Errors in thermoelectric thermometers

Hitherto, no quantitative relationship between errors in thermoelectric thermometers and their causes (i.e. changes in Seebeck coefficient and temperature distribution) has been demonstrated. A theory of the operation of thermocouples has therefore been developed. From this, an experimental method is derived, which distinguishes and measures the effect of various factors on a given type of thermocouple, and which enables the results to be used to calculate the performance of the thermocouple under given conditions. The effects of as-received inhomogeneity, plastic strain and heating up to 650°C are given for chromel-alumel, and in less detail for other common materials. It is shown that it is possible to predict drift and other instabilities in chromel-alumel under defined conditions, and reduce the uncertainty of measurements by an order of magnitude in favourable cases. It is also shown that the conventional concepts of calibration which are widely used in the application and investigation of the performance of thermocouples, are generally fallacious.