Simulation of nonideal behavior in integrated piezoresistive silicon pressure sensors

Abstract

The signal formation in piezoresistive silicon pressure sensors has been studied for many years and is very well understood meanwhile. Nevertheless, there are higher order effects of non-ideal behavior such as nonlinearities and hysteresis effects which limit the possible accuracy of such sensors considerably. In high volume applications like in automotive area it is worth to think about minimization of such effects in order to improve performance and also to enhance yield and to reduce costs. In this paper we report on a method to describe the above mentioned higher order effects of non-ideal behavior directly in terms of process- dependent parameters and therefore on a method to correlate process tolerances directly to output non-idealities and yield. The method comprises a quantitative value for 'design-capability' which we call 'influence strength' and which describes the quality of a design with respect to the accuracy-specification of the device as well as with respect to process tolerances. The effectiveness of the method is shown by several examples: first we use the method by applying it to effects of non-linearity. Second we present a simple model to describe the thermal hysteresis of integrated bulk-micromachined pressure sensors and we deduce a method to compensate and minimize hysteresis of integrated bulk-micromachined pressure sensor and we deduce a method to compensate and minimize hysteresis effects. Above all we show that it is possible to improve the accuracy of pressure sensors considerably and we compare our model predictions to experimental data.