Stock O2s trimming lean in closed loop
Thread Starter
Teching In
Joined: Jan 2006
Posts: 44
Likes: 0
Stock O2s trimming lean in closed loop **I'm using HP Tuners 2.2.0.0** Well... I have tuned my VE and MAF and my car runs better than ever before.... except when I enable closed loop again, then it bucks a little while cruising and really annoys me. It seems like the stock 02s want to trim it to about 15.1:1 and I have verified this with my wideband while cruising. I think it's because of the longtube headers,my mustang did this too. Is there any way to offset what the stock 02s see as stoich?
I already tried adjusting the rich/lean switching voltage and also the stoichiometric reference from 14.68 to 14.22... neither worked.
I already tried adjusting the rich/lean switching voltage and also the stoichiometric reference from 14.68 to 14.22... neither worked.
The rich/lean switching voltage does move the AFR
target, you may need to move it below 300mV or
above 700mV before you see much effect. You have
to mess with the cells that pertain to the "airflow
mode" you are in at idle / cruise.
The switch thresholds are the only thing that will
change the closed loop mixture permanently. Any
other air or fuel change will be trimmed out. But the
threshold is the trim-to.
target, you may need to move it below 300mV or
above 700mV before you see much effect. You have
to mess with the cells that pertain to the "airflow
mode" you are in at idle / cruise.
The switch thresholds are the only thing that will
change the closed loop mixture permanently. Any
other air or fuel change will be trimmed out. But the
threshold is the trim-to.
Lower mV makes it trim to a leaner mixture. Look at
this pic and you can see that it will take a pretty low
mV to break away from stoich on the lean side and
pretty high mV to bend it rich. Also that there will be
a fair bit of environmental / history variation "push"
on the closed loop mixture when you get away from
the straight-vertical (especially on the rich end,
see that 750mV could be anything from still-stoich
(500, 750C EGT) to 13.5:1 if it's real hot.
If you are seeing lean readings off the wideband you
have a few other things to check. One, let it idle for
a while and then look at average NB sensor voltage.
Ought to be whatever you set in the appropriate cell.
That's the whole idea of closed loop. Now, some tune
files have a 450mV or 500mV switchpoint and some
have 300-350mV at low gas flows. Maybe you are
skewed lean because of something like that. The way
to judge whether closed loop is in control, is to look
at voltage vs target.
Then you have the issues of exhaust gas / sensor
temperature (often with headers cooling the gas),
and air in the exhaust (cam overlap, exhaust leaks,
misfiring) skewing the relation of voltage to mixture.
Most of these however skew the voltage low and
would result in trimming to a voltage-matches-
target, but rich result.
Working on a blown 6.0 truck I moved the idle AFR
from about 12:1, up to about 14:1 by pushing the
thresholds down to about 250mV. Don't take that chart
as gospel, only as illustration. Chart says that nothing
happens until 150mV but my experience disagrees.
If you use bidirectional controls to open the loop and
command mixture, and just record the airflow, NBO2
volts and WBO2 reading you can map out the present
O2 voltage that corresponds to the mixture you like.
Realizing that WBO2 meters have their own error
sources; calibration drift, ground offsets if you're
reading it into the scan cable I/Os, sensitivity to any
air in the exhaust (esp. if you are using a tailpipe
probe or your WBO2 bung happens to be close to a
cutout or less-than-airtight band clamp). Eliminating
sources of error & variability before you tune, may cut
down on the cut-and-try.
this pic and you can see that it will take a pretty low
mV to break away from stoich on the lean side and
pretty high mV to bend it rich. Also that there will be
a fair bit of environmental / history variation "push"
on the closed loop mixture when you get away from
the straight-vertical (especially on the rich end,
see that 750mV could be anything from still-stoich
(500, 750C EGT) to 13.5:1 if it's real hot.
If you are seeing lean readings off the wideband you
have a few other things to check. One, let it idle for
a while and then look at average NB sensor voltage.
Ought to be whatever you set in the appropriate cell.
That's the whole idea of closed loop. Now, some tune
files have a 450mV or 500mV switchpoint and some
have 300-350mV at low gas flows. Maybe you are
skewed lean because of something like that. The way
to judge whether closed loop is in control, is to look
at voltage vs target.
Then you have the issues of exhaust gas / sensor
temperature (often with headers cooling the gas),
and air in the exhaust (cam overlap, exhaust leaks,
misfiring) skewing the relation of voltage to mixture.
Most of these however skew the voltage low and
would result in trimming to a voltage-matches-
target, but rich result.
Working on a blown 6.0 truck I moved the idle AFR
from about 12:1, up to about 14:1 by pushing the
thresholds down to about 250mV. Don't take that chart
as gospel, only as illustration. Chart says that nothing
happens until 150mV but my experience disagrees.
If you use bidirectional controls to open the loop and
command mixture, and just record the airflow, NBO2
volts and WBO2 reading you can map out the present
O2 voltage that corresponds to the mixture you like.
Realizing that WBO2 meters have their own error
sources; calibration drift, ground offsets if you're
reading it into the scan cable I/Os, sensitivity to any
air in the exhaust (esp. if you are using a tailpipe
probe or your WBO2 bung happens to be close to a
cutout or less-than-airtight band clamp). Eliminating
sources of error & variability before you tune, may cut
down on the cut-and-try.
