anybody have a tried and true method to recalculate o2 switch points??
10-05-2005, 09:42 AM
#2
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
I would be suspicious of any direct matching, the
wideband samples and its output is often smoothed
making the WBO2 output you look at, time-lagged
relative to the NBO2 voltage which is always bugging
around at closed loop. If you are at idle and things
are moving slow, then maybe. But with so many
electrical and chemical lags & averagings & so on,
it's probably not going to be a very clean matchup.
Plus, the electrical basis is just not the same. The
NBO2s are looked at for threshold crossings and a
time-averaged 1/0 accumulation is done by the PCM,
changing the voltage pivot will make not much of a
change in the PWM-averaged (if you will) "output
voltage" when the edges are fairly sharp. Plus what
you see out of a logger is highly decimated, 1/8 of the
real acquisitions are exported if you're running 10Hz
frame rate.
I think you might be better off with an oscilloscope
and look at the real, real-time activity of the NBO2s
rather than the datalog representation. You want any
pivot voltage to be within the range of switching swing.
Depending on the "proportional" dithering scale and
time constants, and the sensor temp, the swing can
be narrow or wide, near-linear or very digital-looking.
If the NBO2 output does not reliably cross the switchpoint
then you will be set up for mis-trimming. This envelope of
response, which you might think of as an "eye diagram",
needs the pivot voltage to lie inside it for good switch
data integrity. Acquiring the extents of the eye would
be tedious for sure, not to mention the difficulty of
exercising all environmental / driving-modal corners.
So the short answer (which of course comes after sitting
through the long-winded one) is, you probably don't have
a lot of room to move it without consequences (seen a
0.3V minimum suggested) and you may be chasing after
artifacts more than actual discrepancies if you are trying
to sync up NBO2 and WBO2 instruments.
wideband samples and its output is often smoothed
making the WBO2 output you look at, time-lagged
relative to the NBO2 voltage which is always bugging
around at closed loop. If you are at idle and things
are moving slow, then maybe. But with so many
electrical and chemical lags & averagings & so on,
it's probably not going to be a very clean matchup.
Plus, the electrical basis is just not the same. The
NBO2s are looked at for threshold crossings and a
time-averaged 1/0 accumulation is done by the PCM,
changing the voltage pivot will make not much of a
change in the PWM-averaged (if you will) "output
voltage" when the edges are fairly sharp. Plus what
you see out of a logger is highly decimated, 1/8 of the
real acquisitions are exported if you're running 10Hz
frame rate.
I think you might be better off with an oscilloscope
and look at the real, real-time activity of the NBO2s
rather than the datalog representation. You want any
pivot voltage to be within the range of switching swing.
Depending on the "proportional" dithering scale and
time constants, and the sensor temp, the swing can
be narrow or wide, near-linear or very digital-looking.
If the NBO2 output does not reliably cross the switchpoint
then you will be set up for mis-trimming. This envelope of
response, which you might think of as an "eye diagram",
needs the pivot voltage to lie inside it for good switch
data integrity. Acquiring the extents of the eye would
be tedious for sure, not to mention the difficulty of
exercising all environmental / driving-modal corners.
So the short answer (which of course comes after sitting
through the long-winded one) is, you probably don't have
a lot of room to move it without consequences (seen a
0.3V minimum suggested) and you may be chasing after
artifacts more than actual discrepancies if you are trying
to sync up NBO2 and WBO2 instruments.
10-05-2005, 09:53 AM
#3
well...I thought that IF I get the VE correct to 14.63
then maybe(immediately after) I look at the o2's...and get them where I dont have too far of a deviation from the wideband...
example being my o'2 show stoich and my wideband shows lean..so I move the switch point until its similar in results??
and I usually log only the o2's and wb so I can get the fastest possible frame rate
its kind of like using afr error %..only Ihave to do the comparison by eye???
either way...
there has to be a method to correct the o2's for when you put on longtubes and it reads differently because of location
then maybe(immediately after) I look at the o2's...and get them where I dont have too far of a deviation from the wideband...
example being my o'2 show stoich and my wideband shows lean..so I move the switch point until its similar in results??
and I usually log only the o2's and wb so I can get the fastest possible frame rate
its kind of like using afr error %..only Ihave to do the comparison by eye???
either way...
there has to be a method to correct the o2's for when you put on longtubes and it reads differently because of location
10-18-2005, 09:02 AM
#4
Sound, did you ever figure out how to set the switchpoints for longtubes? Or perhaps someone has a tried and true table for those of us with headers. These switchpoints make a bigger difference than I had originally though. Moving them 100 mv off stoich results in a -5 LTFT difference on the low end. That leaves signifigant room for error if these are not calibrated for LTs...
10-18-2005, 10:08 AM
#5
I have had success with 2 cars.... I have been working on a spreadaheet for everybody to use..(LTFT/STFT based)
still gotta test it out again...
not going to release a bunk product....
but I have figured out something of sorts.....
just need some time...and I havent had any lately with a new little one around here
still gotta test it out again...
not going to release a bunk product....
but I have figured out something of sorts.....
just need some time...and I havent had any lately with a new little one around here
Last edited by soundengineer; 10-18-2005 at 10:14 AM.
Trending Topics
10-18-2005, 12:43 PM
#9
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
It would be -very- interesting to see wideband reading
vs a series of O2 switchpoints (and maybe EGT but not
so many people have that handy). The O2 curves I have
all seem kinda cartoonish, and maybe there is some
funny stuff in the real world at lower EGTs that should
be better represented.
vs a series of O2 switchpoints (and maybe EGT but not
so many people have that handy). The O2 curves I have
all seem kinda cartoonish, and maybe there is some
funny stuff in the real world at lower EGTs that should
be better represented.
