I.Q. Instruments - Thermocouples explained simply

THERMOCOUPLES

A thermocouple is formed any time two dissimilar metals touch each other. When the temperature of this junction is different to the temperature of other parts of the metals an EMF is generated. Fortunately for a lot of people working with metals this EMF is very small (usually measured in µV/°C) so there is not much chance of receiving an electric shock but it is sufficient to cause corrosion problems (which we won't tackle here).

This property is used in industry to measure temperatures, especially if these temperatures cannot be measured by other techniques.

The two metals are usually formed into wires and welded together, however crimping, soldering or even just twisting the wires together gives results. There is no theoretical limit to the wire diameter and thermocouples can be made in just about any size required.

Any two metals can be used in theory but in practice we need something which is robust, stable, has a large enough signal to use, and is affordable. The most popular thermocouples are listed later.

Over the years the EMF output of each standard thermocouples has been plotted and tables are available for calibration purposes. To calibrate a thermocouple instrument it is only necessary to inject a millivolt signal to match the tables.

WHAT'S THE DIFFERENCE?

A thermocouple is a differential device and the output is related to the DIFFERENCE in temperature between the hot junction and the cold junction.

Thermocouple tables are based on a cold junction at 0°C. If the cold junction is not at 0°C (which is usually the case) a correction has to be applied.

To determine the output of a thermocouple using tables, use the formula:-

Vo = Vh - Vc

Where Vo is the output voltage we are looking for, Vh is the voltage shown in the tables for the hot junction, and Vc the voltage shown for the cold junction.

For example: Assume a type J thermocouple has its hot junction in boiling water at 100°C and the cold junction is at an instrument which is at a room temperature of 25°C.

The output of the thermocouple measured at the instrument is therefore :-

5.268mV - 1.019mV = 4.249mV.

ACCURACY

The accuracy of a thermocouple (actual output vs. value in tables) depends on the composition of the alloys making up the wires. This is typically of the order of ±3°C at 100°C for a standard type J thermocouple and ±1.5°C for specially selected types.

This is not all that great if we require extremely high accuracy. So why do we use them?

Easy. They are easy and cheap to make and can measure temperatures from around -200°C to +1800°C depending on the wires used.

THERMOCOUPLE TYPES

TYPE MATERIAL RANGE °C
Continuous OUTPUT µV/°C

B PtRh6%/PtRh30% 100~1600 6

E Chromel/Constantan 0~1000 56

J Iron/Constantan 0~700 52

K Chromel/Alumel 0~1100 40

N Nicrosil/Nisil 0~1100 38

R Platinum/PtRh13% 0~1600 10

S Platinum/PtRh10% 0~1600 10

T Copper/Constantan -185~300 38

W5 Tungsten/TungstenRh5%

Fred Philpott B.Sc. (Hons. Lond.)
IQ Instruments CC
South Africa

MORE TO BE ADDED AS TIME PERMITS

Return to Home Page

TYPE	MATERIAL	RANGE °C Continuous	OUTPUT µV/°C
B	PtRh6%/PtRh30%	100~1600	6
E	Chromel/Constantan	0~1000	56
J	Iron/Constantan	0~700	52
K	Chromel/Alumel	0~1100	40
N	Nicrosil/Nisil	0~1100	38
R	Platinum/PtRh13%	0~1600	10
S	Platinum/PtRh10%	0~1600	10
T	Copper/Constantan	-185~300	38
W5	Tungsten/TungstenRh5%