Techniques for calibrating a linear position sensor

Abstract

Techniques are described for sensing a position of an object located within an enclosure over a position range. The techniques include generating an expected output signal of each of a plurality of magnetic field sensors disposed along an outer surface of the enclosure, receiving actual output signals from the sensors, wherein each actual output signal indicates a relative proximity of a magnetic target coupled to the object to the corresponding sensor. The techniques further include superimposing the expected output signal over the actual output signals, and iteratively shifting the expected output signal over position relative to the actual output signals and comparing the shifted expected output signal to the actual output signals, until the expected output signal compared to the actual output signals corresponds to a substantially minimized error parameter. The position of the object may then be determined based at least in part on the shifted expected output signal.