Ve table tuning opinion
Ve table tuning opinion Curious as to what everyone uses,
Add vs multiply by % when copying your LTFT histogram and pasting into your VE table? Ive read both and not sure which is the best way. Thanks
Add vs multiply by % when copying your LTFT histogram and pasting into your VE table? Ive read both and not sure which is the best way. Thanks
TECH Enthusiast
Joined: Apr 2008
Posts: 654
Likes: 0
From: Dickinson, Tx
Whatever you do throw out any readings if you were heat soaked. I try to only street tune my car when it's early in the morning or late at night. If you do it in the morning then later when it's hot out you'll be chasing your tail unless you change the bias table to lean more towards ECT than IAT.
I do half unless it's below 5%, then I do by percent. But I lean towards the leaner side on the lower tables for part throttle and fatter on the WOT tables, 95kpa and above. Meaning I try to be 2 to 3+ for part throttle and -2 to -3 for WOT.
EDIT: I do not use LTFTs though. I strictly use my Wideband error %.
I do half unless it's below 5%, then I do by percent. But I lean towards the leaner side on the lower tables for part throttle and fatter on the WOT tables, 95kpa and above. Meaning I try to be 2 to 3+ for part throttle and -2 to -3 for WOT.
EDIT: I do not use LTFTs though. I strictly use my Wideband error %.
Thanks for the reply Daniel. Im about to throw the LTFT stuff out myself. My car is a track only car so I only make 1/8th mile passes. I feel like at the least I need to be tuning with STFT and LTFT disabled. I do have a wideband in the car so it would probably be best to just use wideband error%.
Trending Topics
LS1 Tech Stories
The Best V8 Stories One Small Block at Time
6 Common C5 Corvette Failures and What's Involved In Repairing Them
Pouria Savadkouei
Retro Modern Bandit Pontiac Trans AM Comes With Burt Reynolds' Autograph
Verdad Gallardo
Top 10 Greatest Cadillac V Series Performance Models Ever, Ranked
Pouria Savadkouei
Top 10 Most Powerful Chevy Trucks Ever Made!
Hennessey's New Supercharged Silverado ZR2 Has 700 HP
Verdad Gallardo
Coachbuilt N2A Anteros Is an LS2-Powered C6 Corvette In Italian Clothes
Verdad Gallardo
Awesome K5 Blazer Restomod Comes With C7 Corvette Power
Verdad Gallardo
10 Camaros You Should Never Buy
10 LS Engine Myths That Refuse to Die
Verdad Gallardo
My opinion is that if you believe in the paste-histogram
fairy then you have no business tuning.
There's too much other stuff embedded in the LTFTs
especially if you're still stock on the trim cell boundary
settings. You can't do a good job tuning the highly
granular VE table when your LTFT covers a RPM span
from 2500-6000RPM. Fuel pressure error, O2 sensor
aging and heat issues, history, all that kind of thing
which is "external" garbage will get pulled into your
VE table if you allow this kind of unthinking "feedback".
If I were of a mind to trust statistics, I'd turn off the
LTFTs and use the STFTs, which are much quicker
to come around as you change load-point. Still there
is a lot of residue from transitions (start at zero) that
you really can't make go away.
This is why I still prefer to do it the hard way - pick
sections of the log where I know what's going on, that
the data isn't garbage, pull the stuff that's worth acting
on into Excel where I can sort and sift it as I please,
paste updates into the table and interpolate between
the new, believed-good data points where no new data
is being put. Some of it you have to eyeball extrapolate
(the family of curves and response surface views are
nice for context / sanity).
But if that all sounds unappealing, feel free to cut twice,
measure once.
fairy then you have no business tuning.
There's too much other stuff embedded in the LTFTs
especially if you're still stock on the trim cell boundary
settings. You can't do a good job tuning the highly
granular VE table when your LTFT covers a RPM span
from 2500-6000RPM. Fuel pressure error, O2 sensor
aging and heat issues, history, all that kind of thing
which is "external" garbage will get pulled into your
VE table if you allow this kind of unthinking "feedback".
If I were of a mind to trust statistics, I'd turn off the
LTFTs and use the STFTs, which are much quicker
to come around as you change load-point. Still there
is a lot of residue from transitions (start at zero) that
you really can't make go away.
This is why I still prefer to do it the hard way - pick
sections of the log where I know what's going on, that
the data isn't garbage, pull the stuff that's worth acting
on into Excel where I can sort and sift it as I please,
paste updates into the table and interpolate between
the new, believed-good data points where no new data
is being put. Some of it you have to eyeball extrapolate
(the family of curves and response surface views are
nice for context / sanity).
But if that all sounds unappealing, feel free to cut twice,
measure once.
How are you able to look at the histogram and tell what is and isn't good data?
In cruise cells I can get a few thousand count, while in some higher rpm higher load cells I only get a little bit less than 100. That's why when doing VE tuning it's good to get a nice long drive with multiple pulls of the same speed and throttle. That way you get more data.
I usually only count a cell to be true if the count is at least 25, sometimes even 40. If it's under 10 it's garbage data.
I prefer to use Lambda with a wideband...
But if I absolutely have to I use STFT's with a cell count of 30.
But if I absolutely have to I use STFT's with a cell count of 30.
Well, I prefer to operate in LAMBDA... 14.7 is AFR for a Gas Stoich equation...
Lambda always equals 1
E85 Lambda = 1.0
C16 Lambda = 1.0
Gasoline 91 octane = 1.0
If im using STFT I try to error on the negative side of zero... about -3.3 to -5.5% to avoid added fuel when entering PE.
But again I prefer to log with a wideband in Lambda
Lambda always equals 1
E85 Lambda = 1.0
C16 Lambda = 1.0
Gasoline 91 octane = 1.0
If im using STFT I try to error on the negative side of zero... about -3.3 to -5.5% to avoid added fuel when entering PE.
But again I prefer to log with a wideband in Lambda
The more counts you have in each cell, the more likely it's a good number as it has time to average out. Small amount of data could pick up a random lean/rich condition that isn't really true over a long period of time so that's not good data.
In cruise cells I can get a few thousand count, while in some higher rpm higher load cells I only get a little bit less than 100. That's why when doing VE tuning it's good to get a nice long drive with multiple pulls of the same speed and throttle. That way you get more data.
I usually only count a cell to be true if the count is at least 25, sometimes even 40. If it's under 10 it's garbage data.
In cruise cells I can get a few thousand count, while in some higher rpm higher load cells I only get a little bit less than 100. That's why when doing VE tuning it's good to get a nice long drive with multiple pulls of the same speed and throttle. That way you get more data.
I usually only count a cell to be true if the count is at least 25, sometimes even 40. If it's under 10 it's garbage data.
Well, I prefer to operate in LAMBDA... 14.7 is AFR for a Gas Stoich equation...
Lambda always equals 1
E85 Lambda = 1.0
C16 Lambda = 1.0
Gasoline 91 octane = 1.0
If im using STFT I try to error on the negative side of zero... about -3.3 to -5.5% to avoid added fuel when entering PE.
But again I prefer to log with a wideband in Lambda
Lambda always equals 1
E85 Lambda = 1.0
C16 Lambda = 1.0
Gasoline 91 octane = 1.0
If im using STFT I try to error on the negative side of zero... about -3.3 to -5.5% to avoid added fuel when entering PE.
But again I prefer to log with a wideband in Lambda
Thanks for the reply Daniel. Im about to throw the LTFT stuff out myself. My car is a track only car so I only make 1/8th mile passes. I feel like at the least I need to be tuning with STFT and LTFT disabled. I do have a wideband in the car so it would probably be best to just use wideband error%.
So disable ltft and stft and use afr error
But if you have a wideband, use the AFR error histogram instead of STFTs.
That is unless you want to run open loop, but that wasn't part of the original discussion.