Thermocouple ranges with multimeters, why no small mV ranges?

[dangerousamateur]

Hi guys,

I've been reading on this site for some time, this seems to be the place to hangout when it comes to electronics


Recently a friend lent me his Fluke 179. Using it with thermocouples I wondered why this and a lot of other cheaper devices are able to measure such small voltages.
0,1°C with a K-Type thermometer are some 4µV.

OK, not with super accuracy, but the smallest DCmV range has 100 µV resolution (600mV).

Just curious do these meters have a seperate circuit for thermocouples?
Why isn't there some 60mV range, if the hardware is obviously capable?

[SeanB]

Mostly noise, the 60mv of the thermocouple is into a lowish impedance, so the noise is lower. If you try to do this with the typical DMM input impedance of 10M or more your input circuitry would be swamped with noise coupled into this high impedance, while with the thermocouple with essentially zero resistance this is not a problem.

[Kleinstein]

Most of the auto ranging handheld DMM actually have an 10 M resistor in series at the input, when reading voltages and is part of the input protection. This resistor gives quite some noise and also makes it sensitive to leakage currents. So the resolution is limited.

The better bench top meter use a different autoranging and protection scheme and thus allow higher resolution. This usually includes a relay aand higher supply and is thus less practical for a battery powered system.

Special TC meters can use less protection (you don't expect them to survive some 400 V AC) and do not need automatic range switching. So they can be build lower noise and more stable.

[dangerousamateur]

Thank you for your answers.
What you wrote makes sense to me.

Do you know how they do it in this case , is the input resistance switched?
The input selector switch has the same position for mV and TC, you only push the yellow button.

[Kleinstein]

With a dedicated mechanical switch for a mV range the amplifier can work non inverted. This allows for high impedance (e.g. > 1 Gohms, though not very well defined) and low noise. Modern AZ OPs have no problem resolving down to the µV range, but the input range (e.g. +-2 V, maybe +-15 V) is limited. Higher voltages would need a relay to switch in a divider - so auto-scaling gets more complicated. At the very low noise end it is a compromise between low noise and low input current. So the lowest voltage noise input is no longer suitable for high impedance (e.g. 1 M) sources.

TC measurements also needs cold junction compensation. That is an extra thermometer (NTC, PTC or diode) to measure the temperature going from the TC wires to copper.