The thermocouple is the most widely used temperature sensing device. It converts thermal energy into electrical energy. It dates back to 1821 when Thomas Seebeck found that joining two wires made of different materials generates an EMF when heated at one end. The amount of heat is directly proportional to the amount of output EMF or voltage. This is a simple, rugged device that can be made inexpensively. It also has the advantage of being able to measure a 4500°F range of temperatures. It is, however, unstable and needs a reference junction to make sure that its output is useful.
A thermocouple measures temperature by using a pair of wires made from dissimilar metals. At one end of each wire, they are joined together, often by welding them. The differing thermoelectric characteristics of the wires generate a very small voltage between their free ends, from which the temperature of the joined ends can be derived.
No power supply is needed for a thermocouple, but the voltage that it generates is extremely small (measured not just in millivolts, but microvolts) and very nonlinear, requiring hardware and/or software to convert it to a temperature value. Laboratory equipment or integrated circuit chips are available for this purpose.
Different types of thermocouples are available to measure different temperature ranges, and each type has its own characteristics, requiring appropriate conversion. The types of metals used in a thermocouple are based on intended operating conditions, such as temperature range and working atmosphere. Different thermocouple types have very different voltage output curves. When a replacement thermocouple is required it is important that the thermocouple type used in the replacement matches the original.
A “raw” thermocouple looks very unimpressive, as it merely consists of two wires welded together at one end.
A schematic symbol that is often used to represent a thermocouple is shown in Figure 1. Because this component does not consume current, the plus and minus signs do not mean that power should be applied to the wires. The positive sign indicates which wire will generate a higher voltage than the wire with the negative sign.
Thermocouple working principle
When one end of a piece of wire is maintained at a temperature that is different from the other end, the temperature gradient along the wire creates a small electromotive force that manifests itself as a difference in electrical potential between one end of the wire and the other. This is known as the Seebeck effect, named after the man who discovered it. The magnitude of the potential will depend on two factors: the temperature difference between the ends of the wire, and the type of wire that is used.
Figure 2 illustrates the concept. Part 1 of this figure shows two wires, named A and B. The left ends of the wires are heated to the same temperature, TX, while the right ends remain at a cooler temperature, TY. Because the wires are composed of different metals, the voltage drop across each wire will be different.
To make this model useful, some factors must be eliminated. In part 2 of Figure 2, the hot ends of the wires have been welded together.
This now guarantees that they share the same temperature and the same voltage, VX. We do not yet know what these x values are. In part 3 of the figure, the cold ends of the wires are clamped in an isothermal block, which keeps them at an equal temperature, still represented as TY. The block is not electrically conductive, so the cold ends of the wires still have different voltages, VA and VB. We cannot measure these voltages directly, because they are relative to VX, which is unknown. However, a voltmeter can measure VA and VB relative to each other.
The voltmeter will have its own voltage gradient on its wires, and possibly a temperature gradient too, but both of these wires are made of the same metal (probably copper) and share the same temperature gradient. Therefore, their effects will be equal.
A mathematical relationship exists between the temperature gradient and the voltage difference in each thermocouple wire. Suppose KA is a constant or function that enables the voltage difference in wire A to be determined from its temperature gradient, and KB serves the same function for wire B. Suppose TDIF is the difference in temperature between TX and TY. We may state:
KA * (TDIF) = VX – VA
KB * (TDIF) = VX – VB
By subtracting the second equation from the first and rearranging the terms, we get:
TDIF * ( KA – KB ) = VX – VA – VX + VB
The two VX terms cancel out, leaving VB – VA on the right. We can recognize VB – VA as the voltage difference measured by the meter. Call it VM. So:
TDIF = VM / ( KA – KB )
This enables calculation of the temperature difference between the ends of the wires, based on the meter reading and the conversion factor for each wire, which can be found experimentally. Because TY is being held at a known, constant value, we can determine the value of TX:
TX = TY + TDIF
When the thermocouple was first invented, the cold ends of the wires were placed in a bath of ice and water, forcing them to acquire and maintain a known temperature of 0 degrees Celsius.
The advent of accurately calibrated thermistors made it possible simply to measure the temperature of the cold ends. In this way, a thermistor enables a thermocouple to work. This prompts the question: why not just use the thermistor to measure TX, and throw away the thermocouple? The reason is that a thermistor has a more limited range, seldom being used for temperatures above 150 degrees Celsius.
Note that the “hot end” of the thermocouple wires does not actually have to be hotter than the “cold end,” even though those terms are commonly used. The equation to find TX Works just as well if TY is higher than TX. The temperature difference will simply have a negative value instead of a positive value.
Because “hot” and “cold” are misleading terms, modern documents generally refer to the “measurement junction” and the “reference junction” of the wires. Note, however, that the wires are not actually joined with each other at the reference junction.
A common misconception is that voltage is generated where the wires are joined at the measurement junction. This is not correct. The voltage is a function of the temperature gradient between the measurement junction and the reference junction in each wire. Therefore, the way in which the wires are joined is irrelevant, provided there is an electrical connection between them. They can be welded, soldered, brazed, or crimped together.
How to use a thermocouple
Where a thermocouple is used in a laboratory, typically each wire is insulated, and they terminate in a plug that is inserted in a meter. The reference junction is hidden inside the meter, along with some electronics to decode the temperature data. The meter must have a setting that is appropriate to the specific type of thermocouple being used so that the conversion factors are correct.
Because the type of metal in each wire must be consistent all the way from the measurement junction to the reference junction, other types of wires cannot be used to extend the reach of a thermocouple. Any extension must use wires made from the same metals. Connectors, also, must have pins and sockets that match the types of metals in the wires.
If you want to learn more about thermocouples, you can check this awesome book.
Principles of electric circuits conventional current
Electronic Measurement and Instrumentation
Electric Motors and Control Systems
Electronic Components Volume 3