Zettl Group Page

Research Highlights

Electrostatic Graphene Loudspeaker

Qin Zhou and A. Zettl

Department of Physics, University of California at Berkeley
Materials Sciences Division, Lawrence Berkeley National Laboratory
Center of Integrated Nanomechanical Systems, University of California at Berkeley
Berkeley, CA 94720, U.S.A.

The following media files are intended for public access, and may be reproduced as long as proper credit is given: "Courtesy Zettl Research Group, Lawrence Berkeley National Laboratory and University of California at Berkeley." All rights reserved © 2013.

Introduction

Efficient audio sound transduction has a history dating back millions of years. Primitive insect singers generate loud and pure-tone sound with high efficiency by exciting resonators inside their body. Male crickets generate chirping sounds via stridulation, where the scraper edge of one wing is rubbed against the ribbed edge of the other wing. Advantageous structural properties of the wings (relatively large, low-mass flexural membranes) allow extremely efficient muscle-to-sound energy transduction. In a human context, unnatural (i.e. non-voice) sound production has been explored for millennia, with classic examples being drumheads and whistles for long-range communications and entertainment. In modern society, efficient small-scale audio transduction is ever more important for discrete audio earphones and microphones in portable or wireless electronic communication devices.

For human audibility, an ideal speaker or earphone should generate a constant sound pressure level (SPL) in the frequency range from 20 Hz to 20 kHz, i.e. it should have a flat frequency response. Most speakers available today reproduce sound via a mechanical diaphragm, which is displaced oscillatorily during operation. The diaphragm, with inherent mass, restoring force (i.e. spring constant), and damping, essentially constitutes a simple harmonic oscillator. Unlike most insect or musical instrument resonators which exhibit lightly-damped sharp frequency response, a wide-band audio speaker typically requires significant damping to broaden the response. Unfortunately, "damping engineering" quickly becomes complex and expensive, with inevitable power inefficiencies.

An alternative approach to response spectrum broadening is to reduce both the mass and spring constant of the diaphragm so that inherent air damping dominates and flattens the response peaks. Moreover, with ambient air serving as the dominant damping mechanism, most input energy is converted to a sound wave, which makes such speakers highly power efficient. For these reasons, the ideal audio transduction diaphragm should have small mass and a soft spring constant, and be non-perforated to efficiently displace the surrounding air. Electrostatically-driven thin-membrane loudspeakers employing an electrically conducting, low-mass diaphragm with significant air damping have been under development since the 1920's (the first were made of pig intestine covered with gold-leaf), but miniaturized electrostatic earphones are still rare. One key reason is that the per-area air damping coefficient significantly decreases when the size of the diaphragm falls below the sound wavelength. Hence, for small speakers a thinner and lower mass density diaphragm is required to continue the dominance of air damping. Such a diaphragm is difficult to realize. If conventional materials such as metalized mylar are made too thin, they invariably fatigue and break.

Graphene is an ideal building material for small, efficient, high-quality broad-band audio speakers because it satisfies all the above criteria. It is electrically conducting, has extremely small mass density, and has been used to construct mechanical resonators. Here we demonstrate the construction of a miniaturized graphene-based electrostatic loudspeaker with excellent frequency response across the entire region for human ear, with performance matching or surpassing commercially available audio earphones.

Additional information can be found in a feature article in MIT Technology Review, found here.

Illustrations

If you use any of the following images, please include the credit "Courtesy Zettl Research Group, Lawrence Berkeley National Laboratory and University of California at Berkeley."

	Schematics of the graphene-based EDGS speaker. A graphene diaphragm, biased by a DC source, is suspended midway between two perforated electrodes driven at opposite polarity. The varying electrostatic force drives the graphene diaphragm which in turn disturbs air and emits sound through the electrodes. The light mass and low spring constant of the graphene diaphragm, together with strong air damping, allow for high-fidelity broad-band frequency response. Such a speaker also has extremely high power efficiency.
	(a) 7 mm diameter graphene diaphragm suspended across annular support frame, (b) actuating electrodes, and (c) assembled speaker.
	Frequency response of two miniature audio speakers: a) Graphene diaphragm EDGS speaker (this work); b) Commercially available Sennheiser® MX-400 magnetic coil speaker. The EDGS speaker performs noticeably better than the commercial voice-coil speaker at high frequencies, both in terms of maintaining high response and avoiding sharp resonances (the slow oscillations in the EDGES curve are due to sound wave interference in the space between the speaker and microphone and they depend on the relative position of the speaker and microphone, but the main trend is consistent). In the low frequency region, both EDGES and the MX-400 perform well. The decrease in the response curves at very low frequency are largely due to limited capability of the microphone and the inefficient coupling between the speaker and microphone.
	Vibration velocity of graphene diaphragm in EDGS v.s. frequency, measured by Laser Doppler Velocimetry (LDV). Such a measurment is useful because it eliminates extrinsic effects (e.g. acoustic structural design, sound card, microphone response), and represents the “pure” response of the graphene diaphragm itself. Within the error limit of the LDV setup, the response curve appears to be quite flat, indicating that graphene serves as an ideal key component for loudspeakers.
	An artist's rendering of the EDGS animation (mp4 format) static image (PNG format)

Recordings

If you use any of the following media or images, please include the credit "Courtesy Zettl Research Group, Lawrence Berkeley National Laboratory and University of California at Berkeley."

The following recording is made in the following condition:
The source files are 128 kbps mp3 music files, played through Creative® Audigy ZS2 sound card, voltage amplified by ~30x using operational amplifiers (AD797), played through EDGS, and recorded by a close-by SONY® ICD-SX700 Digital Voice Recorder. The raw left channel recordings are compressed into 128 kbps mp3 files.

Recordings of songs played through EDGS

Somebody that I used to know, by Gotye (128kbps mp3 format)
Sound of silence, by Simon and Garfunkel (128kbps mp3 format)
Your song, by Elton John (128kbps mp3 format)
Yellow, by Coldplay (128kbps mp3 format)
Hotel California, by the Eagles (128kbps mp3 format)

Acknowledgements

This work was supported in part by the Director, Office of Energy Research, Office of Basic Energy Sciences, Materials Sciences and Engineering Division, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231, which provided for graphene growth and characterization; by the Office of Naval Research under grant No. N00014-09-1066, which provided for graphene transfer and electrode manufacture, and by the National Science Foundation under Grant No. EEC-083819, which provided for design, construction, and testing of the device. The authors thank Yung-Kan Chen and Prof. David Bogy for assistance with LDV measurements.

External Links

The original arXiv article published online on March 10, 2013.

Last modified: Mon April 9 18:45:00 Pacific Daylight Time 2013