US2413062A - Apparatus for the production of music - Google Patents

Description

Dec. 24, 1946.

APPARATUS FOR THE PRODUCTION OF MUSIC Fs-i C01. II CoLIII B. F. MIESSNER Filed Nov. 1, 1941 Col. IV

4 Sheets-Sheet l Dec. 24, 1946. B. F. MIESSNEYR 2,413,052

APPARATUS FOR THE PRODUCTION OF MUSIC Filed Nov. 1, 1941 4 Sheets-Sheet 2 Dec. 24, 1946. B. F. MIESSNER APPARATUS FOR THE PRODUCTION OF MUSIC 1941 4 SheetsSheet 3 Filed Nov. 1,

INVENTOR.

Dec. 24, 1946. MlEsSNER 2,413,062

APPARATUS FOR THE PRODUCTION OF MUSIC Filed Nov. '1, 1941 4 Sheets-Sheet 4 l 152 N110 141/ fl145 INVENTOR. Be inEMfessner BY Patented Dec. 24, 194$ UNITED STATES PATENT OFFICE APPARATUS FOR THE PRODUCTION OF MUSIC Application November 1, 1941, Serial No. 417,468

4 Claims.

This invention relates to the p uc of music, and more particularly to the production of tones of bell or chime nature, In carrying out the invention there may be employed mechanical vibrators, which may for example be either of rod form or of the tubular form commonly used for orchestral chimes; and the vibrations of these vibrators may be translated into electric oscillations and those oscillations in turn into sound. Such translation of course has the effect of removing all limitations on the volume of output sound, and of permitting the use of small Vibrators when desired. I also employ it to great advantage in controlling the harmonic structure and other qualitative characteristics of the output tone. While in carrying out such control it is possible to take advantage of certain broad principles already known, their application to vibrators of this character involves special problems which arise because of the nature of the vibrators, their inherently inharmonic series of partials, and other peculiarities.

One special problem arising in connection with the production of some tones of the nature abovementioned is the provision of a proper hum tone, of lower frequency than the apparent fundamental tone and of lower amplitude, and of proper rate of damping relative to the rates of other tone components. By my invention I am able to provide such a hum tone in thoroughly regulable amplitude and damping rate.

Another special problem arising in the production of these tones, and particularly in the production of bell tones, is the provision of a sulficiently compact series of components which, while containing many inharmonic intervals, will be sufiiciently characterized by harmonic or approximately harmonic intervals to yield a pleasing tone sensation. By my invention I am able to achieve this result to an even more perfect degree and with much greater certainty, as well as with far lower size and cost, than in conventional bell manufacture.

These problems are typical of many which, together with their solutions, will be apparent from the detailed description of my invention.

It is an object of the invention to provide improved and simplified means and methods for the production of tones of the types described.

It is an object to produce improved tones of the type described.

It is an object to provide means and methods for obtaining a series of components properly related to each other, in various respects, for the production of such tones.

It is an object to provide improved vibrator means in instruments adapted for the production of such tones.

It is an object to provide improved means for translating the vibrations of the vibrator means into electric oscillations.

It is an object to provide improved means for controlling the vibrations of the vibrator means and the oscillations translated therefrom.

Other and allied objects will more fully appear from the following description and the appended claims.

In the description of my invention hereinafter set forth reference is had to the accompanying drawings, in which:

Figure 1 is a table showing certain relationships between the partial components of vibration of vibrators which may be employed with my invention;

Figure 2 is a side view, principally in elevation but partly in section, of one embodiment of my invention;

Figure 3 is a cross-sectional view taken along line 33 of Figure 2;

Figure 3a is a cross-sectional view taken along line 3a-3a of Figure 3;

Figure 4 is a family of curves illustrating the manner of vibration of a vibrator such as shown in Figure 2;

Figure 5 is an elevational view of a slightly modified form of vibrator which may be em- I ployed;

Figure 6 is a cross sectional view taken along line 66 of Figure 2;

Figure 7 is a front elevational view of a further modified form of vibrator;

Figure 8 is a view, principally in section, of an alternative embodiment of my invention;

Figure 9 is a vertical cross-sectional view taken along line 99 of Figure 8;

Figure 10 is a cross-sectional view taken along line lfll0 of Figure 11;

Figure 11 is an enlarged top plan view of the translating device 6i (and a fraction of the vibrator 52) of Figure 8;

Figure 12 is a family of curves illustrating the manner of vibration of a vibrator such as shown in Figure 8;

Figure 13 is a view similar to Figure 9 but illustrating a modification of the showing of that figure;

Figure 14 is a fragmentary view of the vibrator of Figure 8 together with a modified form of translating device which may be employed;

Figure 15 is a similar view illustrating a still further modification of the translating device;

Figure 15 is a vertical view, partly in elevation and partly in section, of the vibrator of Figure 8 with certain still further modified translating devices and associated circuits; and

Figures 17, 18 and 19 are cross-sectional views respectively taken along lines illi, i8l3 and I9l9 of Figure 16.

The vibrators with which I have particularly illustrated my invention may generally be classed as beams. More specifically, I have illustrated them as of two typesrods firmly secured at one end and tubes flexibly suspended at one end (the other end in either event being free) though it will be understood that no unexpressed limitations of the vibrators to these particular formations or manners of support is intended. The two types abovementioned respectively behave, vibrationally, as a clamped-free beam, and as a free-free beamthe first being, and the second not being, constrained against rotation at the supported end. The vibration of each type is c'omplexi. e., contains a series of partial components. The partial series for the two types are capable of joint discussion, subject to certain specific distinctions which will be pointed out.

Thus the tabular Figure 1 of the drawings shows in column II the relative frequencies of vibration for the several partials of a clamped-free beam. It is proper to take it also as applying to the free-free beam, subject to the following qualifications: (1) that the second partial of the freefree beam is then very minutely higher in relative frequency than shown for the second partial in the table, and (2) that the first partial of the free-free beam is then not an actual vibration within the beam, but is the pendular swing of the beam as a unit about its suspension, at a low frequency determined by the mass and center of gravity of the beam and not necessarily directly related to the relative frequency value shown for the first partial in the table.

It will of course be understood that the table is intended simply to show the relationships between th frequencies of the partials of any one vibrator, and is not concerned with frequency relationships between different vibrators. It will further be noted that the relative frequencies have been arbitrarily referred to that of the third partial as unity.

While the use of mechanico-electro-acoustic translation with the beam vibrators may be made to emphasize the lowest partials, so that one of them becomes the determinant of apparent pitch of the tones, I have found that this is usually musically undesirableamong other reasons, because of the wide frequency spread between such a partial and those above it. As a generality, the third partial is usually the lowest which it is desirable to employ, and frequently it is desirable to go still higher in the series, for the fundamental of the output tone. One or more partials lower than the fundamental may then be employed to create the hum tone-but, as above mentioned, this must be of rather limited amplitude. These considerations will become more fully apparent in connection with the descriptions of typical embodiments of my invention.

In Figure 2 I illustrate an embodiment of my invention employing clamped free beams. Herein I! is a massive rail, preferably metallic, to the front of which the individual vibrators (one of which appears as l2) are fastened through respective lugs (one of which appears as I3) into which the top ends of the vibrators are securely gripped. Each vibrator may typically be a flat strip of steel-i. e., rectangular in cross-section though preferably with slightly rounded corners, as seen in the cross-sectional Figure 3with its smaller dimension, or thickness, being that which appears in Figure 2. Secured to and depending from the rail H may be a vertical apron It, typically of thick wood, on

which the mechanico-electric translating devices,

or pick-ups, may be mounted. While any desired means may be employed for setting the vibrator into vibration, I have by way of example illustrated a hammer It arranged to strike the vibrator, and-adapted to be propelled by some conventional action mechanism such for example as a piano action.

While there may be employed any known form of translating device operating in spaced relation to the vibrators (some of which for example are shown in Patent No. 1,929,027 to me and in Patent No. 1,906,607 to Jacobs), I have illustrated herein translating devices of an electromagnetic type. Thus one translating device appears as 2|, comprising coil Ziiinset into the apron l6, and magnetic member IS passing horizontally through the coil and apron and through a metallic rail i? on the back of the apron, the magnetic member 1 9 being adjustable horizontally and lockable in any desired position by means of set screw H3 in rail H. The magnetic member may comprise a permanent magnet; but preferably its forward portion Isa will be of soft iron, secured to the forward end of the magnet proper.

Preferably in the carrying out of my invention there Will be employed in association with each vibrator a second translating device placed at a different position longitudinally of the vibrator. Accordingly in Figure 2 there appears a second translating device 2|; it may be entirely similar to the device 2|, and various parts of it bear designations similar to corresponding parts in the device 21 but furnished with prime marks.

In determining the proper positions of the translating devices longitudinally along the vibrator I take account of the peculiar way in which the vibrator tends to vibrate. In Figure 4 this is developed, for the several partials, in greatly exaggerated form. The several curves herein designated as L2, etc., represent, individually for the correspondingly numbered partials, the configuration which the vibrator tends to assume at the peaks of its vibration at those partials respectively (the total vibration of the vibrator being of course a superposition of curves respectivelyof the forms of the several ones shown). Actually this data is presented for a vibrator provided with some weight at its extremity, as

U hereinafter described; but in the portion nearer the base or lug Is the change in the several curves effected by the presence of the weight is not large.

Preferably the pick-up 2! will be positioned relatively'near the base of lug l3for example,

at to times the length of the vibrator away from the lug, where it will translate all partials up to those quite high-numbered ones which have nodes at about the pick-up position. The pickup 2i may be positioned relatively further from the lug 13, for example at about .55 times the length of the vibrator therefrom. The pick-ups may be electrically connected together-if desired, through respective potentiometers 22 and 22'--in such a phase relationship that the oscillations respectively translated by the two pickups from the first (and from the second) partial vibration will be effectively opposed against each other. The amplitudes of the composite ocillations translated by each pick-up may be regulatedeither by adjustment of the efficiency of the pickups (for example, by adjustment of the spacings of the ends of magnetic members l9 and I9 from the Vibrator) or by the potentiometers or by both-so that either one of the first and second partials may be balanced out from the combined pick-up output, and the other greatly attenuated.

The arrangement just described is well adapted for the production of a tonein which the third partial is to form the fundamental (i. e., the frequency apparently determining the pitch of the tone), for it will be seen from Figure 4 that the same phase of pick-up inter-connection which causes an attenuation of the first and second partials will, with the pick-ups at the positions mentioned, cause a reinforcement of the third partial. At the same time an interdevice reinforcement will occur for some of the partials higher than the third.

The adjustment of the oscillation amplitudes may be made to eliminate the second partial completely; the first will be greatly attenuated, and in its attenuated amplitude will in many cases be of no consequence (in view of the further fact of its extremely low'frequency). On the other hand, the adjustment may be carried beyond balancing out of the second partial, toward complete balancing out of the first; in this way there may be introduced as much second partial as may be desired for a hum tone.

The embodiment of Figure 2 has so far been discussed without special reference to a weight which is shown secured to the vibrator adjacent its free extremity. This weight may for example be of the form illustrated in Figures 3 and 3a: a block provided with a longitudinal cut liia adapting the block to fit over the thickness dimension of the vibrator, and preferably having slightly wider cuts if?) made inwardly from its extremities so that only a restricted central portion will be clamped in contact with the vibrator by a set screw Hie-the latter feature avoiding any appreciable local stiffening of the vibrator and any likelihood of rattling between vibrator and weight. Very slight shifts of the weight are useful for close final tuning of the vibrator; but the primary purpose of the weight, and the purpose for which it is apportioned, is the improvement of the inter-partial frequency ratios.

It may be mentioned that it is a desirable refinement to make the cuts lllb of slightly different depths, so that the central clamping portion is slightly de-centered. It thereby is made possible to reverse the relationship between the center of gravity of the weight and the center of its clamping to the vibrator, which may occasionally be desirable for the most perfect joint achievement of the ratio improvement above mentioned and of precise tuning to the frequency desired for the vibrator.

The frequency ratios shown in column 2 of Figure l, for the unweighted beam, will be seen to be characterized by an almost complete absence of harmonious relationships to the third partial frequency (harmonious relationships being expressed by the ratios of reasonably small integers to each other). It is true that the frequency ratio of the fourth to the third partial is almost 2; but it is noticeably non-exact. The frequency ratio of the second to the third partial is 112.8; and this odd relationship militates against the pleasantness of the use of the second partial as a hum tone. I have observed, however, that by the use of a weight of the proper size, the frequency ratios may be changed to those shown in column 3 of Figure l. Herein the ratio of secondto third-partial frequency is changed to an integral 1:3 relationship; at the same time the otherwise inexact fourth-to-third relationship becomes a quite exact 2. Further, the useless 3.25 relationship between fifth and third partials becomes 3.35 (almost exactly 10/3) and the relationship between fifth and fourth just half that much (i. e., 5/3) -musically, the interval of an A to a C therebelow. Among the higher partials, while the eighth loses its integral relationship to the third, both the seventh and ninth take on such a nature which they lacked in the unweighted vibrator.

While the value of the required weight may be determined by test in any particular case in the light of the foregoing description of the results made possible, I may mention that in cases which I believe typical I have found that a weight of the order of of the mass of the vibrator has been found to give the above-described results, when positioned with its center of gravity very nearly at the extremity of the vibrator.

It has been known heretofore to thin the vibrator in a region closely adjacent its base, to effect some improvement of frequency relationships over those characterizing the simple vibrator; a vibrator so thinned is shown in the detail Figure 5. But I find preferable the use of the weight as disclosed (or, as a permissible variation, a joint use of weight and of slight thinning); there is thereby provided better control of the improvement of frequency ratios as well as of the precise final tuning of the vibrator-beside the assurance that the proper effects on all the characteristics involved will be obtained jointly. I may mention that when employing the weight with a vibrator which has been appreciably thinned adjacent its base, the proper position of the weight may shift somewhat from the free entremity toward the base, as illustrated in Figure 5.

I have above discussed the typical case of the use of the third partial as the fundamental of the output tone. While this is musically useful, there are two factors each of which may at times militate against its pleasing nature: 1) the nearest hum tone therebel'ow is removed therefrom by a 3:1 ratio (assruning the ratio improvement disclosed above) which, though harmonious, is a long interval (musically, an octave plus a fifth) and (2) while the third partial is a predominant part of the tone from shortly after inception on, it tends at and immediately following inception to be obscured to a considerable extent by higher components.

Indeed, under som circumstances the pitch of the tone seems to be consistently estimated at a value which does not appear as any of the partials, but instead is half the frequency of .the fifth partial. I believe there are two possible explanations for this: (1) that, since this value of half the frequency of the fifth partial is almost exactly the same (in the case of the improved ratios above described, quite exactly the same) as the difference or beat frequency between the sixth and fifth partials, it is supplied, as a missing fundamental, from those two partials, by a well understood physiological and psychological action on the part of the listener; and (2) that while the fifth, a prominentpartial, may establish the note family (i. e., the A family, or B family, or the like) to which the note belongs, the judgment of the octave-level to which the note belongs is distorted downwardly by one octave by reason of the presence of lower tone components (of which there are none below the pitch-establishing fundamentalin the case of a tone having only harmonious components).

In any event, I have frequently found it pleasing and satisfying to adopted half the fifth-partial frequency as the tone fundamental, or apparentpitch determinant, In this event, assigning this a value of unity, the inter-partial frequency relationships (on the improved-ratio basis) become those shown in column 4 of Figure 1. While the fourth-partial frequency appears to be badly related to the new fundamental for a first overtone (by 1.2:1, or 6:5, which is the musical interval from C to A thereblow), this is not the case in the production of a bell or chime tone-of which an important characteristic has been pointed out to be the presence of a quint tone at just this frequency relationship to the fundamental. The fifthand sixth-partial frequencies become related to the new fundamental in the highly desirable 2 and 3 ratios. And of the lower partials, the second and third each bear theoretically good relationships (respectively 1:5 and 3:5) to the new fundamental.

To emphasize the now particularly important fifth and sixth partials, as well as to more greatly minimize the third (which now should at most be merely .a hum tone), it is desirable to re-position at least the more outwardly of the twotranslating devices, or pick-ups. Typically it may be brought nearer to the base of the vibrator; and accordingly I have shown in Figure 2 a translating device 2! which may be substituted for the device 2| by means of a switch 23. The device 2!" has been shown by way of example as positioned at about .4 times the length of the vibrator from the base. From Figure 4, and it being borne in mind that the translating devices'are electrically connected for opposition of oscillationstranslated from l-o-wer partial vibration, it? may be seen that opposition will occur for-the first, second and third partials simultaneously, whileaiding will occur for the fourth, fifth and3sixthwhich latter are to be prominent ingredients of the tone.

A typical adjustment of the relative outputs from the pick-upgfi and 2 l may be the adjust ment which fully balances out the second partial; This adjustment will be attended by some net translation-from first partial; but this is so particularly low infrequency relative to the new fundamental, that in many cases the mere inherent frequency-transmission characteristic of the system to which the oscillations aresupplied may be sufficient to insure its absence from the output tone. The mentioned adjustment will also be attended by some net translation o-f-third partial, which will be utilizedand reliedon-to form the hum tone. It will be appreciated thatthis hum tone will bea sub-quint -thatis, .of a frequency an octave below the quin above mentioned as formed by the fourth partial.

While a tone-producing system so arranged and adjusted is very ,satisfactoryin. many instances, two possible disadvantagesfor certain purposes: may be mentioned: (1) the. adjustment which: eliminates or sufficiently. minimizes the second-'- partiaL component may. leave an undesirably 8 strong amplitude of third-partial component; and (2) the sub-quint frequency of the third partial, while useful as a hum tone frequency, is frequent-ly considered not to be the ideal one for a sole hum tone.

The first disadvantage abovementioned may be obviated by further repositioning of at least the more outwardly of the-translating devices. Thus I have shown in Figure 2 a translating device 21" which may be substituted for 2| or 2| by the switch 23. This device Zl' may be relatively nearer the translating device 2l--for example at about .275 times the length of the vibrator (the device 2i being for example at about .2 times that length) from the lug [3. Such a positioning of the translating devices results (assuming the adjustment of their outputs to be for-full elimination of the second partial) in almost perfect elimination of the first partial, rather small translation from the third partial, a medium degree of translation of the fourth, and an almost normal translation of each of the next several partials.

The second disadvantage abovementioned-the lack of a hum tone of frequency other than the sub-quint (which latter has now been reduced in amplitude)may be obviated by the derivation of a bum tone of any desired frequency from a separate or supplementary vibrator simultaneously excited with the main vibrator 12. Without intending any unexpressed limitation as to the form of the separate vibrator or as to the manner of its excitation simultaneous with that of the main vibratorboth of which may obviously be varied within wide limitsI have nevertheless shown an embodiment of what I consider an extremely simple and effective arrangement for the stated purpose. It comprises the securing t the main vibrator l2 (preferably very near the base or lug I3, so as to have a negligible infiuence on the vibrational characteristics of the rail II to hold a translating device 3i (similar to 21) in operative relationship to the supplementary vibrator 32. The output of this transe lating device, regulated if desired not only by the spacing of its magnetic member from 32 but also by a potentiometer 30, will of course be connected with the joint output of the translating devices previously described. Obviously the regulation of the output of the device or pick-up 3| provides a complete regulation of the amplitude of the added hum tone relative to that of other tone components. Furthermore, the possible infinitesimal postponement of influence of the exciting means on the supplementary vibrator (resulting from the indirectness of transmission of the exciting-means action) quite parallels the slow development of the hum tone in the bestv bells.

The partial of the supplementary vibrator (itself a clamped-free beam) which is preferably relied on as the hum tone is the first partial. This has the joint advantages (1) of rendering the-supplementary vibrator of minimum length forthe thicknessemployed (which thickness it self may be made very'small, so that the whole mass is in turn very small) and (2) of causing the first overtone of the supplementary vibrator to fall at a relatively high frequency (i. e., at least of the order of 6.25 times the supplementaryvibrator-first-partial frequency), which will usually occur in the range of several times the frequency being used as the fundamental of the output tone developed (except for the contributions of the supplementary vibrator) from the main vibrator.

And I may point out the preferred thinning of the supplementary vibrator near its base (for example, just above the bend abovementioned) sufliciently so that its second-partial frequency becomes harmoniously related to the frequency beingrelied on as the fundamental from the main vibrator. Probably the preferred hum tone (and therefore the first-partial frequency of the supplementary vibrator) is an octave below (i. e., 4; times) that fundamental frequency. If then the second-partial frequency of the supplementary vibrator 32 be made 8 times its first-partial frequency (readily done by thinning as abovementioned), it will provide a component of 4 times the frequency assigned as the fundamental of the note being producedthereby supplying a very useful overtone carrying the series of consecutive harmonic overtones onward from the 2- times and 3-times already provided (by the main vibrator) to include the 4-times. On the other hand, for example, the second-partial frequency of the supplementary vibrator 32 may be made 7 times its first-partial frequency, so that it will provide a component of exactly 3 times the frequency assigned as the note fundamental; a limited development of such half-integer components has been sometimes thought desirable in the creation of .the best bell or chime tones.

In terms of the fundamental produced as above described by the main vibrator, the total partial series in the output tone (up to the frequency of the tenth partial of the main vibrator) becomes the asterisk denoting derivation from the supplementary vibrator. This series is characterized by (l) a most desirable hum tone of .5, (2) a useful further hum tone of the sub-quint frequency of .6, (3) a perfect series-1, '2, 3, and, if desired, 4of harmonically related components, (4;) a quint tone in precisely proper relationship (1.2:1) to the fundamental, and (5) a series (all above 4-times the frequency of the fundamental) of the dissonant tones which must always be to some extent present to create the typical effects of bell or chime tones.

Of course in the foregoing series the fundamental (designated by 1) is a fictitious or supplied component. In many instances this will not be found a shortcoming, because of the strength of the tendency-above discussed-to supply it. But in those cases wherein it may be desired that it be an actual component, it may be supplied by an additional, simultaneously excited vibrato in the same manner as the added hum component above discussed. Accordingly in Figure 2 I have illustrated a second supplementary vibrator, designated as 33, Whose first-partial frequency maybe chosen to form the l in the series. It may be secured to the main vibrator in similar manner to that above described for 32, and may extend upwardly alongside that supplementary vibrator close enough thereto so that the single translating device 3! will be effective for both supplementary vibrators. Differential adjustments of the translations from the two supplementary vibrators may be made by bending of these vibrators; and this expedient may if desired be relied on for all adjustment of the translations from these vibrators, dispensing with the need for adjustability of the magnetic members in the translating device 3 I.

Figure 2 illustrates a further expedient for the control of the partial-frequency components from the main vibrato-r which may be employed when desired. This is the selective damping of the vibrator, to cause the rapid decay of certain components without appreciable effect on others. Such selective damping may be effected by placing on the vibrator, at a point which is appreciably non-nodal for the components to be affected but is substantially nodal for those which are to remain substantially unaffected, a damping structure wholly carried by the vibrator. Such a structure has been shown at 2'! in Figure 2. As seen in this figure (aided by Figure 6) it may for example comprise a piece 28 of damping material (for example, of the synthetic rubber neoprene which is characterized by appreciable viscosity, or of the still more viscous and less resilient Visccloid) bent into the form of a U about the vibrator, and a member 29 surrounding the piece 28 and clamping it in place. For the illustrated vibrator the member 29 may for example comprise a generally U-shaped spring, preferably formed with inward bulges 29a a short distance from the respective open ends of the U-these bulges performing the clamping function, and they and the inside central portion of the member 29 forming three regions of contact of that member with the piece 28. The damping structure has a length (i. e., dimension longitudinally of the vibrator) preferably of quite srna11 order-v its function not being to impede the vibratory deformation of the vibrator within that length, but rather to take advantage of the inertia of the mass of its elements. Because of this inertia, small though it be, the damping structure tends not to move vibratorily; and the vibratory forces transmitted to it from the vibrator, forced to pass through the piece 28 of damping material, cause an energy dissipation in the latter.

Those forces, and hence that dissipation, are obviously supplied only at the frequencies of those partial components which do not have substantial nodes at the point of attachment of the damping structure; hence the selectivity of action. The choice of the material of the piece 28 and of the total mass ofthe damping structure will of course determine the degree of the action. When that degree is very small, the effect of the structure on the partial components not having substantial nodes at the point of attachment will be simply that of noticeably increasing the rates of damping of those components in the output tone. When that degree is larger, however, those rates of damping will be so great that the components may not be identified over the general clang or other impulsive impression created at the intials (of those up to-and including the ninth) euceptingthe fifth, sixthand ninth, each of which hasa substantial node in this region. Thus in the arrangement and use ofthe Figure-2 structure as last above described, the effect of the damping structure is to aid in the suppression of thefirst, second and third partials, and to reduce the quint (fourth partial) and some of the higher dissonant partials (e. g., seventh and eighth), while leaving essentially unimpaired the fifth and sixth partials which are being relied on for the production of the fundamental and of the 2-times and 3-tirnes frequencies (and of course leaving unimpaired the action of the supplementary vibrator means). It will vbe understood, however, that this particular employment of the damping means, while .a peculiarly favorable one in certain instances, is primarily intended to be simplyillustrative of its use to cOntrol the output tonequality, .byattenuating certain partials (especially low ones) not desired in t e output tone.

The structure of Figure 2, insofar as it combines, with translation from .a plurality of points along a vibrator, a flattened cross-sectional form of the vibrator (e. g., a form inherently having a preferred .plane of vibration), is one that I have found broadly of great .uti1ityand for a reason not immediately apparent. This .reason vis grounded in the propensity of any vibrator which has similar vibratory propensities both in the plane of intended vibration and in all other planes containing the vibrator (e. g., of any vibrator round in cross-section), to shift the plane of its vibration between that plane in which the vibration was originally caused and the plane at right angles thereto-this shift being frequently aslowly oscillatory one. Clearly this shiftingaction, even with a single translating device, will cause the translated output from the vibrator to be modulated at the low frequencylof .the shifting; and to this extent only there is more or less apparent the desirability of employing with a single translating device a vibrator structure characterized by a preferred plane of vibration (suchasshownfor a stringinPatentNo. 1,933,298 tome).

.But I have discovered that such shifts of vibration between the two planes may occur independently for the respective several partial components of vibration; and since the adjustments ofa pluralityof translating devices and of their outputs, for particular net translating actions-as to the several partials, are influenced by overall dynamic conditions, the perfection of those net adjustments for their purpose may be very seriously impaired .by the shift of vibration plane. And I have found, as may followfrom the foregoing once that .hasbeen appreciated, that the use .of a vibrator characterized by a preferred plane of vibration,-by reason of its elimination or substantial elimination of the shifts of plane, frequently effects an unexpectedand great improvement in the perfection ofthe netor balanced action of a pluraltiy of .translating devices respectively associated with different parts ofthe vibrator.

While this discovery andexpedient is not limited to the case of the :beam vibrators, it is of direct applicability thereto, and it is for this reason thatI have-preferred, and haveillustrated in Figure 2, the .rectangularly cross-sectioned vibrator.

It Will be understood, however, that ,manyof the aspects of my invention are not limited to the use of such vibrators, and that other forms of clamped-free beam-such for example as sim- 12 ple round rods-maybe employed in substitution for the vibrator above described. I have accordingly illustrated in Figure 7 a partial front elevational view of a round-rod vibrator 42 intended for optional substitution for the vibrator !2 of Figure 2.

The weight, when that is employed, for the round rod may for example take the specifically different form of a cylinder 35, having a bore of which the approximatelycentra1 portiontfia fits the rod, and of which the outer portions 35b-are slightly oversize. The damping structure for the round rod may;for example take the specifically different form of .a cylindrical bushing 38 of the dampinglmaterial and .a cylindrical member .39 surrounding 38 and clamping it to .the rod though the member 39 may be omitted if the mass of 38 be itself large enough for the desired damping action.

Supplementary vibrators-may be employed with the round vibrator, if desired, in amanner analogous to that discussed above for the rectangularly cross-sectioned vibrator. Accordingly I have shown secured -to the rod 42, 'near the lug l3 into which it is gripped, the base endsof two supplementary vibrators 32 and 33, which may respectively serve the functionsserved'by 32 and 33 in the structure of Figure 2. By way of example two specifically different features have been illustrated "in Figure -7 in connection with the supplementary vibrators-though it will be understood that thesefeatures are in no way limited to the round, as distinguished from rectangularly cross-sectioned, main vibrator. One of these features is the formation of each supplementary vibrator as a relatively thinner element, for example of piano wire, and the relatively more appropriate use of .small terminal weights (32a, 33a), rather than a thinnin of these vibrators near their bases, .to effect such control of the second-.toefirst partial frequency ratios of these vibrators; such weights may for example be of solder. The other feature mentioned above is the extension of these supplementary vibrators not upwardly as in Figure 2,.butdownwardly,

in general parallel to and alongside the main vi- I brator '42, so that the translating device 2| active for the main vibrator may at the same time serve as the translating device for the two supplementary vibrators (the latter for example being bent appropriately to adjust the translation from each without disturbance of the adjustment of magnetic member l9, which latter adjustment will have been effected as found necessary in connection with the translation from the main vibrator) Hereinabove there have been described various expedients for the elimination or attenuation of various partial components-especially of low partial components which are inherently separated by relatively wide frequency intervals and for that broad reason, among others, are frequently of small utility in the production of the most acceptable bell or chime tones. Thus for example in the typical series of components for an output tone set forth in detail above-the first two partials of the main vibrator are for practical purposes absent, the third is present in at most greatly attenuated amplitude (supplying a fsub-quint hum tone), and thefourth is at least somewhat attenuated.

A .result of this character may be achieved .by the use, alternatively or additionally to the expedients above disclosed, of a high-passfilter in the output circuit of the translating means for the main vibratorwhich translating means may be limited if desired to a single translating device such as 2| (as by omission of the others or adjustment of their potentiometer 22 to render them ineffective). Such a high-pass filter I have indicated in Figure 2 as 42; it is shown connected in the output circuit from the potentiometers 22 and 22, prior to the merger with that output circuit of the output circuit from potentiometer 30 for the supplementary vibrator means.

In Figure 2 there has been illustrated, though thus far not mentioned, a helical spring 26 having its upper extremity secured in any convenient manner to the bottom of the vibrator l2, and its lower extremity secured to some stationary member 25 therebelow. A useful function of this spring is to attenuate the first-partial vibration of the vibrator l2a vibration which is usually an unused one, as has already appeared. This mechanical attenuation of the vibration itself is usually unnecessary insofar as mere elimination of the first partial from the output tone'is con cerned, when expedients such as already described are employed; but it is useful both in avoiding undue swings of the free extremity of the vibrator (correspondingly reducing the spacing which that must have from other objects), and in suppressing any possible tendencies toward modulation of the translation of higher partial components at the first-partial frequency. The influence of such a restraining means as 26 on the frequencies of the partial components (theoretically a. raising influence) sharply decreases with increase in the number of the partial; and with a relatively weak tensioning of the spring this influence will be alto ether imperceptible excepting on the frequency of the first partial, which is ordinarily not being used in any event.

The description proper of the embodiment of Figure 2 may be completed by noting the cascaded elements following the translating devices and their respective potentiometers, abovementioned. These elements may comprise the circuit G into which the outputs of the several translating devices (i. e., the output terminals of the filter 43 and potentiometer 36 when those are employed) are connected, for example in series relationship; the control system M, in which any and all known forms of control for the composite oscillations in electronic musical instruments may be employed; the amplifier 45; the volume control 46; the further amplifier 41; and the loudspeaker or other electro-acoustic translating device 48. It will be understood that the elements consecutively numbered from 43 through 43, as Well as the potentiometers 22 and 22 and 36, may if desired serve in common for a plurality of vibrators such as any one of those illustrated or described, respectively tuned to the several notefrequencie of a musical scale.

The consideration of this embodiment may be completed by calling attention to the fact that, while for the sake of illustration the embodiment and typical modifications thereof have been disclosed in considerable detail, and various typical adjustments and optional features separately considered, there are thereby intended no unexpressed limitations-such as to the particular partials to be elicited or suppressed, components to be added by supplementary vibrators, dampings to be effected, particular forms of elements, or the like.

While my invention has so far been described in connection with clamped-free beams, it will become apparent that it may also be carried out in connection with free-free beams. Because of the suspension which permits the free-free nature of vibration of the latter beams (i. e., a suspension which does not constrain the vibrator against rotary movement about its support) the mean position of such a beam in space may vary oscillatorily-the oscillatory variation of position being for example the pendular swing of the vi brator about its support, above termed the first partial of its vibration in the generic View of the actions of the two types of vibrators. While this first partial might be utilized if desired, 'it is aimost always an undesired partial; and preferably its complete elimination, in the mechanicoelectric translation (from each individual translation, when more than one is effected), is insured by mounting the pick-ups on the vibrator, so that they as well as the vibrator undergo the firstpartial movements.

One of the simplest forms of pick-up which may be employed for mounting on the vibrator is a piezo-electric translating unit. When this is so mounted on the vibrator as to be subjected by the vibrator vibration to an oscillatory elongating and shortening tendency (e. a, secured to the vibrator at all points along the unit in a direction which is longitudinal both of the unit and of the vibrator), it may be a simple, single crystal of piezo-eleotric material (such as Rochelle salt) with electrodes on its two sides forming its electrical terminals, I have generally preferred, however, to employ a double-crystal unit of the so-called bi-morph type-comprising two crys tals secured together so that bending of the unit involves elongating and shortening tendencies in the respective crystals, and having an electrode between the crystals forming a first terminal and respective outer electrodes electrically connected together to form the second terminal, an oscillatory bending of the unit serving to cause a corresponding voltage to appear across the terminals. Such a unit is effective either when mounted as abovementioned (then because of unequal clongations and shortenings of its two crystals) or broadly when mounted in any manner which causes an oscillatory bending of the unit in its longitudinal direction. Accordingly, I shall first illustrate an embodiment of my invention ernploying such bi-morph piezo electric units translating devices; but it Will be understood, as will hereinafter be the subject of typical illustra tion, that many of the aspects of my invention are generic to other forms of translating devices as well as these pieZo-electric units.

In Figure 8 I illustrate a free-free beam vibrator comprising a metal tube 52 suspended by a cord 53 from appropriate supporting pins 5|. This tube may for example be of the material and type commonly used for the purely mechanico-acoustic production of chime tones, though no unnecessary limitation in this respect is intended. Following conventional practise for such tubes, it may if desired be provided with a plug 52a secured within its top extremity, above the level at which the cord 53 passes through the tube; this plug may for example be of insulatin material, such as the phenol-resin composition known as Bakelite. Also following conventional practise, the tube or vibrator 52 may be subjected to impulse excitation, to place itinto vibration, by a hammer 5*? moved by 'a conventional means (such as an electromagnetic action) not shown. Attention is called, however, to the fact that, as shown in Figure 8, the hammer proper is secured to its shank 54a, as through.

with the vibrators of earlier figures.

a bushing 54b lockable in any desired position longitudinally of the shank, so that it may strike the vibrator either in the very end portion of the latter (e. g., opposite the plug 52a) as conventionally and as illustrated in full in Figure 8, or at other points within the upper portion (for example, the top quarter) of the length of the vibrator.

In connectionwith the vibrator 52 of Figure 8 I have shown two translatin devices SI and 6! which may be considered as respectively fulfilling, in a general sense, the functions of the devices 2! and 2! of Figure 2. These devices have been shown as bi-morph piezo-electric units of the type above described. Preferably, as a protection against mechanical damage and atmospheric disintegrating effects, the translating units will be positioned inside the tubular vibrator, when that is large enough in diameter (as assumed for the particularly illustrated vibrator) so that the units can be introduced and positioned therewithin by the aid of appropriate tools. The device 6l-illustrated on enlarged scale in Figures 10 and 11-will be seen to com prise the two crystals Bla, the interconnected outer electrodes Blb forming one electrical terminal, and the central electrode Bic forming the other electrical terminal. Because of the curved (e g, concave) shape of the surface to which the unit is to be secured, it may first be cemented to a thin slab Gld (for example of Bakelite) Whose opposite surface is curved to fit, and will be cemented into intimate permanent contact with, the interior vibratorsurface. The translating device GI may be a unit in all respects similar to the unit 6%. (And it will be understood that the optional use of a simple, single rather than bimorph, piezo-electric unit would in each unit involve simply the omission from each illustrated unit of that one of the crystals and that one of the electrodes 6 lb which i further from the surface of the vibrator.)

The determination of proper positions of the translating devices longitudinally along the vibrator involves considerations generally analogous to those above pointed out in connection There is, however, a specific difference in the configurations which the vibrator assumes at the peaks of its vibration at its several partial frequencies, between the clamped-free and the free-free vibrators. Those configurations for the latter (considered without any weight analogous to IE) are presented as the family of curves I02 through ill! in Figure 12. The specific differences from Figure 4 (other than those second-order ones incidental to the contemplations of presence and absence of a weight in the respective figures) are confined to the region of the vibrator which, for the clamped-free vibrator, is adjacent the supported endthough in this region they cannot be neglected. In spite of the specific differences just mentioned, if in the Figure-8 structure the pick-up 6| is positioned at (i. e., so that its longitudinal mean point is at) approximately .266 times the length of the vibrator, and the pick-up 6i is positioned at approximately .55 times that length, from the supported end of the vibrator, then there may be obtained an action generally similar to that of the first-mentioned adjustment of the Figure-2 structure.

Thus the electrical outputs of the pick-ups 6i and 61 may be connected together through respective potentiometers 22 and 22' in such a phase that the oscillations respectively translated by the two pick-ups from the seccnd-partialvibration will be opposed against each other. (This may in a typical case permit the connecting together of corresponding terminals of the two units (it and 6! and accordingly, if desired to save conductors, the central electrode of each unit may be electrically connected to the vibrator immediately adjacent the unit, and a common conductor'employed from the vibrator to one side of each of the potentiometers. The other terminal of each unit may be electrically connected to the respective potentiometer by a respective conductor passing through the plug 52a.) The amplitudes of the composite oscillations translated by each pick-up unit may be regulated by the respective potentiometer; and the regulation may for example be such that the second partial is fully balanced out from the combined pick-up output. It will be seen from Figure 12 that at the same time there will be an inter-pick-up reinforcement of the third partial, as well as of several of the partials higher than the third. Thus this arrangement is well adapted for the production of a tone in which the third partial is to form the fundamental. Furthermore, an offbalance adjustment for the second partial, in either direction from true balance, will provide a second-partial hum tone of any desired amplitude relative to the rest of the tone.

The expedient of weighting to improve the inter-partial frequency ratios may be employed with the free-free as well as with the clampedfree vibrators. Accordingly in Figure 8 I have shown, adjacent the bottom extremity of the vibrator 52, a Weight '55, of generally cylindrical form surrounding and adjustable along the vibrator. The bore of the Weight may have an approximately central portion 550; fitting the vibrator, and slightly oversize portions 55b above and below 55a; a set screw 55c passing through 55a may serve to hold the weight in place on the vibrator. weight, having been detailed in connection with the earlier embodiment, need not now' be repeated.

In general, there are applicable to the freefree vibrator the considerations detailed above with respect to the use of the respective partials of the clamped-free vibrator. Along lines already discussed for that case, it may be desirable to assign a frequency other than that of the third partial-for example, a frequency of half the fifth partial-as the fundamental or pitch-determining frequency. As in that case, it will then usually be desirable to carry out the translation in such a manner as to reduce the prominence of the third partial in the net translated oscillations, so that it at most supplies a hum tone. And as in that case, this may be accomplished by properly re-positioning the translating devicestypically, by placing them at a pair of positions of which one incident will be a closer spacing to each other. Thus the translating device further from the supported end (59) may be replaced by one nearer the other device; and while this may be nearer by any desired degree, with corresponding effects (as brought out in connection with the device 2 l in Figure 2), I have preferred for most purposes, and have illustrated in Figure 8, a quite near position, analogous to that of the device 2 I in Figure 2.

In Figure 8 the device so to be substituted for Bi appears as 5V; it may for example be a piezo-electric unit such as already described. Typically it may be positioned at .333 times the Further considerations regarding the length of the vibrator (when the device '6l is at .266 times that length) from the supported end of the vibrator. The substitution is indicated as eifeoted by the switch 63. Assuming an adjustment of the outputs of El and 6V so that the second partial is balanced out, it will be seen from Figure 12 that in the net translated oscillations there will be a restricted translation of the third partial, a medium translation of the fourth, and an almost normal translation of each of the next several partials. Thus (as in the case of the tone from the Figure-2 embodiment insofar as that is supplied by the main vibrator) there will be obtained a subuint tone of restricted amplitude, a supplied fundamental, a quint of medium amplitude, and a good development of further partials, of which the first two will be of approximately Z-times and 3-times, respectively, the frequency of the fundamental. And it will of course be understood that the adjustments here mentioned in connection with the Figure-8 structure are just as susceptible of shift for the production of special effects, as are those of the Figure-2 structure as above described,

When it is desired to expand the harmonic development just mentioned to include another hum. tone and/or a real (in addition to a suplied) fundamental-or, indeed, any other componentor components-there may be utilized the expedient disclosed above in connection with the Figure-2 structure: the supply of any or all those additional components from supplementary vibrator means excited simultaneously with the main vibrator. While there is again intended no unexpressed limitation as to the form or manner of simultaneous excitation of the supplementary vibrator means, I have in Figure 8 by way of example shown those means as separat from the main vibrator and subjected to separate excitation simultaneous with that of the main vibrator.

Thus at St I have illustrated a rail, to which through lug there may be secured a clampedfree supplementary vibrator 58 in the typical form of a round rod. The first-partial frequency of the supplementary vibrator may for example be half the frequency assigned as the fundamental of the main vibrator 52, so that it will contribute a sub-fundamental hum tone; and the supplementary vibrator may if desired be slightly thinned near its base (as at 58a) to bring its second-partial frequency to 3 /2 or 4 times the fundamental frequency assigned to 52 (as above di cussed for the supplementary vibrator 32). The supplementary vibrator 58 may be arranged to be struck by a hammer E l, whose shank 64a is actuated simultaneously with the shank 54a of hammer 54, for example through electromagnetic action means well understood in the art and not herein necessary to show. Translation from the supplementary vibrator 58 may be eifected in any convenient manner-typically by means of an electromagnetic translating device or pick-up 60 supported in spaced relation to the vibrator and for example having its output connected across potentiometer (from which the oscillations translated from the supplementary vibrator will be merged with those from the potenticmeters 22 and 22 for the main vibrator, as in the Figure-2 structure) One or more further supplementary vibrators, such as second one 59, may be secured to 58 to be simultaneously excited and to provide one or more other tone componentsfor example, an actual fundamental for the output tone. A

'form which 59 may take is that of a relatively thin vibrator having one end secured to 58 near lug 5i, and extending downwardly alongside 58 so that the two in common will influence the translating device t6. Control of the oscillation 5 output from 59 relative to that from 58 may of course be effected by bending of 59, as already discussed for the supplementary vibrators of earlier figures.

With the free-free main vibrator there may of course be employed, alternatively or additionally to the expedients already disclosed for the reduction or elimination of lower partials, the high-pass filter means which were disclosed in connection with Figure 2. Accordingly I have shown, connected in the output circuits from pctentiometers 22 and 22 as in Figure 2, the highpass filter 43, which may be employed when desired for the function already made apparent.

I have thus far disclosed, as applied to the free-free vibrators, the use of expedients for the reduction or elimination of partials which are similar to those disclosed at an earlier point herein for the "clamped-free vibrators. There are, however, other expedients for that reduction or elimination which, while not in all respects limited to the free-free vibrators, are especially easy to carry out therewith. These include, for example, the elimination of an undesired low partial by the placement of the translating device quite exactly at a node for that partial, and the accompanying elimination of another low partial by the excitation of the vibrator at a node for that other partial. (In general, neither of these manners of elimination will be relied on for the 5 first partial of the free-free vibrator, in the terminology herein employed; but the mounting of the translating device or devices on the freefree vibrator has eliminated the first partial,

rendering the manners of elimination just mentioned particularly available for other low partials.)

Reference being had again to Figure 8, there will be seen a translating unit 62 mounted on the vibrator 52 at a lower position than those of the translating units 6!, ii i and BI'. For this point there may be chosen one which is a node (or almost a node) for one of the low partials to be eliminated (or attenuated). For example, for elimination of the second partial this point may be approximately .78 times the vibrator length from the supported end of the Vibrator, which will be seen from Figure 12 to be a node for the second partial (that is, in the absence of weight 55, a slight re-positioning-readily determined by test-being necessary for full elimination when any given weight is employed). This translating unit 62 may for example be of a form such as already described for the unit 65. It may be substituted for 6! by the switch 54; and at the same time contributions from either 6| or 6l' may be avoided by moving the switch 63 to a blank switch-point 53a. With such an adjustment, of

course, the translating unit 52 becomes the sole translating means effectively associated with the 65 main vibrator.

Further, the hammer 54 may be re-positioned along its shank 54a, for example to approximately the position shown in Figure 8 in dotted lines. This position may be chosen so that the hammer will strikethe vibrator at a point which is a node (or almost a node) for another of the low partials to be eliminated (or attenuated). For elimination of the third partial this point may be approximately .13 times the length of the vibrator from the supported end of the vibrator.

With the structure adjusted as just described, the fourth partial of the main vibrator 52 will be the lowest partial to appear in the oscillations translated from that vibrator. While this partialmight be relied on as the fundamental of the output tone, it is probable that in typical cases it may be more desirable to assign as that fundamental a frequency of half that of the fifth partial, as to which the fourth partial will become a quint, the fifth partial an octave, and so on as above described. ihen if there is desired a hum tone and/or a real (as distinguished from a supplied) fundamental, supplementary vibrator means such as 5858 may of course be relied on. A hum tone may, however, be obtained without the use of the supplementary vibrator means (e. g., with potentiometer 39 adjusted for ineffectiveness thereof), by slightly displacing the translating unit or the striking point from that which fully eliminates the respective partial-e. g., by displacing the hammer upwardly along its shank 54a, so that it strikes above the node for the third partial by whatever distance is required for development of a desired third-partial (in the typical case abovementioned, sub-quint) hum tone.

And it will be understood that the adjustments just suggested are not primarily limitative, but rather are exemplary of eliminations, attenuations and control which may be effected by joint regulation of striking point and single translating point. Further, it will be appreciated that the regulation of striking point offers a measure of additional control even when a plurality of translating devices are employed. For example, with the devices 65 and BH in use as above described, the last-mentioned adjustment of hammer 54- would serve fully or almost fully to eliminate the residual small amplitude of third-partial (e. g., sub-Quint) hum tone; or the hammer might be placed at (or nearly at) a node for 1 the fourth partial, to eliminate (or attenuate) the otherwise medium amplitude which that partial (for example, furnishing a quint component) would otherwise furnish.

Damping means such as described in connection with earlier embodiments may be employed with the free-free vibrators. The damping means may take the form of a structure El comprising a ring d8 of a damping material such as abovementioned surrounding the vibrator, and a metallic ring 69 surrounding the ring 58 and serving to build up the total mass of the structure to that necessary for the desired degree of action. By way of example, the structure 6'! has been shown as positioned at approximately .066 times the length of the vibrator from the supported end, at which point the fifth and sixth partials have substantial nodes. Accordingly these partials will be little affected by the damping structure, which will however be quite effective as to the other partials.

If the means which supports the ffree-free vibrator-e. g. the cord 53-be made of material characterized by considerable energy absorption, or damping influence, it itself will act as a damping means. As illustrated in Figures 8 and 9, thiscord is connected with the vibrator at a point about .03 times the vibrator length from the vibrator end; since this point is a node only for some very high (and usually fairly weak) partials, the use of an appreciably energy-absorptive material in this case would result in a rapid damping of all the more important vibration components. While that may be desirable in certain instances, it will more frequently be desired to achieve a more selective action, with minimum damping of certain selected partials. Ac-

cordingly I have shown in Figure 13 the support of the vibrator, by a cord 53' of yieldable and en- 5 ergy-absorptive material, at a point similar to that employed for the damping structure 6? in Figures 8 and 9-az1 arrangement which may be made to approximate the efiects described above for the vibrator equipped with that damping structure, without the actual use of such structure.

In the embodiment of Figures 8-9 (specifically, in Figure 9) -e is iilustrated a feature permitting an alternative achievement of a simultaneous supporting and dampin function. This feature is to be found in the screw or other member 69a secured in and project ng outwar y from the ring iii in the damping structure This member 569a may be employed as a rigid support for the vibrator 52, e cord 53 then being omitted; in spite of the ngidness of the free' free vibratory nature is preserved, by reason of the presence of the yieldable material 68-which will, however, cooperate with the mass of the vibrator in providing a c'rong damping action on all partial components not having substantial nodes in its region. -11 this case, since the entire weight of the vibrator is supported from the ring 59 through the piece es of damping material, it may be desirable to cement the latter both to the vibrator and to the ring 69 (rather than to rely merely on frictional gripping).

With the free-free vibrator there may be employe-d a helical spring 65 corresponding to the spring disclosed connection with the embo iment of Figure 2. The spring may have its upper extremity secured in any convenient manner to a peripheral point at the bottom of the vibrator 52, and its lower extremity secured to a suitable stationary member 65. The functions and frequency influences of the spring 35 will be understood to be similar to those of the spring 25 already described-particular note being made of the especial value of the restraint against undue pendular excursions in the case of the free-free vibrator, in which. case they otherwise may tend to attain considerable magnitude. This is particularly true, and the restraining means of particular value, when the hammer 55 is arranged 50 to strike the vibrator at a point considerably re moved longitudinally of the vibrator from the region of support of the vibrator.

In the embodiment of Figures 89 I have shown-specifically by illustration in Figure 9- a further feature which may be employed when desired. This is the provision of a hammer 54 arranged (on a shank Eta, along which it may be adjustable for purposes analogous to those for which hammer 5 5 is adjustable) to strike the vibrator in a plane at right angles to that in which the hammer 54 operates. The translating devices, being arranged for effectiveness as to vibrations in the plane of operation of hammer 54, will not efficiently translate the vibrations of the vibrator at and immediately following an impact of hammer 54 against the vibrator; but as those vibrations progress they will (beside retaining components in the plane of the actually-employed hammer 54') develop and retain components in the plane of effectiveness of the translating devices. Accordingly upon excitation of the vibrator by the hammer 5" there will be produced a tone of gradually rising initial amplitude, rather than one of the usual impactive inception, The hammers 54 and 5 may of course be either simultaneously or selectively operated, according to the type of tone it is desired to produce. When supplementary vibrator means are employed and excited simultaneously with the main vibrator, the full translation of impactive components from the supplementary vibrator means coupled with a translation from the main vibrator with small or no impactive components (achieved as just described) makes possible a selective control of inception characteristicswhich may be considered a counterpart of the selective control of the termination (damping) characteristics as by means already described.

I have so far described the embodiment of Figures 8-9 with particular reference to internal ly mounted pick-ups. and to pick-ups of the piezoelectric type. For reasons already apparent, these have certain advantages, The broader aspects of my invention are not, however, limited to use with such pick-ups. Instead of the piezo-electric pickups there may be employed other pick-ups, for example of one or another of the types operating by the vibratory change in relative positions of two spaced elements (one of which may or may not be the vibrator proper, as desired). A mounting of the pick-ups external of the vibrator may of course be employed. And if desired the mountings may be such as to provide for the adjustment of the pick-ups longitudinally of the vibrator.

Thus in Figure 8 I show, arranged to be sub stituted'for the pick-up Si or 62 by the switch 64, a pick-up which is external of the vibrator and which is of the electromagn tic type. Further by way of example, it has been shown as arranged to be positioned where desired longitudinally of the vibrator; it has, however, been specifically shown at a longitudinal position similar to that of the pick-up 62 above described.

Reference being had to Figure 8, there will be seen a ring 10 surrounding the vibrator 52; this ring may for example have an internal diameter slightly larger than the external diameter of the vibrator, but it is held in any desired position along the latter by a screw ll threaded through the ring into abutment against the vibrator. The ring 10 may be of magnetic material, such as iron or steel. Secured in the ring 10, for example diametrically opposite to the screw 1 I are the extremities of a generally circularly formed light leaf spring 12, this spring extending upwardly from one of its secured extremities and curving first away from the w'brator, then downwardly, then toward the vibrator, and finally upwardly to its other secured extremity. Secured to the central part of the spring and extending therefrom into close spaced relationship to the ring 19 is a pole member 14, preferably of soft iron. Surrounding this pole member is a. coil (schematically shown) and secured to the spring and extending outwardly therefrom in alignment with the pole member 14 is a short bar magnet 16 the pole member, coil and magnet together forming an electromagnetic pick-up 73.

The pick-up 13 is designed to operate by reason of vibratory variation of the spacing between the ring 16 (which of course vibrates with the vibrator) and the adjacent end of the pole member 14. While the latter (and with it the balance of the pick-up 13) is also supported by the vibrator, that support is through the intermediary of the spring 12; and the pick-up, because of its inertia, will tend to remain stationary, this being permitted by a vibratory fiexure of the spring 72 in accordance with the vibrator vibration. The

Cir

f small, it nevertheless will introduction into the translated oscillations of unwanted extraneous frequency components from the resonant characteristics of pick-up and spring may be avoided by such expedients as the mechanical tuning of the combination of pick-up and spring to a very low frequency, for example below that efiectively transmitted by the amplifier 45.

It may be pointed out that to avoid an unwanted damping action when a sizeable mass is supported by the vibrator, that mass should in general be connected with the vibrator through means free of any significant energy-dissipation tendencies. This was achieved in the case of the piezo-electric pick-ups above described (even though they may be of relatively small mass) by reason of the substantial rigidity of the slabs (e. g., Bld) and cement through which they are secured to the vibrator. On the other hand, this is achieved in the case or the pick-up 73 by the support of that pick-up through the spring I2 which, while yieldable, may be substantially free of energy-dissipating effects. If this low frequency be properly chosen and the excitation of vibration strong enough, it may be relied on for the production of a vibrato effect.

The support by the vibrator of a pick-up element which is to move vibratorily with respect to the vibrator need not necessarily, however, be restricted to a substantially dissipationless supporting, if the mass of that element be relatively negligible-for then its damping effect on the vibrator will be negligible. Thus in Figure 8 I show another electromagnetic pick-up 17, in which there is employed a small armature supported by the vibrator through the intermediary of a very appreciably dissipative element. This pick-up T! has been shown, by way of example, as not arranged for adjustment longitudinally of the vibator; and the position in which it is illustrated is an approximately complementary one to that of the pick-up 13 (or 62)i. e., at a point approX- imately .22 times the length of the vibrator from the supported end, at which point the relative strengths of the several partials (at least for the unweighted vibrator) tend to be similar to those at the position of 13 (or B2). The pick-up 1'! may be electrically substituted for SI or 62 or 13 by the switch 64.

The pick-up 11 may include a pole member 78 terminally secured to the vibrator and extending horizontally therefrom. For magnetizing this pole member there may for example be supported on it one end portion of a small bar magnet '19. Surrounding the pole member '!8 beyond the magnet 19 is the schematically shown coil 80. Onto the outer end of the pole member i8 may be cemented a small and relatively thin disc or pad 8|, preferably of very yieldable and permissibly resilient material; and onto the outer surface of the pad 8| may be cemented a small and light armature 82 of magnetic material, such as a thin iron armature.

While the inertia of the armature 82 is very be sufficient so that the vibration of the vibrator is of somewhat attenuated amplitude as it appears in the armature; the attenuation represents vibratory compression and expansion of the pad ii, and consequent vibration of the armature relative to the pole member in accordance with the vibration of the vibrator. While this type of pick-up may be characterized by somewhat lower efficiency than that last described, it has the advantage of permitting critical or far over-critical damping in the I relationship to the vibrator itself.

23 pad-and-armature system, with consequent Eli-1 ination of any spurious frequency components without reliance on any mechanical tuning.

Another form of pick-up which operates by the vibratory change in relative positions of two spaced elements and which may be employed is an electrodynamic piolr-up'in which the two elements are respectively a looped conductor (i. e., a coil) and a magnetic-field-producing element. Thus in the fractional Figure i l I show a bar magnet 8 secured. to the vibrator 52 and extendhorizontally therefrom, and a coil 85 coaxial with and freely surrounding the magnet B l-the coil 85 for example being supported from the vibrator by light leaf springs 86, which may be mechanically tuned with the mass of the coil to a very low frequency. In view of the inertia of the coil it will tend to remain stationary when the vibrator is vibrated; and the resulting vibratory variation in the relationship of the coil to the field-producing magnet will cause the induction in the coil of vibration-rcpresenting oscillations.

It will be appreciated that in the electrodynaniic piclr-up just described a substantial mass is supported by, and movable relative to, the vibrator, by means of substantially dissipationless elements. An electrodynamic pick-up may equally well, however, employ a very small mass movable relative to the vibrator and supported by means characterized by substantial dissipation. Thus the fractional Figure I show a small post 39 secured to the vibrator 52 and extending horizontally a short distance therefrom to support a coil 9-0 coaxial with the post. Onto the outer end surface of this coil may be cemented a disc or pad ti (similar for example to the pad 8.! above described), and onto the outer surface of the pad may be cemented a miniature magnet 92 (for example of the highly efficient iron-aluminum-nickel-cobalt alloy commonly known as Alnico) magnetized in a horizontal direction. Vibrations of the vibrator will appear in the magnet in somewhat attenuated amplitude; the attenuation representsvibratory compression and expansion of the pad, and thus vibration of the magnet relative to the coil, with consequent induction in the latter of vibration-representing cscillations.

Still another form of pick-up which operates by the vibratory change in relative positions of two spaced elements and which may be employed an electrostatic pick-up-the simplest embodiment of which comprises an electrode in spaced Since the electrode is readily made of very small mass, the presonce or absence of dissipation in the supporting means between it and the vibrator will in the usual case not be important. In Figure 16 I show secured to the vibrator 52, at several points therealong so that several partial compositions are selectively available, a number of electrostatic pick-ups. A very simple one appears as the light metal sheet or electrode ill, curved, into a cylindrical section preferably of slightly greater radius than that of the outside of the vibrator. The center of this electrode ill may be insulatedly secured to the outside of the vibrator by means of the screw H2 and the thin insulating washers H3, a flexible conductor lit being electrically connected with the electrode immediately adjacent the washers.

Another flexible conductor H9 may be electri cally connected with the vibrator in any convenient manner, as schematically illustrated in Figure 16. Between the conductors H0 and H4 there may be connected, in series, a voltage source H6 and a high resistance H5; and across H5- i 15, through a condenser Hi, there may be connected the input of the cascade i-561-8 abovernentioned.

If the voltage source Hi3 be of steady or direct voltage, there will be charged through resistance H5 the capacity between the vibrator and the electrode 5 l i; vibratory variations in this capacity will cause a variation of the voltage across it, and the variations in voltage will be trans mitted through the condenser i l"! to the cascade abovementioned. On the other hand if the source lit be of oscillations of sup-eraudible frequency, a current of that frequency will flow through I l5 and 5 l6 and through the capacity between vibrator and electrode i l l vibratory variations of this capacity will modulate this current at the vibrator vibrational frequencies. A voltage of the somodulated superaudible-frequency oscillations will appear across lit-l l6, and will be transmitted through the condenser to the cascadein which the amplifier it will in this instance be operated as a demodulator as well as an amplifier.

The electrode of an electrostatic pick-up may of course mounted internally of the vibrator if desired. Accordingly I show another electrode E2! in the form of a light metal sheet curved into a cylindrical section preferably of radius slightly smaller than that of the inside of the Vibrator. The side-to-side center of this sheet may be connected to the vibrator by a small vertica1 strip 222 of cement; and as an additional securing means there may be stuck across the center of the inner surface of the electrode, and onto the vi.- brator surface thereabcve and therebelow, a Vertical strip 523 of tape. A flexible conductor I24 may be connected to the center of the electrode. By a switch lilo this conductor may be substii-uted for the conductor 2 ii, thereby substituting the pick-up last described for that previously described.

Another form of electrostatic pick-up may, as illustrated in Figure comprise a light metal sheet l3l acting as electrode, cemented to the outer surface of a "pad it? of very soft, yieldable material whose inner surface is cemented to the vibrator. A conductor 234 may be connected to this electrode, and maybe substituted for conductors i it and 123 by the switch 558.

Another construction illustrated in Figure 16 comprises a strip electrode i i-i having one extremity secured (as by cementing) to a block .12 of insulating material carried by the vibrator, the electrode being curved so that its principal portion extends along the vibrato in very close spaced relationship thereto. The showing of electrode E li includes preferred thin coating i i-3 of rubber on its inner surface adjacent the vibrator, to prevent actual short-circuiting in the case of maladjustment Or the like. There is also illustrated in Figure 16 another strip electrode ll, differing from t ll in that, instead of being supported in cantilever, it is supported at both extremities on respective insulating blocks [52. It too may have a thin coating (list) of rubber on its inner surface, if desired.

Connections to the electrodes h ll and Lil may be made by conductors Hit and i, respectively. In Figure 16 these have been shown as electrically connected together, to illustrate the joint use of two electrostatic pick-ups. In this case the relative contributions of the two may be adjustef. by differential adjustment of the spacings of the two electrodes from the vibrator, as well as by choice of their respective areas. The phase re lationship between the two may be most simply controlled by appropriately choosing the sides of the vibrator on which the two electrodes are respectively positioned; as illustrated, they are on opposite sides, for opposition partials Which a in similar phase. conductors M4 and I54 may be H4, I24 and I34 by the switch combined electrodes I 4: and I translating means.

While I have described my invention in terms substituted for ISO, to render the 5| efiective as the of particular embodiments thereof, and with reference to two typical forms of vibrator (clampedfree and free-free beams), I intend thereby instrument including a beam for vibration: a Weight for altering the ratios between lower partial frequencies in said vibration, a longitudinally restricted 2. In a musical instrument: the combination of a beam vibrator supported in the neighborhood of one extremity thereof and from the vibration of Which output sound is translated, and means under tension secured to the other extremity of said vibrator.

Images