Wayne M. Wright

50 kHz capacitive micromachined ultrasonic transducers for generation of highly directional sound with parametric arrays

In this study, we examine the use of capacitive micromachined ultrasonic transducers (CMUTs) with vacuum- sealed cavities for transmitting directional sound with parametric arrays. We used finite element modeling to design CMUTs with 40-mum- and 60-mum-thick membranes to have resonance frequencies of 46 kHz and 54 kHz, respectively. The wafer bonding approach used to fabricate the CMUTs provides good control over device properties and the capability to fabricate CMUTs with large diameter membranes and deep cavities. Each CMUT is 8 cm in diameter and consists of 284 circular membranes. Each membrane is 4 mm in diameter. Characterization of the fabricated CMUTs shows they have center frequencies of 46 kHz and 55 kHz and 3 dB bandwidths of 1.9 kHz and 5.3 kHz for the 40-mum- and 60-mum-thick membrane devices, respectively. With dc bias voltages of 380 V and 350 V and an ac excitation of 200 V peak-to-peak, the CMUTs generate average sound pressure levels, normalized to the devices surface, of 135 dB and 129 dB (re 20 muPa), respectively. When used to generate 5 kHz sound with a parametric array, we measured sound at 3 m with a 6 dB beamwidth of 8.7deg and a sound pressure level of 58 dB. To understand how detector nonlinearity (e.g., the nonlinearity of the microphone used to make the sound level measurements) affects the measured sound pressure level, we made measurements with and without an acoustic low-pass filter placed in front of the microphone; the measured sound levels agree with numerical simulations of the pressure field. The results presented in this paper demonstrate that large-area CMUTs, which produce high-intensity ultrasound, can be fabricated for transmitting directional sound with parametric arrays.

Propagation in air of N waves produced by sparks

Weak sparks, of length 0.5–1.0 cm and energy per discharge 0.01–0.1 J, served to produce intense acoustic transients resembling N waves. Amplitude decay and waveform elongation were studied, for propagation distance up to 2 m, through the use of a wideband capacitor microphone with essentially uniform response from dc to 1 MHz. Within the range of propagation distances for which the first (compression) phase of the N wave was completely formed, the duration of this compression phase T and its amplitude ps were found to agree with the theoretical relations T=T0[1+σ0 ln(r/r0)]1/2 and ps =(r0 ps 0/r)[1+σ0 ln(r/r0)]−1/2, where σ0 is a parameter that depends upon the values of ps and T at a reference distance from the source r0. The time required for the amplitude of the head shock to increase from 5% to 95% of peak value was observed to vary from 0.45 μs (imposed by the microphone response) to greater than 2.0 μs as the wave traveled outward and as its amplitude decreased. Finally, the microphone was calibrated through use of the variation with distance of measured values of T; this new method has led to calculation of a free‐field sensitivity that agrees within ±1 dB with the results of other calibrations.Weak sparks, of length 0.5–1.0 cm and energy per discharge 0.01–0.1 J, served to produce intense acoustic transients resembling N waves. Amplitude decay and waveform elongation were studied, for propagation distance up to 2 m, through the use of a wideband capacitor microphone with essentially uniform response from dc to 1 MHz. Within the range of propagation distances for which the first (compression) phase of the N wave was completely formed, the duration of this compression phase T and its amplitude ps were found to agree with the theoretical relations T=T0[1+σ0 ln(r/r0)]1/2 and ps =(r0 ps 0/r)[1+σ0 ln(r/r0)]−1/2, where σ0 is a parameter that depends upon the values of ps and T at a reference distance from the source r0. The time required for the amplitude of the head shock to increase from 5% to 95% of peak value was observed to vary from 0.45 μs (imposed by the microphone response) to greater than 2.0 μs as the wave traveled outward and as its amplitude decreased. Finally, the microphone was calibrat...

6F-4 50-kHz Capacitive Micromachined Ultrasonic Transducers for Generating Highly Directional Sound with Parametric Arrays

Capacitive micromachined ultrasonic transducers (CMUTs) were designed, fabricated, and characterized to generate highly directional sound in air with parametric arrays. Finite element modelling was used to design the vacuum-sealed CMUTs to operate at frequencies of approximately 50 kHz and to generate pressures of approximately 139 dB (0 dB = 20 muPa RMS) at the face of the transducer. CMUTs with 40-mum and 60-mum thick membranes were fabricated and characterized. The characterized devices had center frequencies of 46 kHz and 55 kHz and bandwidths of 2.0 kHz and 5.4 kHz, respectively. Driven with a 200-V peak-to-peak excitation voltage, the devices generated up to 115 dB and 107 dB of pressure at 3 m, respectively. The devices were used to produce a narrow (8.7deg) beam of 5 kHz sound, which at 3 m was 58 dB.

Specularly scattered sound and the probability density function of a rough surface

The coherent component of specularly scattered underwater sound is sensitive to the probability density function (PDF) of displacements of the rough surface. For the specular reflection of diverging waves, the coherent component and the PDF are shown to be related by the Fourier transformation. Laboratory measurements of sound scattered at a partially shadowed nearly Gaussian model sea surface show the coherent component is much larger than would be expected for a Gaussian PDF. Fourier transformations of the measured PDF, on inclusion of a shadowing correction, gave the coherent component. Fourier transformation of the coherent component yields a surface PDF similar to the measured PDF with shadowing correction.

Generation of sound within a closed cell by an alternating current in a straight wire

The suitability of using a wire carrying an alternating current as a laboratory sound source has been examined. An analysis is outlined in which Joule heating of the wire is assumed to be communicated to the surrounding air, which is enclosed within a chamber having dimensions comparable to the sound wavelength. The predictions are compared with measurements of the sound field set up inside a cylindrical cell (15‐cm long × 5.3‐cm diam), in response to a biased ac in a straight wire parallel to the cell axis, over a frequency range (3.5–8.5 kHz) which contains the three lowest resonant modes that are independent of the axial coordinate. This acoustic field varies with source wire position according to theoretical predictions for each of these (two azimuthal and one radial) modes. An average electrical source power of 33 mW/cm in ♯40 Nichrome wire results in a maximum sound pressure within the cell of approximately 5 mPa at the 3.9‐kHz first azimuthal resonance, which is less than 0.1% of that which should ...

The design and characterization of capacitive micromachined ultrasonic transducers (CMUTs) for generating high-intensity ultrasound for transmission of directional audio

A directional source of audio sound created using a parametric array, sometimes called an audio spotlight, generates a sound beam that is much narrower than the sound beam generated by a conventional source. These directional sources require the transmission of a modulated high-intensity ultrasonic carrier wave. Capacitive transducers are well-suited for parametric array audio applications because they can efficiently generate high-intensity ultrasound with a relatively wide bandwidth. CMUTs with vacuum-sealed cavities are particularly advantageous because they lack squeeze-film damping, which increases bandwidth but reduces displacement, and because their sealed cavities and permanently attached membranes make them relatively robust. In this paper, we present the basic design constraints of CMUTs intended to generate low-frequency high-intensity airborne ultrasound. In addition, we describe a new method for fabricating these CMUTs that results in uniform cavity depths and a thick insulating oxide layer. Measurement of a fabricated devices input impedance and small-signal displacement demonstrates the success of the new fabrication method and shows good agreement with theory.

Scattering of Acoustic Impulses by a Cylinder in Air

An experiment is described in which A waves from weak sparks in air were scattered from solid cylinders. Creeping waves, which correspond to acoustic energy that propagates along the periphery of a diffracting object, were clearly distiguishable from incident waves and specular reflections.

Focusing of N waves in air by an ellipsoidal reflector

An experimental investigation of the focusing of intense, airborne pressure pulses by an ellipsoidal reflector is reported. Short-duration N waves were generated by weak sparks at the near focus F1 of four different prolate ellipsoidal surfaces. Reflection then concentrated the rays at the far focus F2. Measurements were made with a wide-band microphone, primarily along the axis of the reflector. A few transverse measurements in the plane of F2 were also made. In the axial measurements the reflected (upright N wave) and edge-diffracted (inverted N wave) components of the signal are distinct at points distant from F2, approach each other as the focus is neared, and merge at the focus. At the focus the waveform is U-shaped, or cusped. Beyond F2 a “droopy” N wave emerges and, where the edge wave can be resolved, it now arrives first and is phase inverted relative to its prefocal version. The more intense reflected wave does not experience phase inversion. Some features of the observed waveforms are consisten...

Profile of Repeated Shock Waves in a Tube

Measurements of the waveform of a plane progressive sawtooth wave in a 2‐in.‐i.d. air‐filled tube have been made with a wide‐band condenser microphone mounted flush in the rigid end plate of the tube. The source, a horn driver at the other end of the tube, was excited with tone bursts to avoid standing waves. The frequency was 3.4 kHz and the sound‐pressure level 154 dB. Under these conditions, the predicted shock formation distance is about 5 ft and the sawtooth distance about 8 ft. Measurements at 12 ft did not, however, confirm the expected perfect sawtooth shape. Although the bottom of each shock was sharp, the top was well rounded. Total delay between the bottom of the shock and the peak was about 25 μsec, i.e., about 10% of the wave period). At greater distances the delay was longer. The delay cannot be ascribed to poor microphone response. Tests with N waves in free space indicated a microphone rise time of less than 1 μsec. The rounding of the shock peaks apparently is caused by tube wall effects....

Focusing of an N wave by an ellipsoidal reflector: An experiment in air

An airborne experiment to model the intense pulse field developed by a lithotripter is in progress. The reflector is half an ellipsoid, machined from an aluminum block and having the following dimensions: major axis 280 mm, minor axis 140 mm, eccentricity 0.866, distance from aperture plane to either focus 121 mm. An electric spark at the interior focal point generates an N wave pulse. The reflected wave in the exterior region is observed with the aid of a condenser microphone having a very wide bandwidth. Of particular interest is the sequence of axial measurements from the aperture to the exterior focus and beyond. Expectations based on linear theory were that the pulse would start out N‐shaped near the aperture, become U‐shaped in the neighborhood of the focus (because of phase changes characteristic of three‐dimensional focusing), and take on the shape of an inverted N beyond the focus. Measurements confirm expectations for the prefocal and focal regions. Beyond the focal region, however, the shape is...