sizes.

sig not found
Forum Post / Reply
You must log in before you can post or reply to messages.

 

Start New Topic Advanced Search