Correction of nonlinearities in series structure-based resistance thermometry readouts.

Abstract

Series structure-based resistance thermometry readouts offer several advantages for multi-point temperature measurements. However, because of the diversity of nonlinear error sources and differences among channels in such readouts, existing nonlinear error correction methods are ineffective. In view of this situation, a nonlinear error correction method based on error source analysis is proposed. The proposed method first determines the impacts of error sources by analyzing the circuit architecture. The contributions of the common-mode rejection ratio and the mismatch between positive and opposite exciting currents are then eliminated using resistance bridge calibrators. Finally, the residuals are fitted to various polynomial functions. The results of experiments show that correction based on the proposed method results in a maximum nonlinear readout error of 1.87 × 10-5, compared with 4.01 × 10-5 using the classical method. Thus, the proposed method of nonlinear error correction is effective for series structure-based resistance thermometry readout.