JPrevost

Grumpy, I'm doing it. I've got just about all that I need, if not, I'll go get.
What is the exact way you hooked up the vss? Could you spend a little time and just draw up a small schimatic to show?
I'm starting with my extra 8746 ecm. Considering I'm starting my own source coding I figured it's about time I do this.
I have a 10amp 13.8 volt power inverter, would that be enough to power the ecm? I don't know if I'll be able to get some crappy injectors or not but couldn't I just use LEDs while testing the source code?
I'm hoping to make this all fit into a small 2'x1'x8" space so I can travel with it.
Also could we set a date for the pow wow?
Thanks

Grumpy

Mines pretty much built in a ecm case.
I just used some 55s for the rpm and vss.
Also went nuts and used flip-flops so the they were 50% duty cycles.
Dats the one part that's lasted forever, and I've lost the skematics for em. If you posted to GMECM I'm sure one of the EEs would give ya a hand.

May 25+26
On the 24th gonna convoy over to Mendelsons, dunno what time on that yet.

On this version I'm just running leds, instead of injectors, if you want to run injectors you really need to use a Tractor battery to have the amps and then run a trickle chagrer when your firing injectors.

GofasterFirebird

iTrader: (3)

ok, I am sorry, what is the point?

JPrevost

The point is to be able to watch changes you make in an eprom before you take your 1.5 ton vehicle on the road. One thing I hated doing was making some changes, taking it out on the road and having it hesitate really bad, sometimes it would even stall in extream situations. Also if (not likely when) you do source code / assembly coding you never know what'll happen and it's always a good idea to check things out. For example, I want to program my ecm to control my fan. I hope to eventually have the fan parameters as such; coolant on and off temp, vehicle speed on and off. So when the car is running down the highway at 55+mph the fan turns off and just spins. I'm also planing on going to a cowl hood mid summer and I think it would be cool to have a speed dependant cowl flap. It won't be a problem so long as I get some time to tinker with this stuff.
Basically you can watch what a car would do without actually leaving moving your butt .

Grumpy, how are you displaying the timing?
Also the GM-ecm list is down, been that way since Sunday but I'm sure you already knew that cause you're the moderator .

GofasterFirebird

iTrader: (3)

ok, I see, but say I hook up this box thing, how do I know if my car is gonna stall when I put the chip in my car??

Also, what is meant by source code and why do you alter it? Is it so you can add stuff like a speed activated cowl valve?

Momar

When you use a program that is on the market to modify the chip you only modify stuff thats already there. If you can edit the source you can write entirely new stuff into it. For instance his speed activated cowl. He can make it so that when any sensor he chooses is at any reading he chooses it activeates a relay possibly send out varying current(not sure on that one) or such so that it can control just about anything electronic you could imagine.

Ben

GofasterFirebird

iTrader: (3)

got the source code part, thanx


but what about my first question????

Momar

I dont totaly understand that or how much detail you can tell from it, but It does appear to tell you when each injector would be firing, and he has lights for important sensors. I imagine that they come on when certain requirements would be met by the sensor or something.

Grumpy

Will you know for certain?.
No.
But, you can eliminate alot of errors.

In my ecm I've found some stuff that's not even in the code they have settings for.

You can see how the commanded AFRs change when you make a change rather then have to make a WOT pass. Once you see how the commanded AFR and actual line up then you can pretty well predict what to do where to get what result.

You can monitor the duty cycle of the injectors at any and all points of the table. Rather then having to sort thru datalogging.

An ecm bench is like a WB.
You don't have to have it.
But, they make GREAT TOOLs for getting the system right.

and this is all just about the oem code, changing anything takes you to a whole new level of tuning.

Grumpy

Generally you want to use the ecm to ground something rather then supply the current to power it up. It's also alot safer if a wire shorts to ground.

Grim Reaper

I only have a brief moment, but I am thinking this thread should be made "sticky" also. What are you guys thoughts.

For those that wonder "what's the usefulness of a test bench"? The usefulness is limited by your own imagination. Definitely for anyone diving into "Source Code" a test bench offers a great way to see that certain routines are being executed as you think (or hope they should be).

Many years ago, I was a software developer and one of the key things we always did (when they didn't have "debugging Editors like today", we would use the program to give "Hellos" or "expected values" to prove that the newly implemented code was infact executed and done as expected.

E.G. You are developing a routine and want to prove that the routine (say a Highway Mode Spark Advance Routine) was executed and went through as expected. And, under certain conditions, you want to make sure it is not. On a regular computer you'd just display "I am here" on a Display - but you don't readlly have a display with an ECM.

So, you use a field that has minimal significance (say the RPM/MPH Gear displayed on most scan tools) and you display "specific values" to ensure it is doing what you are expecting under various scenerios. Then, you apply the "specific voltages" to simulate various "conditions" and you see if the ECM sets those "expected values" to the "RPM/MPH" Gear value in your scan tool.

AlexJH

I'd like to see this sticky, but I'm an EE student so this stuff interests me.

This summer one of my projects is building a testbench, so I'm sure I'll have more to say about it when I start scrounging for parts.

JPrevost

The gm-ecm mailing list is finally up and running . I got a reply and a circuit diagram from Jim Cottrill. This is it:

Scott_92RS

Would it be possible to hook up a serial LCD to the ECM and control it via custom source code? I have no clue when ot comes to ASM, but wouldnt you be able to set some kind of data string, then call a routine that sends that text to the LCD? That would be a great addition to the test bench. Also, are any schematics publically available for teh whole bench?

JPrevost

http://www.diy-efi.org/gmecm/projects/aldl160/

Scott_92RS

HOLY CRAP! Thanks!

monte-ss

Hey grumpy, what did you use to build the test bench or where did you get the plans for it? I did a search for anything like that here but didnt really turn anything up. I agree this would be much easier than taking a chip out and not being sure it will actually work.

thanks

Grumpy

There were no plans.
The first series about how to get one to work were failures. When it dawned on me to use a series of 555s, then it got alot easier. some early stuff was spinning a distributor with a drill motor to try and figure out what signal things really needed. Then selecting pots to replace sensors. Then choicing resistors to use with the pots to get the min and max values so as not to trip diagnostic flags etc.

Just a whole lot of playing around.

monte-ss

ok so the way I see it is, its just a computer with 12 v source a 555 timer, with a possible potohm to vary the signal to represent rpms for the dist, and then a few more pots for the sensors like you were saying, and just LED's for your outputs, how would this tell you what ammount of timing is added or how fast/long the injectors are on. Im curious it seems like that would work, but a lot of wiring or atleast the ecm and the plugs cut off if thats what you used.
the kind of things Im looking for are, what values of the Cap's and resistors onthe timer, and what potohm's with what resistance values did you use for the remainder of the inputs, after that I think I can get it.

JPrevost

I don't know how to fake the ecm for the rpms but I know that all the other sensors are either switchs (usually ground, look at wiring diagrams for your ecm) and the rest are pot.ohm controlled but be sure to use the 5 volt power supply from the ecm.
To get the ecm to display actual final values for such things like spark advance and pulse width you need to modify the bin file to output those values in the aldl stream. I wish I had the source code for winaldl but that isn't going to happen any time soon.
Right now I have the bin modified to display final calculated spark advance in place of the eprom id.
Does anybody have any circuits for the rpm?
Anybody know where to get some old wiring harness so I don't have to directly solder to the ecm? I believe all ccc (computer controled carb) thirdgens have the same wiring harness as my TBI computer.

SATURN5

any updates? Heading there myself tomorrow for some stuff to wire a buddy's Triumph chopper.. cheers, Bob

JPrevost

So what is the signal needed for the ecm to see engine rpm?
Is it the same circuit as the VSS?

hectorsn

The ecm gets a square dc signal from the module but I don't know exactly what kind of wave and voltage. However, I think the easy thing to do would be to mimic the pick-up coil signal, which is an ac signal, and feed it to the module and so forth. I think the pick-up signal could be the same as the vss.

Grumpy

Easiest to use a 555 just to trigger an ignition module. that way the ecm sees the right signals on the right lines. Some cals sample the module connections. Same with EST module. the voltage drop across a LED is fine for triggering the module.

same basic circuit as the VSS you'll have to play with values some to get the right range.

monte-ss

I have an old radio shack engineers notebook on 555 timers, I copied these 2 parts out that ought to help, one is a graph giving figures on what capacitor and resistance values to acheive certain frequencies, with the right pot, you can get the frequency to change. Since the distributor turns half as fast as the crank and you want one ossicilation per spark, you would need the frequency to be 4 * any given RPM for example at idle of 600 RPM you would need the frequency to be at 2400 Hz and up to 6000 RPM to be 24000 Hz.

monte-ss

got the pics backwards, but the last one is how the 555 timer is hooked up, and if you are needing to use 2 555 timers you can use one 556 chip which has 2 555 timers inside it and only needing one 5V input and one ground. Here is the chart that was supposed to go in the last post. Ill have to rescan it, the pic is too big to fit, Ill repost it here shortly

monte-ss

not sure what the deal is but I cant post the pic, it keeps saying too large but its smaller than the allowed value, unless the server counts both pics together, but if anyone wants it email me at jpotvin@rap.midco.net and Ill send it to you

TP355Z

i know i'm probably off topic (usually am) but where could I find an article on building the interface link between the cars ECM an a PC? I've got the ALDL program and have the 8746, i just need the link to start pulling information (right?). Too - I did dig for an article but i came up short (i printed the how-to ecm programming one tho, i'm reading that now) - so its not like i'm just asking.... I'm sure JPrevost will find an amount of irony in seeing me here...(even if i'm not programming my own chips i'm still pulling info! ARGH!!! lol). Thanks in advance.

TP

JPrevost

http://w1.601.telia.com/~u60113744/s...dl/winaldl.htm

TP355Z

thanx jon

SATURN5

Ok..getting started on the bench. But still have a couple questions.

1. Can I Ohm my Knock sensor and replace with a resister to ground?

2. I don't remember seeing how the O2 was set up. Can I use a pot to swing from 0v to 1v?

3. Grumpy, any values for the pots? Heading to Mendleson's on Friday for parts.


I think I have the rest under control. Will post some pics as soon as something's ready, doing all the planning and getting the case prepped. Going to use a 165 case/connector and wire up with a pigtail to go between the ECM and bench. Going to include the aldl converter inside too. Lots of drilling and soldering ahead. cheers, Bob

JPrevost

http://www.diy-efi.org/gmecm/projects/ecm_testbench/
this is a good link for this thread

Grumpy

You should see the new one. <G>
Much cleaner, and now that it's specific to one ecm, I can do a full labotomy on it.

And with the Romulator, no chip buring.

Grumpy

JPrevost

When can I check it out? I'm only what, 1.5 hours away. My test bench is still in pieces. All the circuits have been made but it isn't organized. I'm one of those kinds of people that needs all or nothing. Since I didn't have all the same switchs or trim pots I just left it sit back in NJ. Every time I think I've got the time I end up wrenching on somebody's car. Someday when I take a break from school and get a job....

Z28GEN3

access forbidden- darn it!

where else can i find info on the test bench, like how to build one?

Grumpy

Hmm, you were a no show at the last invite <G>

You got my number, and email address, might just one or the other, and contact me. <G>

Grumpy

Look for Saturn5 's address, and write him, he was making printed circuit boards for them.

91L98Z28

wow. i just realized I completely forgot to get my order in last time for a board..

the thread here indicates that he is out of stock on boards:

https://www.thirdgen.org/techbb2/sho...hreadid=172255

i'm still interested in one, though..am perfectly willing to pay first if it helps get another batch made.

that makes two of us..

HaulnA$$

Count me in too. If anyone also has a copy of the diy-efi project file, I would truly appreciate a copy.

Z28GEN3

IM in, much easier to do this then hook up the battery each time i make changes.

SATURN5

And out of nowhere... Sorry, I haven't been on the computer lately.

And now that some results are back, I can relax a bit and let the cat out. My wife was diagnoised with stage 1 cervical and intestinal cancer and has been undergoing chemo. She had an exam after her first run, and shes responding well. It looks like one more round of chemo and that should be it.. Hopefully.

Yes I am currently out of boards, however if you all can wait a couple weeks while they are made, and I get 5-6 boards prepaid, I'll set up another run. Sorry but I can't front the cost this time.

Cost will be the same, $55.00 shipped.

The only thing is I am up to my elbows in life, so I will only offer bare boards with documentation and parts list. I just can't swing parts kits or assembled boards.

You can PM me here, or email me at rawardsr@ameritech.net.

Thanks, Bob