Linearizing the Output Characteristic of GMR Current Sensors Through Hysteresis Modeling

This paper presents a new method for increasing the measurement accuracy of giant magnetoresistive (GMR) sensors. The method used is to linearize the output characteristic by numerically eliminating the hysteresis-the main error source. A simplified mathematical model of the hysteresis is derived from the T(x) model and tested in a software simulation environment by comparing the results with the real measurements. The model is then discretized and implemented on a fixed-point digital signal processor (DSP). An algorithm based on the model is developed in order to eliminate error propagation during the measurements. The 75% improvement in accuracy and the linear output characteristic delivered by the DSP verify the validity of the proposed method. The successful final results demonstrate that modeling hysteresis is a convenient and very effective way to increase the accuracy and the measurement range of GMR sensors in practical applications.