Abstract

The force-sensing resistors (FSRs), like the other types of sensors, present a series of errors that make their use limited to applications that do not require high levels of precision. Among the most outstanding errors of this type of sensors are the hysteresis, non-linearity and drift errors. A methodology that seeks to reduce these errors by analyzing alternative correlations to those proposed by manufacturers based on simple linear correlations is performed. This paper presents the results of a research aimed at the characterization of a Tekscan A201-1 sensor, in which several correlations that combine the voltage and capacitance readings, were used to obtain an improved response signal that reduces hysteresis errors, non-linear and derives in FSRs. In this study, eight sensors to which a predefined force pattern was applied were analyzed. The obtained results allowed to reach a significant reduction of the error in comparison with the linear regression method proposed by the manufacturer of these force sensors. Finally, the correlations evaluated and the reductions of errors obtained by the methodology used are presented.