Coax cable used for DC instrumentation

ll

of

ar

Engineering's goal is to design a system until it is 'good enough' and often, and in many applications and locations, coax can be interchangeable with STP. Upon that premise I can agree with your claim.

However, in general, STP will be better: Floating mike, somewhat reasonable, except...

1 if a person holds the mike, there is a connection to 'gnd', which creates a gnd loop. 2 if the shield is not terminated as a total and complete envelope wrapped around ALL internals of the mike, there is unbalanced capacitance, which allow noise sources to inject signals into different circuit locations inside the mike. In other words, direct pickup from those noise sources.

[about a 5 mV signal at a distance from the measuring node]

That isn't the RIGHT information; what is the output impedance? If it's like a thermocouple (one kind of pyrometer), coaxial cable is CONTRAINDICATED because the outer shield and inner conductor will have different temperature lags in case of a chill wind on the apparatus.

If it's a thermocouple, with low impedance (under 10 ohms), then unshielded twisted pair will probably work fine; don't ground ANYTHING (the high temperature insulation is failure-prone anyhow) except next to the meter. Use Cat5 wire (it's cheap) and consider dedicating an extra pair, for preamp power, at each node. Either a splice, or a preamplifier, requires a small box for protection. Small meaning the size of a connector (print server computers inside a DB-25 connector shroud are not uncommon).

If it's a temperature measurement, with slow response, the measurement end can take several readings and average (or even integrate over an exact interval) to get rid of AC errors; or you can apply an overall grounded shield to reduce AC pickup.

Signal sources have four important characteristics, GIVE ALL FOUR YOUR ATTENTION. Signal amplitude (5 mV); source impedance (totally unknown); bandwidth (DC); resolution (totally unknown).

Of the four, we answer-ers only got one (the 5 mV), and a half (if it were REALLY DC, you could measure once in January and just refer to that number until June).

"William Sommerwenker"

** Shame Wanker Bill has no idea what the " very good reason " is.

..... Phil

"William Sommerwanker"

** He won't be complaining.

His signal is a DC voltage.

.... Phil

And, as we all know, DC signals are immune to noise and inteference.

"William Sommerwanker Lying Turd "

** There will not be any with a 35 foot run of RG59 co-ax.

The receiving device very likely has a LPF incorporated and dead easy to add one of it does not.

So piss off to HELL

- you PITA know nothing wanker !!!

..... Phil

There's another issue we haven't discussed, and that's electrochemical reactions. BNC connectors aren't well-sealed against the environment. You don't "stuff" getting into the connector and corroding the connection.

An XLR connector offers no only a balanced line, but the contacts have greater surface area, and you can seal off the connector will silicone.

e

Does your "sensing device" support differential inputs.. I would think that would determine if twisted pair is the solution or not.

That's not the whole story; if the source is floating (not locally ground-referenced) the twisted pair is a good solution for non-differential input, as well.

The RG terminations refer to the RADIO FREQUENCY and pulse properties of the line. If you are only measuring voltages which change slowly (not more than 50Hz components) ANY coax will do. Think of it as a screened test lead ( of course you should earth the outer screen) Hope this helps P.W.