Does the E38 reference the Stoic AFR table, non-flex fuel equipted?

[Michael_D]

I'm getting conflicting information on this. I was under the impression that unless the vehicle was flex fuel equipped, changing this table's values would not change lambda and the O2's would think 14.7 was lambda 1. I have recently been told by a very experienced tuner that this is not the case. He told me that I could change AFR to whatever I wanted lambda to be, and the ECM would reference this table and adjust fueling to bring AFR to the value in the table.

I'd like to run 14.2 AFR for pure gasoline, verse 14.7. Or at least give this a whirl, as an experiment to see if fattening up the part throttle, lower rpm cruising would help with some reversion induced surging.

I tried this once before, while monitoring my wideband output, and it continued to report 14.7, even though I changed the Stoic AFR table to 14.2. Fuel trims did not appear to have changed much, if any. But, I did not re-program the wideband lambda 1 setting to 14.2 before I logged this particular run. I thought it would show actual AFR, but now I believe my thinking was in error.

[Iam Broke]

The ECU is going to control to stoich regardless of fuel in closed loop, the difference between 14.2 and 14.7 AFR hardly affects fuel trims. IIRC each 10% ethanol affects LTFT's ~2.5%.

Since you didn't adjust the AFR on your gauge to = stoich in your tune, no difference in gauge display was noted. I use the stoich table to run E85, non flex fuel at 9.83xxx and it def affects the fueling all over.

Tune in Lambda where 1.0 = stoich regardless of fuel and it's less confusing. .8 PE lambda = 1.25 PE in the GM tables.

Your O2 sensors are really Lambda sensors, AFR is just a number based on fuel type.

~14.7 is stoich for pure gasoline, ~14.2 is for E10.

[Michael_D]

OK, so just for clarification, as I'm not too bright.... If I change the Stoic AFR table to 14.2, the ECM is going to reference that value as Lambda 1?

And I suppose I'm going to have to go through the entire calibration process after I move it to 14.2; MAF and VE anyway?

I wish I would have started learning the whole tuning process using Lambda. I'm old school, where AFR actually meant something. I understand the concept of lambda, it's just a mental block for me that I need to get through.

[Iam Broke]

A. Yes, and it will add a bit of fuel compared to 14.7 so you may see your LTFT's shift slightly negative in closed loop compensating for the change as your fuel actual stoich value didn't change.

2. It's only going to shift ~2%, so prob not worth the bother. Multiply your entire MAF freq by .98 if you notice the ltft's go negative more than a few % to pull some fuel out in CL.

D. I learned to tune in Lambda, so AFR confuses me. I'm just old, no school.

[Michael_D]

Thanks for your input. I'm about to flash a new tune to the car and give it a whirl. I'll compare LTFT's between the two and see what happens.

Umm....almost afraid to ask this for the sake of really showing how dense I am.... Oh well, no point in worrying about that at this point.

I am a bit confused by your statement of possibly seeing LTFT's go negatived. If I have the MAF and VE tables calibrated to 14.7 being stoic, or lambda, and I move this value 14.2, wouldn't the trims go positive to add fuel? 14.2 is richer than 14.7, so I'm kinda scratching my head wondering why the trims would go negitive??

I have always thought that when the STFT's (ot LTFT) go negative, fueling is being reduced. Do I have this backwards?

[Iam Broke]

You are forcing slightly more fuel by changing the stoich of the tune to 14.2 from 14.7 AFR.

In closed loop, the 02's control to stoich regardless of fuel type, so since you are commanding more fuel, it'll cause the trims to go negative & pull some out (since the fuel type never changed).

Hope that helped.

[Michael_D]

I just don't get it. I need to re-read Banish's books I reckon. Seams backasswards to me that fuel would be trimmed to enrichen the mixture commanded. I'm over thinking this.

But you were correct with your forecast. My logs indicate that what you said would happen, trims went negative.

I took another log with the AFR changed to 14.2. Same road, same ambient temp, same humidity. I dropped both LTFT error histo's of both MAF and VE into Excel. The "average" error is:

MAF 14.7 AFR Stoic: + 2.26
MAF 14.2 AFR Stoic: - 0.52
VE 14.7 AFR Stoic: + 1.40
VE 14.2 AFR Stoic: - 1.38

I miss my screwdrivers and feeler gauges right now.

I may just be under an illusion of wishful thinking, but the car seamed to behave a bit better with the new Stoic setting. The typical 1500 -2200 rpm range that just continues to give me fits was a bit less obnoxious.

[Iam Broke]

Damn, the maf changed a little over -2.5%. I'm not a good teacher, perhaps someone else can explain it more clearly.

[DSteck]

(14.7-14.2)/14.7 = 0.034, or 3.4% which is an increase in fuel delivery (as evidenced by your trims going from positive to negative).

Stop thinking about AFR and only think about lambda.

Here is a very abbreviated and simplified version of the PCM's logic:

-Read airflow... MAF or VE. Lets say it reads 5 lb/min of airflow.
-Look up commanded EQ ratio (1 for part throttle)
-Look up stoich value for fuel (14.2 for example)
-Determine desired fuel delivery (5 lb/min airflow, 1 lambda on 14.2 stoich)

(5 lb/min air) / (1 lambda * 14.2 air/fuel) = 0.352 lb/min fuel

-Take desired fuel flow rate and convert to an injector pulse width using injector data, engine speed, applicable modifiers, and black magic
-Pulse the injector
-Repeat

The stoich value only defines how much fuel in relation to air is required to achieve 1 lambda. When going from 14.7 to 14.2, you're telling the PCM to get richer by 3.4% at ALL conditions. This is why your trims went negative. AFR is a fairly useless measurement. All wideband operate in lambda, and then convert from lambda to AFR for the gauge display. Why did AFR outweigh lambda in popularity? No idea. The point is... Start using lambda, and your life will become easier if you use different fuels.

And before anyone says it, his trims changed by ~2.8% instead of 3.4% due to the nature of the PI control for the fuel trims and crossing zero. You'll probably see a change almost exact if you don't change signs.

[Michael_D]

Dave, you are the man

This quote is all I needed........

(as evidenced by your trims going from positive to negative).

I have been thinking that when the trims go down, that fuel delivery was cut and that when they went up, fuel delivery was increased. Now I know the opposite occurs, and increase means fuel "trim".

I know I need to work Lambda. I am moving that direction.

I wouldn't put too much analysis into my LTFT error comparison above. NONE of the trims were disabled in either run, so you are seeing the average that includes DFCO/CFCO and all the rest you get in closed loop.

Seeing how my commanded AFR is now 14.2, and I had previously calibrated WOT with a wideband, I'm asuming that I now use the 14.2 value to set PE ratio? I have been targetting 12.35 AFR and had PE Ratio set to 1.19. Now it would be 14.2/1.15 = 12.35, correct?

[DSteck]

Stop.

Using.

AFR.



Set your PE value to 1.1744 and be done.

[Michael_D]

I know Dave, I know... Everything in the tune references lambda, not AFR. Multipliers, equations, formulas, etc.... I know, I get it.

One can't simply ignore AFR though. If you do not have an understanding of proper fuel to air ratios, and what to target, you could burn things up pretty quick.

By setting the PE value to 1.1744, I can only assume, (because I do not know) that when in open loop, the ecm ignores the value in the Stoic AFR table?

[Iam Broke]

Stop.

Using.

AFR.

[DSteck]

Stoich value is always used.

Stop using AFR. It is not important. Set it to the value you want, and put it out of sight, out of mind. Set your wideband up to report lambda and log commanded EQ ratio. Make them agree.

[MikeOD]

I know Dave, I know... Everything in the tune references lambda, not AFR. Multipliers, equations, formulas, etc.... I know, I get it.

One can't simply ignore AFR though. If you do not have an understanding of proper fuel to air ratios, and what to target, you could burn things up pretty quick.

By setting the PE value to 1.1744, I can only assume, (because I do not know) that when in open loop, the ecm ignores the value in the Stoic AFR table?

The O2 sensors in the car DO NOT sense AFR, they sense Lambda...they ALWAYS target Lambda 1.0...the Stoich AFR setting in the PCM is telling the equations what Lambda 1.0 equals...if you set it to 15.0, 14.0, 13.0 closed loop will find 1.0 and deliver the same amount of fuel no matter what. If you stick a wideband in the car, you'll find that as long as closed loop is enabled, the car is going to find the same AFR on it's own, no matter what you put in as Stoich...it just might take a moment to get there, and if you put a ridiculous number in, then the fuel trim may not be able to fully correct it.

The PE numbers are an "Equivalency Ratio"...similar idea to Lambda...Lambda is current AFR divided by stoich AFR for the fuel you're using...so a richer number in Lambda is less than 1.0. Equivalency ratio as used by the PCM in a silghtly different way...final target AFR = stoich AFR divided by EQ ratio...so when Dave says set PE equal to 1.1744...that's the exact same target amount of fuel as Lambda 0.8515.

Because Lambda 1.0 has a different mass of fuel as the fuel composition changes...there has to be some reference to composition and this is how they do it. If you know your fuel is 14.2:1, set the stoich value to that so the PCM math works the way it's intended to, then if you're tuning from the fuel trims, just target 0% change (+ or -)...trims can add or subtract...

If you see a 10% fuel trim, and the stoich value is correct, and the injector calibration is correct, then you're measuring the airflow with a 10% error, and thus also delivering the wrong spark advance...as you correct the MAF/VE to bring the trims to 0%, you're also now delivering the correct intended spark advance to the engine.

If you're tuning from a wideband, make sure you're correctly determining what voltages are what with regards to your fuel...widebands are also Lambda sensors...some controllers allow you to tell them what stoich is, and then when the sensors reads lambda 1.0, they will show you the correct AFR on the gauge...regardless of that though, they have an analog voltage output that you pipe into HPTuners...look through the wideband literature to figure out what the actual lambda is for that analog output, and you're good to go.

[Michael_D]

MikeOD, thank you very much for the detailed response. I am self taught through books, experiments and those on this sight who are patient enough to answer my stupid questions. And a lot of surfing this sight as well.... I've read, re-read and re-read your post, and am having trouble understanding a couple things.

The O2 sensors in the car DO NOT sense AFR, they sense Lambda...they ALWAYS target Lambda 1.0...the Stoich AFR setting in the PCM is telling the equations what Lambda 1.0 equals...if you set it to 15.0, 14.0, 13.0 closed loop will find 1.0 and deliver the same amount of fuel no matter what. If you stick a wideband in the car, you'll find that as long as closed loop is enabled, the car is going to find the same AFR on it's own, no matter what you put in as Stoich...it just might take a moment to get there, and if you put a ridiculous number in, then the fuel trim may not be able to fully correct it.

Understood. Narrow band O2 sensors output a voltage indicating oxygen content. AFR will vary with fuel type at lambda 1. I do not follow what you are saying about the stoic value though. Are you saying that it doesn't matter what number I put into the stoic afr table, and that trims will still trim fueling in closed loop to reach actual lambda 1, even if I intentionally want a richer fuel mixture and change the stoic afr value?

The PE numbers are an "Equivalency Ratio"...similar idea to Lambda...Lambda is current AFR divided by stoich AFR for the fuel you're using...so a richer number in Lambda is less than 1.0. Equivalency ratio as used by the PCM in a slightly different way...final target AFR = stoich AFR divided by EQ ratio...so when Dave says set PE equal to 1.1744...that's the exact same target amount of fuel as Lambda 0.8515.

Understand this as well. PE is just a multiplier to lambda 1 to enrichen the mixture when PE permissives are met. I'm back to my comments above though. When in open loop, PE enabled, is the ecm calculating an enrichment rate away from lambda 1? Or, does the ecm look up the AFR stoic value first, then use a corrected +/- lambda number away from 1, then apply the PE value to that number? [1 / 1.035(14.2)] / 1.1744 = .823 Lambda

If so, then I need to use the stoic value to figure out the PE multiplier I want to target a 12.35 afr, no? If not, then I do not use the stoic AFR value, and use 14.7 to find the PE multiplier?

Because Lambda 1.0 has a different mass of fuel as the fuel composition changes...there has to be some reference to composition and this is how they do it. If you know your fuel is 14.2:1, set the stoich value to that so the PCM math works the way it's intended to, then if you're tuning from the fuel trims, just target 0% change (+ or -)...trims can add or subtract...

Again, back to my original question, can I arbitrarily change this in order to enrichen closed loop operation? I want closed loop fueling to be richer than stoic.

If you see a 10% fuel trim, and the stoich value is correct, and the injector calibration is correct, then you're measuring the airflow with a 10% error, and thus also delivering the wrong spark advance...as you correct the MAF/VE to bring the trims to 0%, you're also now delivering the correct intended spark advance to the engine.

I'm under 3% error in all cells that I can actually populate with data, with most under 1% error - I do not have access to a dyno, so this has been a challenge.

If you're tuning from a wideband, make sure you're correctly determining what voltages are what with regards to your fuel...widebands are also Lambda sensors...some controllers allow you to tell them what stoich is, and then when the sensors reads lambda 1.0, they will show you the correct AFR on the gauge...regardless of that though, they have an analog voltage output that you pipe into HPTuners...look through the wideband literature to figure out what the actual lambda is for that analog output, and you're good to go.

Yes sir, did that.

[MikeOD]

Understood. Narrow band O2 sensors output a voltage indicating oxygen content. AFR will vary with fuel type at lambda 1. I do not follow what you are saying about the stoic value though. Are you saying that it doesn't matter what number I put into the stoic afr table, and that trims will still trim fueling in closed loop to reach actual lambda 1, even if I intentionally want a richer fuel mixture and change the stoic afr value?

Yes, the fuel trims can correct up to 25% in either direction, and they will always target lambda 1.0. In closed loop, you cannot control the target AFR, you can only tell the computer what that AFR is at lambda 1.0, that's what the stoich AFR table does...it's a reference for the PCM, not an output control. You have a very minor amount of control via the target voltage for the narrowband O2 sensors...but if you lookup the response curve for a narrowband, you'll see that over most of their voltage output range, they span a very very small lambda range that is essentially 1.0...something like 60-65% of their entire output range is basically lambda 1.0, and in the other voltages above and below, they're very inaccurate/non-repeatable.

Understand this as well. PE is just a multiplier to lambda 1 to enrichen the mixture when PE permissives are met. I'm back to my comments above though. When in open loop, PE enabled, is the ecm calculating an enrichment rate away from lambda 1? Or, does the ecm look up the AFR stoic value first, then use a corrected +/- lambda number away from 1, then apply the PE value to that number? [1 / 1.035(14.2)] / 1.1744 = .823 Lambda

You're overcomplicating it. PE mode is an equivalency ratio to lambda 1.0. Generally with a GM PCM, if a fuel trim is adding fuel in closed loop to get to lambda 1.0, it will be remembered when going into PE, and generally if it's pulling fuel first, it will be ignored in PE, although some newer controllers always apply the trim to PE wether it's positive or negative.

If so, then I need to use the stoic value to figure out the PE multiplier I want to target a 12.35 afr, no? If not, then I do not use the stoic AFR value, and use 14.7 to find the PE multiplier?

Again, you're overcomplicating it. Everyone always learned what AFR to target when pump gas was always ~14.7:1...that AFR number doesn't matter, and technically never did matter. The way I learned...I always targeted 12.5:1 to start, then played to see if the engine would tolerate going leaner to find a little more power. 12 years later, I can't find 14.7:1 pump gas anymore...can't buy it, it's not available around here...it's more like 14.2 now. That means if I still decide to target 12.5:1, I wind up with an EQ number in the PE table that is 1.136, or 0.88 lambda...that is starting out a little too lean, and I don't have as much room to play as I did in the past...sometimes I had to add fuel to even make a clean pull. That's when I realized Lambda was what mattered. When pump fuel was 14.7:1 and I always started a tune at 12.5:1, I was actually targeting 1.176 for an EQ ratio...or 0.85 Lambda. Now that stoich fuel is 14.2:1, I still target 1.176 EQ, which is actually something in the 12.07:1 range. I use the same numbers in the PE table that I always did, I just set the stoich reference in the PCM to match the 14.2 fuel I actually can buy. I set the reference number, I forget it's there, and I tune in lamba...in a car with an alcohol sensor in the fuel system, I don't even set the stoich AFR...I just open the table and make sure someone hasn't overwritten it with bad data, and if they have, I put stock data back into it...after that, again, I forget it's even there, and I think in lambda.

Again, back to my original question, can I arbitrarily change this in order to enrichen closed loop operation? I want closed loop fueling to be richer than stoic.

No, you can't. If you want to run closed loop, your closed loop fueling will be at stoich, you can't change that. The only computers that can run closed loop at something other than lambda 1.0 are computers using wideband O2 sensors instead of narrowband O2 sensors (which ours cannot do)...they can target any lambda value they want to. Because of the way a narrowband sensor works, you're always going to be at 1.0. Given the target O2 sensor switching voltages I mentioned earlier, you MIGHT be able to get it to run something as low as 0.98 lambda or high as 1.02 lambda (just pulling numbers out of space as a rough guess to show you that you cannot effect much change at all)...

I'm under 3% error in all cells that I can actually populate with data, with most under 1% error - I do not have access to a dyno, so this has been a challenge.

Anything under 4-5% error is fine...the cells you see with 1% error now will not show only 1% when the weather changes anyways...or when you buy gas from a different brand at a different station.

[MikeOD]

I'd like to run 14.2 AFR for pure gasoline, verse 14.7. Or at least give this a whirl, as an experiment to see if fattening up the part throttle, lower rpm cruising would help with some reversion induced surging.

Do you have a big cam in the engine? How much timing are you running at lower RPM and lower airflow? Do you have long tube headers on there? Post up the tune and some of the setup details.

[Michael_D]

Thanks again Mike. As you can see from your response, and those before you, there is a different opinion whether or not you can change targeted closed loop air fuel mixture ratio. After I moved it to 14.2, my fuel trims went negative, so I thought that meant changing the value did in fact force the ecm to run a richer than stoic air/fuel ratio (my fuel is not E85).

Car is an 08 Z06. Cam has 73 deg overlap at .006 and .674 lift. AR Headers, 1.7/8. Ported FAST and OE TB new (not drilled) FIC 60# injectors. Ported heads with swirl damn removed. Timing is pathetic. My premium fuel is 90 octane. I've tried moving timing from as high as 48 to as low as 25 deg in the cruise rpm at .28 and lower cyl air mass. I even installed a Velocity ring. The damn thing just feels like it's missing under 2000 rpm, and I'm at a loss what it could be. It's not a cam surge or reversion type of miss either. I've even changed plugs and wires. Running TR6's now. Tried gaps from .030 to .050. Logged cylinder missfire several times, ended up changing three coils that were missing more than the others.

I'm theorizing that at lower rpm, the engine is running lean, because the O2 sensors are seeing an overly rich fuel mixture, and adjusting trims to bring fuel mixture to lambda 1, resulting in a lean condition at the combustion chamber.

I really do not want to run open loop. Elevation and weather changes way too much and quickly where I live. I need trims.

Here's my last file.

[MikeOD]

Thanks again Mike. As you can see from your response, and those before you, there is a different opinion whether or not you can change targeted closed loop air fuel mixture ratio. After I moved it to 14.2, my fuel trims went negative, so I thought that meant changing the value did in fact force the ecm to run a richer than stoic air/fuel ratio (my fuel is not E85).

Mine isn't E85 either, it's E10 and yours probably is too.

The change in fuel trims shows you how closed loop with narrowband O2 sensors works, you targeted a richer AFR and the trims pulled fuel when they went negative and put the final AFR right back where it was before you made your change to the stoich value.

I'm posting from my phone now, and then watching the hockey game but I'll take a look at your file as soon as I can.