Selection of MOSFET Sizes by Fuzzy Sets Intersection in the Feasible Solutions Space

Abstract

A fuzzy sets intersection procedure to select the optimum sizes of analog circuits composed of metal-oxidesemiconductor

field-effect-transistors (MOSFETs), is presented. The cases of study are voltage followers (VFs) and a

current-feedback operational amplifier (CFOA), where the width (W) and length (L) of the MOSFETs are selected from

the space of feasible solutions computed by swarm or evolutionary algorithms. The evaluation of three objectives,

namely: gain, bandwidth and power consumption; is performed using HSPICETM with standard integrated circuit (IC)

technology of 0.35μm for the VFs and 180nm for the CFOA. Therefore, the intersection procedure among three fuzzy

sets representing “gain close to unity”, ”high bandwidth” and “minimum power consumption”, is presented. The main

advantage relies on its usefulness to select feasible W/L sizes automatically but by considering deviation percentages

from the desired target specifications. Basically, assigning a threshold to each fuzzy set does it. As a result, the

proposed approach selects the best feasible sizes solutions to guarantee and to enhance the performances of the ICs

in analog signal processing applications.