Elena Gaura

Overview

There are various obstacles to the achievement of perfect sensing devices, with the imperfections tending to be more prevalent in low cost MEMS sensor designs simply because the design features required to eliminate or lessen these imperfections tend to add to the design complexity and therefore manufacturing cost.

Whatever the sensing application, designers usually want measurement accuracy at the lowest possible cost. Whilst no measurement is ever obtained under ideal circumstances, the effects of interfering and modifying inputs, non-ideal sensing devices, process variations and time variations can be reduced either by changing the system design or by adding new elements to it. Filtering (at the input or intermediate stages of the signal conditioning chain) and negative feedback, added to compensation (in the analogue or digital domain) are the most common processing functions applied to the sensor read-out with the scope of improving sensor signal quality, following the process of calibration or measurement of the wsensor's characteristics. The relative "ease" with which these add-ons are produced to accompany MEMS sensors (monolithically integrated or within multiple chip solutions) justified the use of technology, and encouraged the use of microsensors as replacements for traditionally produced products. At the same time, the improved sensor signal quality was often seen, or more precisely sold as enhanced sensor functionality and brought about, in the authors opinion, two of the most controversial (and confusing) terms in the sensing field, the words smart and intelligent as applied to sensors.

The chapter will discuss briefly sources of...