Engineer's Notebook: Temperature measurement (bulk).
an important topic I thought I would devote some effort to it.
Earliest method liquid in glass: evacuate a thin cylinder & filled with Mercury or alcohol.
There's a reservoir at the bottom to store most of the liquid.
1/7

As temperature goes up, liquid expands, goes higher up the cylinder. Typically calibrated using ice baths & boiling H2O which under ideal conditions are exactly 0° & 100°C, or close enough.
2) bimetallic spring. 2 dissimilar metals w/diff expansion rates are stamped together
2/7

as temperature changes one expands more than the other, causing a needle to deflect one way or the other. Cheap and durable but not accurate.
3) Electronics. several methods of differing cost and accuracy:
a) seebeck bimetal. Junction of 2 diff metals implies small DC voltage
3/7

Different metals for different ranges. Accurate and inexpensive
b) Platinum RTD. resistance of wires change nearly linearly based on temperature. measure the resistance, get the temperature. Very accurate over wide ranges, very expensive.
c) Thermal diode. Diode current/volt.
4/7

characteristics change over temperature. By measuring tiny voltage changes you get temperature. Inaccurate not calibratable (software calibration) changes with time. Unbelievably expensive, pennies.
d) Pyrometry. Photosensor or array measures light coming off the object.
5/7

typically IR but can be microwaves of several different frequencies specifically used to measure the temperature of individual molecules. Creates a current that can be easily measured, but often varies in tremendous range.
Goes from inexpensive to millions of dollars!
6/7

4) alternate methods.
one of my faves is crayons which melt under specific temperatures you draw horizontal line on the surface you need measured and if it exceeds that temperature the crayon melts! Very cheap, very inaccurate, but fun!
Hope you enjoyed.
Have Questions? Ask!
7/7