Improved mathematical operations based calibration method for giant magnetoresistive current sensor applying B-Spline modeling

Abstract

Nowadays, giant magnetoresistance (GMR) based sensor technology is widely used in numerous industrial applications. However, an inexpensive and efficient method for compensating the nonlinearity of GMR spin valve sensor is still under the industrial expectation, which limits the applications of the GMR sensors. By applying the B-Spline modeling, excellent results were achieved; nevertheless, a calibration with several hundreds of measured data points over dozens of temperatures had to be conducted. In this paper an improved mathematical operations based calibration method for GMR spin valve type current sensors applying B-Spline modeling is introduced. This method allows a calibration for each further sensor applying B-Spline modeling with only two temperatures and per temperature only five measurement points. In comparison with the original calibration effort for B-Spline modeling, the new calibration method with minimized efforts showed an outstanding measurement accuracy of ±0.6% at a nominal current of 50A in a temperature range from −20°C to 80°C. Due to the unexpected linear temperature behavior of the magnetic parameters, mainly of the uniaxial anisotropy to the exchange direction anisotropy ratio (Ku/Ked), only 5 measurement points at two different temperatures were necessary for our new calibration method.