Rotor design for capacitive sensors

The shape of the rotor is an important accuracy issue of capacitive sensors for measurement of angular and linear position. Depending on the evaluation algorithm, two different shapes are commonly used; these rotors produce trapezoid or sinusoidal capacitance patterns. However, capacitance patterns which allow the construction of more accurate and more robust estimators of the desired parameter can be used. In general, position estimation is a nonlinear process, therefore the most accurate estimator is computationally expensive. In this paper an approach for robust optimal design of sensors with suboptimal estimators is presented. The optimal design is obtained by a tight coupling of statistical and field numerical methods taking into account uncertainties arising from inaccuracies of the topology, electronic component tolerances and measurement noise. It is shown that all these factors have substantial impact on the optimal shape of the rotor.