Oxygen sensors

[Bohdan.L.Bodnar at att.com]

| How I saw it done in the controller in the 1991 GMC pickup truck
| is the O2 chip has an internal voltage divider that places the output
| of the O2 sensor at 0.450 volts until it warms up. Once its internal
| resistance drops below a certain point, the size of the O2 amplitude
| indicates primarily temperature.

> ok, so if its cold the output voltage stays between, say, .3V and .6V
> for an extended period of time. Once it exceeds these thresholds, we
> declare it's hot enough to use.
>
> If we aren't trying to meat emission standards, I suppose a simple
> warmup period is enough. (?)

My $0.02 worth...

The usual procedure is to (1) monitor coolant temperature and (2) run a
watchdog timer.  In some systems (e.g., Chrysler's EFI) the computer also
monitors something like vehicle speed;  this information is used to "diddle"
the timer's setting on the fly.

If the coolant temperature stays below a certain point for a programmed time,
a "stuck cold" fault code will be set (note that most contemporary computers
will also monitor the bounds on the coolant temperature and will have fault
codes for out-of-bounds conditions) and A DEFAULT value used.  A lot of people
with scan tools will read the coolant temperature and assume that the computer
is seeing this temperature;  in fact, the computer is merely outputting the
value it is currently using.

If the temperature reaches a certain operating point (usually, around 180F)
the computer will start varying the a/f ratio and see whether the O2 sensor
responds with a *binary* signal (i.e., close to 0 and close to 900 mV) which
varies.  Modern computers will not only monitor the response, but also, given
a mathematical model of the engine, associate a response time window with the
varying.  Put differently, the computer is looking for a response (O2 sensor
output) given an input step function (a rich or lean mixture) and a model of
the system.  This is why on carbureted engines in closed loop the M/C
solenoid's dwell is varying and why on EFI engines the total pulse width of
the injectors is varying (on peak-and-hold injectors, the "peak" time is fixed
-- around 1.5 ms -- whereas the "hold" part will vary).



>                                      John S Gwynne
>                                         Gwynne.1 at osu.edu
>

Regards,

Bohdan Bodnar