• The site migration is complete! Hopefully everything transferred properly from the multiple decades old software we were using before. If you notice any issues please let me know, thanks! Also, I'm still working on things like chatbox, etc so hopefully those will be working in the next week or two.

Why LTFT, when STFT should be quicker

IndeedSS

SS-DD Level Member
After installing my IC and dropping from a 3.25 to a 2.8 pulley, I've been looking at the tuning aspect more closely.

I've seen the write-ups where tuning requires many long drives while trying to get the LTFT as close to zero as possible.

Then someone suggests using STFT+LTFT which allows a tune to be dailed in quicker. But if the LTFT's haven't learned then would that skew the tuning some?

And now to the point. Why not just disable the LTFT learning and go solely off of STFT. Once the STFT's are close to zero, and the LTFT's enabled, shouldn't the LTFT's be on the money?
 


LTFT is a long term average to keep things on par. STFT is quick instant changes which are always going to vary whether tune is perfect or not. The pcm wants to know that moving negative will lean it out and going positive will add fuel so it will constantly check this until it is satisfied. LTFT keeps everything close over time and isnt going to account for a quick change which is good. The pcm also relies on the LTFT numbers for other factors such as a referrence for PE fueling, if it is + or - for a long time then so is PE as it reverts back to this. If the car has a vacuum leak or other problem causing things to be off and you go WOT then the car's fueling will be off and possibly cause damage. Tuning with just STFT would be near impossible since it moves around so much whether the the tune is on the money or not. If it is WAY off to start with then it will help get you on the right path sooner, but once the trims are close to zero it is always going to hover around a little bit because there are too many varying factors such as temp, humidity, fuel variations, etc and the pcm always wants to see 02 crosscounts below and above 450mv so will do what it has to so it can constantly make that happen.
 
So using the LTFT+STFT custom PID is really the best method? As I understood it, the only reason the Wbody group didn't use STFT was there wasn't a way to disable LTFT learning? Of course we can do it with the HPT. I didn't have a PT long enough to know if they do or not.
 
I disable fuel trim learning and reset the fuel trims the moment I start the scan. With that, the LTFT stay at zero and I can tune straight from the STFT average. It has worked flawlessly for me so far, and saved a lot of time waiting for fuel trims to learn.
 
By dissabling the LTFT yes it will be faster to get things in line but once LTFT is at zero or as close as possible remember that STFT is always going to hunt around a little bit and is never going to sit at zero constantly unless in open loop
 
With the combo of having a cam and dropping from a 3.25 to a 2.8 pulley the trims are both hunting a little even at idle. Good news is NO KR and the LTFTs are locking at 0 when at WOT. I want to smooth the idle just a bit. It may be a nice experiement. I always can fall back on the current tune, if things don't go as planned.

May get around to throwing the WB on this weekend.
 


Back
Top