Here again in my slow but steady (and somewhat long winded) quest for emissions nirvana.

I used a digital voltmeter to test O2 sensor function, hot idle, meter between sensor wire and negative battery post. I did read the informative thread a few days back, plus an article linked in the FAQ about using an oscilloscope to get a true picture of sensor performance, but I don't have one of those at home. Let's say for a moment that the flashing numbers on my voltmeter ranged from 45 to 75 millivolts and only touched 125 once in a two minute period. Without the extra detail, is it safe to say this thing is shot? Mr. Haynes tells me that 45 is the threshold for popping a code.

If the sensor is shot, and I order a new one, I'll have a few days to wait. Would I get better performance if I unplugged the harness from the old sensor and let the ECU run in default mode for those few days?

Maybe I should also order a Bentley while I wait.

Thanks all.
--
Jim - '92 Red 245 191K, and sometimes an '85 245 275K

http://www.f-body.org/oldfaq/html/tech/sect2.html#oxysensor

So I put this thing in a vise, hook it up to a voltmeter, and burn it with a torch. My wife will think I've lost my mind.

--
Jim - '92 Red 245 191K, and sometimes an '85 245 275K

Yes, that is the test. If you do it, you can call yourself a bona-fide automotive researcher instead of someone just looking to hit on the right replacement part by guesswork.

When I have done it, I've used an analog meter, but not the one typically found in a toolbox. It is an old VTVM or vacuum tube volt meter with a very high input impedance. The needle response is sufficient to see how quickly the voltage responds to moving the sensor from the blue tip of the flame outward, but not that useful (to my eyes) monitoring the closed-loop switching in the car-- I use the scope for that.

In this image, you can see a way to get around clamping your sensor in a vise to make the ground connection. The odd connector on the coaxial cable is the old style used on the input to VTVMs. This arrangement was an aborted attempt to do CO adjustment on my non-lambda k-jet car, before finding the EGR pipe tap in the manifold was not quite the same thread as the lambda sond.

An analog meter also works well. Internal dampening will cause an analog meter to read more of an average. A scope will show fast transients. The Lanbda control box probably has some dampening.

Nope, an analog meter cannot be used to monitor the O2 sensor, it "loads" the output and the reading will show a meaningless 0 volts.

.... unless it is a VTVM (Vacuum Tube Voltmeter).

Rich

How would you use one? Do you mean damping, as in sluggish response, or dampening, as in leaky windshield dripping on ecm?

No, it isn't safe to say it is shot. It might be working, but reporting what it sees.

I think there is a link in the FAQ, or text even, describing how to evaluate a sensor out of the car, using a propane torch flame. I can tell you from personal experience that works quite well.

Bentley is nice to have, but it won't give you any more magic than Haynes on this one.
--
Art Benstein near Baltimore