NefMoto

heya!
Just a quick question here: Can I use KFLF map to fix some problematic areas, where I need more fuel to reach 14.7:1 AFR? I know that I can do this by underscaling MAF but I would prefer just to fix fueling...

Thanks

Yup. Absolutely.

Question is, what is wrong with your fueling that you need to?

Is it REALLY a fuel system problem, or is your MAF measurement off?

I think it is fueling problem... I changed injectos to bigger ones (from saab 2.0t), FKKVS filled with 1s, KRTKE calculated and TVUB got from datasheet. At idle I get -10% correction and at part throttle it goes to +10% and more. MAF and air intake is stock.

BTW: IS there a way to force ECU into closed loop operation at WOT?

This is a very dangerous way to do things as a small misread from the oxygen sensor (which is a quite cheap item) having a hickup can completely destroy your engine with knock/pinging before the system even reacts to the fact that the engine all of a sudden is knocking heavily so... Don't do this unless you are 100% aware of the risks it might incorporate.

Open loop at WOT is usually never an issue if you have your fuel calibrated properly so thiniking about going closed loop at WOT is usually trying to cover other mistakes in the calibration. :-)

I have wideband sensor so it isn't that cheap... but what i want is to use my o2 sensor to calibrate fueling for wot and then I could again reset everything back to open loop

Yes, set your req lambda to 1 :P

So, to do that i have to set 1s in LAMFA and TABGBTS > FFs? Any more maps that could affect requested lambda? ;)

The manufacturing cost for a wide band sensor is about $15-20 US so... it is actually quite cheap. You can't calculate with the price you pay at the tuner shop :-)

What I mean is that even wide band sensors will transmit odd values if the resolution in time is speedy enough so if you don't filter the signal properly (i.e. use an average reading over a few samples) a single measurement error (that occurs often actually) will give you serious headache if you are not careful. If this wouldn't be the case, all cars would run CL under all driving conditions in stock mode, byt they don't...

Why not fix the problem?

If your fueling was accurate before you installed the injectors, and now it is off since installing them....it's pretty obvious what the problem is.

Why fudge it with KFLF?

Not that I am endorsing forcing closed loop, but I don't think lambda 1 is required for closed loop on wideband ECU's. Wideband ECU's in closed loop adjust lamsoni to lamsbg versus narrowband that target usvk to USR.

Lambda 1 at full load = shit melting

in wideband ecu's it is always closed loop even in wot. try to play with injectors closing times, and fkvvs for idle. kflf doesn't do much at low loads and rpm. i had the same problem when i changed my injectors.

fkkvs for idle repairs
kflf for above 30% load departures

im still of two minds about "departures".

i know we've all agreed that accurate load is (mostly) irrelevant, but i'd still rather ONLY fix fueling system problems with KFLF, and load problems in MAF correction.

nyet's previous discussions have highlighted how KFLF / FKKVS are partially related, and thus the only way to maintain consistency is to pick one fueling repair methodology and exclusively change that map.  were you not to follow this, he believes that you'd run into a problem where changing one map requires changing the other map due to map-to-map co-interactions.  i'm writing this just to clarify;  i believe he is correct and i also think KFLF is the map to be selected.

this is because of the linear relationship we know:  fuel output = krkte * load + tvub       in other words, tvub should be sufficient to solve idle problems and if load is correctly scaled, krkte should handle the primary function as an accelerator pump, thus making WOT sufficiently fueled.  that means, if you use the LTFT centering method, your idle and WOT (open-loop) should be clean enough, leaving only part-throttle corrections necessary as would be done in KFLF...

(just some errata for those interested in the fine-tuning of me7...)    on this subject, i'm working on a matlab / excel process which will take raw log data and return the correctly interpolated KFLF correction table based on a mathematical model.  obviously you'd have to write that table, reflash the car, and again recenter TVUB and KRKTE.  this would be rinse, repeated, until fuel corrections are very low <generally speaking>, as can be seen by minimal STFT corrections during logging.

I was mostly thinking KFKHFM :)

But your point is well taken.

In my humble opinion, if you swap injectors and have fueling problems your issue lies in the function RKTI.

If you change your MAF setup, your issue lies in GGHFM.

I also believe that KFLF should not be touched and that if any other fuel corrections need to be applied it should be via FKKVS.

Paraphrasing the Application of HFM correction map KFKHFM in %GGHFM -
The residual error (lambda deviation from 1) are interpreted as mixture errors and must be compensated by means of the characteristic map FKKVS in %RKTI.

Just sayin' I think 99% of (non idle) fueling deviation issues are on cars with modified intakes ...

Wow, I missed this post. Very cool.

while reading other people's posts, i belive FKKVS might be the map to use instead of KFLF. regardless, the correct log data needs to be taken to match whichever table is used.  (obviously)

Nehalem:
Did you finish this?