LM34 as external temperature probe

[GDSever]

I was planning on turning a couple of my LM34 temp sensors into temperature probes and started thinking about the inductance that all the wire running out to them might create - Should I be worrying about putting decoupling / bypass capacitors across the +5/GND lines on both ends?

Its not a critical reading (not a control point), however I would like to limit the amount of noise.

 

[moxhamj]

Inductance won't be an issue as the voltage varies vary slowly. Neither will capacitance, and you could add caps if you like and it will certainly help filter noise.

The best solution is to not let the noise get into the wire in the first place, and a simple practical way to do that is to use twisted pair wire. The datasheet recommends this http://www.national.com/mpf/LM/LM34.html

A cheap source of twisted pair wire is cat5 cable. You might only use one of the 4 pairs, but it may still end up cheaper than using a single twisted pair, because cat5 is mass produced. Plus you get a robust sheath, and that is always a good thing

 

[westaust55]

If you go the way of twisted pair cable such as the CAT5 as suggested by Dr Acula, then if the cable lengths are going to be long then you could use a core from another pair connected as a compensation loop. This would enable you to measure the voltage along the active pair and compensate for any volt drop along the wires.

This scheme with three wires (for one compensation loop) and sometime even 4 wires (for a compensation loop on both the positive and negative side of a signal) is typically used for resistive temperature devices (RTD’s) such as Pt100 RTD’s in industry for temperature monitoring.