Simple Methods to Measure the Additive Error and Integral Nonlinearity of Precision Thermometric Bridges

Abstract

Determination of the accuracy of AC bridges with the high precision adjustable transformer voltage dividers based on the strongly magnetic coupling coils is referred in this paper. Metrological model of the temperature measurements by resistive sensors connected to these bridges is presented. The additive, multiplicative and linearity components of error in temperature measurement are considered. Method for determining the zero of the bridge error is developed. It is based on the non standard connection mode of the equipotential pairs of current and voltage terminals of standard resistance to the bridge. It is estimated that up to 1 MΩ of this resistance, the effective resistance obtained on the bridge input is less than 10− 10 Ω. Four-terminal standard resistors of 0.1 Ω up to 1 MΩ are used in experiments to find bridge zero readings. Results indicate that the additive error of the tested precision bridge is about 0.5 LSB and is independent from the nominal standard resistance value. The conditions under which in temperature measurements remains only the linearity bridge error are formulated. Proposed is the unconventional method of measure and estimating the bridge nonlinearity named as dichotomy method. It applies algorithm based on the division of the measurement range and then obtained subsequent intervals always in half. Graphical interpretation and the analytical expression for the nonlinearity error are given. Sets of paired four terminal reference standard resistors are proposed for use in control. The resistance of each of them separately and of the given serial physical connection of them both has to be measured by tested bridge. The reasons affecting the accuracy of the physical realization of resistance summation is discussed. By calculations and experimental verification is find that in measurements by dichotomy method the bridge linearity error of 0.1 ppm or less can be discovered. Conclusions and final remarks are included. Both methods are simply and can be easily implemented in any metrology lab and be used also in automatic calibrators.