Question on running closed loop

[okie]

Tuned with a properly calibrated NGK wideband w ntk sensor. Mods on GTO are headers and intake for the most part. I very carefully modified the ve table. I very carefully modified MAF while in open loop also. I am new but read up and took my time over a couple of months. Everything looked fantastic.....then I put back into closed loop....and then my
actual AFR tanked in part throttle on the wideband down to 14.2-14.3. It doesnt matter what I do (maf hz) since it appears my 02 sensors are running the show now. The wot afr still looks very good however 12.6 for the most part and then down to 12.4. I guess I am a little dissapointed that I spent so much time just to have my 02 sensor drive my afr down and ltft are not slightly negative.

I have gone over the tune for too long. I have learned alot(not enough apparantly) but I was hoping for a better outcome. Since I believe slightly negative ltft are the objective, I believe you might be better off tuning part throttle closed loop by LTFT vs versus wideband. You might as well pull the wideband and hang it on the wall while tuning part throttle if slightly negative
LTF trims are your goal.

[S2H]

Tuned with a properly calibrated NGK wideband w ntk sensor. Mods on GTO are headers and intake for the most part. I very carefully modified the ve table. I very carefully modified MAF while in open loop also. I am new but read up and took my time over a couple of months. Everything looked fantastic.....then I put back into closed loop....and then my
actual AFR tanked in part throttle on the wideband down to 14.2-14.3. It doesnt matter what I do (maf hz) since it appears my 02 sensors are running the show now. The wot afr still looks very good however 12.6 for the most part and then down to 12.4. I guess I am a little dissapointed that I spent so much time just to have my 02 sensor drive my afr down and ltft are not slightly negative.

I have gone over the tune for too long. I have learned alot(not enough apparantly) but I was hoping for a better outcome. Since I believe slightly negative ltft are the objective, I believe you might be better off tuning part throttle closed loop by LTFT vs versus wideband. You might as well pull the wideband and hang it on the wall while tuning part throttle if slightly negative
LTF trims are your goal.

o2's are very heat range dependent...
get them too hot and they tend to rad funny...
not a whole lot you can do...
either live with it being rich at part throttle...
or run in open loop
you can try to push the 02 switch points around in the tune...but really you cant do a whole lot to them due to some hard coded stuff in the tune(the table will move farther than the hard coded stuff will actually do)
and if you get too far from center you will just be chasing the o2 when it gets hotter or cooler as exhaust temps will push anything that is not 450mv quite far out of range very easily...

please post a log with your tune...
and have the wideband and the narrowband running at the same time...
I'm curious to see just how far off your o2's really are

[eficalibrator]

Do you, by chance, have "E10" gasoline? A change in assumed versus actual stoich points would explain your condition.

[S2H]

Do you, by chance, have "E10" gasoline? A change in assumed versus actual stoich points would explain your condition.

I thought stoich was stoich?
meaning our widebands measured remaining oxygen....
if it reports stoich then it doent matter what fuel we have
gas reports lambda 1.0 at stoich
e85 reports lambda 1.0 at stoich
our wideband will report the value that is represented by out formula...its not telling us our actual AFR...its telling us Lambda 1.0= a certain voltage....
which is why I can tune e85 to lambda 1.0 for part throttle and it will show up as 14.7 on my wideband....even though the actual mixture is going to be 9.7ish

if my narrowbands pushed it inline then my wideband should have read the same
if my narrowbands say lean...then my wideband should say lean until it gets back inline by using fuel trims...

the only time it should be different is like I said...
location of the o2's and a different heat range than stock

[boomer]

I'm having the same issues as you. my part throttle AFRs are always around 14.3-14.5. i have my AFX wideband kit with NTK sensor. since i am one who wants to use closed loop, i gave up on using my wide band to tune part throttle AFRs and decided to trust my narrow bands for only this. from here, this is what i did:

Use my stock narrow bands to tune my LTFTs to about -2% from about 2400 Htz to about 8000 Htz (all while in closed loop). after i have that part of the MAF dialed in, i force open loop and disable fuel trims exc. then i use my wide band to tune my MAF curve 8000 Htz+. when doing this, i log MAF AFR error % and only copy and paste from 8000 Htz+ to equal my commanded PE AFR.

this seems to work just fine. LTFTs are staying right where i want them, and my WOT AFRs have been staying very close to my commanded WOT AFRs as well. i know the G8 PCMs are different from the GTOs but i thought i would try to chip in

[eficalibrator]

Stoich is stoich for a single fuel. But E10 is stoich around 14.2 versus 14.7 for pure gasoline. Depending on what seasonal blend he has at the moment, it may skew the chemistry. I'm just making the point that it's important to know for sure what's in the tank before making assumptions about the source of a 2-3% error since this is easily within the range of possibilities for "pump gas".

Also, the NTK(ECM) widebands will show 14.56 on the display at lambda=1.00 because they're calibrated against Indolene clear, the emissions test fuel. This can account for the first 1% error if the PCM is targetting 14.68 for stoich.

[okie]

By gosh I am running E10 from Quick Trip! I think your on to it! Quick Trip runs a full 10% ethanol also....so I should not be targeting 14.7.

[okie]

Also, the NTK(ECM) widebands will show 14.56 on the display at lambda=1.00 because they're calibrated against Indolene clear, the emissions test fuel. This can account for the first 1% error if the PCM is targetting 14.68 for stoich.


I didnt know that either. Also another good point.

[okie]

Hey, I am excited to know that I am not losing my mind! Thanks for the input!!!!


I was beginning to wonder.

Also, I enjoyed reading your book. Thank you.

[eficalibrator]

Good to hear you're getting somewhere on this.

If you liked the book, you'll love the class(es).

[okie]

I will know for sure tomorrow but I know you hit the nail on the head on this one.
I will never forget this one. It cost me too much of my time. I am sure I would love your classes. I'm also an engineer(petroleum) but work daily modeling engines and natural gas compression on the computer. Right now this is a hobby. Give turbo some cheese and a scratch him behind the ears for me!

[okie]

It was the E10 driving the AFR down to 14.1 in closed loop. Rebuilt ve table today. Adjusted for E10. Scaled the maf sensor. All LTFT are now between 0 & -3. Locked at zero at wot. All part throttle afr is 14.1. 12.6 at wot with no kr. Life is good. Thanks for the help.

[eficalibrator]

I hate to say "I told you so," but...

Congrats on getting the tune sorted out better.

[flea]

Stoich is stoich for a single fuel. But E10 is stoich around 14.2 versus 14.7 for pure gasoline. Depending on what seasonal blend he has at the moment, it may skew the chemistry. I'm just making the point that it's important to know for sure what's in the tank before making assumptions about the source of a 2-3% error since this is easily within the range of possibilities for "pump gas".

Also, the NTK(ECM) widebands will show 14.56 on the display at lambda=1.00 because they're calibrated against Indolene clear, the emissions test fuel. This can account for the first 1% error if the PCM is targetting 14.68 for stoich.

The base AFR table on the 05'-06' GTOs (unlike many other calibrations) indicates the ECM does not have the ability to adjust the commanded AFR based on ethanol concentration. Funny that GM does not seem to think a 4% error (0% EtOH vs 10% EtOH) just from the fuel type is a big deal.

Out of curiosity I put togther a table comparing the two.

[flea]

The other variable that drives me nuts is that every station I've encountered selling E10 states "the fuel may contain up to 10% ethanol"

All this makes it virtually impossible to know precisely what you have to work with from a tuning standpoint.

1. Inherent WB error?
2. How much EtOH is actually in the fuel?
3. LT headers throwing off O2s, but by how much? refer to #1 and #2
4. All this gives me a headache!

[talsayed]

I was going to ask about E10.

Thanks flea.Thank you guys for the explanation,we usually don't use these stuff in our region (E10).

[flea]

I know I'm on a rant .......

Variable #5: Is the response of OEM O2s and WB sensor the same to differing concentrations of EtOH? If not, then how different are they?

[Big Loser2]

So is it safe to ASSume that if you are running a NGK WB and the only fuel you can get "May contain up to 10% ethanol" that an AFR error of 4% with the occasional spike is to be expected? If so this will make me sleep a lot better at night. I would also like to bump just to see if anyone has a response to flea about the OEM O2's.

[zratedGTP]

So this may account for some of my tip-in knock that I get once and a while. Doing a new tune to test right now.

[RWTD]

The other variable that drives me nuts is that every station I've encountered selling E10 states "the fuel may contain up to 10% ethanol"

All this makes it virtually impossible to know precisely what you have to work with from a tuning standpoint.

I'm totally with you! One of my biggest pet peeves about the fuel industry!