US2472714A - Piezoelectric sound pressure microphone - Google Patents

Description

June 7, 1949. F. MASSA 2,472,714

. PIEZOELECTRIC SOUND PRESSURE MICROPHONE Filed April 16, 1945 7 I 3 s 3 6 VII/ l. I V 7 51,

DEGREES PHASE Sill/'7' I F'RE 9. IN KC.

Jnventor Patented June 7, 1949 UNITED STATES PATENT OFFICE rmzoerec'rm'c SOUND PRESSURE MICROPHONE Frank Massa, Cleveland Heights, Ohio Application April 16, 1945, Serial No. 588,692

4 Claims." (Cl. 179-110) My invention is concerned with an improved point before the microphone was introduced.

Microphones have been designed which achieve this overall requirement over portions of the audio-frequency range. The purpose of this invention is to describe a new type of construction which achieves the basic requirements of a good 7 The convenience of having a really flat fre- :q'uency response characteristic is practically selfevident. Although variations from flatness of a microphone response are generally tolerated in the acoustic laboratory, .even though considerable time may be employed in correcting for the microphone variationsover the frequency :range, there are certain'measurements that can- :not :be made when such variations exist. Among such measurements are those involving distortion measurements on sound generators. The apparent harmonic content in the sound Wave would be very greatly dependent on the degree of variation from flatness of the measurement microphone. Attempts to compensate for irregularities :in a .microphone response by means of electrical networks, in addition to beinga generally cumbersome j ob, introduce electrical phase shifts in the system that, in themselves, may cause further errors when using themicrophone for the measurement .of "wave shape .of the sound wave. An important criterion, therefore, for .a good measurement standard is the extent of the frequency range over which its response is trul .flat.

2. Large dynamic range To .be generally useful, the electrical output of a standard shouldbe linear with respect to the sound pres-sure over wide ranges of amplitude. .A good standard should not berestrictedin its field of use only to soundpressure measurements has hitherto urement of extremely high sound pressures such as would exist near the throat of a horn loud speaker or inside enclosures such as sections of sound filters .(mufilers) or conduits. To realize a large dynamic range, it is of primary importance that the mechanical constants of the vibrating system be inherently linear for large variations in applied pressure.

3. Small physical size An important criterion that must be met bya good working standard is that its physical size should besmallas compared with the wavelength of the sound pressure being measured. If the structure is large enough to cause appreciable difiraction, the microphone will show sensitivity variations with angle, and when placed in use it will require a knowledge of the direction along which .the sound wave is travelling as Well as accurate alignment of the microphone with respect to the axis along which its calibration was made. Another disadvantage of a large physical size is the possibility of standing waves being established between the sound source and the microphone, which may make sound pressure measurements at high frequencies impossible.

For sound fields whichare not free progressing waves with a known direction of propagation, there is no assurance or being able to measure true sound pressure unless the microphone is physically small compared with'the wavelength. "For example, "if sound pressures are to be measured within an enclosure, such as the determination of pressure distributionalong a conduit or in various compartments of a noise filter, it is impossible to do it accurately with a microphone of such physical size that its active area integrates a number of different pressure components acting over its entire face.

In some cases, even for work in the lower range of frequencies, it ls necessary to have a microphone'which is physically very small, even though the wavelength is relatively large. A typical example is one in which the sound pressure measurements are to be made inside relatively small chambers in which the insertion of the micro phone must not appreciably upset the physical dimensions of the enclosure.

4. Smooth electrical impedance Even though a microphon may have a very flat frequency response characteristic, it is important that its electrical impedance characteristic be such that it easily lends itself to use with conventional electronic circuits; otherwise it might become impractical to make use of its flat open circuit sensitivity. A generally uncle sirable impedance characteristic is one in which large variations in'magnitude occur over relatively small ranges in frequency. A very desir- 3 able type of electrical impedance for the microphone is one represented by a single circuit element over the entire audio-frequency range so that it will be easily possible to make simple adaptations of th microphone to input circuits for special uses. special uses include the desirability of being able Some of the more common a true plane surface to the sound wave by the to easily attenuate the microphone sensitivityuniformly over the entire frequency range at the input to the grid circuit to prevent overloading in cases where extremely high sound pressure measurements are to be made. simple circuit element will make it easy to vary the time constant of the microphone circuit, if desired. Equally simple will be the possibility of introducing filter sections at the input stage to permit special measurements in limited frequency regions when improved signal to noise ratio may be necessary in measuring very weak sound fields.

5. High mechanical impedance One of the most important requirements for a sound pressure measurement standard is that its mechanical impedance shall be extremely high over the entire audio-frequency range. The importance of this particular requirement is not generally appreciated. It is usually assumed that as long as a free field calibration of a microphone is known, it may be used as a good working standard. The fallacy in this reasoning is that the microphone only gives the correct interpretation of sound pressure when used under the identical environment in which the calibration was made. If the microphone is used under conditions other than free field (assuming free field calibration), there is no assurance that the microphone will measure true sound pressure except in the frequency region in which its mechanical impedance is high.

, If the mechanical impedance of the standard microphone is not higher than the mechanical impedance of the environment in which it is used, the insertion of the microphone at the point where the sound pressure measurement is to be made may greatly disturb the sound pressure which existed at the point before the microphone was installed. Under such conditions, the measurement obtained might be in very considerable error.

One of the best means of achieving the important requirement of high mechanical impedance is to employ a stiffness-controlled mechanical system. Of the numerous attempts which have been made to produce a good stiffnesscontrolled working standard, the miniature condenser microphone described by W. M. Hall in the Journal of the Acoustical Society of America, vol. i, page 83 (1932), and also by Harrison and Flanders in the Journal of the Acoustical Society of America (Supplement to July, 1932) has been generally accepted as best meeting the basic requirements. My invention overcomes the limitations which are still present in the miniature condenser microphone and thereby approaches the idealized requirements for a working standard to a much greater degree, as I shall show.

Other desirable features of my invention will become clear after reading the objects and specifications which follow:

An object of my invention is to provide a microphone which measures true sound pressure in a sound field Without disturbing the pressure wave phone was introduced.

Also, a

- pressure sensitive portion of the microphone, thus completely eliminating any cavity ahead of this surface such as is caused by clamping rings in conventional types of microphone structures.

A further object of my invention is to produce a mechanical vibrating system in a sound pressure microphone which is true stiffness controlled over the entire audio-frequency range extending well beyond 15,000 cycles.

Another object of my invention is to provide a microphone which behaves as a simple vibrating mechanical system in which the fundamental mechanical resonance occurs very much higher than 20,000 cycles.

Still another object of my invention is to provide a microphone in which the electrical impedance of the unit is practically that of a pure condenser over the entire audio-frequency range to at least 20,000 cycles.

A further object of my invention is to provide a microphone whose generated voltage is proportional to sound pressures over very large ranges of sound pressures exceeding millions of dynes per square centimeter.

Another object of my invention is to produce a structure for use as a sound pressure measurement standard in which the stability of the system is very high and not subject to change with time and room temperature variations.

Still another object of my invention is to produce a true pressure microphone standard which is very easy to use and convenient to operate.

Another object of my invention is to provide a microphone which is very simple to use as a measurement standard and not subject to the necessity of requiring polarizing voltages for its operation.

A further object of my invention is to provide a method of assembly of the component parts of the Still another object of my invention is to pro- 1 vide a preamplifier of such design that the microphone may be placed in a sound field with minimum obstruction offered to the sound.

Another object of my invention is to provide a novel means of shock-mounting the microphone unit from the amplifier housing.

The novel features that I consider characteristic of my invention are set forth with particularity in the appended claims. The invention, itself, however, both as to its organization and method of operation, as well as advantages thereof, will best be understood from the followin description of several embodiments thereof, when read in connection with the accompanying drawings, in which- Fig. 1 is a partially cut-away view looking end-on at the microphone.

Fig. 2 is a longitudinal section through the microphone taken along 2-2 of Fig. 1.

Fig. 3 is an enlarged view showing the details of the terminal connections of the microphone.

Fig. 4 is an endview showing the assembly of h. group of piezo-electiic crystal plates as employed in the microphone construction.

Fig. isa side view of Fig.4.

Fig. 6 is a .view showing a single crystal plate of Fig. 5 with a conducting surface or electrode applied to a portion of the crystal face.

Fig. 7 shows the crystal plate of Fig. '6 after electrical contact strips are attached.

Fig. 8 shows an importantstep intheassemb1y procedure for the microphone in which very accurate alignment of the crystal assembly to the base plate is automatically maintained.

Fig. '9 is an end view of a portion of the device shown in Fig. 8.

Fig. is a graph which gives the data showing the deviation from true stiffness control at the higher audio'frequencies of :a small stretched -diaphragm as comparedwith no deviation for'the microphone describediin this specification.

Referring more particularly to the figures, Rig. .1 .showsan end view of "the microphone assembly, and Fig. 2 is a longitudinal section through the assembly. The basic components which make up the microphone are a thin diaphragm i, an assembly of piezo-el'ectri'c crystal plates 2, a rigid :base 3, and an outer housing '4. The crystal leads 5 pass through a pair of electrically insulating bushings '6, which are forced fit into the base 3 A pair of conducting terminal pins Tare driven mto the bushings B establishing permanent connection to the crystal leads 5. The outer housing 4 is a cylindrical shell that is :made a forced fit over the undercut shoulder 3 of the 13886 3. corners is cut through the opposite end of the housing 4, as indicated by the cut-away portion of l, to result in a small clearance between the crystal assembly 2 and the sides of the opening.

One of'the basic features of the construction of the microphone is the obta'inment :of a high "degree of mechanical precision in the assembly so that the final performance will be stable over a :long period Of time. To realize this precision, the constructionis builtup -around the base piece Q whose surface :to which the crystal assembly .2 is cemented is made perfectly flat. The shoulder 3 on base 3 is machined exactly at right angles to the flat surface on which the crystal assembly is cemented. Starting with these two accurate surfaces on the base3, the entire microphone as- :sembly is built up, as will be described later in connection with Fig. 8 and Fig. 9.

To bring out the electrical terminals 1 for con- .nection to the crystal leads, a pair of insulating bushings 6 are forced into the base 3, as shown in the enlargeddetail in Fig. 3. Each of the flex- ..ible leads 5 passes through a hole in one of the insulating bushings 6 anda conducting terminal .pin 1 is driven into the hole in each of the bush- :in-gs 6, establishing permanent contact with the lead 5 and simultaneously providing a. pair of external contact tips through which external connection may be established to the microphone.

Fig. 4 and Fig. 5 are top and side views of the crystal assembly 2 which is the active element of the microphone. I prefer to employ an even number of crystal plates in the crystal assembly so that the two outside plates will then be at common potential and can be kept nearest to ground potential when in use, if desired, thus reducing the magnitude of the leakage capacity -:between the crystalassembly and ground, which is advantageous in minimizing sensitivity loss in A rectangular opening with chamfered of crystal material will be discussed later.

'the microphone'circuit. Six'plates 8, 9, i0, H,

12, and iii are indicated in the particular crystal assembly shown in Figs. 4 and 5. Although I prefer to use several slabs of crystal connected in parallel in order to realize a suiilciently high value of capacity for the assembly, it is also possible to employ :a single ithi'ck crystal plate for the active element. If this is done,the capacity of the element will be very low and the leakage capacity from one side of the crystal to the housing will introduce a shunting sensitivity loss across the crystal. The low value of capacity will also introduce otheradifil'culties in the ampliyfier circuit to permit hat response to very low frequencies.

Common potential faces of the crystal are marked in Fig. .4. The six-"common sides are-brought out to the right, as shown, by three common'leads 15. The topand bottom leads l5 pass over insulating strips l5 'whichare attached to the crystals, asshown in Fig. 5, and join with thecommonrcenter lead I5to connect .to the flexible conductor 5. Asimilar construction is followed for the leads :of opposite polarity which arebrought-out to the left of Fig. 4. Four crystal leads 15 are necessary, and three insul'ating pieces I6 are used for the .lefthand connection, as is obvious from :the details of Fig. .4. The two end crystalsfi and 13 are chamfered, as shown, to avoid the necessity .of cutting right angle corners in the housingrl of- Fig. 1, which, if done, would make the .radial distance of the four corner regions over which the diaphragm is ce-- mented too "small unless the outside diameter increased. 7

The construction of the individual crystal plates is'shown for a typical plate 10 in Fig. 6

and Fig. I. The piezo electric crystal plate 10 is an expanderplate in which a pressure applied along one axis causes a charge to be developed along aright angle axis. .-;Typical types of crystal plates that may be employed are 45 X-cut or 45 Y-cut Rochelle salt, 45 Z-cut primary ammo- .niumphosphate, 01"71" ZX-cut quartz. The choice In preparing the crystal assembly 2 the plates, for

exampleplate M3,:are cut of such dimensions to produce the desired sensitivity, impedance and resonant frequency for the microphone. On a pair of opposite faces, a-conducting film M is applied, preferably with margins near the top and bottom edge, as shown, so that the stray leakage capacity produced in the microphone between the crystal assembly 2, the base 3.rand the diaphragm -l is reduced toa minimum. The conducting film -l=4 may be applied by the method described in my co-pending application, Ser. No. 522,196, filed February V10, 1944, now abandoned, or by the evaporation of gold or other metals in a vacuum, or by cementing thin conducting foil to the bare crystal, or by painting a conducti'ng cement over the desired surface. In order to bring out suitable electrical connection to the crystal electrodes, the thin contact strip 15 is attached .over an edge of the electrode 14, as shown in Fig. 7. .A strip of similar thickness 1-511. is placed along the opposite edge of the electroded surface, as shown in Fig. 7, so that when the plates are assembled as agroup, as shown in Figs. 4 and 5, each plate will remain parallel to each of the other plates.

When the crystal :plates are assembled as a group, as shown in Figsmi and 5, the plates are cemented and held in a .fixture which keeps all surfaces of the assembly mutually perpendicular. .A cement aloaded-with-a metallic powder to in- 7 crease the conductivity is preferably employed when assembling the crystal plates so that the strips I 5, I which make common connection to its neighboring electrode surface will make the connection with negligible contact resistance.

After the crystal assembly 2 is complete, the surface on the end which is to be cemented to the base 3 is made perfectly fiat by dressing it off while the assembly is rigidly held in a fixture which holds the long sides of the crystals exactly at right angles to the end surface being dressed. The next stage in the construction consists in cementing the prepared end of the crystal assembly 2 to the flat surface of the base 3. To maintain extreme accuracy in performing this operation, the fixture shown in Fig. 8 is employed. After the crystal assembly 2 is approximately located on the base 3, with cement positioned between the crystal assembly and the base, the guide I! is carefully placed over the crystal and the crystal adjusted until the guide l1 engages the shoulder 3 of the base 3. Locating means, not shown, are provided so that a predetermined location is maintained for the position of the guide piece I"! around the periphery of the base 3. A top view of the guide I! is shown in Fig. '9 indicating the four fiat guide surfaces H which clear the crystal assembly by a small margin, thus causing very accurate location of the crystal with respect to the base 3.

With the guide I! in place to maintain accurate alignment of the crystal assembly 2 with respect to the base 3, the entire sub-assembly is placed in a small rigid press consisting of a framework 2| and a screw 20 which applies pressure through the plate I9 and the pressure pad l8 to cause intimate contact of the crystal surface to the base 3. While the assembly is under pressure, the cement is allowed to set to produce an intimate bond between the entire area of the end of the crystal assembly 2 and the base 3. After the assembly of the crystal and base is completed, it is possible to visually inspect the joint in order to observe that the work has been properly carried on. The importance of having a good intimate and continuous joint between the crystal and the base is due to the fact that any imperfection in this joint is reflected through as a change in the mechanical impedance of the system with resultant variations in the performance of the microphone.

Having accurately made the assembly of the crystal to the base, the housing 4 is pressed into position around the crystal assembly, as indicated in Fig. 2. Due to the accurate alignment of the crystal assembly on the base, the rectangular opening on the diaphragm end of the housing will allow a uniform small clearance, preferably less than "1 5", to remain between the crystal assembly and the edges of the opening, as shown in the cut-away portion of Fig. 1. The reason for having such a small clearance is that when the diaphragm l is assembled in place there will be no appreciable areas of diaphragm which are unsupported and which might give rise to spurious variations in the response at the higher frequencies.

In order to insure a perfect assembly of the diaphragm l to the surfaces of the crystal assembly 2 and the flat portion of the end of the housing 4, I find it advantageous to make the crystal assembly 2 slightly longer than necessary and to dress off the excessive crystal material by rubbing the end of the assembly on a fiat abrasive surface until the exposed crystal and housing surfaces are brought into the same plane. In the case of quartz which is more difficult to wear down, it is possible to attach securely a thin sheet of Bakelite or similar substance on one or both ends of the crystal assembly and during the final dressing operation the exposed Bakelite surface may be more easily brought into accurate alignment with the plane of the exposed flange surface of the outer housing to prepare the unit for the assembly of the diaphragm.

The diaphragm l is relatively thin and does not contribute to the mechanical constants of the vibrating system. It merely acts as an enclosing sound sensitive surface for the microphone. After bringing the exposed crystal surface and the housing flange into the same plane, the surfaces are coated with cement and the diaphragm l, which is likewise coated with cement is put in place, as shown in Fig. 2. The complete assembly is then mounted in the same type of press 2| as shown in Fig. 8 and the unit is allowed to set under pressure, thus completing the assembly. In the choice of material for the housing 4, I prefer a metal having a coefficient of expansion nearly equal to that along the length of the crystals so that there will be a negligible static stress set up along the axes of the crystals due to changes in temperature. For a microphone employing primary ammonium phosphate as the crystal element, a housing of cadmium would be the most desirable since the coefiicient of thermal expansion for cadmium is practically identical with the coefficient of expansion of the crystal along its length. Zinc would also be usable with a coefficient of expansion approximately 10% less than the coefficient for the crystal. For the case of quartz in which the coefficient of thermal expansion is very low, one of the nickel alloys having low temperature coefficients could be selected for the housing material.

It is to be noted that the method of attaching the diaphragm does not require the use of a clamping ring; therefore, this construction completely eliminates any cavity ahead of the diaphragm such as occurs in conventional diaphragm assemblies. The absence of such a cavity prevents the occurrence of any acoustic resonance that would otherwise be present in the response in the higher frequency region.

A microphone built as shown in Fig. 2, in which the outside dimensions of the housing are approximately diameter by la? long, offers a fairly good compromise between smallness and sensitivity. For such a structure a clearance of the order of between the crystals and the housing opening will give satisfactory performance. It is no trouble to halve the dimensions of the structure if a still smaller physical size is desired in cases where most of the work is to be carried on above 15,000 cycles. If this is done, however, both the capacity and the sensitivity will decrease, which means that it will become more diflicult to work with the amplifier circuits at the extremely low audio frequencies. This, on the other hand, would be of no serious concern if the unit were to be used primarily at the higher audio frequencies.

Although several types of crystals may be used in the microphone, I prefer to employ primary ammonium phosphate as a satisfactory substance for most general purpose applications. X-cut Rochelle salt would have higher sensitivity but its electrical impedance and mechanical stiffness would vary with temperature which would offer some trouble in cases where filter circuits would mins be desired at'theinput tothefirst Another disadvantage. of Rochelle. salt-is. its 'rela. tively low melting point: of approximately 135? El, which would limit the fieldof applicationof the "standard. Quartz. would have a muchhigher. melting point than the primary ammonium phos phate; however, a microphone employing quarts of the same physical dimensions and designed for the same capacity willl havc'approximatcly 15 db. less sensitivity than primary ammonium phosphate.

' The melting point of primary ammonium phosphate is'above 350 F. and a microphone emplaying this crystal. can be used to at least 200- F". without difficulty. Since this particular crystal material shows a definite volume conductivity which is a: iunctlon of the impurities in the subs stance, it is lmportant to choos'e'the material for high resistivity if fiat response is desired down to cycles. Using primary ammcnium Phosphate in the microphone above described, in. which the overall dimensionsarc diameter by H long, gives an electrical capacity equal to approximately 100 mi. and av sensitivity of about 26 microvoltsfdyne/cm. pressure applied to the diaphragm throughoutthe entire audio-frequency range.

The use of the piezo-electric construction just described has several advantages over the small stretched diaphragm condenser microphone for use as a sound pressure measurement. standard. Due to the limitation on the. strength of materials, it is not possible to stretch a practical size diaphragm to resonate above 20 kc. Even at 15 km, a steel diaphragm /2" diameter requires a peripheral tension. of about 70,000 lbs/sq. 111., while for an aluminum or an. aluminum alloy diaphragm, a tension of. about. 25,0001 lbsJsq. in. is necessary. These stresses are: not only dangerously close to. they yield point of the materials, but

the possibility of setting lip-such a high tension and keeping it from changing over long. periods of time and over wide ranges of. room temperature variations is extremely difficult.

As a result of the relatively small mass of a stretched diaphragm. as compared with'thecffective mass of the vibrating crystal in the microphone herein described, the small condenser microphone deviates to a larger degree from true stiffness control below its resonant frequency than would be the case if its vibrating mass were larger. An indication of the amount of the deviation from a true stiffness may be learned from the change of phase angle of the mechanical impedance of the system with frequency. For a simple vibrating mechanical system, it is well known that the phase angle is given by where:

A true stiffness-controlled vibrating system would show a value of s from Equation 1 equal to --90 over the entire frequency range of operation. In order to compare the microphone described in this specification with a typical miniatore stretched-diaphragm type. the. amount. of phase shift. from; a pure stifiness reactance has been computed fromEquation 1 assuming that the stretched diaphragm has an elf'ective area of 1 sq. cm., an eflective mass of .001 mm, and a resonant frequency of 12 kc. The result is shown by the dotted line in Fig. 1d. The heavy solid line at the degree phase shift line in Fig. 10 showsthe computed phase'shift for the con: stants of the, piezo olectric microphone herein described. From. Fig. 10 it. is evident that the stretched diaphragm type deviates from the ideal stiffness-controlled system at frequencies well within the v audiofrcquenoy range, whereas the described microphone shows no deviation at all throughout the entire audio range: to '20 kc. The computation of the dotted curve has. neglected themechanical damping. resistance in the system which, if taken into account, would show a greater phase, shift; error than indicated. If the stretched diaphragm is larger in diameter or resonates below 12 be, which is more typically the case with practical structures, the. phase, shifterror will be alsosmuch worse than indicated by the dotted curve-in Fig. 10.

The maximum s und. pressure which can be measured by the. stretched. diaphragm type; of microphone is. limited by the relatively small per.- missible deflection-or the. diaphragm before non! linearity setsln. These non-linearities occur for soundpressures of the order of athousand dynes/cm. where the alternating amplitudes. will already amount to a few per cent of the. fixed diaphragIn:v spacing... In. the described microi phone employing either primary ammonium p osphate crystals or quartz, he lineari y: s limited only by the linearity or the complianceot the crystal substance; and since these materials obey Hookes law to pressures of the order ofth usands of pounds: per square inch, there is no practical limitation: to the use of themicrophcne up to sound pressures of many milliondynes/ cmF. Because the resonant frequency of the micro. phone assembly has been placed very much. higher than 20 kc. by employing a crystal length less than one inch. the electrical impedance of the microphone is essentially that of a simple condenser over the entire audio-frequency range beyond 20 kc. The advantage of having such a simple circuit element for the microphone impedance permits great flexibility in the use of the microphone in combination with special input circuits.

Since the voltage generated in the crystal element is proportional to the pressure acting on the diaphragm over the entire audio-frequency range to beyond 20 kc. and also because of the very accurate assembly procedure which has been described, the resultant microphone output is absolutely uniform with respect to pressure acting on the diaphragm over the entire frequency range to 20 kc. For the diameter unit, some difiraction occurs at the higher audio frequencies which may be of no serious concern for the large majority of uses. An-

other microphone built along the same lines disclosed, but having an outside diameter of will practically eliminate most of the diffraction problems throughout the frequency range to the neighborhood of 20 kc. When such an extremely small physical structure is necessary for very special high-frequency sound pressure measurements, the resultant loss in sensitivity will have to be tolerated.

In applications requiring pressure measure- "11 ments at relatively high temperatures, the use of quartz as the active crystal assembly is desirable, again producing a somewhat lower sensitivity, as previously described.

Although I have chosen certain specific procedures for producing a microphone possessing new advantages as a sound pressure measurement standard to illustrate the basic features of my invention, it will be obvious to those skilled in the art that numerous departures may be made in the specific details for executing the required functions, and I, therefore, desire that my invention shall not be limited except insofar as is made necessary by the prior art and by the spirit of the appended claims.

I claim as my invention:

1. In the method of making a microphone comprised of expander crystal means which comprises the steps of establishing at one end of said crystal means a plane perpendicular to a direction of expansion of said crystal means, mounting said crystal means with its said plane face against a plane face of a base, connecting a housing to said base with said housing around said crystal means, the unattached end of said r housing lying approximately in the plane of the unattached end of said crystal means, facing-01f the unattached end of said housing and crystal means until they are brought into a common plane, and attaching a planar diaphragm to said crystal means and said housing at said end which has been faced-off.

2. In the process of manufacturing a transducer, the steps of connecting piezo-electric crystal means to a base, afilxing to the free end of said 7.

crystal means a layer of material which has a higher mechanical impedance than the crystalline material, connecting an enclosing housing to said base, the free end of said housing defining a plane approximately parallel to the surface of said afiixed material, facing oif said affixed material until it lies in the same plane with the end of said housing, and connecting a diaphragm to the free end of said afiixed material and to said housing to seal said housing.

3. In a microphone, a base member having a 1 2 planar face, a plurality of expander plates of piezo-electric material connected together and mounted on said base with an axis of expansion and contraction normal to the plane of said face, a housing extending around said expander plates and mounted on said base, the end of the housing away from said base lying in a plane parallel to the plane defined by the end of the plurality of expander plates, a layer of material attached to the end of the plurality of expander plates, the

mechanical impedance of said layer being greater than the mechanical impedance of said plurality of expander plates, the non-attached surface of said layer of material lying in a plane defined by the end of said housing, a diaphragm, and means connecting said diaphragm to said layer of material and to said end of the housing away from said base, whereby the outside surface of said diaphragm presents an unbroken plane to the peripheral edge of said housing.

4. The invention set forth in claim 3 further characterized in that insulated-terminal means are provided in said base member for establishing electrical connection from one side of said base member to said plurality of expander plates mounted on said base member.

FRANK MASSA.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,813,461 Nicholson July 7, 1931 1,939,302 Heaney Dec. 12, 1933 2,096,826 Schrader Oct. 26, 1937 2,138,036. Kunze Nov. 29, 1938 2,181,132 Kallmeyer Nov. 28, 1939 2,388,242 Arndt Nov. 6, 1945 2,393,429 Swinehart Jan. 22, 1946 2,405,604 Pope Aug. 13, 1946 FOREIGN PATENTS Number Country Date 477,740 Germany June 20, 1929

Images