On The Futility Of Narrowband Sensor Based Tuning
#1
Moderator
Thread Starter
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,604
Likes: 0
Received 6 Likes
on
6 Posts
On The Futility Of Narrowband Sensor Based Tuning
I got my HPTuners and wideband to play together through
my "ruggedized interface box" and did some logging & tuning
today. I captured a fairly large number of datapoints for the
wideband and the narrowband O2s.
Everybody says the narrowbands are worthless. Looking at the
scatter-plot you can see why it's true. For the target 12.8:1
AFR my narrowbands (stock manifolds, yet) vary from about 830
to 910mV. Or conversely tuning to the "center" 870mV might be
giving you anywhere from 15:1 to 10:1.
And this is just the variation on one, fully warmed up car over
the space of 10 minutes.
my "ruggedized interface box" and did some logging & tuning
today. I captured a fairly large number of datapoints for the
wideband and the narrowband O2s.
Everybody says the narrowbands are worthless. Looking at the
scatter-plot you can see why it's true. For the target 12.8:1
AFR my narrowbands (stock manifolds, yet) vary from about 830
to 910mV. Or conversely tuning to the "center" 870mV might be
giving you anywhere from 15:1 to 10:1.
And this is just the variation on one, fully warmed up car over
the space of 10 minutes.
#2
Jimmy, the very same findings (with hard numbers) were posted around here several months ago to prove a point to our old forum 'expert' TeamZR1.......you just try to convince him....I couldn't.
#3
TECH Enthusiast
iTrader: (17)
Join Date: Jan 2002
Location: Michigan
Posts: 652
Likes: 0
Received 0 Likes
on
0 Posts
Ahh yes... Those of us "in the know" about the **need** for wideband setups certainly have a one-up on those others that are too "narrow-minded" (pun-intended) to wake-up & understand the worthlessness of stock narrowband O2's at WOT....
Props to my LM-1.
Props to my LM-1.
#4
FormerVendor
iTrader: (1)
Join Date: Feb 2002
Location: Waldorf, MD
Posts: 3,059
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Fetumpsh
Jimmy, the very same findings (with hard numbers) were posted around here several months ago to prove a point to our old forum 'expert' TeamZR1.......you just try to convince him....I couldn't.
You done it now. This thread will reach 20 pages when he sees this post
#5
Moderator
Thread Starter
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,604
Likes: 0
Received 6 Likes
on
6 Posts
The only way I could see the narrowbands working out would
be by collecting a lot of data and working to the center of the
distribution. That, perhaps, might have some meaning. But any
single data point has nearly none.
I don't have too much interest in convincing people of things;
that's for the data to do. Or not. It was useful for me to see,
just how much to trust things for myself. I was hoping for a
better single-car WB:NB correlation, really, but there it is.
be by collecting a lot of data and working to the center of the
distribution. That, perhaps, might have some meaning. But any
single data point has nearly none.
I don't have too much interest in convincing people of things;
that's for the data to do. Or not. It was useful for me to see,
just how much to trust things for myself. I was hoping for a
better single-car WB:NB correlation, really, but there it is.
Trending Topics
#10
Originally Posted by HumpinSS
You done it now. This thread will reach 20 pages when he sees this post
#11
TECH Addict
iTrader: (10)
i think you are missunderstanding us when we say you can tune on a narrow band o2...
the narrow bands are super accurate at 14.7 right? well the ve table represents a perfect 14.7 to be a 0 long term and short term in that cell, right?/ ok, so if you make the car stay in open loop with no power inrichment you can drive around and record normal driving and see how far off the voltage is from 14.7 then you just chang the ve table with your results until you no longer have to make changes.. "this is after you have gotten all your ltrims and short trims near 0 in every cell" so
now, after setting up the whole table to 14.7, then you can reactivate all the other stuff you had to turn off, and if you did it right what ever ratio you want at wot, you just put in the pe vs rpms and it will be right on with what you command..
the narrow bands are super accurate at 14.7 right? well the ve table represents a perfect 14.7 to be a 0 long term and short term in that cell, right?/ ok, so if you make the car stay in open loop with no power inrichment you can drive around and record normal driving and see how far off the voltage is from 14.7 then you just chang the ve table with your results until you no longer have to make changes.. "this is after you have gotten all your ltrims and short trims near 0 in every cell" so
now, after setting up the whole table to 14.7, then you can reactivate all the other stuff you had to turn off, and if you did it right what ever ratio you want at wot, you just put in the pe vs rpms and it will be right on with what you command..
#12
Originally Posted by wait4me
i think you are missunderstanding us when we say you can tune on a narrow band o2...
the narrow bands are super accurate at 14.7 right? well the ve table represents a perfect 14.7 to be a 0 long term and short term in that cell, right?/ ok, so if you make the car stay in open loop with no power inrichment you can drive around and record normal driving and see how far off the voltage is from 14.7 then you just chang the ve table with your results until you no longer have to make changes.. "this is after you have gotten all your ltrims and short trims near 0 in every cell" so
now, after setting up the whole table to 14.7, then you can reactivate all the other stuff you had to turn off, and if you did it right what ever ratio you want at wot, you just put in the pe vs rpms and it will be right on with what you command..
the narrow bands are super accurate at 14.7 right? well the ve table represents a perfect 14.7 to be a 0 long term and short term in that cell, right?/ ok, so if you make the car stay in open loop with no power inrichment you can drive around and record normal driving and see how far off the voltage is from 14.7 then you just chang the ve table with your results until you no longer have to make changes.. "this is after you have gotten all your ltrims and short trims near 0 in every cell" so
now, after setting up the whole table to 14.7, then you can reactivate all the other stuff you had to turn off, and if you did it right what ever ratio you want at wot, you just put in the pe vs rpms and it will be right on with what you command..
#14
Originally Posted by Fetumpsh
Your method is fine for part throttle tuning, but WBO2's must be used for WOT......I 'hope' you are not suggesting running the engine at 14.7 at WOT to get the VE correct.
It should be ok if you shoot for 875-900mv.
#15
TECH Regular
Join Date: Apr 2003
Location: Atlanta, GA
Posts: 413
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Fetumpsh
Your method is fine for part throttle tuning, but WBO2's must be used for WOT......I 'hope' you are not suggesting running the engine at 14.7 at WOT to get the VE correct.
I believe he is saying that if you have your VE Table dialed in correctly (ALL Cells) with minimal STFT/LTFT corrections (As the VCM is very accurate at keeping the A/F @ 14.7). You can enter the desired multiplier in the PE table and your commanded A/F ratio is semi-achieved.
#16
Moderator
Thread Starter
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,604
Likes: 0
Received 6 Likes
on
6 Posts
Suppose you were able to make a diagnostic tune
with the spark pulled back to whatever 14.7:1 fuel
would tolerate without detonation. Presuming you
were cool and low compression enough to stay below
the diesel limit (even run 100 octane for the experiment)
you might be able to get high flow (at least to some
RPM point) data while keeping the narrowbands as
a decision value in a cellwise binary search. This I
think would lead you to a broadly accurate VE table
in only some manageable number of passes through
the drive & retune cycle.
with the spark pulled back to whatever 14.7:1 fuel
would tolerate without detonation. Presuming you
were cool and low compression enough to stay below
the diesel limit (even run 100 octane for the experiment)
you might be able to get high flow (at least to some
RPM point) data while keeping the narrowbands as
a decision value in a cellwise binary search. This I
think would lead you to a broadly accurate VE table
in only some manageable number of passes through
the drive & retune cycle.