Re: O2 sensor Primer

[Mitch Loescher <loescher@neosoft.com>]

>	All oxygen sensors are alike basically.  They compare the oxygen
>content of the exhaust gas to a reference gas housed in a small chamber
>inside the sensor.  The difference between the two oxygen levels will
>cause a small voltage to develop between the two gases.  The reference gas
>and the general construction is designed so that the voltage is 0.5v when
>the engine is burning at stoichiometry(approx 14.7:1.) A richer mixture is
>indicated by a greater voltage while a leaner mixture is indicated by a
>lower voltage.  O2 sensors are extremely accurate at or near stoich, but
>they loose their accuracy as the mixture deviates from stoich by more than
>10%.  There are many factors that affect their accuracy, but that is a
>discussion that is far more advanced than I can lead. 

They are constructed of a small thimble shaped piece of Zirconium Dioxide
coated on the inside and outside with platinum.  The outside of the thimble
sits in the exhaust stream the inside sees basically air in the reference
side.  The platinum sides are electrodes with one on ground, and one going
to the computer.  Any more wires are heaters, because the sensor has to be
hot to work.  

The ZrO2 ceramic is permeable to O3- ion, which exists in small quantities
in air at elevated temperatures. So, these ions diffuse through the cell
taking the charge with them due to the difference in concentration and
voila, a small battery (technically an electrochemical concentration cell)
whose output voltage is related to the difference in O2 concentration.
Because rich exhaust is deficient in O2, the cell reads about a volt.  When
lean, the cell reads about 0.1 V.

The problem with the cells is that the response curve (voltage=F( air/fuel
ratio) ) is shaped like a cliff, so if the voltage is low, you know you're
above 14.7, but don't know how far (at least don't know very accurately),
and if the reading is high, you are below 14.7 but again, you can't tell how
far.  That's why you can see if the sensor is alive by looking at the
reading oscilate back and forth.  The computer doesn't try to measure the O2
content absolutely, it just bounces back and forth across 14.7, so it'll
always be close.  If you want to control the mixture near 12 or so for
performance, you'll need a different sensor type...the response of a ZrO2
cell will be totally flat from 10 to 13.

 - Mitch Loescher