need some help with high fuel trims

[Sharkey]

some may remember my post from 6 months ago regarding high fuel trims and setting lean codes under deceleration on my 05 cobalt ss/sc. ive been slowly plugging away trying to get this problem sorted out with little luck. i started a few weeks ago with ve table tuning through disabling the mass air flow sensor and zeroing out the adder tables as per the how to on cobaltss.net. i got most of the afr error taken care of, except a few areas. i later tried again, after logging and retuning i found my fuel trims went way up all over, i ended up flashing in my old tune and was scratching my head as to what i did wrong. fast forward to yesterday, i decide to do some more tuning, took a log with the maf active and tried to adjusting using the long term fuel trim histogram. it made a decent size change, some of the fuel trims moved around, however im still getting high fuel trims under deceleration, up to 19.5% and setting a p0171.

ive been having the p0171 problem for a long time, ive had a couple different cold air intakes, different header and exhaust setups, different injectors, a canned tune, etc since this problem has started happening, never a change. when i purchased my 60lb injectors i purchased a tune, the shop told me we would remote tune till the car was right, then after a few exchanges of tunes left me with a mess of rolling idle and never helped figure the lean code out like they said. ive gotten some of that stuff fixed.

now i know what your thinking, ive got a vacuum leak. thats what ive been after when the car was less modded when it first started happening. ive been through every hose, line, clamp, gasket, ive used the acetylene/propane trick on numerous occasions, cant find a vacuum leak. ive had everything between the air filter and head flange apart, cant find any source of a vacuum leak. to me it would seem like a mechanical problem like a vacuum leak, i just cant find anything. the car had been tweaked by the previous owner, he didnt really know what he was doing, however i drove the car for a year or 2 before this issue started.

im posting my last log, maf is active here, i apologise for it being so long. fuel trims under decel hit 19.5%. i flashed it last night and watched my trims on my scan gauge on my 35km drive to work, by about half way there i already had set the p0171. i believe im making a mess of the tune, the tune in theory should be very rich under high vacuum, at least i think so.

any thoughts or insite would be great, im getting frustrated with it.

[GM COBALT SC SS]

i havent looked at much other than your ve table. but the values are way too low in some areas and high in others. you can see this with the 3d image of the ve table. what you need to do is hit the cells where you see the leans spikes to account for the problem area when you copy, then special paste (multiply and subtract by value) onto the tune file.

what it looks like happened is you only hit a few cells when you first logged and only copied and pasted those values. or, when you copied and pasted, it only changed a few values. it has done that to me before. i copied all the cells and when i pasted it only changed the ones in the first vertical row. i guess it was a glitch that only happened once, but it could have happened to you. just watch out for that.

so pretty much when you ve tune, log as many cells as possible, then copy, then paste special, multiply by/subtract from in the ve table on the tune file. you probably know most of this already, but i just wanted to throw it out there to make sure

edit: oh yeah, and dont forget to smooth your ve table once your finished. you can use the smooth table functions offered or the preferred method is hand smoothing

[Sharkey]

the problem im having is its adjusting the same cells every time, and the same cells are seeing the fuel trim going up to 19.5%. im richening and richening and nothing is changing. doesnt my ve table look kind of odd? i looked at some other tunes from the repository and none of them had values nearly as high as mine on the left side of the table.

im looking at my logs and ve table thinking something has to be wrong here. is it possible ive got a sensor not reading correctly under decel conditions? the cobalt is my first attempt tuning a stock pcm, ive had extensive experience with megasquirt, but thats 100% speed density and a heck of a lot simpler. what i do know for sure is that my ve tables with the megasquirt never looked anything like mine, including the turbo ecotec on megasquirt. what im getting at here is im just really not familiar with is what the data looks like on a properly running lsj.

to those of you that know the lsj well (i know there are enough of you guys here), in my data log does everything looks as you would expect as far as sensor inputs?
-map under decel
-map under boost (2.7" pulley, if it doesnt seem right maybe a leak after the blower eg cracked intake)
-maf under decel
-cylinder air mass (every so often ill notice a spike to 1.62)

another thing to add is ive noticed some random stuff with fuel trims. last night for example i noticed when i stopped at a light my fuel trim at idle went as high as 12%, then slowly worked its way down to its normal +/-2. last week when i filled up with fuel most of my fuel trims moved way rich (-12% range) right away, then within 5 minutes of driving went back to normal.

i was trying to remember when this problem originally stared happening, im thinking it was either when i did my dual pass end plate or when i put my clear image midlength header on. ive had my intake off since doing the dual pass, found nothing amiss there. i have no exhaust leaks, im just wondering about the 02 sensor placement. do you think the o2 placement would have much bearing on it? i could swap it around with the wideband sensor for testing without too much of an issue.

im posting another log, a little shorter. towards the end of the log it shows in traffic and lots of the high trims. sorry for no wideband, cant figure out why its not logging. works when i log with the laptop but not the cable.

[laser_racer]

Well one problem you have is you didn't fail the maf. Under engine diagnostics airflow tab then 0 the maf fail high. Set your p0101, 102, and 103 to mil on first error so maf will be failed all the way.

[Sharkey]

i havent been doing that in this tune, but i was doing that and tuning based on wideband afr error (histogram 6). i wasnt getting anywhere really with that, it was making changes but the error wasnt changing very much under decel (it was in other areas). i ran out of time at that point and enabled the maf and all other tables that i had zeroed. i had read that if your happy with the maf tuning you can tune based on fuel trims without disabling the maf using histogram 3 (ve-long term fuel trims). i had done that once, it made changes to the ve table but i didnt get anywhere, fuel trims in those cells still way high.

did you by chance look at the ve table? what are your thought on it?

[laser_racer]

i havent been doing that in this tune, but i was doing that and tuning based on wideband afr error (histogram 6). i wasnt getting anywhere really with that, it was making changes but the error wasnt changing very much under decel (it was in other areas). i ran out of time at that point and enabled the maf and all other tables that i had zeroed. i had read that if your happy with the maf tuning you can tune based on fuel trims without disabling the maf using histogram 3 (ve-long term fuel trims). i had done that once, it made changes to the ve table but i didnt get anywhere, fuel trims in those cells still way high.

did you by chance look at the ve table? what are your thought on it?

Looks rough but to tell you truth I never even bothered with our ve table. Everytime I decide to work in it I just go inside and do something else lol.

[Sharkey]

thats what every tune ive downloaded looks like. i understand that in reality with the lsj ve tuning shouldnt really be needed. all most do is increase boost and thats all maf tuning. this is why im so confused with whats going on. ive got values in some spots that are 60% richer than stock and it still sets fuel trims of 19.5%.

so for argument sake, saying ve tuning shouldnt really be needed, what are your thoughts on what might be going on? remembering back to when i was tuned by zzp, they said they saw some weird things going on with my maf readings (the tune they sent me to start with was very lean at idle) but they were able to work around it. do you see anything odd? is it possible the maf has an issue?

[GM COBALT SC SS]

it would run okay on purely your maf except that dynamic airflow tab has the high rpm disable set to 4,000 rpm and reenable at 3,900 rpm. so youre gonna have a ve and maf blend up till 4,000 rpm. and since your ve table is so far off, it will skew your trim far off as well

if you want it to run pure maf, which others have had success with, you set the disable to 500 rpm and the reenable to 400 rpm. if you look at the tune, its under the airflow>dynamic> tabs.

hope this helps.

[GM COBALT SC SS]

you can compare your spikes in your ve table to mine here. it is not as smooth as it should be, but im gonna tweak it more when i have time.
finals are next week!

maybe comparing the two can help you some. its on the gms2 3.065 inch pulley at about 100 feet above sea level.

[Sharkey]

so if i were to run a pure maf tune, how will this affect my fuel trims? i was always under the impression that the car was running mostly off the maf and only used the ve table in the event of a maf failure, as well as used the ve table to reference for fuel trims.

i had a look at your ve table, looks much like most of the ve tables ive looked at, and what i started with. ive butchered mine from basicly what you have to try and get the fuel trims in check with no improvement.

[cobaltssoverbooster]

ive found that by watching the fuel trims in correlation with the manifold kpa you can adjust the "inj fr vs kpa vac" table to help with the fuel trims. gets it closer and easier to adjust.

anyone else find this true?

[Sharkey]

had another look at it tonight, looked through my old thread and decided as per laser_racers recommendation back then i decided to start over. i have used his recommended 60lb injector settings as well as changed a bunch of other little stuff around. im going to run maf only.

one question i have is the maf has already been calibrated, i believe it was fairy close. am i going to have to start over on that now that ive changed everything, or is the maf calibration not affected by anything but the maf itself?

if somebody could have a look over this tune as soon as they can id really appreciate it. i just want to make sure i havent made any major mistakes.

[laser_racer]

I would just take and do a log of this tune on your car just to make sure how the maf is looking.

[Sharkey]

ok sounds good. one more question, when tuning maf (using maf frequency vs afr error) should i have my adder tables zeroed like i would when tuning the ve table?

[Sharkey]

loaded the tune in and it ran better right from the start. ive worked most of the error out of the maf table. the maf table changed quite a bit, was rich at idle, and after a few revisions i tried in boost it was lean. aside from one spot in boost, it seems dialed. and it pulls a lot harder than it used to.

im going to head out for another drive after correcting the maf table once more and if everything looks alright ill move on. now that im running maf only will i need to do ve tuning to get the fuel trims in check? or do they now base off the maf instead of ve? just wondering because you say you dont ve tune.

[laser_racer]

Well let me put it this way to tune for afr error you only do that during wot pulls when in pe mode. Otherwise you use fuel trims for idle, cruise, and partial throttle acceleration. The fuel trims will be the error in the table. Also the short pulse adder is used for idling the car so leave that table be. You get what I am saying? If not let's say anything under 6k Hz you use fuel trims to adjust and over that use afr error.

[GM COBALT SC SS]

ive found that by watching the fuel trims in correlation with the manifold kpa you can adjust the "inj fr vs kpa vac" table to help with the fuel trims. gets it closer and easier to adjust.

anyone else find this true?

have not tried this myself. but i might have to look into doing this when i put on the 60s.

so you make a histo with the inj fr vs kpa vac and log fuel trims on it??

[cobaltssoverbooster]

Yea but to adjust I had to include the multiplier do I was adjusting the totals

[Sharkey]

Well let me put it this way to tune for afr error you only do that during wot pulls when in pe mode. Otherwise you use fuel trims for idle, cruise, and partial throttle acceleration. The fuel trims will be the error in the table. Also the short pulse adder is used for idling the car so leave that table be. You get what I am saying? If not let's say anything under 6k Hz you use fuel trims to adjust and over that use afr error.

ok i get what your saying. so far all ive been doing is tuning the afr error in the maf table. seems to work ok as last i looked the fuel trims seemed to be decent, except at idle, they were in the -12 range. car was running great so i decided id put things back on (dfco and 3 other adder tables) and now its rich in boost. i guess i should have tuned maf error with those tables on.

ill post a log tomorrow.

[laser_racer]

Make a log of idle, cruise, partial accel, and wot. Do a log with all of those things in one log and make it decently long or long. Afterwards post the log and the tune you were running at that time so we can see how its going.