Fuel trim question

[Ben Charles]

All ears

[Higgs Boson]

All eyes. And I am not sure that I would call changing stoich value AFTER MAF and VE are tuned properly a "hack."

[DSteck]

It is a hack. Here's why.

Ok, you tune the car in open loop with stoich set to 14.68. Let's assume your MAF functions perfect at all operating conditions. MAF and VE are dialed in PERFECT, meaning that no matter what, you're always hitting 1 lambda except during PE, where you're hitting 0.85 lambda (using that as an example).

This translates to 14.68 AFR during normal operation and 12.49 AFR during PE.

Let's say that during normal operation, in one instance you're moving 10 lb/min of airflow. With a 14.68 stoich point and 1 lambda, this means you're using 0.681 lb/min of fuel. While in PE at 0.85 with 14.68 as programmed stoich, you're moving 50 lb/min which translates to 4.00 lb/min of fuel consumption.

Now you put the car in closed loop and change stoich to 14.2. You end up with 0% trims (again, ideal world). There have been NO changes to the VE and MAF calibrations. In the same operating condition that you're moving 10 lb/min of air during non-PE, with 0% trims, that means you're using 0.704 lb/min of fuel, which is a 3.4% increase in fuel consumption. During PE at 50 lb/min of airflow with a 14.2 stoich AFR programmed in (and for 0.85 lambda, that'd be 1.1765 PE multiplier), you'll end up commanding a 12.07 AFR which means 4.14 lb/min of fuel consumption, which is again a 3.4% increase in fuel consumption. This means your wideband that is set up to show stoich is going to show 0.822 lambda at WOT instead of 0.85 because your existing airflow model was set up using a leaner stoich value.

If you tune in open loop, and switch to closed loop and get huge trims, then your wideband is either way out of calibration or doesn't work well. The other possibility is your O2 sensors have drift, but I doubt that's the case. Changing the programmed stoich value impacts fueling EVERYWHERE. You shouldn't screw with the stoich AFR value just to get trims correct AFTER you've "tuned" in open loop.

You need to check your wideband. Tweaking stoich to fix fuel trims is a hack, just like manipulating the PE table instead of tuning the MAF and VE. It's the same thing as cheating the fueling multipliers instead of fixing the airflow model. The fact of the matter is that we have no way to determine stoich of the fuel in our tank unless we want to spend thousands of dollars sending it off to a lab. Whatever stoich value you tune the car on is what you need to leave it at. If your wideband doesn't show stoich while the car is in closed loop, then you have an issue with accuracy. Both of my NGK AFXs show stoich (well, they show oscillation around stoich due to closed loop) during closed loop, but the average deviation from stoich comes out to 0%. When I tune a car in closed loop to 0% trims, and then switch it to open loop, it still shows 0% error from commanded.

That's fact. Not conjecture.

[Higgs Boson]

Ok, thanks. I'm not going to argue with you. I understand the theory but I also understand what works on the cars I tune and apparently other people's tunes as well.....

VE and MAF are airflow models not fuel flow models. They are curves to represent the nature of the mechanics of your engine. If the shape is correct, they are correct. If the whole table is 5% high or low then it is. The shape is more important. That's why the stoich table is editable.

HPT, if it is so dangerous and wrong to use all the tools in the box, please remove the stoich box in the next beta to save me from myself!

Look, DSteck, I am REALLY sorry if my philosophy is a little different than yours. Call it a hack if you want, but I can show you the pro tune that WAS in my car and you will see a hack job.... It's a big wide world and a lot of people in it. You and I can acheive the exact same result with only a minor difference in style or method and everything will be ok. I bet if you and I both tune the same car our curves will look the same, even if one is a touch higher or lower overall. And we are talking about a small amount.

Besides, if the stoich of his fuel is right at 14.2 and that's what the stoich box says to have trims line up, that seems BETTER to me than it saying anything else. But, I'm smoking something.

[axekick]

I'm not saying if one way or the other is right. I just want to see if I'm understanding this. Adjusting the stoich table after putting it back in closed loop to get trims to zero is simply a global adjustment to account for a discrepancy between the wideband and the narrobands.

That doesn't seem bad as long as the discrepancy is linear and independent of AFR.

[DSteck]

I'm well aware that the MAF and VE are airflow models and nor fuel flow models. The stoich table combined with the injector calibration data is what gives you fueling. Picking a random stoich value to make up for discrepancies is not correct. Please explain what you do about your WOT fueling going rich when you do this.

When you mess with the stoich value, you start to affect transient fueling. I don't care if you had a "professional" tune before. Nobody can define who a pro is... Some people think just because it's a shop tune, it is professional.

I'm still waiting to hear what wideband you're using. Like I said, when I tune cars in open loop and enable closed loop, trims hover around zero. Shitty equipment will cause problems, and frankly I trust narrowbands 100% to tell me I am at stoich. This is why 99% of the time I tune part throttle with closed loop. I know that my injector data is correct, and I trust my equipment, so I calibrate the MAF and VE because I cannot measure stoich of the fuel.

If my gun sight is slightly off, I fix it instead of aiming just left of my target.

The stoich table is not editable for the purpose of people being lazy and doing a ghetto global adjustment. It's there for people running different fuels with different stoich ratios. Kind of how the PE table is there to adjust commanded fuel and not actual fuel (which is what hackabilly idiots do).

Greg Banish would have a stroke if he saw this thread.

[edcmat-l1]

I trust narrowbands 100% to tell me I am at stoich. This is why 99% of the time I tune part throttle with closed loop. I know that my injector data is correct, and I trust my equipment, so I calibrate the MAF and VE because I cannot measure stoich of the fuel.

Exactly. At the end of the day, the PCM could care less what your WB says, all it knows is how far off the trims are. Nothing wrong with tuning part throttle with trims. Works great for me all the time.

Kind of how the PE table is there to adjust commanded fuel and not actual fuel (which is what hackabilly idiots do).

I'm on the fence with the whole PE table "raping". Now, I don't typically "rape" PE tables, but take in to account the GM PE tables. Are they "raped"? They sure look it. But, this is a subject for another thread.

Greg Banish would have a stroke if he saw this thread.

He probably hasn't posted yet cause he read it, and now he's done locked up!

[Higgs Boson]

You are really blowing this out of proportion. It is nothing like hacking the pe table. Wot fueling should be richer if your stoich is richer. It is fine. I'm using a good wideband, if fact, I think one recommended by you in other threads.

I can't respond anymore to this thread. Your attitude is ridiculous and too uncivilized for me to continue. Sorry, bud.

By the way, are you even going to respond to the op to help fix his issue since my working fix was spot on or is attacking me just too much fun for that? Also, my car is not some nearly stock car, easy to tune. I had to retune it from the ground up , 408 with nitrous, etc.....put down 725 rwhp with my hack tune. I suspect it will blow up any moment.

[IDRIVEAG8GT]

You are really blowing this out of proportion. It is nothing like hacking the pe table. Wot fueling should be richer if your stoich is richer. It is fine. I'm using a good wideband, if fact, I think one recommended by you in other threads.

I can't respond anymore to this thread. Your attitude is ridiculous and too uncivilized for me to continue. Sorry, bud.

By the way, are you even going to respond to the op to help fix his issue since my working fix was spot on or is attacking me just too much fun for that? Also, my car is not some nearly stock car, easy to tune. I had to retune it from the ground up , 408 with nitrous, etc.....put down 725 rwhp with my hack tune. I suspect it will blow up any moment.

Eh.......

[DSteck]

I'm on the fence with the whole PE table "raping". Now, I don't typically "rape" PE tables, but take in to account the GM PE tables. Are they "raped"? They sure look it. But, this is a subject for another thread.

It was explained to me that the PE table is set up how it is for durability in the event some guy is WOT in fourth gear up a 30% grade. That's why it is all goof-wonky... But also because the MAF is calibrated on a flow bench and the calibration engineer can't modify it.

diatribe

You want me to talk about the mega power street cars I do with narrowbands? Is that supposed to be worth street cred?

Nothing is being blown out of proportion. I'm just pointing out that your tweaking stoich approach is amateur hour crap and nobody should follow that advice. If your NGK AFX is deviating, then either you need to calibrate it or use a better power and ground. I hope you aren't using a cig lighter adapter. Both of my AFXs jive with the narrowbands in every car unless something causes electrical noise or there's a poor ground.





To the OP... Verify that your wideband is worth it's salt, and just tune the car in closed loop except for PE conditions while using normal stoich value.

[edcmat-l1]

But also because the MAF is calibrated on a flow bench and the calibration engineer can't modify it.

Ding Ding Ding!

[streetsweepergto]

Good thread

Valid points

[Higgs Boson]

I didn't state my dyno number for street cred, just that my car is not just a basic maf curve tune. Nothing more. I am also not disputing your method, just your aggressive attitude. You tune cars all the time, I tune them once in a while. I am not using a cig lighter adapter, lol. I have already said that I have read a lot of your posts and probably followed some of your advice. You don't have to get your panties in a wad....my "diatribes" are no longer than yours so no need to be disrespectful.

I have helped some people with their tunes, I have received a lot more help in return. I'm not stealing your business, we probably have similar interests so just chill out.

I have had plenty of tunes where the nb's line up just right and some just dont. Funny how the stoich change was the right amount do nt you think? Just retard coincidence? I never said it was the best way, just that it's what works sometimes. Actually I stated my tuning method, which should have had the disclaimer "for these situations."

Anyways.....have a nice evening.

It was explained to me that the PE table is set up how it is for durability in the event some guy is WOT in fourth gear up a 30% grade. That's why it is all goof-wonky... But also because the MAF is calibrated on a flow bench and the calibration engineer can't modify it.



You want me to talk about the mega power street cars I do with narrowbands? Is that supposed to be worth street cred?

Nothing is being blown out of proportion. I'm just pointing out that your tweaking stoich approach is amateur hour crap and nobody should follow that advice. If your NGK AFX is deviating, then either you need to calibrate it or use a better power and ground. I hope you aren't using a cig lighter adapter. Both of my AFXs jive with the narrowbands in every car unless something causes electrical noise or there's a poor ground.





To the OP... Verify that your wideband is worth it's salt, and just tune the car in closed loop except for PE conditions while using normal stoich value.

[DSteck]

Just retard coincidence?

Pretty much. I'm still waiting for you to explain why that exact number corrects it without screwing up fueling in open loop. If the wideband and narrowbands don't agree, you have another issue. End of story.

I'm not wound up at all. I just don't want people doing things the wrong way because of an unfounded irrelevant adjustment.

[Higgs Boson]

I'm not wound up at all. I just don't want people doing things the wrong way because of an unfounded irrelevant adjustment.

That's fair. It just seems like you create an immediate assumption based on one statement that goes against your philosophy. From there on out, said person is nothing but poop. Seems like a rough way to go about life, but hey, I'm a smoking retard.....

Why don't you tell me why changing that number will screw up open loop? Shouldn't you check it before you quit tuning? Seems like there is always a final test drive on the street and if any adjustments need be made they should be.

My intention was to help him correct his trims for his fuel based on something I have seen work not teach him how to tune.. You could have offered an alternative, maybe better solution but instead chose to attack. It doesnt seem productive. Just makes ya feel good for the day. Don't worry I have been there too.

[DSteck]

I choose to hightlight when people are wrong so it doesn't propogate. Why I didn't really answer the OP is something I don't feel like explaining.

I already explained why it screws up open loop. Go ahead and do your whole process... tune open loop, then enable closed loop and modify stoich. Go drive with 0% trims. I bet anything your wideband shows that you're off. The fact of the matter is you have a wideband accuracy issue if you're ending up with deviations like what you've talked about. Like I said, as soon as you go into open loop with what you've done, it's all going to show up somewhere in the neighborhood of 3-4% rich. What will you do then? Adjust the entire airflow model by that error, then re-enable closed loop and bump stoich down even more to fix trims?

I have never had to fudge values elsewhere in the tune to fix fuel trims after tuning in open loop.

[Higgs Boson]

Like I said, I normally don't change it. Sometimes I do. It seems reasonable to require explanations where you are not obligated. Anyway, have a good night.

[edcmat-l1]

Higgs, look at it from a reverse perspective. Get the computer in closed loop, dial in the VE and/or MAF with your trims. THEN see what the WB says.

Your car's computer does not know the stoich of the fuel being used. All it knows is what you put in as a stoich point. That is "1" in lambda, and everything else fuel control related revolves around that "1".

The problem with changing the stoich point to come up with the right trims, is then everything else you did prior, is off by that same amount. So, do over. Then check with the wideband, and off course you will be off again. You'll be chasing your tail.

Or I have another way for you to go about it. Set your stoich FIRST. Then tune using your WB. THEN reenable closed loop, and see how far it's off. I'll betcha it's off the same percentage.

I've used plenty of different kinds of widebands. Some off as much as a full point around stoich. I've learned not to trust them so much. I definitely don't use them to tune part throttle. I'll say it again. The computer could care less what the WB says, all it cares about are the narrow bands. And same as Dsteck, I'll go with the NBs over the WBs for closed loop tuning any day.

[Higgs Boson]

Higgs, look at it from a reverse perspective. Get the computer in closed loop, dial in the VE and/or MAF with your trims. THEN see what the WB says.

Your car's computer does not know the stoich of the fuel being used. All it knows is what you put in as a stoich point. That is "1" in lambda, and everything else fuel control related revolves around that "1".

The problem with changing the stoich point to come up with the right trims, is then everything else you did prior, is off by that same amount. So, do over. Then check with the wideband, and off course you will be off again. You'll be chasing your tail.

Or I have another way for you to go about it. Set your stoich FIRST. Then tune using your WB. THEN reenable closed loop, and see how far it's off. I'll betcha it's off the same percentage.

I've used plenty of different kinds of widebands. Some off as much as a full point around stoich. I've learned not to trust them so much. I definitely don't use them to tune part throttle. I'll say it again. The computer could care less what the WB says, all it cares about are the narrow bands. And same as Dsteck, I'll go with the NBs over the WBs for closed loop tuning any day.

I hear you loud and clear. I usually tune the closer to stock stuff with the narrowband trims, I agree they do well there. I had zero luck with them on the more radical combinations.

[nmp0098]

One more thought: you can get a graduated cylinder and rubber stopper for about $10 and know exactly how much ethanol is in your fuel. Everything I find in MI and OH is between E6 and E10. If you don't tune with the correct stoich for the fuel you are using, the cylinder load calculations will be incorrect by a small amount. This will influence the advance lookups among other things, and may cause you to put a slight hack in the advance table. I typically get my fuel from a single station that consistently has E10. It's nightmarish running "E85", which varies from E70 to E85 with no change in labeling. But, I do love the smell of ethanol in the morning whenever I'm running it.