At 09:31 AM 8/23/2011, John Lauterbach wrote:
>Here are the details.  O2-sensor not connected.  Voltage from O2-cable
>coming from engine is +0.46 volts.  Again, O2-sensor is not connected.
>I have tried with two different ECUs and get same result.

This is normal.  The ECU applies a bias voltage to the line so it can
tell when the sensor heats up and comes on line (and swamps the bias signal).

>With engine off, resistance between O2-cable coming from engine and
>ground is 140K ohms.  I have a spare FI cable assembly and get same
>resistance from that cable when a ECU is connected.

That sounds like a not-unreasonable value.

>If I connect O2-sensor as intended, engine runs rich and stalls.

Using a digital meter,* describe the behavior of the O2 sensor while
disconnected from the ECU and engine running, blipping throttle, fast
idle and such.
I'm expecting that you're going to see output of zero or close to it
under all running conditions.

Connect the sensor (ok to do while engine running) - I expect you'll
still see zero or very low output.   This would indicate either bad
sensor or oxygen getting into the exhaust through misfiring cylinder
or exhaust leak.

*Oscilloscope would be great, analog meter would also be great but
would need sensitivity of  at least 40-50 kohms per volt and set on
maybe three-volt scale.  An indicator based on an LM3914, such as the
one that Ken Lewis sells would also be excellent.  Since you probably
don't have any of these, a regular digital meter on two-volt scale is
usable though unwieldy.  If it has a fast-acting bar-graph as part of
the display that would help.

Yours,
David