An adaptive measurement algorithm to improve the fault tolerance of a capacitive angular measurement system

Capacitive angular sensors have to cope with small, but varying capacitance values, which result both in gain and offset error of the measured input signals. As a consequence ratiometric measurement methods have been introduced, which account for these errors by obtaining the minimum and maximum value of all variable capacitances of the sensor together with the actual, angle dependent input signal. This approach can detect neither permanent failures like broken wires or shorts nor excessive signal variations reliably. Additionally they require several complex numerical operation and are susceptible to numeric resolution limitations. We present an alternative approach which reduces the number of mathematical operations significantly, yet showing enhanced stability to cope with large sensor impurities. Finally the proposed algorithm offers various means of implementing fault tolerance into a system. A missing rotor or a faults segment capacitance can be detected without having to introduce any absolute threshold values.