[Shawndude]

I have always wanted to monitor the ignition timing advance/retard on my car. With OBDII cars it's dead simple, but no such luck for us. This feature would let us optimize fuel grade, see when we get pre-ignition (by monitoring retard) and increase boost with more safety.

How do we make this thing? Anyone seen one available we could purchase? Standalone is not an option for me.

I've looked through the articles on how our ignition works, and I think I got the grasp of it.
(page 9 and up, page 11 very usefull):
http://www.cygnusx1.net/supra/library/techarticles/h23.pdf


The easiest way I've come up with, is to use the G2 hall effect sensor, and then via inductive pickup on #1-6 spark plug, run it into a processor and calculate the timing.

I don't like the inductive pickup much, so we can use the IGT output line, and have the processor look for the first output after the G2 goes high, and time that relative to RPM. The processor would have to disregard the next five triggers after this. I'm no programmer, so I'd have to ask my friends if that is a problem.

Alternatively, I could just get a RS-232 data acquisition boards (they're cheap), and feed the input into the computer for display, and data logging. It would be nice to have a stand alone unit though.

Any ideas, inputs, nudges in the right direction? Just kind of thinking out loud, so feel free to throw out suggestions my way.

 

[shiva]

shawndude,

are you sure the g1 and g2 sensors are hall effect? i think they're just reluctor pickups so even if you use IGT instead of an inductive pickup on the plug wire, you still have some conditioning to deal with. i'm pretty sure you'll need to condition the g1 or g2 signal for it to be useful for any adc. but, the good news is that should be easy! there are a few possibilities:

1) national semiconductor makes a chip specifically designed to trigger from reluctor signals. it's called the lm1815 (you can get the data sheet at national.com), and ought to be pretty "plug 'n play" as it has a mode where it uses an adaptive threshold-- it continuously changes the threshold based on the voltage of the signal. it might take some messing with but it should work. it costs about $3 from digi-key, and i think you can still get free samples from national.

2) if it doesn't, people have had success dealing with reluctor pickup conditioning with simple comparator circuits with a low threshold and a voltage clamp

3) old toyota ignitors (before TCCS or any sort of elec. spark advance, such as one off an '82 supra), and GM's old HEI module (easily available for $10-$30), might be able to handle conditioning for you. they would get a reluctor signal from a distributor sensor, and spark at each zero crossing of the signal. that is, the it would connect its coil - output to ground. so, you could easily have a +12v trigger based on a reluctor pickup. there's no guarantee it'll work but there's a good chance it will


and for IGt, the processor would of course have to look for the falling edge of the signal.

also, do you think there is a chance where spark will occur before the g1 or g2 pickup? it seems unlikely that toyota would design the position of the pick ups that way, though the fact that there are two and that it needs to know which cylinder is firing means the computer does do a lot of interpolation. the way i understand it is the next Ne crossing after g1/g2 indicates 10deg BTDC piston #1/#6. so g1/g2 could correspond to any position between 40deg and just over 10deg BTDC of piston #1/#6, probably much closer to 40deg, but who knows for sure.

that might make the coding a bit tougher, if you actually care about measuring that sort of timing because it will obviously be under very light load, which, perhaps you dont

the rpm signal would have to be conditioned too probably, right? national makes a frequency to voltage converter, i think its the lm 2907. i dont know how it works really. but after that it'd probably be easy to hook to a computer.

and... why stop with timing? pressure, o2 sensor voltage, maybe even egt. that'd be sweet, man. get a bunch of pressure and temp sensors, tap pipes all over your intake, and help solve some of the old pressure drop debates. with a cheap data acq. outfit that interfaces well and smarties like you on the forums we could all really learn quite a bit

shiva

 

[Ckanderson]

my brain hurts

 

[Shawndude]

Wow, thank you for the input Shiva. I'll definitely have to study this a lot further.

I'm kind of leaning toward getting a acquisition package, something like shown here:
http://www.bb-elec.com/bb-elec/literature/CatalogPages/pg52.pdf

Lots of choices out there though.

Wanna hear something really funny? I thought about just hooking up the timing light, strapping a web-cam to the timing cover and just making a mpg movie of it by feeding it into the computer. Heck, I could use two web-cams, one to show the RPM, one to show the timing.

How's that for a cave man approach.

Edit: Dang, I could focus one camera on the whole instrument cluster, and I have a voltmeter that reads frequency. Hook that up to the AFM meter to get the airflow, strap it to the dash, and I've got butload of approximate information. Water temp, Boost, RPM, Speed, air flow, oil pressure, ignition timing. Use another voltmeter for throttle position, and I'm styling.

Ha ha ha, we've got tons of web-cams here, I'll throw this together one of these days, for shits and giggles.

 

[Bishop92T]

I think you will find timing goes off the "chart" on the timing cover. Hehe, a friend of mine drove into some wet grass and brake torqued the car to build full boost while some other knucklehead checked the timing. End result was the timing was too far off the scale to see.

 

[SuprAng]

how bout a simple knock monitor? you can buy those from hks and others... aslo the safc2 has the option...

Ang

 

[dbsupra]

i have ALWAYS wanted something to look at timing. i decided i would have to settle on standalone or just look at how much it is retarding timing if any. this would mean a j&s system, which would be proactive if knock was heard ($800ish for setup) or a knock monitor gauge from casper (~$100). the casper wouldnt do anything but tell me if timing was being retarded tho, but much cheaper setup. to mix it up further i had thought about using a apexi itc, but setting that up would just be guess work using how much i took out on the afc.
let me know what you guys come up with- im interested.

dave

edit- the casper electronics unit would tell how much the timing is being retarded, not just the fact it is being retarded. just thought i would clarify

 

[Shawndude]

I don't think any knock monitor, such as the Casper unit would be very useful. It will only allow you to see if the engine is knocking, but you'd still wouldn't know if the engine is pulling timing, due to the knock.

In theory the engine should be always at the verge of knock, advancing and retarding back and forth. It should be knocking very minutely all the time!

(page 8 in particular) http://www.autoshop101.com/forms/h40.pdf

Only if things get so bad, that the ECU retards timing all the way, and then it knocks you would be able to see substantial knocking on a knock monitor.

You could run cheap gas, without knock but your timing would be retarded let's say 5 degrees, over higher octane gas. That is the information I want to see.

Do you guys agree with this?

Speaking of knock monitors, I looked at the HKS "knock monitor" and it is nothing but a microphone!!
http://www.autodesignhaus.com/tuners/hks/ipm.htm

It has a filter feature which can be turned off, which I can replicate on my computer, but I'm quite sure the ECU is doing a good job retarding the timing to keep it out of knock. I would just like to see how far it is retarding it, and if there is anything I could do to make it happier (higher gas octane, water injection, cooler intake, better intercooler cooling etc etc).

Here is a link to basically duplicate the HKS knock monitor without the filter feature, for about 20 bucks:
http://www.autospeed.com/cms/article.html?&A=0353

The filter would be a band-pass 6-8 khz. You could build that fairly easily also, but I don't think you'd really need it.

Damn, why are my posts always so long?

 

[Shawndude]

I think you will find timing goes off the "chart" on the timing cover. Hehe, a friend of mine drove into some wet grass and brake torqued the car to build full boost while some other knucklehead checked the timing. End result was the timing was too far off the scale to see.

Hah, I'm glad nobody won a Darwin award in that fiasco!

Yes the timing marks would be way off the marks on the cover. I think timing can get as high as 40 degrees or so?

So I would just mark extra lines on the harmonic balancer with white paint, and then I'd be able to see it on the cover scale. Let's say I mark it 15 degrees off the stock notch, then I would just add the cover reading, plus 15 degrees (or whatever).

Edit: Thanks for the input everyone!

 

[chevyeater-on-sf]

Ah, if only we had a comm link to the stock ecu.

Your knock sensors are nothing more than a type of microphone as well. They pick up normal engine noises just as well as knock. The magic is in the tuning of the ecu to differentiate between normal noise and knock. Some knock can get by them and some noises or vibrations can fool them into thinking there is knock when there is not.

 

[derK]

me no smart

 

[Silver88Turbo]

Sorry to wander off topic slightly but is it possible that the knock sensors could pick up the vibrations from very slight rod knock? I spun a bearing not too long ago but the engine seemed fine to me before it happened except sometimes at WOT the ECU would pull timing until about 4750 rpm when it would advance it and there would be a sudden burst of power....like what felt to be almost 100ftlb difference. The only thing I can attribute this to would be the knock sensors picking up the slight rod knock and interpreting it as pinging. EGT's where fine and I wasnt boosting high and using premium gas so Im 99% sure it wasnt detonation.

And I think we should be working on a effecient way to control timing and not just observe it . One step at a time though I guess...

 

[dbsupra]

i would have to disagree shawn. sure what you are talking about neither of those devices will do. what you want is exactly what i have been looking for. i was just making mention of what was out there right now and what they do. sure the casper gauge wont do what you want it to but i dont see it being worthless. if its knocking then the ecu is retarding. it has a a segment display for how much knock. the more knock the more timing is being pulled. how much timing? see thats the problem. but i would have more of a clue correlating the data from the monitor vs having nothing to go on but faith and best guess from how much im pulling from the s-afc. i hope you hear where im coming from.

dave

 

[ZombieSS]

I can build the device your talking about. Thanks for the LM1815 part number shiva, i've been looking for a IC that did this.

The trickiest part of doing all this is the timing required to write the output to an lcd since the writing to an LCD is a slow process. I'm going to order some LM1815's and play around with them to see if I can get a clean signal from the disty. I've always had trouble when trying to get good signals for doing anything RPM created and have tried several of my own conditioning circuits and software with mixed results.

Measuring the timing after you have the initial trigger signal and an RPM is fairly trivial since most microprocessors have a really good resolution.

If there is interest in a device like this I think I might build one. I've kinda been wanting to do this myself, but with a twist. I want to control the timing on a boost vs rpm scale. That would allow us to have a much greater degree of control over the timing on the car.

 

[Shawndude]

dbsupra, I agree with you, and I didn't mean to to imply it was worthless. I appreciate any input from you guys. It's speculation at this point, but I think you'd see the knock sensor always showing slight knock. I know Larry_A had the gauge, maybe he can enlighten us. He explained some of it on this thread:
http://www.supraforums.com/forum/showthread.php?threadid=121811
________________________

Silver88Turbo, a rod knock can pull your timing, if it happens to occur at the wrong time, and the wrong frequency. The computer knows when ignition takes place (and what piston) so it only judges the sound during that time. It has lots of filtering to do, and the system is fairly fragile. But I do not know any more than what is discussed at this site, so by reading it you'd probably know better than I do.
http://www.autoshop101.com/autoshop15.html

I think the e-manage has pretty good control of the ignition, although it's blind. A big thread on it in the tech section you can check out.
http://www.supraforums.com/forum/showthread.php?threadid=96432

Bishop92T is the man to talk to if you can't grasp something.
_________________________________________

I talked to my software programmer co-workers, and they think it's a walk in the park to make this happen.We can make it a standalone, or have it computer loggable. I haven't decided which way to go yet.

Tonight I will use a dual-input lab scope, that has fairly good processing power, to see the details of its operation.

Once again thanks for all the suggestions.:bigthumb:

 

[Shawndude]

Awesome ZombieSS! I agree, thanks shiva for the LM1815 link, it looks ideal for this application, and helps tremendously.

This data acquisition unit looks useful enough for this, and has good support. I'd have to have my programmer friends look over the details, but maybe you'd know if this would do the trick, for computer interfacing?

http://www.labjack.com/labjack_u12.html

Cheers.

 

[MKNTRAXX]

wow.. this looks very interesting.. I'll keep an eye on it.. and when something is developed count me in.

 

[Michael90T]

As usual shawndude you got a great idea!! you build a few I will pick one up!

hell you know me I love flashy light and monitoring equipment

 

[bdekoning]

what i THINK shawndude is trying to figure out is too complicated for the work... what WOULD be a great idea though would be to get a good knock monitor and something like an apexi S-ITC so one could safely tune their own timing maps... substantial hp gains there..... (but very dangerous)

 

[Shawndude]

Some decent progress.

I will test a GM HEI reluctor module once I get my hands on it, since the local parts store didn't have it. Thanks Shiva for mentioning it, I think it will work perfectly from the documents I've seen on it. If not, I have the LM1815 chips on order.

I'm meeting with an Electronics Engineer next week, and we'll get going on the circuitry.

Almost for sure it won't be a standalone, it will get logged to a computer. Much more useful that way, since we can correlate it to air flow, RPM, throttle input, etc.

And just to clarify, I have no intentions of selling this or manufacturing it for others. I will be more than happy to provide all the schematics or any information learned from it.

Cheers.