MOSFET-Only filter design automation based on polynomial regression with exemplary circuits

Abstract

In this work, a novel design automation method is presented for easy construction of MOSFET-Only or MOSFET-C filters. In contrast to previous works in the literature, the proposed algorithm is simple since it uses polynomial fitting models obtained from SPICE data in order to reformulate small signal parameters and some DC characteristics of MOSFET's. Polynomial regressive models facilitate to interact between design constraints (power, area, speed, pole frequency etc.) and design variables (operating voltages, currents and device dimensions). In this regard, two novel agile MOSFET-Only filters produced with the proposed automation process are presented in order to qualify the automation flow. The produced filters are simulated in LT-SPICE and CADENCE-SPECTRE environment by using 0.18 mu m TSMC technology. These agile filters can easily be used in a wide range of applications such as encrypted communications or very high frequency circuits. In addition, it is shown that there is a good agreement between proposed design automation flow and simulations.