In response to Ecomike's comment stating my testing procedure is in error regarding the output of an O2 sensor with the key on/engine off, I'll stand by my "guidance". Now, first, I stated in another post that I had minimal experience with Renix O2 sensors, but was referring to Jeep HO and similar O2 sensors (like the ones GM has been using since the mid '80s); I've had extensive experience testing the more common sensors. When you back probe the sensor output lead, and turn the key to "ign", you are looking at the ECM generated reference voltage--not the O2's output. Obviously, with engine not running, there will be no output.
If you read "0" volts in this circumstance, as Ecomike suggests you should, that would indicate a problem with either the wire, or with the ECM. That's why it's a good first step in testing the ECM circuit.
And, no, it's not necessary (or recommended by me) to pull the O2 sensor out of the exhaust to test it; however, if you have a couple of sensors laying about, for whatever reason, it's easier to "bench" test them than to install them in the exhaust just for that purpose.
I don't care what kind of meter you want to use for testing electronic circuitry, it should be high impedance. For testing low-end devices, like resistors, capacitors, transformers, most any analog multimeter will work, and for some cases, as previously discussed (the TPS for example), the analog meter is preferred. That said, the analog meter requires more interpretive skills than a digital meter to use correctly and accurately; also the meter is most accurate in the upper 1/3rd of the select range. There are no issues with parallax or meter movement error to consider with the digital multi-meter. The problem with the digital meter is that it uses circuits to convert from analog to digital, requiring it to sample detected voltage at some predetermined rate (such as to times per second). I might add that anytime you're using a scanner to look at engine sensor data, the operation is the same as using a digital meter; the ECM is converting the sensor data to digital and sampling the converted data at some predetermined rate. This sampling requirement reduces the digital meters effectiveness in testing varying DC voltage, or varying resistive values, such as seen when testing a TPS for discontinuity (bad spots); a relevant reading may be lost between samples.
As far as high or low impedance goes, most people in the business would call 40k ohm/volt low impedance. Testing an O2 on the 3-volt range would be the same as putting 120k resistor in parallel with the sensor--not good. The digital meter (the common variety most of us have) is like placing a 10 meg-ohm resistor in parallel. With the circuits we test (automobile ECM/PCMs), 10 meg-ohms will not affect the readings in a material way.
The best I can tell searching for info on the E-bay meter you sited as using over the last 35 years, the Micronta 22-210, the unit is a 30k ohm/volt multimeter--not good, in my opinion, for testing active electronic circuits or components: "22-210 30,000 OHMS/VOLT MULTITESTER".
With regard to using nitrogen or CO2, neither will accomplish the objective of richening the mixture to test O2 response. Since the test I described required the engine to be running, the O2 heater circuit should also be active, by default.