Dependability aspects in capacitive angular measurement systems

Today´s smart sensors allow for the implementation of ample computing resources to improve their performance without increasing chip area and thus production costs significantly. What makes them more complex and at first sight less robust is in fact also a chance to improve their fault tolerance. In this paper we present a fault analysis for a capacitive angular sensor designed for automotive applications. Based on the assessment of potential faults and effects degrading the performance, we discuss constructive and operative measures and design decisions that have been taken to enhance the robustness of the system. A key component is a self-calibrating carrier frequency measurement principle, which is used to compensate common drift effects as well as to filter out electromagnetic disturbances. Electrical faults can be detected and lead to a fail stop behavior of the system. It is shown how these fault tolerance mechanisms are implemented in both a passive (i.e., by virtue of system design) and active way through appropriate data processing algorithms. The results of the case study are considered typical for capacitive sensors, but seem applicable also to similar problems.