Does this tuning logic make sense ?

[realcanuk]

After weeks of reading, learning and logging I need to understand if my thinking makes sense.
I started by dialing in the maf using the trims and got things in line. I read that doing it with the AFR error on the wideband is better so I redid it that way. When I get the error down to +-2 everywhere, the trims are not in line anymore when I turn them back on.
I'm thinking that if the O2's are looking for 14.7 by default, but my commanded is 14.40, then the trims and afr error can never really line up. By tuning with the wideband afr error (with the trims off) I am basically getting the afr to 14.40 everywhere. If the trims want 14.7 then they cannot match.
If this is correct, what is better, to tune to 14.4 with the afr error, or use the narrowbands to 4000 rpm and the wideband error for higher and PE ??

Thanks for any input as this is driving me crazy.

[realspeeddan]

In closed loop the commanded fuel target is stoich (lambda 1.0) (before fuel trims). If you haven't adjusted the stoich point in the tune for ethanaol, I don't understand how you are commanding 14.4 on the tune. In closed loop it will be commanding stoich which is 14.6x with 0% ethanol.

If you want your fuel trims in closed loop to match your open loop mapping then you have to have your wideband set for the same stoich point the car's computer is using.

Or you could eliminate the confusion by having a wideband that reads lambda and start thinking in terms of lambda instead of air/fuel ratio. The car targets lambda 1.0 in closed loop, so do your open loop part throttle tuning with your wideband targeting lambda 1.0.

[realcanuk]

In closed loop the commanded fuel target is stoich (lambda 1.0) (before fuel trims). If you haven't adjusted the stoich point in the tune for ethanaol, I don't understand how you are commanding 14.4 on the tune. In closed loop it will be commanding stoich which is 14.6x with 0% ethanol.

If you want your fuel trims in closed loop to match your open loop mapping then you have to have your wideband set for the same stoich point the car's computer is using.

Or you could eliminate the confusion by having a wideband that reads lambda and start thinking in terms of lambda instead of air/fuel ratio. The car targets lambda 1.0 in closed loop, so do your open loop part throttle tuning with your wideband targeting lambda 1.0.

I have changed the stoich for 14.40. I may try doing it in lambda but assuming using AFR I am just wondering is it better to use the afr error and not worry about the trims... or vice versa. It seems since the afr error is calculated from 14.4 the two cannot line up.

[Higgs Boson]

Put your commanded afr in the tune back to whatever it was stock.

Tune in Lambda.

Profit.

[realcanuk]

Put your commanded afr in the tune back to whatever it was stock.

Tune in Lambda.

Profit.

I hear you, but why does everyone say to change it for E10 ? Is it not worth doing ?

[realspeeddan]

I have changed the stoich for 14.40.

OK, so you changed it to 14.4 in the cars computer, but what is the stoich point for your wideband? For example, the stoich point on my NGK Powedex wideband is 14.5 and it cant be changed. So If I tune the car until my wideband reads 14.4, I'm really tuning so that it is richer than stoich. This is true no matter what the fuel is. For example if I am running e-85with a stocih point of 9.85 and my NGK wideband says 13.8 the car is really running rich, not lean, because it 13.8:1 is richer than the wideband's stoich point of 14.5.

The reason we are not answering the question of wheather you should go by your fuel trims or by the wideband is because if everything is working properly the fuel trims should match the wideband. Find out what the stoich point is for the wideband, tune the car until the measured afr matches the stoich point of the wideband and your fuel trims should be plus/minus 3% when you go back to closed loop.

[Higgs Boson]

I hear you, but why does everyone say to change it for E10 ? Is it not worth doing ?

Not really, IMO. You can check it at WOT and adjust if necessary, that is open loop.....use lambda. CL will adjust for the difference.

Also, if LTFT is enabled and the last thing they did was add fuel, that will carry over into PE/WOT/OL so it won't be lean.

The O2 sensors look for stoich, not an AFR number. Best way to keep it real is to use lambda.

[tpcaz]

When I adjusted my MAF curve using WB AFR error vs MAF freq I was able to get the actual AFR very close to commanded. However, once I enabled trims again, they were off. So I plotted LTFT vs MAF freq and used that to make the final adjustments to the MAF curve. After doing that a couple of times, the fuel trims settled down and my AFR wasn't jumping around anymore and stayed closer to commanded AFR for both cruising and WOT.

[realcanuk]

OK, so you changed it to 14.4 in the cars computer, but what is the stoich point for your wideband? For example, the stoich point on my NGK Powedex wideband is 14.5 and it cant be changed. So If I tune the car until my wideband reads 14.4, I'm really tuning so that it is richer than stoich. This is true no matter what the fuel is. For example if I am running e-85with a stocih point of 9.85 and my NGK wideband says 13.8 the car is really running rich, not lean, because it 13.8:1 is richer than the wideband's stoich point of 14.5.

The reason we are not answering the question of wheather you should go by your fuel trims or by the wideband is because if everything is working properly the fuel trims should match the wideband. Find out what the stoich point is for the wideband, tune the car until the measured afr matches the stoich point of the wideband and your fuel trims should be plus/minus 3% when you go back to closed loop.

Stoich for my wideband is 14.7. It seems to me if I adjust using the error from the wideband I would get the far to 14.4 as commanded but if the trims are looking for 14.7 they will be off. I think I will either play around with using lambda on the weekend or I will use the trims to tune everything and then use the wideband for wot.

[realcanuk]

When I adjusted my MAF curve using WB AFR error vs MAF freq I was able to get the actual AFR very close to commanded. However, once I enabled trims again, they were off. So I plotted LTFT vs MAF freq and used that to make the final adjustments to the MAF curve. After doing that a couple of times, the fuel trims settled down and my AFR wasn't jumping around anymore and stayed closer to commanded AFR for both cruising and WOT.

my experience exactly when using wb afr error.

[realspeeddan]

Stoich for my wideband is 14.7. It seems to me if I adjust using the error from the wideband I would get the far to 14.4 as commanded but if the trims are looking for 14.7 they will be off. I think I will either play around with using lambda on the weekend or I will use the trims to tune everything and then use the wideband for wot.

If stocih for your wideband is 14.7, then you need to tune the car until the wideband reads 14.7. Period.

If the car is running on E10 then the actual stoich poiunt is 14.09. When the car is ACTUALLY RUNNING AT 14.09 YOUR WIDEBAND WILL READ 14.7. So you have to tune until the wideband reads stoich, which in your case is 14.7.

If you tune it until your wideband reads 14.4 then you are tuning it to run richer than stoich , and you will always have larger fuel trims when you go back to closed loop.

This is why you have to think in terms of lambda.

[realspeeddan]

... but if the trims are looking for 14.7 they will be off....

The O2 trims are not looking for 14.7. They are looking for ACTUAL stoich.

It doesn't matter if you have E-85, E-15, E-10 or straight gasoline in the tank. They will switch from rich-to-lean trims or lean-to-rich trims when the afr crosses the stoich point. So with E-85 they will switch from rich/lean when it crosses 9.85 actual (14.7 on your wideband). On straight gasoline the trims will switch when you cross 14.67. This switching of the fuel trims will happen when it crosses the ACTUAL stoich point regardless of the of fuel the computer thinks it is running on and regardless of where the computer thinks the stoich afr ratio is.

[realcanuk]

The O2 trims are not looking for 14.7. They are looking for ACTUAL stoich.

It doesn't matter if you have E-85, E-15, E-10 or straight gasoline in the tank. They will switch from rich-to-lean trims or lean-to-rich trims when the afr crosses the stoich point. So with E-85 they will switch from rich/lean when it crosses 9.85 actual (14.7 on your wideband). On straight gasoline the trims will switch when you cross 14.67. This switching of the fuel trims will happen when it crosses the ACTUAL stoich point regardless of the of fuel the computer thinks it is running on and regardless of where the computer thinks the stoich afr ratio is.

Interesting. I thought the narrowbands looked for 14.7 regardless of the fuel used. I still have a lot to learn about this stuff.

[realcanuk]

If stocih for your wideband is 14.7, then you need to tune the car until the wideband reads 14.7. Period.

If the car is running on E10 then the actual stoich poiunt is 14.09. When the car is ACTUALLY RUNNING AT 14.09 YOUR WIDEBAND WILL READ 14.7. So you have to tune until the wideband reads stoich, which in your case is 14.7.

If you tune it until your wideband reads 14.4 then you are tuning it to run richer than stoich , and you will always have larger fuel trims when you go back to closed loop.

This is why you have to think in terms of lambda.

One more question. If I can make my wideband read stoich as 14.4, then I should tune for 14.4, right? Seems easier if that can be done.

[eficalibrator]

Dan is right. O2 sensors do not have any clue what kind of fuel you're burning. They just measure oxygen content and return a voltage (or current for widebands). They are LAMBDA sensors.

Best practice is to clearly define the stoich point in the tune so that it matches your fuel. (~14.1ish for e10, 14.68 for e0, 9.85 for e85) Then all factory open loop fuel tables are in units of relative fueling (either lambda or EQR, depending on how your display is configured), GM uses EQR as a default.

You wideband is measuring lambda. Many cheap widebands convert back to AFR for gasoline, but this ASSumes a stoich point of about 14.7. The best thing you can do here is immediately convert the signal back to lambda by dividing by the reference ratio for your wideband. Better yet, get a good wideband that actually reports lambda.

When calibrating fueling errors, always use delivered lambda divided by target lambda. This will eliminate the confusion associated with various ethanol blends and the resulting AFR errors.

[realcanuk]

I want to thank everyone for their input. With your help and some more reading I finally got it all straight in my head.
I believe the best way to things, especially if you tune multiple cars is to use lambda. In my case, because I am only learning for my own car and interest, I decided to stay with AFR.
it all worked out very well. I changed the settings so that my wideband uses 14.4 (my commanded afr) as stoich. I then dialed in the maf using afr error and then again using stft's and both ways matched up great. When I got the trims to good levels, my afr error was less than 1% everywhere. On top of that my wideband actually reads around 14.4 where before it was reading 14.7.
I see no issue with doing it this way and am very happy with the results.

[realspeeddan]

Yeah, I see nothing wrong with doing it that way, as long as you understand what is going on.

For all my talk of tuning in lambda, my wideband still displays 14.5 for stoich. I've been doing it that way for so long that I "think" in 14.5:1=stoich AFR terms. I just know that 14.5=lambda 1 on my equipment, even when I'm tuning an E-85 car. As long as you know how the gauge really works you will be just fine.

[Higgs Boson]

Dan is right. O2 sensors do not have any clue what kind of fuel you're burning. They just measure oxygen content and return a voltage (or current for widebands). They are LAMBDA sensors.

Best practice is to clearly define the stoich point in the tune so that it matches your fuel. (~14.1ish for e10, 14.68 for e0, 9.85 for e85) Then all factory open loop fuel tables are in units of relative fueling (either lambda or EQR, depending on how your display is configured), GM uses EQR as a default.

You wideband is measuring lambda. Many cheap widebands convert back to AFR for gasoline, but this ASSumes a stoich point of about 14.7. The best thing you can do here is immediately convert the signal back to lambda by dividing by the reference ratio for your wideband. Better yet, get a good wideband that actually reports lambda.

When calibrating fueling errors, always use delivered lambda divided by target lambda. This will eliminate the confusion associated with various ethanol blends and the resulting AFR errors.

Interesting that you say to change the stoich point in the tune.

Does GM (or whatever other factories) tune on E0? They are always a 14.7 or 14.6 it seems.

Why not calibrate them at 14.1 out of the factory since you can't really even get E0 anymore? Now E15 is approved as well. :-( :-( :-(

[wyochimneysweep]

A lot of pumPs in north Dakota are still e0. We also got e10, e30, and e85 pumps

[Higgs Boson]

A lot of pumPs in north Dakota are still e0. We also got e10, e30, and e85 pumps

fair enough. i am just basing it on the national average, corn is a big hit.

maybe they are calibrated for the chinese market. they are still experiencing their industrial revolution, cars exported there are tuned for coal.