Suppression of Oscillator Bias Voltage Phase Noise via MEMS Resonator Arraying


Microelectromechanical system (MEMS)-based oscillators are in the heart of many of our electronic devices today, forming the timing basis for our increasingly higher frequency circuits. High-performance, low-noise oscillators are critical to meeting the standards of today's communication protocols. Effects of noise from all inputs of a circuit to the final output spectrum should be understood to make better oscillators, and this report specifically considers the effects of noise on the bias voltage supply needed for strong electromechanical coupling to sustain resonance. We investigate how arraying multiple MEMS resonators---putting multiple copies of the same resonant structures in parallel and mechanically coupling them so they resonate together---for use in an oscillator can potentially lead to better resilience against phase noise induced by this source of noise compared to a single resonator.

We set out to first measure the sensitivity of high-Q wineglass-disk-based MEMS oscillators to bias voltage noise to determine which types of noise (noise from direct mixing versus noise from resonant frequency modulation) dominate at which frequency regimes. This establishes a baseline that we can compare against prior work in this area done on a different resonator structure. Then we construct oscillators around different sized arrays to determine the effects of arrays on this sensitivity and see if they can provide better noise performance.

We show how there is improved close-to-carrier phase noise performance at the cost of potentially increased power consumption due to increased loading capacitance for an array compared to a single-device-based oscillator. Without increasing the power, measurements show degraded far-from-carrier noise rejection for arrays due to lower output oscillation level.

Publication date: 
May 13, 2022
Publication type: 
Master's Thesis
%0 Thesis %A Ni, Jeffrey %T Suppression of Oscillator Bias Voltage Phase Noise via MEMS Resonator Arraying %I EECS Department, University of California, Berkeley %D 2022 %8 May 13 %@ UCB/EECS-2022-119 %U %F Ni:EECS-2022-119