Questions about 34970A, modules, and relay lifetime ...

[mossygreen]

I have a 34970A and 2 34901A modules that I purchased about 12 years ago as my employer was folding. The things have sat idle in indoor conditioned space since that time. I am about to do a little test and got to looking in more detail at what I had. Everything powers up, passes self tests, responds to GPIB commands, thermocouples give "correct" readings, etc.

Fundamentally I'm trying to understand what to expect w/r to relay lifetimes in this application. Reading the manual and datasheet give some info, but I lack the practical understanding of what unloaded vs. very lightly loaded means, etc. The plots of the impedance of the relays are very steep so it looks like things could just fall apart in the span of a few thousand relay switch cycles. However, I don't see people talking about cycle counts for used modules, etc.

The mux itself and one of the two modules was used in some thermal testing. Using the SCPI interface I get this for the DMM's relay counts:

        2667053,1424003,2667338

The more lightly used module gives this for the channels:

        2874293, 2874291, 2874264, 1, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 0, 0, 0, 0, 0

and this for the other relays within the module:

        0, 0, 0, 0, 2874285, 0

17 of the channels have no use but three have ~2.8M cycles on them. If these were used for thermocouples and RTD's, am I likely okay for a while for the same kind of usage?

I can see some diagrams in the service guide with a suggestion as to how to disassemble a 34901 module and there are part numbers for the relays. Is that something people do themselves, or do most people trade-in their used modules and get refurbished ones?

How long do the internal DMM relays last? Is their aging just calibrated out?

Any guidance appreciated ...

[CaptDon]

We have a large number of these HP acquisition units where I work. Some have been in constant use for at least a decade. It seems our most common failure is loose connections and sometimes having to pull the module out and plug it back in which I will guess is only to breakdown oxidation at the edge connections. There was always a concern about 'dry switching' with less than 10ma of current flow but I think the internal relays use some sort of wiping action to get around dry switching since we never really seem to have any problems even with thermocouples which have microamps of current flow.