Volume 21, Issue 3, May 1949
Index of content:
21(1949); http://dx.doi.org/10.1121/1.1906490View Description Hide Description
The propagation of plane waves in a piezoelectric medium has been studied. A simultaneous solution of Maxwell's field equations and Newton's law of force gives the necessary conditions for wave propagation in the infinite medium. By piezoelectric coupling, every electromagnetic wave has associated with it an elastic wave. Nonpiezoelectric crystals are, at best, birefringent electromagnetically and trirefringent acoustically. In piezoelectric media, the electromagnetic and elastic modes are coupled, thus giving five phase velocities for both the electromagnetic and acoustic waves. These media are therefore electromagnetically and acoustically pentrefringent. The slower electromagnetic waves are the field variations accompanying the elastic waves, while the fast elastic waves are the mechanical deformations accompanying the electromagnetic wave.
The displacement in the elastic wave will differ in phase by ninety degrees from the electromagnetic wave, its amplitude being proportional to the electric field strength and inversely proportional to the frequency of vibration. The flux of energy will, in general, be in directions different from that of phase propagation. The generalized Poynting vector will be the sum of two vectors representing separately the flux of electromagnetic energy and the flux of elastic energy.
Analysis was made of propagation in a particular direction in sphalerite. The orientation of the vectors involved has been calculated for the fast and slow modes of vibration.
A study of specific boundary value problems gives interesting results. The piezoelectric effect gives a contribution to the reflection of electromagnetic waves at an interface. This results in a small correction to be applied to the familiar expressions for the reflection coefficients. This correction might be used to determine piezoelectric constants, providing it is possible to measure accurately the reflection coefficients and other constants involved.
Slabs of piezoelectric media may be set into vibration by an incident electromagnetic wave. At particular frequencies the crystal may be set into resonance. At resonance for the elastic wave there is almost perfect reflection of the incident electromagnetic wave. It should be possible to detect these resonances by absorption measurements.
The piezoelectric slab, if set into resonance by an acoustic wave, emits a weak electromagnetic signal.
21(1949); http://dx.doi.org/10.1121/1.1906491View Description Hide Description
The absorption of a plane wave of sound by a sphere is computed. The calculations are based on the assumption that the complex ratio of sound pressure at a point on the sphere's surface to the normal component of particle velocity is a constant independent of the direction of incidence (“normal impedance assumption”). Absorption measurements were made on hair felt‐covered spheres placed in a reverberation room, and were compared with the computed absorption by means of the reverberation room statistics appropriate for spheres. The theory and measurements both show that absorbent spheres can have absorption coefficients greater than unity. The discrepancies between theoretical and experimental coefficients seem to indicate that the normal impedance assumption is not valid for the hair felt used in the experiments.
21(1949); http://dx.doi.org/10.1121/1.1906492View Description Hide Description
Improvements to an ultrasonicinterferometer are discussed. As a result a greater sensitivity to acoustic reactions is obtained and the reproducibility of the data is greatly improved. Velocity measurements in dried ‐free air, dried , and dried are given. Amplitude absorption coefficients for , , and are also included. Measurements are made over the temperature range from 9°C to 36.6°C and over the pressure range from 82.17 cm Hg to 0.45 cm Hg. The frequencies extend from approximately 500 kc to 2.16 mc. A maximum value for the absorption in , attributed to molecular absorption, is located at an f/p ratio of approximately 10 mc/atmos.
21(1949); http://dx.doi.org/10.1121/1.1906493View Description Hide Description
Supersonic velocities and adiabatic compressibilities of a series of mixtures of water‐ethyl alcohol‐sodium chloride have been found. In general, maxima of sound velocities and minima of adiabatic compressibilities were obtained. With increase of salt concentration, the depth of the minimum of compressibility decreased and shifted towards the value for alcohol free‐salt water, and for 3N concentration of salt, no minimum at all remained. The phenomenon has been explained as caused by the depolymerization of water by sodium chloride. The explanation supports earlier results on the water alcohol system.
21(1949); http://dx.doi.org/10.1121/1.1906494View Description Hide Description
The motional impedance of a magnetostrictive element is determined by a unique method involving embedding of the element in a block of Permafil, a non‐magnetic plastic. The method used can be adapted to determine the motional impedance of any odd‐shaped magnetostrictive element. The possible application of this method to the determination of magnetostrictive constants is pointed out.
21(1949); http://dx.doi.org/10.1121/1.1906495View Description Hide Description
An analysis of the substitution method of measuring the open circuit voltage generated by a microphone is given which shows that the “normal” substitution voltage equals the open circuit voltage for all types of acoustic measurements and for any value of electric impedance loading the microphone. It is shown that the method recently proposed by some authors of removing the acoustic load from the microphone when applying the substitution voltage results in a substitution voltage which does not equal the open circuit voltage. It is also shown that a formula for the response of a transducer derived for a system in which the microphones are open‐circuited may be used when the microphones are terminated by finite electrical impedances, by replacing the generated open circuit voltages in the formula by the corresponding “normal” substitution voltages.
Consideration is given to the restriction in the definition of the pressure response of a transducer made necessary by the fact that the pressure on a microphone diaphragm is a function of the electrical impedance terminating the microphone.
An experiment is described which involves a microphone coupled to a chamber, the acoustical impedance of which is high relative to that of the microphone. The results of this experiment agree with the conclusions of the analysis.
21(1949); http://dx.doi.org/10.1121/1.1906496View Description Hide Description
The earlier work on the comparison of the JRB and the Massa Artificial Ear couplers has been continued to seek the explanation of certain differences observed. The later work, for the most part, has been undertaken on couplers under active consideration for adoption by the American Standards Association.
The difference in pressure in 6‐cc couplers as measured by the Western Electric 640‐AA and the Massa M‐101 microphones appears to be identified with the strong earphone coupler resonance peak rather than directly with the shape of the coupler. The effect is apparently associated with the diaphragm impedance of the microphone more closely when the microphone is in use as a measurement device than when merely present, but inactive, in the coupler, except in the case of the M‐101 microphone whose presence gives rise to an unexplained increase in pressure.
For general earphonecalibration work, the results obtained by means of any of the coupler systems will be satisfactory. For some physical purposes, however, the effects may lead to significant errors.
21(1949); http://dx.doi.org/10.1121/1.1906497View Description Hide Description
The conditions under which titanateceramics may be made to have piezoelectric properties suitable for application to transducers are described. These conditions involve a selection of the titanate material to give a Curie point well above the working temperature,polarization from an external voltage source within certain limits of potential gradient and time duration, and mechanical structure. Most common transducers have a mechanical impedance far below that of a solid block of ceramic. Multilamellar strips have been developed to provide the mechanical impedance required in a phonograph pick‐up, microphone, or the like. These consist, in general, of a metal strip with a thin plate of ceramic soldered to one side or with thin plates soldered to both sides. As the multilamellar strip cools below the solidification point of the solder, the ceramic becomes pre‐stressed in compression. This strip is then used in flexure in a transducer. It has many unique properties, especially for applications involving extreme conditions of humidity.
Applications of this strip for a phonograph pick‐up and to a microphone are shown.
Precautions involved in controlling production of a pick‐up are discussed.
21(1949); http://dx.doi.org/10.1121/1.1906498View Description Hide Description
Two sources of non‐linear distortion in a dynamic loudspeaker are considered in this discussion; both are related to the magnetic characteristics of the driving mechanism The first type of distortion arises because of a force of attraction between the voice coil, carrying a current, and the iron of the field structure. This force varies as the square of the current and produces second harmonic distortion. The force may be related to the space rate of change of self‐inductance of the voice coil as it moves in the air gap. The magnitude of the distortion produced in this way may be several tenths of one percent, and is greater for low frequencies and large currents. This distortion may be reduced by proper proportioning of the voice coil and field structure and by using a short‐circuited winding on the field structure.
The second type of distortion arises due to non‐uniformity of the magnetic field in which the voice coil moves. This effect is well known qualitatively, but equations are given here for its quantitative evaluation. These equations indicate that the distortion characteristically is less than one percent, and is greater for large amplitudes of motion. If the voice coil is centered in a symmetrical field, only odd‐order distortion is produced. If the field is not symmetrical about the voice coil, even‐order distortion is also present. This distortion may be reduced by proportioning the voice coil and field structure so that the mean field in which the coil moves remains as constant as possible.
21(1949); http://dx.doi.org/10.1121/1.1906499View Description Hide Description
In order to isolate some of the factors affecting the internal tuning of brass instruments several differently tapered bells were made and a trumpet made up such that the different bells could be easily attached and detached. Tests were then run on the internal tuning using both a physical measure of the trumpet air column resonant frequency and a group of five players. The results were analyzed statistically and indicate that changes in the shape of the bell taper as great as those used for these tests caused significant changes in internal tuning, and that these changes were functionally related to the type of taper. Secondary results of the analysis gave measures of the magnitude of the differences among players and trials and showed the relationship between the player average and the physical measure of air column resonance.
21(1949); http://dx.doi.org/10.1121/1.1906500View Description Hide Description
The Composer's Music Box is an electro‐acoustic tone generator producing single tones. It is played in much the same way as a musical score is written. It has a pitch range of about an octave and an intensity range of about 40 db. Electrical circuits provide nearly constant tone structure independent of the fundamental frequency, and provide considerable variation in attack and decay times.
21(1949); http://dx.doi.org/10.1121/1.1906501View Description Hide Description
If we assume that in the evolution of the mammalian ear nature has made the best possible use of physical and acoustic principles, we are led further to assume (1) that the middle ear provides maximal sensitivity combined with optimal quality and (2) that the middle ear is so adjusted as to minimize noise pick‐up from the speaker's body during talking. In terms consistent with these assumptions, it is possible to explain: why mammals have two vocal cords; why a bony rod lies on the eardrum; why the point of rotation of this bony rod is at the edge of the eardrum; why the mass of the ossicles appears so exaggerated; why the eardrum is conical in shape; why the membrane of the footplate of the stapes is so small; why the footplate of the stapes has an elliptical form; why a joint is necessary between the stapes and the incus; and why an animal with thin head bones has a bulla.
21(1949); http://dx.doi.org/10.1121/1.1906502View Description Hide Description
An experimental technique was developed to measure, for intact middle and inner ears, the volume displacement of the fluid in the cochlea produced by a given sound pressure on the eardrum. In order to determine whether it is possible to immobilize the stapes by increasing the static pressure in the cochlea, the transmission characteristics of the eardrum‐to‐round‐window system were measured with the ear under the maximum static pressure that can be produced in the cochlea without bursting the blood vessels. Since no important change in transmission was observed, it was concluded that immobilization of the stapes cannot be produced by an increase in static pressure in the cochlea.
With the same technique the mechanical impedance of the movements of the stapes footplate, the round window, and the fluid in the cochlea were measured singly and combined as in the normal ear. Once the elastic properties of the cochlear partition were known, it was possible to construct an ear model that exhibited patterns of vibration similar to those found in the cochlear partition of the human ear. By changing the shape, the size, and the fluid viscosity in this ear model the relative importance of the different parts of the inner ear were studied.
These experiments made it plain that the pattern of vibration of the cochlear partition is determined primarily by the volume elasticity of the membrane and the fluid close to the membrane. Length, height, shape, and diameter of the cochlear canal do not influence the vibration pattern. A mathematical description of the hydrodynamic properties of the inner ear must take into consideration this independence of the physical dimensions of the cochlear canal.
21(1949); http://dx.doi.org/10.1121/1.1906503View Description Hide Description
In order to determine the way in which mechanical analysis of frequency occurs in the cochlea, a technique was developed for preparing anatomical specimens. This technique made it possible to measure the amplitude of vibrations along the cochlear partition when the stapes was driven with pure tones. It made it possible, further, to determine how the amplitude of vibration at any given place on the cochlear partition varies with the frequency. The relation of amplitude of vibration at a given place to frequency gives a curve very similar to a resonance curve. We can then calculate the logarithmic decrement characterizing the vibrations of the cochlear partition. This log decrement is almost independent of frequency and has a value of d = 1.3. The decay time of the vibrations of the cochlear partition was observed directly with a microscope, and the values obtained in this way agreed with those derived from the resonance curves. All the measurements have been made on preparations of the human ear.
21(1949); http://dx.doi.org/10.1121/1.1906504View Description Hide Description
Continuous recorded speech passages were presented against a background of white noise and against a quiet background. The effect of the noise on the loudness of speech was determined by having listeners adjust the level of speech heard in quiet to sound equal in loudness to an assigned average level of speech presented against selected levels of background noise. In general, it was found that (1) low noise levels produce a proportionately lower depression in the loudness level of speech than higher noise levels, (2) as a first approximation, however, the effect of noise on the loudness of speech is a function of the speech‐to‐noise ratio rather than of the level of the speech alone or of the noise alone, and (3) under none of the experimental conditions did white noise actually increase the loudness of speech (the “Egan effect”).
The assumption that a given level of noise subtracts a constant number of loudness units from speech (regardless of the level of the speech) was found to be fairly adequate in describing the experimental results.
21(1949); http://dx.doi.org/10.1121/1.1906505View Description Hide Description
Some 40 or more years ago, there was evidence of interest among scientists in the sounds of mosquitoes. Despite very crude acoustical apparatus, significant conclusions had been reached by the early investigators. Although very little was done in the next three decades despite the growth in the field of acoustics, interest has recently been renewed because of the world‐wide need to check the spread of mosquito‐borne tropical diseases. Recent experiments have confirmed many early findings and have provided valuable new data.
Mosquito sounds are distinctive; gross differences occur that are readily detected both by listening and by waveanalysis. All fundamental sounds occur in the center of the sonic range (200–1000 e.p.s.); all are rich in harmonics. All sounds are warble‐modulated, some at a single low rate in the order of 5 c.p.s., others at a double rate with the same lower rate and a higher rate in the order 5 times the lower. In some tones, some harmonics are completely interrupted while the fundamental remains undisturbed. Warble amplitudes are quite large; 20 percent is not unusual. The fundamental pitch often drifts; in one case, for example, there was a 25 percent increase in as little as 0.05 sec. Male sounds seem higher pitched than female; sound spectrograms show this to be due more to the difference in harmonic emphasis than to the difference in fundamental pitch. Generally speaking, mosquitoes do not generate or respond to sine‐wave tones.
Mosquito sounds are low in energy level; the total sound power output of a female Anopheles albimanus whose sound was used to bait a mosquito‐killing trap in Cuba was in the order of 10−13 watt. With this low power level it is difficult to obtain recordings with the high signal‐to‐noise ratios customary in high quality commercial recording because of the relatively high noise levels. Despite this handicap, it has been possible to reduce the noise due to microphonics, hum induction, vibration, and airborne and other disturbances to the point where recordings with 40‐db signal‐to‐noise ratio are made frequently.
21(1949); http://dx.doi.org/10.1121/1.1906506View Description Hide Description
- LETTERS TO THE EDITOR
21(1949); http://dx.doi.org/10.1121/1.2142328View Description Hide Description