The Exploitation of Variability and Noise in Micro Devices

Abstract

The strength of Micro Systems Technology (MST) is the ability to fabricate a large number of small devices economically. However such devices tend to have errors caused by the variations of fabrication and inherent noise signals such as Brownian motion or Johnson noise. This paper develops the understanding of Micro Switch Group Sensors (MSGS), which works by exploiting this random noise. In the experimental work reported here, an MSGS comprising of 20 switches has been built using electronic circuits and tested to verify the performance. The output of the device is simply the number of switches turned on, this is then transformed into the expected mean value of the input signal using standard statistical theory. The performance of the devices was shown to be in good agreement with the theoretical prediction. The linearity and the standard deviation of the output signal of MSGS are investigated and it is concluded that an MSGS may be successfully applied as a measurement tool.