Thanks a lot, TX...

I've got the maf-calibration.cfg (scanner setup/config) file here... and FWIW, my Dynamic and MAF airflow are almost spot on. I've got very-very small variances, but nothing I'm losing sleep over.

I did "half" of the SD tune - I got my VE dialed in real well (as well as I could 14.7:1 with the narrowbands), but when I put the MAF back into the equasion, my LTFT's are now up between +3 and +6. I 'could' ditch the MAF, but that requires some rewiring and installing of an external IAT sensor - I might as well just leave it in there.

Are you guys working with recalibrating the MAF from "scratch".... per se... ? Can I get away with just scaling things to bring my LTFT's back inline with what my VE would otherwise display on it's own?

 

well the theory goes as, if you get the VE tuned well via wideband in open loop, then once closed loop is reactivated, the trimming should already be minimal, since the VE produces as commanded. Now I can see this being a problem since O2s are pretty unreliable, but the trimming i saw once i got back into closed loop seemed way too abnormal. i would like to operate in closed loop since there's the benefit of AFR correction, but you see, you can't really tune to 18:1 AFR in closed loop because of trimming. no matter what you tune the VE to, the AFR will be brought back to what the O2s see as stoich AFR via fuel trims.

I tune in open loop SD that way there's no trimiming or MAF to bias the VE and I deal with ONLY the VE table and commanded AFR. That's how you can ensure that your VE is 100% correct (providing your wideband is not giving false values) I'm starting to feel that even if my VE is dialed in perfect with a wideband, the O2s might not agree cuz they suck and could trim however they want according to their calibration.

I think there's alot of confusion about what the VE is actually used for. We don't dial in a VE table to a certain AFR, we dial in the VE as a reference for any AFR, and use one specific AFR to map out the table. Meaning, if your VE is good, it doesn't matter what AFR you want, all you have to do is command and what you ask is what you get. In closed loop you command 14.628, in open loop it's your choice to command what you will in the OLFA table, in PE mode it's also your choice to command. Closed loop is special though since if the VE is not correct, it will adjust to get you back to stoich commanded AFR. But running closed loop is the exact same as all your values being 1.00 in your OLFA table, with the added benefit of 02 monitoring and correction since due to weather and other conditions the motor may not always stay at a particular AFR.

The deeper this hole goes, the more i start to think that even a wideband sensor could be logging false data, and surely either the 02s calibration or operation is biased as well. When you are in open loop, the user is the correction factor as there is no 02 input whatsoever. This means that I am "God" in open loop... Now when i say i want 14.628 at all the areas I've been whining about, I don't get those AFRs and even when I adjust to try and get those AFRs, I still don't. I adjust and adjust and as my VE starts to unsmooth itself, so does my driveability, and I still never appease the rich condition. The thing that is killing me is that If I'm asking for 13.0, I can get the VE to give it to me in its entirety. And since my VE produces 13.0 when commanded of it, it leads me to believe my VE is tuned well. Then when i ask/command stoich AFR of it, it almost does what i want, with the exception of stay rich down low.

So this is the problem I am getting at... something in my tune is preventing the stabilization of AFR at idle/low rpms it seems no matter what fueling changes I make, it just has to run rich. And since i'm the only correction factor, and the way i correct every other cell works just fine, but not on these, something else has to be to blame. What it is I have no idea, but it's not dependent on closed loop settings since b/c it does this is open loop too. Am I making a clearer picture?

 

Sort of. The only way we are going to be able to tune idle (and cruise) is with an exhaust gas analyzer (specifically for CO).

 

I think you'd be ok with doing some scaling from this point on, the point of doing all that math and averaging is to get the MAF calibration as close as possible to the motor and to realign itself to the flow of the MAF housing itself. It seems due to my other discussion that we can't exactly trust our O2s so perhaps some scaling will be necessary to counteract retarded O2 corrections/fuel trimming. The motor's AFR output seems to also be affected by weather so in a new season you might have to play with the MAF again if your trims swing out too much. You can scale MAF curves from the very beginning and skip everything I've suggested, but it just wouldn't be completely consistent across all MAF frequencies. And yes you are building the MAF from "scratch." Your "scratch" data is the dynamic airflow you log to align the MAF calibration with.

In a perfect world, you would tune the VE and command/produce 14.628 out of it. Then you would activate closed loop and your trims would be zero. and then you would calibrate your MAF to report the same airflow that the VE expects/uses/needs that way your trims would still remain at zero. You can see that if the MAF and VE table are not in conflict with each other, then there's no reason to add or subtract fuel, supposing that the VE table has been tuned 100% correctly. (The fuel calculation is a hybrid reference of the MAF readings and VE to my understanding)

But this is not a perfect world, and O2s suck at being consistent, and you'll never have your MAF 100% aligned to dynamic airflow, and the weather will not stay PERFECT. So what we've learned from all this is tuning sucks... I think I've come to the conclusion that O2s are really really really stupid, and you just have to do the best you can with them... wideband is much more accurate, but it still isn't 100% perfect either...

 

that would be mucho nice... I think i'll agree with you the AFR is too easily falsified, manipulated and confused by wideband or narrowband sensors... I guess we just got to do the best we can...

Or just go back to stock...

So is all this STICKY worthy yet??? At least we got one way to calibrate MAF table knocked down (if that even happened)

 

Seriously, we need to use a CO gas analyzer. Any O2 sensor will be wrong. You lean your car out until you hit 1% CO and you are pretty much dead-on 14.7 to 1. It is the only way to tune for a cam. Just nobody figured it out yet, or bother to apply it to the LS1 for this problem. Or the one or two pro tuners didn't want anybody to know. People have been doing this on carbed vehicles and motorcycles for years. I even have a CO analyzer picked out.

 

pricey much? If not too bad I want in... And how exactly would we be able to log it in vs, the VE table?

 

It's $169.99. Gunson Digital Gastester. Pretty good deal, considering. There is no way I know of right now to log this vs. VE. It shouldn't matter. Basically (I think) you could tune idle using this thing and your narrowband O2 tables, and then apply the same values to most of your low airflow modes. The real trick should be if you can get consistent results using a lower switching point, which I think you can. Mine never switched right at idle after my cam no matter what I did, once I get the switch point right they should work fine. Then it will just be a matter of watching the knock sensors (and using our ears) to check to make sure you don't get any detonation. If this was something you could tune vs. VE or MAP, then the O2 sensors would not be laid out vs. airflow mode.

 

soooo exactly how does this work? I thought you said lean her out till you read the right CO content. (1%) So how do you use this determine new O2 switch points?

 

Well...you have to keep adjusting them with lower values (kind of liek jetting a carb) until you get the right CO (1%) value.

 

never touched carbed before... seems interesting though, if it works shout out about it!

 

I can tune the hell out of a carb. By ear too. I tuned my old 305 El Camino with a 600cfm carb, and full-length headers using only my ears, a tach, and a vacuum gauge. The guy I sold it to said he got 20 mpg. And it took me through the 1/4 mile in about 16 seconds with a 2:73 rear, stock heads and cam. In fact, I kicked CarCrafts butt, because they did a similar build-up and couldn't touch that. Not too shabby.

 

This is how you know when you are on to something...when none of the experts chime in...that always happens...

 

Well I certainly am no expert, but the proclaimed "experts" haven't chimed in to anything I had questions about lately. Maybe no one likes me... oh well

BTW, the weather has made crap of my tune...

 

Are you still using proportional idle? Ever since I disabled that it hasn't bothered me but a few %.

 

I've had idle proportional tables off for awhile.

here is what I'm doing now, I'm tuning in SD Mode, Open Loop, trying to get my VE working again. It seems the cold weather pushed me rich all over the place... This doesn't explain the trimming i saw earlier in the thread, but my car is pissing me off cuz it's behaving like it's been detuned. When i originally did all me tuning, it was in about 60-70 degree weather and now it's in the low 40s. What wierd now though, is the same richness I was concerned about earlier is now showing its face... When i was tuning the VE in SD/Open Loop (commanding 13.0) before I did not have this problem, which is why i was so stumped. But now it seems to be consistently running rich at low rpm/idle/low load at any commanded AFR. Now I'm really pissed...

 

Are you sure you didn't throw in a tune with a lower low octane spark table or something?

 

positive i didn't, if anything I've added timing... but I dropped in the exact tune from that warm sunny day (the one that I dialed in perfect 13.0 in open loop/SD) since I've been happiest with that one up to this point, but low and behold, it's got idle surge, low rpm bucking and all this crap now and running rich over the whole table... This exact tune before was PERFECT It's like the weather effectively "scaled" the VE upwards to create rich condition, the AFR was all inline with every cell, just at 12.0-11.8 So I've just been retuning the VE (while breaking in a new clutch) and hunting more of these driveability issues again... When i reactivated closed loop with that Perfect tune a couple weeks ago, the driveability was still great, just the negative trims at idle and low rpm pissed me off... And now it's doing the rich at idle/low rpm thing even when commanding 13.0! Round and round we go!!!!! wheeeeeeeeeeeeeeeeee...

 

That's really odd. I have never had my trims swing that far that it made a difference in SD Closed Loop. Open loop requires a lot of attention though. All it is doing is relying on the fuel values it has. There is no attempt to find the proper fuel ratio. That's why I stay away from open loop tuning as much as possible. Open loop is incredibly more unpredictable in weather changes than Closed Loop SD.

 

I have the same problems - computer thinks I am waaay lean at idle.

G5X3 on 114. I am convinced it is the overlap throwing air into the exhaust. I kept adding fuel as suggested by most, and the car was PIGGGGGGG!! Rich.

I am playing with it a bit, but I think it is something we will just have to live with.