ECU:How does it calculate Timing??

[TomsSVX]

Hi all. Been messing around with the datalogging software in correlation with the select monitor. I took the car out last night for some extra hard runs to see exactly what is happening with the ECU... Here are some points.

1. Spontaneous acceleration and hard driving are fine... Car pulls well but seems to have a minor restriction... Just doesn't feel 100%.

2. 1/4 mile run.... half way through 3rd gear the car falls flat on its face. Monitoring the MAF signal, load, TPS, knock correction... Nothing seems out of place and all seems to be working properly. Only problem is, my advance has gone to ****. At full load it has about 6-8* advance BTDC which is absolutley killing the engine's performance not to mention spiking my EGT's up to about 1500*F. I looked at knock correction (retard) with both the evoscan software and with my select monitor and it registers 0 at all times. But monitoring the advance on both confirms that it is drastically low...

3. Car runs great through normal driving yet I am still having an idle issue.

My thoughts were always that the only way to pull timing out of the base advance is to have a knock condition in which the ECU could pull up to 10 degrees from the advance.... BUT since the retard function on my SSMI and the evoscan software say 0 all the time... how could this be possible?

I am at a loss of thoughts as to why my timing could be so low??? any ideas?

Tom

[Tim]

DId you end up using race fuel as mentioned or did you stick with pump gas?

[TomsSVX]

Yea.. was using 103 octane race fuel as well b/c when I got to the track... my fuel light was on... I originally wanted to get two runs on pump gas and 2 on race fuel at least.... But It was too low to both racing on what I had left of 93

Tom

[oab_au]

I guess this is the blowen engine running 9psi-12psi?.
What do you expect the spark advance should be at full throttle?

Harvey.

[SVXRide]

yeah, more info, please!
-Bill

[TomsSVX]

Originally Posted by oab_au

I guess this is the blowen engine running 9psi-12psi?.
What do you expect the spark advance should be at full throttle?

Harvey.

I would suspect the advance to be around 20-30* BTDC. Yes it is supercharged again running the "9psi" pulley. When I smack the throttle it will jump to 10psi and once i get to around 6500-7000rpms the boost will climb up to 15psi more than likely due to the fact that the cam profiles don't allow the engine to consume enough and it backs up in the intake. BUT just to be clear, the timing advance starts having issues around 3k rpm. Remember this is the totally stock engine it is on and it ran flawlessly N/A.

Tom

[SVXRide]

YT,
Any possible "back doors" to the timing tables? Any sensors that you're not datalogging that could be sending a signal that is causing the timing to be pulled back? I'm wondering if you aren't producing a combination of inputs that are putting the ECU into "self preservation" mode, thus pulling timing.
-Bill

[TomsSVX]

Thats why I am asking this question. Like I said, as far as I know the advance is calculated based on engine load and RPM. The only thing that should be able to draw timing out of the advance should be the knock sensors... SO, what the hell is happening is the question that is plaguing me right now

tom

[SVXRide]

So, let's go back to basic principles....

1. Pre "smart engine control" we'd lock in a static timing and then dynamic timing (under load) was controlled by vacuum (load) and springs/weights (engine speed) in the distributor. We could change the vacuum canister and the springs/weights in the distributor to modify the additional dynamic advance added to the static advance we locked in via bolting down the distributor. Pretty simple, yet effective.

2. Modern day "smart engine control" -- knock sensors that tell the ECU to pull timing to "protect" the engine from detonation, Crank Position sensor(s), Cam Position sensors, MAF Sensor, Throttle Position Sensor, Coolant Temperature, Air Temperature, O2 sensors, +?

3. I'm going to assume that the stock ECU using timing maps that are load vs rpm. If this is, indeed, true, then we've got to nail down exactly how the ECU is computing load and rpm. I'm going to assume again that rpm comes from the speed sensor in the trans? Now, which sensors actually provide input to the load measurement?

4. Last point (for now) - do we actually know the rpm range that our knock sensors actually work over?

-Bill
p.s. and, by the way, what fuel pressure are you running?

[oab_au]

So this is a 10:1 engine running std cams, exhaust system, and standard advance curve?
If this is the case, the advance for this full throttle condition would be about 36/38* from about 3000 on, put 10/13 lbs of boost on this and it is going to KNOCK and the ECU is pulling the spark back, as Bill says for "self preservation".
I can't see the timing needing more than 15* or less.

Harvey.

[SVXRide]

YT,
What was your AFR when your EGT started climbing? Peak EGT occurs at stoich. If you go richer than stoich, EGT will drop (thus my question about what fuel pressure you're running).
-Bill

[TomsSVX]

Originally Posted by SVXRide

So, let's go back to basic principles....

1. Pre "smart engine control" we'd lock in a static timing and then dynamic timing (under load) was controlled by vacuum (load) and springs/weights (engine speed) in the distributor. We could change the vacuum canister and the springs/weights in the distributor to modify the additional dynamic advance added to the static advance we locked in via bolting down the distributor. Pretty simple, yet effective.

2. Modern day "smart engine control" -- knock sensors that tell the ECU to pull timing to "protect" the engine from detonation, Crank Position sensor(s), Cam Position sensors, MAF Sensor, Throttle Position Sensor, Coolant Temperature, Air Temperature, O2 sensors, +?

3. I'm going to assume that the stock ECU using timing maps that are load vs rpm. If this is, indeed, true, then we've got to nail down exactly how the ECU is computing load and rpm. I'm going to assume again that rpm comes from the speed sensor in the trans? Now, which sensors actually provide input to the load measurement?

4. Last point (for now) - do we actually know the rpm range that our knock sensors actually work over?

-Bill
p.s. and, by the way, what fuel pressure are you running?

Bill, you must understand I am monitoring all these sensor simultaneously. With that said none of them seem to be out the realm of normal. Load readings are consistant and accurate along with the MAF, CTS, and all angle sensors. The RPM of the engine is actually pulled from the crank/cam angle sensors. All of these read on normal levels though.... Thus the problem.

i could only imagine that the ECU has the knock sensors working at all times.

My fuel pressure is stock 36psi.

Originally Posted by oab_au

So this is a 10:1 engine running std cams, exhaust system, and standard advance curve?
If this is the case, the advance for this full throttle condition would be about 36/38* from about 3000 on, put 10/13 lbs of boost on this and it is going to KNOCK and the ECU is pulling the spark back, as Bill says for "self preservation".
I can't see the timing needing more than 15* or less.

Harvey.

You have a PM

Originally Posted by SVXRide

YT,
What was your AFR when your EGT started climbing? Peak EGT occurs at stoich. If you go richer than stoich, EGT will drop (thus my question about what fuel pressure you're running).
-Bill

Bill AFR is in the low 13's high 12's when I am at WOT. EGTS rose up to about 1500* as per my peak monitoring of the new gauges... man those things are nifty

Tom

[b3lha]

Subaru have patented their ignition timing algorthms. You can read how it is supposed to work, but the language is hard to understand.

I think the correct one for our ECU is US Patent 5000150.

Go to http://www.pat2pdf.org, enter 5000150 and download it in PDF format.

[SVXRide]

Ah, a little "lite" reading!
-Bill

[SVXRide]

Originally Posted by TomsSVX

Bill, you must understand I am monitoring all these sensor simultaneously. With that said none of them seem to be out the realm of normal. Load readings are consistant and accurate along with the MAF, CTS, and all angle sensors. The RPM of the engine is actually pulled from the crank/cam angle sensors. All of these read on normal levels though.... Thus the problem.

i could only imagine that the ECU has the knock sensors working at all times.

My fuel pressure is stock 36psi.



You have a PM



Bill AFR is in the low 13's high 12's when I am at WOT. EGTS rose up to about 1500* as per my peak monitoring of the new gauges... man those things are nifty

Tom

YT,
Gotcha. Okay, given that I just took the Pearlie to get it MD state inspected and on the drive to the local shop it stalled when I came to a stop and didn't want to go over 3K when I floored it (making me think I've got a MAF issue. Hummm, is there something magical about 3k rpms??), I'm wondering if there isn't something in the MAF input loop with the ECU that could be causing your "goes flat in 3rd gear" . I'd also be interested to see what happens with your egts if you bumped the fp to around 40psi.
-Bill