# How are very low temperatures(-180k) measured?

1. Feb 24, 2010

### aliz_khanz

okie so cyrogenic processor is one thing but there has to be some phenomenon which is used to measure cyrogenic temperature???

2. Feb 24, 2010

### f95toli

I am not sure I understand your question.
All you need in order to measure low temperatures is a thermometer.
A thermometer designed to be used at or near room temperature (300K) will obviously not work very well at cryogenic temperatures, but that is to a large extent because it is not calibrated (and the materials used are not ideal for very low temperatures).
Some temperature sensors that are used at low temperatures (silicon diodes), work quite OK even well above room temperature.

Now, measuring VERY low temperatures (below say 0.3K) is somewhat trickier, but -180 degrees C can be measured even with something as simple as a 50p schottky diode as long as you calibrate it first, all you need to do is to bias it as some current and then measure how the voltage changes with temperature.

edit: Take a look at the following webpage
http://www.lakeshore.com/temp/sen/smindex.html [Broken]

Lakeshore is one of the biggest suppliers of temperature sensors and e.g. their DT470 Si diode is widely used (I am in fact installing one in a probe tomorrow)

Last edited by a moderator: May 4, 2017
3. Feb 24, 2010

### luke1970

Using a piece of equiptment delicate enough to observe the electron movement is one way, and probably the most effective.

4. Feb 24, 2010

### f95toli

Eh? I have no idea what you are referring to here.
There are many different types of thermometers that works well at low temperatures. But single electron transport isn't a very practical starting point for thermometry at any temperature.
It is possible to do perform temperature measurements using what is known as Coloumb blockade thermometry (which is sort of based on single electron transport); but it is a complicated technique and is never -as far as I know- used routinely.

Also, -180 degrees C is NOT a low temperature; it is not even liquid nitrogen temperature. Even a thermocouple will work reasonably well at 77K.

5. Feb 24, 2010

### mgb_phys

The main difficulty in cryogenic thermometers is limiting the heat leak in from the thermometer.
A thermocouple has fairly large, fairly high conductivity wires, if you had a very small cold part the heat leakage down the wires could be significant.

6. Feb 24, 2010

### truth is life

This is what my thermal physics book has to say about the subject:

Beyond that discussion of practical laboratory thermometry, it delves a bit into the theoretical underpinnings of thermometry earlier, during which it hints at how ultra-low temperature measurements are made--via electronic or nuclear spins (which it says are used to below 1 millikelvin).

7. Feb 25, 2010

### aliz_khanz

but can't i find a proper method , on when to claibrate , which special thermometers or diodes are used and how they function ? Thanks for your help !

Last edited by a moderator: May 4, 2017
8. Feb 25, 2010

### f95toli

There are several books about cryogenic thermometry. However, if you want a practical guide I would recommend Pobell's "Matter and Methods at Low Temperatures".

That said, it is not really that complicated. Just buy a sensor suitable for the temperature range you are interested in from e.g. Lakeshore (see the link above). A sensor that follows a standard resistance vs. temperature curve (which is usually accurate enough) costs something like $200. It is convenient to read out the sensor using a dedicated temperature controller, but you can also use e.g. an ordinary resistance bridge or something similar. For temperatures this high even a normal bench-top multimeter might work (self-heating isn't really an issue at temperatures as high as 100K). Last edited: Feb 25, 2010 9. Feb 25, 2010 ### aliz_khanz Thank you so so so much !!!! I really appreciate your help. I am trying to find this book on internet but it seems like i have to pay$25 bucks for this thing and i only want one chapter from it. Thanks again.

10. Feb 25, 2010

### Gokul43201

Staff Emeritus
If you don't need 50mK accuracy and repeatability, you could probably find a Pt RTD for about half that price.