[motorhead1991]

I suppose I should make an "official" announcement for this .

OpenEEC Project

My thinking was that this would allow more people to get in on creating tunes and enabling a deeper understanding of how Ford's engine management works. This would also enable shared ideas, bug fixes (if present) and a collaborative effort to the betterment of said skills.

So far, tvrfan is the only other person involved. If you would like to be a part of this project, say so here, shoot me a PM or ask to join on Github.

Telegram Chat: https://t.me/joinchat/EhTLgVHIxioWidQK0Shs5Q

[tvrfan]

Motorhead,

If you create sub-areas for LISTINGS, BINARIES, STRATEGIES, I will upload my collection of bins, and ask that everyone else uploads theirs too - and I will add my listings, when I'm reasonably sure they are correct .....It does mean someone will have to check for duplicates though (or does github do that ?)

Then we can build up a decent library.

NB. I can't see that Ford will care anymore for an obsolete product !

[motorhead1991]

You should be able to do it. Everyone that's a part of the organization has writing privileges.

Pushing from local to the web is a little more difficult though.

[tvrfan]

Done, for listing and binaries. I'll add an AA listing shortly....

[motorhead1991]

Done, for listing and binaries. I'll add an AA listing shortly....

When you go to push the listings, please do so in a subfolder. It'll help keep things cleaner.

Or, we can create a folder for each binary. That way, when someone downloads it, they don't get everything (just the ones they need)

[motorhead1991]

Big Update:
For those not following along with tvrfan's disassembly thread, the open source SAD now compiles and works from within the Android ecosystem!

Screenshot_ConnectBot_20180423-211219.png

I'm working on a UI currently, so the binary might have to become a library in order to work, unless I can figure the interface out. I think tvr will be of some help on that front as well.

[skunk]

Subscribing.......wish I had something to contribute! Don't let the lack of response deter you guys, most of us have no clue how to help.....keep up the good work!

John

[jeeptech77]

tvrfan i am having trouble using sad
motorhead has been helping me and we got one bin to dissemble bit i cannot duplicate the process with other bins or the same bin
would you have time to help out ?

[tvrfan]

The easiest thing is probably to post/attach the binary, and I'll run SAD against it here.
This will (I hope) give you a fast result and then you can learn more at leasure, and have a starting point.

What error do you get trying to run SAD on your computer ??

[jeeptech77]

this is a stock ford bin from a la3 ecu

[jeeptech77]

im not having errors either sad is not operating correctly or i am not useing it correctly
i have other bins if you want to test them

[skunk]

The LA3 is a 32k bin.....your attachment says its 56k.....oh never mind I see Clit revised to 56k. If you want here is my slightly modified version of the definition(32k). If you open it in excel you can see the revision changes. I would love to see the unknown variables defined. In this definition I added them to unlock the boost limit of the 8UA:

LA3r11.xlsx

If you wish to use it and maintain a 56k bin open in excel, go to config tab and change the checksum end address to 9FFF then save it.

Here is a very old (16k EECTuner) disassembly package for the LA3 to give you something to compare:

Dis8061_LA3.zip

Here I believe are the results:

LA3asm.zip

I may have to give this a try myself!

Late Edit.......Just found another LA3 dir file

LA3_MAF.zip

Please post your findings
John

[jeeptech77]

John i see most of that is off johns tweecer group
do you know how to run the bill lawrence 8061 dissembler ?

[skunk]

Sorry bud....... Its been at least 10 years since I played with it. My memory isn't what it used to be.

John

[tvrfan]

Had a look at LA3, and I see I've managed to create a bug in 3.06 .... I can't immediately fix this either....

Workaround -

OK - some bins do a PUSH(address) and a RETURN instead of a jump. Example is la3

Code: Select all

2020: a1,20,02,10         ldw   R10,220          STACK = 220;
2024: c9,00,20            push  2000             push(2000);
2027: c9,1c,21            push  211c             push(211c);
. . .

2119: 11,05               clrb  R5               WDG_Timer = 0;
211b: f0                  ret                    return;

SAD will stop here and just print ?? after this.
what it SHOULD do is look at 211c, but it does not (and a temp fix I tried in messes it up even more !!)

so for a temp workaround , if you see this kind of pattern in the first bit of code,
create a text file '<name>_dir.txt' (for 'la3.bin' it would be 'la3_dir.txt')

and add a line

SCAN 211c

or whatever the address is ...and save the file, and rerun SAD.

and I'll try to fix this ASAP .

It just so happens I haven't got a bin which does this trick in my test bunch - I'll add la3 in there...

[jeeptech77]

thanks for the update tvrfan
i am not at the it level so im going to have trouble with that work around
the first issue i have is i really need instructions on the setup for files sad wants or emitts
the first go around Anthony was helping me with i did not set anything up he created a LA3_LST FILE i downloaded it it was in the downloads file i loaded the binary hit dissemble and the dissembled binary was on my desktop in a notepad so that is fantastic
but i tried it again i have other binarys i need to be able to open and even the same binary the downloads folder would open when i hit dissemble i could click on the lst file or even in a open spot on the downloads and it would close
there has never been a error emitted from sad with the exception of it could not find the icon on launch
also i am using the sadwin version as i have not been able to get the file location to work with just sad
Anthony suggested closing and reopening it to work around that glitch but it did not work
i even deleted sad folder and reinstalled it and same thing, so thats my findings
i found the setup and it asked for locations for several file types that im green on ( i do not know there purpose ) so i just chose the make a new folder for each
and i still got the same result
i am going to add a modified la3 bin to this so u can try as well
thanks for the update and thanks for what you are doing

[tvrfan]

Have a look at the documents in EEC-disassembler on github (as per my sig below).
It is crossref'ed in Open EEC, but probably Open EEC has out of date version (?).

I tried to make the SAD docs start off as simply as possible, and gradually build up to more complex commands.
So see if the basics help you, and make sense.

SAD should be resistant to crashes, so if you do get a crash, I'm always interested to fix it.

On Windows, I tried to make SAD set itself up with SADwin as a GUI program.

We are working on a Primer doc here, in different thread, but at very early stages.
Welcome to look at that as well, needs flow fixing up a bit, but comments welcome on all docs.

That 56K LA3 file looks to be trash at the end, I think I looked at that (a long time ago) and decided it's really a 32K binary, like A9L is...

Anyway, it's a good addition to my testing suite - I'd forgotten about it, and sod's law says it shows a quite a nasty bug !!

[motorhead1991]

Have a look at the documents in EEC-disassembler on github (as per my sig below).
It is crossref'ed in Open EEC, but probably Open EEC has out of date version (?).

I tried to make the SAD docs start off as simply as possible, and gradually build up to more complex commands.
So see if the basics help you, and make sense.

SAD should be resistant to crashes, so if you do get a crash, I'm always interested to fix it.

On Windows, I tried to make SAD set itself up with SADwin as a GUI program.

We are working on a Primer doc here, in different thread, but at very early stages.
Welcome to look at that as well, needs flow fixing up a bit, but comments welcome on all docs.

That 56K LA3 file looks to be trash at the end, I think I looked at that (a long time ago) and decided it's really a 32K binary, like A9L is...

Anyway, it's a good addition to my testing suite - I'd forgotten about it, and sod's law says it shows a quite a nasty bug !!

Yeah, OpenEEC is a little behind. If you merge the pull request I sent on GH, the Android stuff will be in place and it'll make merging easy later.

[jsa]

thanks for the update tvrfan
i am not at the it level so im going to have trouble with that work around

I am going to make it as simple as it gets, for you.
Make sure your bin is named or renamed to LA3.bin
Download both these files and save them to the same folder as LA3.bin

LA3_cmt.txt

LA3_dir.txt

Copy SAD.exe to the same folder as LA3.bin
Open the folder containing LA3.bin in Windows Explorer
With your Left Mouse Button, drag LA3.bin and drop it on SAD.exe
Review the output files LA3_lst.txt and LA3_msg.txt

Review the two files attached here while looking at the SAD8065_temp.pdf help file that is available with the github download.

EDIT: I tested it on a 32kb LA3.bin file, which matches the first 32kb of the 56kb file you posted above.

[jeeptech77]

Thank you jsa seems to be working a ok now much appreciated .

[motorhead1991]

I added X1L (MAF 3.0 from '91) to the project. It's a part of the same hardware family as m0m2, so the layout is similar. I should have it broken down to the same point rather quickly.

[tvrfan]

When jsa stops finding bugs in my SAD code......(yeah, just joking...it's excellent work in reality) I will post more bins to Open EEC.

In the meanwhile, If anyone has a specific request, either from my collection or by posting a .bin,
I will attempt to decode it and post it to Open EEC. The usual disclaimers apply.
This will probably help find more bugs in the long run anyway

Attached is a list of what I have collected so far.
Note I can't guarantee each one is correct and unmodified - it's just what I have collected.

[jsa]

What are the differences between each of the CARD pair and each of the anti pair?

[motorhead1991]

When jsa stops finding bugs in my SAD code......(yeah, just joking...it's excellent work in reality) I will post more bins to Open EEC.

In the meanwhile, If anyone has a specific request, either from my collection or by posting a .bin,
I will attempt to decode it and post it to Open EEC. The usual disclaimers apply.
This will probably help find more bugs in the long run anyway

Attached is a list of what I have collected so far.
Note I can't guarantee each one is correct and unmodified - it's just what I have collected.

I gather what I can. So far, I have what I support, aside from FLH2 (Not pushing my luck with that one). Soon, I'll have a 4.9SD bin, but I need to free up my QH first.

[tvrfan]

What are the differences between each of the CARD pair and each of the anti pair?

OOPS ! - yes some duplicates in there.
CARDs are duplicates, but ANTI isn't - but don't know where (YET).

Hmmm....I need to put together a little linux script to compare binary files - the standard tools don't work....

Motorhead - should we have a subdirectory/branch in openEEC for bins ? I know there's still quite a few out there for download from various web sites,
so perhaps we don't need to.

[motorhead1991]

What are the differences between each of the CARD pair and each of the anti pair?

OOPS ! - yes some duplicates in there.
CARDs are duplicates, but ANTI isn't - but don't know where (YET).

Hmmm....I need to put together a little linux script to compare binary files - the standard tools don't work....

Motorhead - should we have a subdirectory/branch in openEEC for bins ? I know there's still quite a few out there for download from various web sites,
so perhaps we don't need to.

I was trying to avoid that, simply for convenience. With the current method, someone can download the binary that they need, vs a whole host of them.

[motorhead1991]

UPDATE: For those reading this thread and are concerned about the status of this project, we're still at it.

I still need to refactor the OpenEEC fork of SAD, but Pym has joined us with the source code for SAD806x.

I also have my own things in the works that'll be posted there soon.

[motorhead1991]

As of last week, SAD has been updated. I've also posted two of my recent projects in their own threads.

[danwie]

I've read about the disassembly of the EEC-IV/V on here extensively, and am happy to find this thread.
While is is great to get the tuning curves out of the modules, it seems cumbersome to use these EEC's with custom code or repair them
I was wondering if someone had already attempted replacing the EEC-IV/V on the older engines with a newer custom circuit board that uses a more accessible microprocessor such as an Atmel or PIC.
That way it would be easier to implement the logic and compile from C/C++ rather than fiddle with assembly of the tables in ROM.

Am I missing something obvious?

[cgrey8]

You aren't missing anything. That's just a huge project that nobody has taken on. Years ago, I had aspirations of doing something like that and the effort had a few false-starts. But I think the thing that really let the air out of the effort is seeing the latest from Holley. Their Terminator and Dominator offerings (which are for GM) are pretty much exactly what I wanted to develop. And while $1000 is steep, I can't imagine anything I would come up with having any less of a price tag given the time and effort it would take to both design, manufacture, and distribute the hardware as well as sell and support the software component.

Add to that, I'd only want to be involved in the design and engineering side. I have ZERO interest in the selling and customer service side of that endeavor mainly because I have little patience for people that expect you to drop everything in your life because they gave you money for a product and need some help. Don't get me wrong, I understand that as a consumer, you expect a certain level of support when paying a lot of money for something. But I just don't have the desire to give what people demand when they believe they are owed it.

If I'm going to help people, I'd rather it be on my terms, via forum support where I can choose to respond or not, depending on my interest, mood, and confidence level to talk on the subject.

But back to the original topic of is it possible. It's more possible than ever. There are microprocessors out there that have multi-core ARM (Cortex A-series) Linux/Java capable processors along with 1 or more dedicated real-time (ARM Cortex M-series) processors for doing realtime work. And being integrated in the same chip allows them to inter-communicate, share memory, and peripherals. The opportunities for doing a project like this just keep getting better and better from a hardware standpoint. But to harness all that takes knowledgeable engineers, hardware designers, and software developers. While it isn't impossible for a single person to do, finding the person that has the knowledge, skill, and willingness to take on something like that is tough. So if it is going to be done, it's more likely to be a collaborative effort. And at some point, money will get involved. Test equipment, development software/licenses, and prototyping services all require capital. Now that we are in the era of crowd-sourcing, it's possible someone will start up some GoFundMe project searching for interested volunteers. My suspicion is this is such a niche thing that you aren't going to get many people donating...particularly given this is a field where people in it resist letting go of money for things like a Wideband or a BinaryEditor license.