I just wanted to start a new thread on the subject because Mike J did'nt have a response after months of trying everywhere.

It would be extreemly usefull to have a device that would give us the timing advance in real time via a LCD screenor something else.

Why run 20 psi and put stress on the turbo when 15 psi does the same because of timing being pulled back.



------------------
Michel
My life is crap
http://www.azur.qc.ca/rallye

Hey Man!!!

I don't have a turbo model but I'd like to have such an instrument like that.

I'm sure that some electronic "Wizzard" could easyly make one. Guess that we could use the output from the "Crank Angle sensor and a signal from the "Transistor Pack"???!!!

An LAD display could be made as a visual reference...

I don't have the electronic knowledge but I'd try it if someone would make a scematic.

Any taker???

Ghislain.

yes, please, somebody make one. There would be a huge profit potential.




------------------
Michel
My life is crap
http://www.azur.qc.ca/rallye

Even just monitoring a voltage somewhere that relates to te timing vould be good enough for me. Then we could chart it and learn what voltage means what... kinda like the o2 sensor voltage monitoring.

------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Alt Email: vortexracing@hotmail.com

Let's see, let me talk to a few people on this and we will see what I can do.

From an email I got from Martin Q:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Ok Mike, i'm gunna lay it on ya like this....

The Talon (actually, the 4G63) uses 2 ignition coils to fire the 4 plugs.
They're fired in sets cause the pistons travel in pairs... so that means the
plugs are actually firing twice as much as they need. No big deal, not an issue.

The coils are 'fired' buy hitting them with a 12V pulse. The pulse is controlled
by a 'module', but the signal that tells them to fire comes straight from the
ECU. There's no 'timing' signal per se... it's just a stream of pulses. And if
it want's to advance the ignition, it sends the pulse sooner.

Make sense yet?

It would be pretty hard to monitor the advance with out going inside the ECU,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~





------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Alt Email: vortexracing@hotmail.com

How about comparing that pulse to a TDC signal or something???

------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Alt Email: vortexracing@hotmail.com

jeez, this should'nt be complicated like this. when I hooked up a portable "snap on" scanner on my car of the diagnostic plug besides the fuse box, I could see in real time the timing in degrees.

------------------
Michel
My life is crap
http://www.azur.qc.ca/rallye

Yes you are talking with the ECU to get that info. Talking to the ECU requires knowledge of the protocol and other stuff which is very complicated and that's why diagnostic tools and dataloggers cost so much.

Our probelm is we are trying to find the data outside of the ECU somewhere. That's where the problems arise because it seems an actual advance value does not exist outside of the ECU (as stated by Martin) and so we would have to create one off other signals.

Or if someone can figure out how to talk to the ECU and ask for the timing advance and/or knock sum I'd be plump happy with that too This is exactly what those diagnostic tools and the datalogger do through the diagnostic port. I think they do it all through only 2 pins [send & recieve].

Later...

------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Alt Email: vortexracing@hotmail.com

[This message has been edited by Mike Jackson (edited May 05, 2000).]

Mike J, you mentioned a compare to the TDC signal. This might work but the electronics involved might start to get sticky. There much work to do here.. but before i get rolling i've got some reading to do ! It would be great to have a little mini computer that could spit back all the essential info in real time.

Guess this'll be a summer project.

Q.
Anyone have info on rebuilding the instrument panel/dash-board ? What about someone in the racing/rally circuit? I'd like to have a nice gauge layout (with the above computer) but i can't figure out where to put all the gauges so they don't look cluttered.

------------------
Dk Blue
90 TSI AWD

Martin, well I replaces the whole center console with a custom made sheetmetal one withth help of Craig watson.

I removed the switch/control for the rear defoger, the recirc/fresh rotary knob for the ventilation and relocated the hazard button on the top left corner.

This left me plenty of space to play with since I don't have any radio equipment. I have room for my MASC digital display, remote control, fuses, switches (fog, water injection, water sprayer, ECU, O2 signal) relays (fuel pumps, fog lights). I installed also a power outlet for the Gtech.

I should take a pic eventually

------------------
Michel
My life is crap
http://www.azur.qc.ca/rallye

Here's another idea... mod the ECU to output the knock sum value to the stock boost gauge. I know it's possible cause Todd Day can do it. The stock boost gauge signal is run from the ECU so you can mod what the ECU sends to it. Somehow you can tell the ECU to output the knock sum to the boost gauge rather then the calculated boost value. The ECU knows how to output the knock sum because it will when a datalogger asks for it... now we just have to find out how to output that all the time and route it to the boost gauge. This is over my head on exactly how but there are the basic principles if anyone wants to take them and run

Problem is I think this can only be done by modding the memory of the chip and guess what that means that it would only be possible for EPROM guys If anyone can figure a hardwire way to do it I'd be very impressed.

------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Alt Email: vortexracing@hotmail.com

THere are far too many variables to be taken into consideration. I think the best bet would be to have Todd Day have this as one of his GUAGES in his EPROM upgrage. $300US plus all the other goodies is a pretty good deal.

Jazz

Does anyone know the name of the protocol the ecu uses to communicate, or whether it uses a protocol at all.. I mean it is digital, digital signals are interpretedas I/O, nothing more.. You have to program chips using a special Unit, dont know what they're called, but Omron Makes them here in Canada, my dad does that kind of thing for a living, I could ask him some questions, eventually come up with something that someone else may be able to polish up..

It is likely the communication stream from the ECU is custom, and bears little resemblance to any known protocols.

Modding the ECU requires detailed knowledge of the code, as well as other tools necessary to modify and program the code into an ECU EPROM. Only a few people know how to do that, and they spent a great deal of time and effort to acquire that knowledge. Todd is one of them.

It should be possible to calculate the timing the same way the ECU does - by measuring signals from the crankshaft angle sensor and ignition coils. While the ECU controls the ignition, it has to know the position of the crank to determine when to fire.

This would require a microcontroller that could detect the various signals and calculate the timing, as well as display the information somehow. It would be a bit tricky, as you would have to convert the time between ignition and TDC into degrees of crankshaft rotation, but that wouldn't be terribly difficult.

In other words, Martin is on the right track here. I think that someone who works with microcontrollers would have the best time designing such a device, although there are other ways of doing it too.



------------------
S.
1000 AAQ: 1000q.dsm.org
ECU Primer: members.home.net/costall/ECUprimer/index.html

That is what I referded to. Using the signal from the cranck angle and the coil pack. Those inputs are the two variable needed as reference; Cranck position as basic reference and the coil pack output been the advance signal reference .

Any body knows any microcontroler Guru??? I know one but he's not into cars at all and he's too busy with his own project to get involved.

Ghislain.

I am myself a programer and a lots of my friends are too... a couple of them are into cars too and the others would sure like to learn a new communication protocol...

So maybe we could try to learn how the ECU communicate... In fact I was trying to find out how ECU were working like 10 years ago... but I did not had the knowledge at that time... and I was not finding any documentation on that.

If anybody have any information (even if it seems obvious or stupid) I would gradly appreciate if you could send it to me by email. my email is francois@chaosdev.com Any kind of information would do (links, books titles, magazines, anything!)

If I can find out how it is communicating I guess the possibilities are endless!

And by the way if it can output things so the dataloger can interpret it I really can't see why we could not use that information to display it in any possible way (like a led, a counter, the stock boost gauge, etc)

I'll keep you informed as soon as I learn anything on that!

Oh yeah and I will need your help too cause I might be good with computers but before reading that thread I was not even sure what timming was! so I will need some mecanic guru too!

Thanks!

------------------
Frank
Real thirsty 92 Laser RS TURBO AWD

Here's some back issues of the digest when Todd Day was first figuring out how to talk to the computer.... just incase it helps you.


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~---------------------------------------------
The ENG circuitry is almost impossible to decode. I get lost in a
massive tangle of SMDs with two character labels and laser-cut
resistors. However, I was able to decipher about a third of the
circuitry. Surprise! Except for the protection diode, the third
I could decipher looks almost like the DATA circuitry.

This is curious. Just what is output circuitry doing with input
circuitry? I think that DATA is a mode switch and ENG functions
as both an imput line and output line. You leave it
floating when you want the computer to talk to you. You ground
it and then talk to the computer on the ENG line. Just a wild
hunch, but it's the best I could come up with.

Now the bad news. Since I don't have the slightest clue what the
computer requires (i.e., it's language), I have no idea of what to
try to send to it. Besides, it's just a wild guess.

I'll do some reading of the main manual when I get home, and update
you if I figure anything else out.

-------------------------------------------------

Okay, I went home and looked thru the big manual, and found an interesting
reference to pin 10 in, of all places, the automatic transmission control
unit! So this blows the idea that pin 10 is some kind of data input. But,
it is called "Diagnosis control" in the manual. Strangely enough, although
it is labeled, it is not mentioned at all in the tranny part of the service
manual.

But, the plot thickens! In the schematic for the TCU, it shows the pin
that's supposed to be for reading out the error codes as having the
facilities for reading a signal on that pin!

So, here is what I suspect. Grounding pin 10 puts both the MPI and the TCU
into a special diagnostic mode where you can actually *talk* to them! This
means direct control over the engine computer *and* the automatic transmission
computer! Basically, pin 10 is a direction switch. When grounded, it accepts
data on what used to be the diagnostic output. When floating or high, it
outputs data on that pin.

Here's the bad news. I have absolutely no idea how to talk to the computers.
I only know how to put them into this mode.

But wait! What about the part of the manual where it tells you to ground
pin 10? [This can be found in the Basic Idle Speed Adjustment section of
the manual under Fuel System]. Well, here, I think what happens when you
put the computer into this special diagnostic mode, is that it stops trying
to automatically compensate for changes in its inputs. This allows you to set
the idle speed without interference from the computer.

BTW, I just went out and tried grounding that pin with a semi-warm engine,
and it did nothing. Perhaps the results would be different with a cold
engine.

---------------------------------------------------------

Interesting tidbits I found inside engine computer...

Processor is Mistubishi MH6111 (does not appear to have
pinout of any Motorola 6811 part (damn)). It has an
ASIC, which I have verified, part number 60011. It
has another IC which I've not identified, but I forgot
part number (sorry, I should have written it down).
It does have an EEPROM with a window (27128).

Neato findings: (keep in mind I have a '90 Talon AWD)

Missing resistor R110, which is connected to the turbo
boost gauge output. Perhaps this allows use of different
boost gauge?

Missing resistor R102, which is connected to the air flow
sensor. Perhaps this allows use of different air flow
sensor?

Missing jumper J104, which is connected to the only empty
pin (does not connect on any other '90 models, either).
This is related in someway to air flow section.

Jumpers J102 and J103 apparently select whether Pin 13
acts as fuel cutoff or motor position sensor input (1.8L).

Pin 14 goes to some mysterious part of the air flow sensor,
which does not exist on non-turbo models.

Pin 17 is GND for 2.0L, motor position sensor input for 1.8L.

Missing jumper J106 and resistor R111 have something to do
with purge solenoid.

Pin 104 is +5V for automatic tranny, GND for manual tranny.

If you do open up your computer, you will find what appear
to be five tab regulators. Actually, the one that looks
different from the rest is the only tab regualtor, and it
outputs +5V. The other four transistors are the fuel injector
controllers.

Well, that's about all I could get from looking at the circuit
and studying the schematic. Perhaps all of you with manuals
could spend some time this evening and report back with any
info that may have been added in newer models.

Also, please please please, someone who knows a service tech
very well, pump them for info on the engine computer and how
you talk to it!

----------------------------------------------------------
Once again, I opened up the engine computer to see what I could see.
This time, I was a bit more intelligent about it, though.

All this talk about controlling the car via a handheld PC has got
me hooked into figuring out how it works. Well, I did a bunch of
tracing on the circuit board, but got caught in SMD hell and was
unable to figure some of it out.

The key appears to be the line marked "DATA". I don't have the main
part of the manual with me here at work, but I could have sworn I
saw something to the effect that grounding this line allows you to
make some adjustment to the car (either idle speed or timing). After
reading the electrical manual, this strikes me as odd, since there are
check plugs for doing either of these tasks. The DATA line is not
mentioned in the electrical manual except for the schematic.

Let's back up a bit here. The line marked "ENG" provides the error
codes for sensor failure, or the "heartbeat" pulse if everything is
okay. So what I'm thinking is that DATA is input and ENG is output.

However, this is not true. The second you ground DATA, the heartbeat
from ENG stops. What is it waiting for? I decided to do some tracing
on the pc board to find out.

The DATA circuitry looks pretty much like what is shown for the reed
switch input. Basically, the line is pulled to 5 volts through a
protection diode and then drives a transistor to give a clean logic
level waveform to the CPU.



------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Just call me Runner-Up

I have a pretty good idea on a way to do this but I need a harware guy to work on it with. If anyone is pretty good at this stuff and has the inclination, email me.

Thanks,
Reza

------------------
Reza
97 F2 TSi AWD
FRC FED

It may be better to continue this discussion on the UBB, since more people are likely to contribute that way. Besides, this forum is pretty quiet, so if we keep this thread going for a while it can't hurt anything.

I am both a hardware designer and microcontroller 'guru'. I am currently working on a gauge design that could easily be adapted to timing, once I have the chance to do some programming on it. It's not a trivial task.

For those who wanted to figure out the ECU communications protocol, you don't need to know what timing is. The ECU would just spit the figures out for you to display as you choose.



------------------
S.
1000 AAQ: 1000q.dsm.org
ECU Primer: members.home.net/costall/ECUprimer/index.html

Hello everyboby.

I was gone for the past few days but I made it back.

I was talking to the busy electronic Guru I was refering to and he told me that he will be looking into this project.

Only thing I need for now is; What kind of voltage the ECU comunicate with different sensors??? I'm pretty sure it is 5v. And so far the only "Input" we'll need is going to be from the Cranck/Cam angle sensor and the coil pack.

One more variable that will be needed is be the range to be monitored. IE; -20 deg BTDC to +5 deg ATDC.

I got a NT and never seen the output (Timing curve) with the Data loger. So if anybody has any valuable input to give here it's the time. This will not be happening overnight but I'll post questions here as required in order to make this happend.

I might also start this topic over as it's getting a little long.

Keep in touch and don't be shy to give me your input as this toy will be for all to enjoy...

Ghislain.

The datalogger records timing from 0 to 50 degrees BTDC.

------------------
1992 Eagle Talon Tsi FWD
Superiority after 60'..
Inferiority in 1.5' of mud..

ICQ # 9622983
www3.sympatico.ca/steve.kinnaird

TX for the info Steve.

I was thinking about making a little LED display ( One LED per deg ) but I think that with such a wide range it might be a little unpractical. I'll keep you informed.

Ghislain.

Well it's kinda like the o2 sensor voltage... as in it has a large range but the range at wot is smaller and that's what we care about. I was logging again on the weekend messing with boost and a/f ratio and was able to get my timing as high as 17 degrees (averaging around 15~16) and down as low as 6 (averaging around 8). So you could build an LED display for 0 to 20 degrees if that helps.

------------------
Mike J

-Quadcylla-
92 Laser RS-T, FWD

Vortex Racing
http://www3.sympatico.ca/robin.jackson2/HomePage.htm
Just call me Runner-Up

This is much better.

TX.

Ghislain.

I would prefer having a digital readout. It's more precise when you glance at it. But if you can finally figure out how to display the timing, I would be sooo happy.

Hello again.

You might be right, we would think that a digital Read-Out would be a lot easyer to read. But there is one thing to consider though, it is the fact that if the reading goes up and down too fast you won't have the time to acurately read the number on the display... This is why a linear display would be easyer to read. Different coulor can be use for various stage of the range.

What I'll be doing in the next day or so is to get with the Dude and tell him what we want to acheve, give him what kind of input we have to work with and then we will figure what kind of display would be good for this aplication.

Again I'll keep you guys informed on the development of this toy.

BTW what do you guys want to call it??? Something short and cool!!!

Ghislain.

I am actually working on this project as well. I hope to have something ready fairly soon, but (of course) I am having trouble dedicating time to it.

I'm going to let you guys in on some of the details so that nobody wastes any effort here. I'm not trying to dissuade anyone from making their own stuff, but I seem to be much further along and don't feel right about watching other people spend their time and effort duplicating what I have already accomplished.

The gauge I am building will have both digital and analog displays. It will also have a peak hold feature to record the minimum timing (maximum timing retard). It will be able to measure the entire timing range of the engine, including the negative range (degrees after TDC) that the datalogger does not show. It will be about the size of a Palm organizer.

I have most of the details worked out now and am in the middle of programming the gauge. It is based on a 4k 4 MHz microcontroller and I will be making dedicated PCBs for it.

Since we are on the topic anyway, if anyone has any feedback as to features that the gauge should or should not include, please mention them here.

One detail I am currently missing is if the spark signal from the ECU to the power transistors is active low or active high. Does anyone know the answer to this? (It would save me having to hook up the oscilloscope to my car....)

I presently don't know when the gauge might be available, but I estimate I will be selling it for $50 CAD or so.

------------------
S.
1000 AAQ: 1000q.dsm.org
ECU Primer: members.home.net/costall/ECUprimer/index.html

[This message has been edited by Sean Costall (edited May 29, 2000).]

[This message has been edited by Sean Costall (edited May 29, 2000).]

Oops, one more thing...

Ghislain: My gauge will be called the VISOR. I'd hate to have yours called the same thing.



------------------
S.
1000 AAQ: 1000q.dsm.org
ECU Primer: members.home.net/costall/ECUprimer/index.html

I'd like to order the first one for a 2G, please.

Seriously... having to fire up the notebook, each time you want to check the timing is a pain. And besides... everyone knows that my car needs another digital readout.

Backlit LCD?

Suggestion: check that the 4Mhz is fast enough. I know it sounds plenty... but that's what the MASC was based on, and we suspect that some of the 2G problems related to the 4Mhz not being quite fast enough to keep up with the ECU. In the second iteration of the MASC, he was going to a 20Mhz chip...

------------------
Andrew Bienhaus - Web-Team Dude
1997 Eagle Talon TSi AWD

Mmmm. Order!!!

I'm just working on the development of this widget and I'll be making one as a test to see if it works.

I should be posting the scematics for all to enjoy by making your own.

I'm not an electronic guy at all. Yes I did a few project in the past and I'll be attempting to make this project see the light of day. But I'm not getting into masive production.

Ghislain.

Andrew: I agree that 4 MHz is not as fast as it sounds. However, my gauge is a more specialized device than the MASC, and I anticipate no problems.

The MASC is complicated by the fact that it has to keep track of what is going on, figure out what to do about it, and do it. It also had a big (slow, serially driven, I assume) LCD display on it, a passel of user controls, etc. My gauge only has to keep track of what's going on - the rest is another project - a mini-MASC, I guess. More like a mini-AFC is what I was considering.

I can always boost the speed up if I run into difficulties. As it stands, I have 4 us resolution on the timers - the fastest time I need to worry about seems to be the injector dead time, somewhere around 500 us. Even if I boosted the gauge speed up I wouldn't get any better timing than that, as I've reached the limits of what the micro timers can do for me.

I was up until midnight last night rewiring my prototype. I ran out of I/O lines and had to do some significant rework to squeeze the last bit of capability out of my controller. I hope to get back to programming it tonight.

The LCD will be backlit.

Ghislain: Looking forward to seeing what you can come up with. I'm spoiled, I program micros all day at work.....



------------------
S.
1000 AAQ: 1000q.dsm.org
ECU Primer: members.home.net/costall/ECUprimer/index.html

I'll order the gauge and the mini-afc once they are ready please and thank you

*drool*. so where do I send my money?

Hey Sean. I'll see what I can come up with then I'll ask you a few questions in order to double check the development process of the "Doohicky".

Now I got to deal with my busted ECU caps. It happend today and one of the "Run" is burned. I'll bring it to a shop to see if it (the run) can be fix but I don't think the odds are in my favor... Would be too easy if I could just changed the caps... Bought some at a local shop.

Ghislain.

ill buy a few when u get them done me and 2 or 3 friends would love to have that
dan

Jeez I'm sorry I mentioned it.

Michel: you can send me as much money as you want....

Ghislain: If you're SOL on the ECU I'd be willing to take a crack at it. The damaged traces should be repairable somehow.


Thanks for the encouragement, guys! But I have to say this now (for the benefit of anybody joining this thread late):

I do not know when I will have this product ready for sale.

This is turning into the Energizer bunny of the UBB....

------------------
S.
1000 AAQ: 1000q.dsm.org
ECU Primer: members.home.net/costall/ECUprimer/index.html

[This message has been edited by Sean Costall (edited June 01, 2000).]

Mmmm.

I gave it shot but I was unsucsessfull. I'm not the best soldering person on the face of the earth but I did attempt to fit it and put three new caps on the board.

Now I get a constant 12V on the ECU check. No good at all. I'll buy a used one and convert the old one to an anchor...

Ghislain.

Well my birthday is coming up soon. Will Mom be able to buy me one of those items?

[This message has been edited by Michel Brais (edited July 09, 2000).]

Hello again.

I gave some scematics to the Dude. He had a look at it and sayed it could be done easily. I'm leaving for summer vacation, will be back in two weeks. I should get with him to get this project going.

Sometime it will get done. Not sure about your birthday...

Have a good one anyway Man.

Ghislain.