Date of Award

5-2017

Degree Type

Thesis

Degree Name

Master of Science

Major

Electrical Engineering

Major Professor

Benjamin J. Blalock

Committee Members

M. N. Ericson, Charles Britton, Jeremy Holleman

Abstract

Low-power analog filter banks provide frequency analysis with minimal space requirements, making them viable solutions for integrated remote audio- and vibration-sensing applications. In order to achieve a balance between the length of deployable service and system performance, a critical requirement of such remote sensor networks is low-power consumption, due to the constraints imposed by on-board battery cells.

In this work, the design and implementation of a sub-threshold complementary metal-oxide semiconductor (CMOS) integrated low-power tunable analog filter channel for Oak Ridge National Laboratory is presented. Project specifications required a tunable, high-order, monolithic bandpass filter channel with small chip area and low power consumption. With initial design focusing on the audio frequency spectrum, the 8th [eighth] order filter channel presented in this work provides an effective Q-factor of 4.5 and a minimum dynamic range (DR) of approximately 85 dB, while allowing for tuning across a range of center frequencies from 2 kHz to 100 kHz, with power-consumption of a single 8th order filter channel measured at 155 μW [micro], nominally. An integrated analog Gm-C [transconductance-capacitance] filter topology was selected for this application. Functionally, the high-Q bandpass filter transfer function is implemented via four cascaded filter cells, resulting in a single 8th order filter channel, fabricated in 130-nm 1.2 V CMOS technology, suitable for use in monolithic integrated spectral analysis applications.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS