Cold junction compensation in thermocouples

Introduction to cold junction theory

When accurate thermocouple measurements are required, it is common practice to reference both legs to copper lead wire at the ice point so that copper leads may be connected to the emf readout instrument due to the cold junction. This procedure avoids the generation of thermal emfs at the terminals of the readout instrument. Changes in reference junction temperature influence the output signal and practical instruments must be provided with a means to cancel this potential source of error.

The EMF generated is dependent on a difference in temperature in the junction, so in order to make a measurement the reference must be known. This is shown schematically in Fig. #1 and can be accomplished by placing the reference junction in an ice water bath at a constant 0°C (32°F). Because ice baths are often inconvenient to maintain and not always practical, several alternate methods are often employed.

Techniques to compensates the cold junction

Electrical Bridge Method

This method usually employs a self-compensating cold junction electrical bridge circuit as shown in Figure 2. This system incorporates a temperature sensitive resistance element (RT), which is in one leg of the bridge network and thermally integrated with the cold junction (T2). The bridge is usually energized from a mercury battery or stable d.c. power source. The output voltage is proportional to the unbalance created between the pre-set equivalent reference temperature at (T2) and the hot junction (T1). In this system, the reference temperature of 0° or 32°F may be chosen.

As the ambient temperature surrounding the cold junction (T2) varies, a thermally generated voltage appears and produces an error in the output. However, an automatic equal and opposite voltage is introduced in series with the thermal error. This cancels the error and maintains the equivalent reference junction temperature over a wide ambient temperature range with a high degree of accuracy. By integrating copper leads with the cold junction, the thermocouple material itself is not connected to the output terminal of the measurement device, thereby eliminating secondary errors.

Thermoelectric refriferation method

The Omega¨ TRC Thermoelectric ice pointTM Reference Chamber relies on the actual equilibrium of ice and distilled, deionized water and atmospheric pressure to maintain several reference wells at precisely 0°C. The wells are extended into a sealed cylindrical chamber containing pure distilled, deionized water.

The chamber outer walls are cooled by thermoelectric cooling elements to cause freezing of the water in the cell to work as a cold junction reference. The increase in volume produced by freezing an ice shell on the cell wall is sensed by the expansion of a bellows which operates a microswitch, de-energizing the cooling element. The alternate freezing and thawing of the ice shell accurately maintains a 0°C environment around the reference wells. An application schematic is shown in Fig. #3.

Completely automatic operation eliminates the need for frequent attention required of common ice baths. Thermocouple readings may be made directly from ice point reference tables without making corrections for reference junction temperature.

Any combination of thermocouples may be used with this instrument by simply inserting the reference junctions in the reference wells. Calibration of other type temperature sensors at 0°C may be performed as well. Heated oven references: The double-oven type employs two temperature-controlled ovens to simulate ice-point reference temperatures as shown in Fig. 4. Two ovens are used at different temperatures to give the equivalent of a low reference temperature differing from the temperature of either oven.

For example, leads from a type K thermocouple probe are connected with a 150° oven to produce a Chromega¨-Alomega¨ and an Alomega-Chromega junction at 150°F (2.66 mV each).

The voltage between the output wires of the first oven will be twice 2.66 mV or 5.32 mV. To compensate for this voltage level, the output leads (Chromega and Alomega) are connected to copper leads within a second oven maintained at 265.5°F. This is the precise temperature at which Chromega-Copper and Alomega-Copper produce a bucking voltage of differential of 5.32 mV.

Thus, this voltage cancels out the 5.32 mV differential from the first oven leaving 0 mV at the Copper output terminals. This is the voltage equivalent of 32°F (0°C).