At 04:48 PM 12/22/2011, John Lauterbach wrote:
>Here are the results from voltage test.
>
>Voltmeter +-lead connected to lead from O2-sensor, -lead on bumper.
>Engine not run for three hours and  O2-sensor not connected to lead to
>ECU
>
>0 volts from 0 to 120 seconds

Cold sensor, not running yet.

>0.13 volts at 150 seconds

Lean (could be any degree of lean)

>0.53 volts at 180 seconds

Very close to stoichiometric

>0.30 volts at 210 seconds

*Very* slightly lean

>engine stalls out at 240 seconds

Unknown complication but needs chasing.

>connected O2-sensor to lead to ECU
>
>restarted engine
>
>0.45 to 0.55 volts

ECU is biasing the sensor while sensor warms up.

>0.6 to 0.8 volts after engine at temperature and engine at about 2000
>rpm

ECU has gone into closed-loop operation. Depending on meter, probably
cycling fast enough that you're seeing an average reading, which
suggests normalish operation.  The suggestion is that it might be
biased a bit rich, since .8 volts indicates some unknown rich condition.

>0.4 to 0.7 volts with rapid change within that range with engine at idle

Within the limitations of the meter this is a normal reading.  At
idle the ECU lean-rich-lean cycle is a few seconds long (you can
probably hear the engine note change slightly).  Typical digital
meters average their readings over a period of 1/3 to 1/2 second, so
they can't meaningfully track the transitions.

Again within the limits of the meter this looks like normal operation
of the closed-loop system.  The little gadget that Ken Lewis sells is
based on an LM3914 chip and will give you better data since it
responds instantly, and if you can get hold of an oscilloscope that
will show exactly what's going on with the sensor.

Typically an engine (at least a simple marine engine) that's too lean
will cough when it tries to accelerate, rather than simply
bogging.  I'm guessing that your trouble is actually too rich.  If
you have an air leak into the exhaust or if one cylinder isn't firing
properly the sensor will see a lean condition where none exists and
the ECU will force the mixture rich, sometimes extremely rich.  This
could foul plugs at idle I think, as well as rubbishing your mileage.

Yours,
David