O2 sensor display.

[G. Scott Ponton]

    Ok. Mostly I am just a lurker. I am sure some of you have read a few of
my posts. Up til now some have been received with very little "respect".
This seems to be quite common with anyone who wants to add something to the
discussion which doesn't fit the small area of expertise some on this list
have. Realizing this must sound a little harsh let me try to reduce the
impact to a few egos by saying I am quite impressed by the wealth of info,
time and effort spent on several of the projects which are ongoing topics of
this list.
    There seem to be a lot of misconceptions about how an EGO sensor
actually works. Maybe I am wrong and everyone really does understand but
some of the posts I have read lead me to wonder.  EGO sensors (Palladium
type, which includes wide band) don't actually measure the amount of oxygen
in the exhaust. They compare the oxygen content of the exhaust to the oxygen
content of the air around the sensor. This pretty well negates variences in
oxygen content from one moment to the next.
    The combustion of gasoline (ideally, lets not get all tied up in minor
details here.  The correction factors needed for perfection are not within
the scope of the technology we are currently working with.) does NOT require
a perfect A/F ratio to occur. Carburation has proven that for the past 90 or
so years. Provided the A/F ratio remains between 15.5:1 and 11.5:1 most
engines will run well enough for the average person to never know they have
a problem. Also the EGT, for a particular engine, varies by less than 5%
within these ranges. (different engines may have different EGTs for the same
A/F ratio.)
    A "wide band" EGO "switches" ( using the term "switches" to attemp to
simplify the discussion. Center of range = 14.7:1 approx. Where 14.7:1 is
the agreed upon IDEAL A/F ratio for complete combustion of gasoline as a
fuel.) from rich to lean at approximately the same point its standard
counterpart does. Roughly 1/2 of its operational scale. So what is the real
advantage? It gives you more room to see A/F variances? Consider this if you
will. In an IDEAL situation, from the point where gasoline is too lean for
combustion to the point where it is too rich, the center point is not 14.7:1
!!!!!!!!!!!
    At A/F ratios of 16:1 the misfire rate is nearly 70% at 10:1 it is just
reaching the same 70%. At both ends of the scale an EGO measures the same
oxygen content in the exhaust right? Wrong ............... the oxygen
content in the exhaust of an engine which is too rich is less than one which
is too lean at the same rate of misfire. An engine which misfires by being
too lean dumps an entire load of unburned oxygen into the exhaust. Where as
a engine which is too rich burns some part of the fuel leaving less oxygen
in the exhaust. (Naturely if it is so extreme in either direction as to not
burn at all this changes as the engine becomes an air pump.) So what good it
an EGO sensor when it comes to tuning an engine?  None! It is a tuning aid
intended to reduce emissions. Because the "air" we breath isn't all oxygen
we have to make some allowances for other emissions created by its
combustion.
    With respect to an EGO sensor I have spent many hours testing and
playing with this particular sensor. I have run temp correction and
correlated the voltage to a A/F ratio. This was done by first calibrating a
MAF and an injector set. By using a flow bench then calculating the air
density, the frequency of the MAF was charted. Using an injector flow test
rig allowed the amount of fuel to be accurately calculated. By knowing the
injector pulse width, fuel pressure, specific gravity of the fuel, and the
air density entering the engine it is fairly easy to calculate the A/F
ratio. Now some interesting things were learned with this setup.

1. EGT has a minor effect on the output of the EGO sensor (this next part is
VERY important) within the NORMAL range. This doesn't include extremely lean
and/or rich A/F mixtures. Both of which affect more than just the EGT.
2. A "wide band EGO is easier to read simply because its operating
parameters are spread across a wider range. It still functions on the same
principle as a "normal" sensor (Read palladium here).
3. The sensor does NOT (once again read palladium) switch back and forth
between rich and lean. This is a function of the ECM/PCM.
4. although .5 volts is pretty close to stoich the voltage values are NOT
linear. .01 volt is not the same amount lean as .99 is rich.

    There are sensors which "switch" back and forth. These are a different
composition and are mostly used by the Japanese at the moment. Their
measurement range is very narrow. Less than 1 point change in A/F ratio.

    I hope this will help some on the list. I also expect there will be a
few souls who will take exception to what is written here. So be it. I will
be more than willing to elaborate in much greater detail on the subject of
gasoline combustion in an internal combustion engine, but the post is liable
to be the size of a book in order to "qualify" all the variables which
effect real world combustion. Also many of the variables have a negotiable
effect on this discussion.

Scott