US2221814A - Musical instrument - Google Patents

Description

Nov. 19, 1940. v E. E. REID 2,221,814

- MUSICAL INSTRUMENT Fqled May 25, 1957 5 sheets-sheet 1 fl l INVENTOR I W4 e?? n 'H EEMMET Reidzso 29 av Flr b X I7 v ATTORNEY Nov. 19, 1940. E. E. REID 2,221,314

` MUSICAL INSTRUMENT Filed May 25, 1957 5 Sheets-Sheet 2 EEB /5 2 I INVENTOR E QZ l EEMME? Reid N91/mm WMM ATTORNEY Nov. 19, 1940. E. E. REID 2,221,814

MUS I CAL INSTRUMENT Filed May 25, 1937 5 Sheets-Sheet I5 INVENTOR lal-:YEMMEM* Reid ATTORNEY NOV. 19, 1940. E, E- RElD 2,221,814

MUSICAL INSTRUMENT .Filed May 25, 1937 5 Sheets-Sheet 4' -ATTORNEY .M .m 3 m mim k R T43 man m .m m m M m am 'mm m m mm E. mg rwm am w m am N mm wm@ rmww hm L.- r cmq-. m o Tf/m5: rm :mmm c c c :E m .f lm 23T I om La: 22H9 m32?? EN mi mm mm MNT l: S

new Qn I n nh .mmh :d mn wv sgg D C. m Q x :U

Nov. 19, 1940. E. E. REID MUSICAL INSTRUMENT Filed May 25, 1937 5 Sheets-Sheet 5 INVENTOR m n E N R my; MA Mh Em and the like.

Patented Nov. 19',

UNITED sTATEs PATENT OFFICE 12 Claims.

My invention relates to musical instruments Important advantages that can be realized in'practice through the invention include flexibility or operation; wide range of pitch and volume, covering as many octaves as desired; perfection and variety of tone quality, equalling or exceeding what is possible on the best organs; and ability to play in either the equitempered or the natural harmonic scale, with full variety of notes in all octaves. Without excessive complication, an instrument embodying my invention (as hereinafter described) can be changed at will from the equitempered scale intonation to a natural harmonic scale intonationor any other desired scale of intonation,-and vice versa; and in the harmonic scale, it` can be 'adapted at will to music written in any key(s). The shift of scale or key of musical intonation can be made by the player in a moment, and can be repeated as often as desired during the playing of a composition.

The changing of scale or key of musical intonation according to my invention is readily applicable to any type of instrument in which the pitch of the musical notes or tones depends on the rate or speed of motion of controlling parts, and may be effected by changing the rela: tive speeds of various pitch-controlling rotors. One way of doing this lato change the ratios of gear combinations bywhich the various rotors are driven from one or more common driveshafts. As Ithecorresponding notes in all octaves covered by anV instrument can be produced byone rotary unit, and as only twelve such rotary units are generally needed in a complete instrument, the mechanism for effecting the change-over may be comparatively simple. I have hereinafter explained these and other features of my invention in connection with an electrical type oi' in strument in which rotating devices cooperate inthe generation of electrical impulses which are converted into sounds. But it will be understood `that this is only illustrative, so far as the broader features and principles of the invention are concerned.

One way of producing electrical impulses of suitable nature and frequency for musical pier-7 poses is to vary a magnetic ux affecting a con ductor, such as a coil or winding, at a frequency corresponding to the frequency of soundvibrations of the pitch desired, thus producing an alternating electric current in theA conductor. The frequencies useful in music are those from about 25 to about 6000 or more'per second. Very weak currents can be amplied (by means well known in radio) to actuate a diaphragm (such as used in a telephone, or in a radio or other dynamic loudfspeaker), thus producing sounds of corresponding character. A way of producing flux changes in a coil or winding is to vary the permeability in a magnetic field affecting it,

as by vibrating or otherwise moving a ferromagnet device or member in or through the field with suitable frequency. A rotor embodying `one 10 or more (soft) ferrous metal (iron or steel) members answers this purpose very well; and a disc or other rotor with an end face presenting a series of such members forming a track of varied permeability affords special advantages.

I at present generally prefer to produce it electrically, by passing (direct) current through an exciting "primary" coil or winding associated with the generating coil as secondary. By controlling and varying the exciting current in the primary coil, the intensity of the resulting musical tone from the sound device (loud-speaker.v etc.) can be controlled and varied, and several important results thereby secured. However, like results can be obtained by the well-known means of volume-control employed in radio, which may be used either in lieu ofl excitation control, or concurrently therewith.

The tones or notes of the organ and of other musical instruments are complex, including harmonic" vibrations of frequencies higher than the dominant or fundamental ones, and also (sometimes) a lower frequency vibration, which gives dignity to the tone. It is by the number, relative intensity, 'and other interrelations of acters of different instruments are determined. In the tones produced by .pipes and strings, it is found that if the frequency of thefundamental these component partials thatA the tonal char- Q same electrical instrument Just as desiredf-such as organ diapason, violin, oboe, etc. Alternatively, and with like variety of effects, the relative intensities of the electrical impulses for the various partials may be suitably changed in transmitting them to the sound device, by wellknown means of .volume-control such as employed in radio. If desired, these modes of proportioning the partials may be used together.

While a single loud-speaker or sound device may suilice, it is often advantageous to employ a number of such devices, suitably connected to the electrical output circuit of an instrument. In a large auditorium, or for out-of-doors concerts or the like, a multiplicity of sound devices will often be needed to produce suflicient volume of sound without forcing. Also. an extra sound device may sometimes be installed in a distant gallery or cupola, to give echo organ effects. with arrangements for throwing this or the main sound device into or out cf action, so that they shall be heard alternately.or even both together sometimes. In tine instruments, it may be preferred to use several different sound devices, diiferently constructed or adiusted, for tones of different pitch or quality, so as to produce all tones as correctly and agreeabLv as possible: e. g., one sound device for low notes, and another for high notes; or one for low tones, another for intermediate tones, and yet another for high tones.

The partials of any one note or tone may be produced all from the same tone-track, or from several tone-tracks, or each from a separate tone. track. In employing rotors with end-faces or disc-elements presenting tone tracks of varied permeability. I prefer to provide tracks for all the partials of a tone or note on the endface(s) of a single rotor. Indeed, I mayeven provide yfor two notes an octave apart on the opposite faces of the same disc or rotor. Whatever the type of rotor structured prefer to combine in Aone rotary unit the permeability-varying devices necessary for corresponding notes (including all their partiels) in all the octaves covered by an instrument. This not only insures that these notes shall always have the correct relations to one another, but also simpliiles the change-over between the equitempered and the harmonic scales, or from playing in one key to playing in another, as mentionedl above.

The rotors of my instrument may be driven by any convenient (electric) motor(s). such as a constant speed type of A. C. motor supplied from an A. C. lighting or power circuit with constantfrequency regulation. y For exciting the magnetic iields of electrical impulse generators. (direct) current may be taken from any suitable source, with constant voltage regulation-such as a battery of dry cells, a storage battery. a D. C. light ing or power circuit, or a motor-generator set operated from any suitable A. C. circuit.

Various features and advantages of the invention besides those already referred to will become apparent from the following description of species or forms of embodiment, and from the drawings. Ail the features shown and described are. indeed, of my invention, so far as novel over the prior art.

In the drawings, Fig. l is an elementary diagram of one electrical type of musical instrument embodying my invention.

l}rig.ris:.msspectivsortuteiiview(cartina..-

grammatic) of a generator of electrical impulses mutable for use in the instrument of Fig. l, including a disc type of rotor and a coacting generator device, and provided with means for intensifying the generating action-the rotor disc being shown partly broken away and in section.

Fla. 3 is a similar view of a rotor of somewhat different form and construction, with a plurality of associated generator devices.

Fig. 4 is a similar view illustrating yet another form of rotor and generator device.

Fig. 5 is a diagram illustrating the generation and synthesis of partials to produce a complex musical tone. 1

Fig. 6 is a perspective or tilted diagrammatic view of a disc rotor with a multiplicity of tonetracks and generator devices, adaptable i'or producing the component partials of a complex tone or note. and for use in an instrument (like an organ) having a plurality of playing manuals or keyboards.

Fig. '1 is a diagrammatic side view of a gener- Fig. 10 is a diagrammatic plan-like view illustrating the organization of an instrument adapted to be changed over for playing 'in different scales cr different keys,either an electrical instrument, or one of any other type in which musical pitch is controlled by speed of rotation,-some parts being shown in cross-section, and certain parts omitted for the sake of clearness.

Fig. ll is a diagrammatic side and end view of parts of the mechanism at the extreme right of Figs. l0 and l2 in proper relation to one another, showing some parts that are omitted from Figs. l0 and l2, and omitting certain parts that are shcwnin Figs. l0 and 12.

Fig. 12 is a diagrammatic plan view of parts which are omitted from Fig. 10 because they overlie those there shown, but which appear in Fig. I1.

Fig. 13 is a view at right angles to Fig. l2. illustrating the action of an interlock in connection with the mechanism of Fig; 12.

Pig. i4 is a fragmentary diagrammatic view illustrating a coupler device dinerent fromthcse inFigs. l0andl1,withassociatedpartsoi'the mechanism.

Fig. 15 is a fragmentary view at right angles to Fig. 14.

Fig. i6 is a fragmentary view similar to Fig. 1i, illustrating a variant arrangement for changing over from one scale or key to another.

Elementary general explanation An electrical impulse generator is shown in ll'lg. 1 as including a device P for producing ilux vari- `ations. and a coacting generating device Q comloud-speaker or a telephone receiver: and an amdevice S. While a magnetic field or ux affecting A the generating coil I may be produced by a permarient magnet (core) associated with this coil or winding as part of the generator device Q, I prefer to produce the field flux electrically, by including in the generator device Q an exciting field coil or winding 20, in inductive relation as primary to the coil I5 as a secondary. The activating primary coil is connected by leads 2I in or across a (direct current) power circuit 22, at `points i, gi. The coils I5, 20 may be wound on a common core 24, which is preferably of soft ferromagnetic (paramagnetic) material, as soft steel or iron. 'Ihough illustrated in Fig. 1 as a straight bar witha conically pointed active end, the core 24 may in practice have any preferred form. To

produce flux variations laffecting coils I5, and 20 by variations of the' magnetic permeability in their field, a device, track,or series-of-members P affording -such permeability variations may be associated with the generating device Q, with provision for relative movement of the devices P and Q. As here shown, the device P consists of a serrated or sinuous ferrous metal strip or wire 25 near one (pointed) pole end of core 24. A

` playing key K is shown in Fig. 1, which when ope'rated (depressed) closes a normally open (and self-opening) switch 26, either in the total .secondary circuit between coil I5 and sound device S, or in the-total primary circuit between coil 20 and the source of exciting current,or both. Itis convenient to distinguish portions of these total circuits 4including the tone control switch(es) 25 as secondary and primary playing circuits 21, 28: as marked in Fig. 1, they are interposed between one lead I6 and a point of connection It to the output circuit I1, and between one lead 2|' and a point of connection i to the power circuit 22. Circuit 21 may also be appropriately called a tone circuit."

With switch(es) 26 vclosed and current flowing in the primary coil 20, relative movement of the regularly serrated or corrugated wire 25 and the generating device Q at uniform speed produces a series of electrical impulses (an alternating current) in the secondary coil I5, playing or tone circut 21, and output circuit I1; and this current is amplified and converted into a musical tone or note by amplifier X and sound device S. I usually prefer to corrugate the Wire 25 to such a shape that the alternating secondary current is truly sinusoidal and produces a pure tuning-fork tone in the sound device S; but other shapes may be used, especially in large instruments. Itis generally more convenient to move the device P (the wire 25), rather thanthe generating device Q. The frequency of the secondary current and the pitch of the tone depend on the speed of movement of the wire 25 as it travels in varying proximity to the generating device Q, as well as on the pitch of the corrugation of the wire 25; while the intensity or loudness of the tone depends on these same factors and on the strength of the exciting primary current, and can be changed at will by changing this exciting current. The intensity or loudness of the tone can also be changed, of course, by changing the proportion in which the secondary current affects the sound l device, S,-as by a variable resistance .29 shunt- Various rotors and generator devices As already suggested, it is advantageous to revolve the permeability-varying device P'adjacent the generating device Q, and to embody or present it at an end or lateral face of a rotor R (whichis preferably of disc form), as a track extending around the axis of rotation. Using the corrugated wire 25 of Fig. 1 (or the like) for the device or track P of the rotor R, I prefer to mount this wire 25 at the face of a (non-magnetic) carrier disc-element, as by embedding the wire (wholly or partially) in a groove 3U concentric with the axis 3I, Fig. 2. As here indicated, and partly revealed by the partial cutting away of the carrier marginthe wire 25 is of rectangular (square) cross-section, is bent into a cir- 'cle or hoop, and is also bent in corrugations or serrations at right angles to the general plane of the hoopv and of the carrier face. The circumference of the hoop 25 is an exact multiple of the pitch of the corrugations, so that an exactly integral number of corrugations is included. Preferably, the hoop 25 is a continuous ring, Without break or joint of such character as to affect its permeability;-or else any break or joint is located in a depressed bend, where its influence on the magnetic field of the device Q will be minimized. y Using a carrier of suitably soft metal like brass, copper, or aluminum, the wire 25 may be secured in the annular groove 30 by slightly forcing in the sides of the groove, either all the way around, or at suitable intervals as indicated at 32.

The electrical impulses or alternating current produced in secondary coil I5 by rotation of the soft ferromagnetic tone-track consisting of hoop 25 adjacent generator device Q may be considerably strengthened by inductively magnetizing the hoop adjacent the device Q to a polarity the reverse of that in said device Q due to its exciting primary coil 2G. For this purpose, a magnet device or inductor 33 may be placed close adjacent the hoop 25 at a point directly (diametrically) opposite said device Q with respect to the axis 3 i, 'Ihis device 33 is shown as a magnet bar with one (conically pointed) end close to the hoop 25 as it revolVes,-either a permanent magnet bar ofv steel, or a soft ferromagnetic bar inductively magnetized by a coil 34 wound around it, and connectlble to any suitable (direct current) circuit. Thus the polarity of hoop 25 adjacent the device Q is the reverse of that induced in said hoop adjacent the device 33. If the bar 33 is a permanent magnet, the coil 34 may be useful to remagnetize it to proper strength occasionally'.

The non-magnetic carrier of rotor R is shown in Fig. 2 as a flat disc of uniform (axial) thickness sufficient to embed and secure the corrugated wire 25, with reasonable strength of metal at the bottom of the groove 3i). The non-magnetic carrier disc of the rotor R shown in Fig. 3 is neither fiat nor of uniform thickness, but conical-faced. Sometimes it may be desirable to make the carrier of double thickness, more or less,-in effect embodying two disc elements hack to back,-with a serrated ferromagnetic hoop 25 in eachlateral or end face, and a correspending' generator device Q for each hoop. In

this case, the thickness may be sufficient to obi/i ate any objectionable cross" influence of either hoop 25 on the generator device Q associated with the other hoop. But when the musical terrelations of the devices Q, Q render such cross influence unobjectionable, or when such cross iniiuence can be precluded by making the 'two hoops 25 of very different radii, the minimum thickness is merely that required for mechanical strength. Aside from considerations of space and cost, there is no maximum limit of thickness: i. e.. one or two disc elements may be embodied in a carrier of any thickness or axial length whatever. Fig. 3 also illustrates a variation in the continuous serrated hoops 25, 25, which do not consist .of bent wire as in Figs. 1 and 2, but of dat strip metal, with straight edges at the groove bottoms and serrated or corrugated outer edges adjacent the devices Q, Q. As here shown, these hoops 25 are of different radii.

Fig. 3 also illustrates an additional and more powerful generator device Q associated with one of the serrated hoops 25. This device Q has its core 24 bent to a horeshoe or U-shape, with conically pointed ends spaced apart along the path of travel of the ring 25 a distance equalrto (an integral multiple of) the pitch of the hoop corrugations, Both the core ends being located close to `the path of the crests of these corrugations ,the magnetic circuit in the core 24 is completed through the hoop when the hoop crests are adjacent the core ends, and is broken when the hoop depressions are opposite the core ends. In other words, the ring 25 completes the magnetic circuit of the core 24 only periodically, as

different crests concurrently pass the core ends.

This great change in magnetic permeability produces correspondingly strong electrical impulses in the generating coil I5, shown on one arm of the core 24. 'I'he exciting or activating coil 2Il is shown on the bend of the horseshoe core 24, where its eii'ect is the greatest.

In the operation of the impulse or alternating current generators wth straight core-bars 25 that are shown in Figs. 1, 2 and 3, it is mainly the crest members or portions of the wire or hoop 25 moving close to the generator devices Q that produce changes of magnetic permeability and flux in Athe fields of the coils I5, resulting in electrical impuplses in these coils. Hence the depressed portions between the crest members might just about as well be omitted, so far as generation of the desired electrical impulses is concerned. Accordingly, Fig. 4 shows a permeability-varying circular tone-track P composed of a series of` equally spaced, separate, soft ferromagnetic members 35 forming inserts set into recesses or holes in the non-magnetic carrier disc of the rotor R. 'Ihese members 25 correspond in number to the crest members or portions of the hoops 25 of Figs. 2 and 3 that would be used to produce the desired musicaLtone. As here shown, the inserts 35 are in effect round plugs xed in holes that extend clear through the carrier disc, so that the members 25 are presented at both disc faces. Also, the generator device Q is shown as diderent from those in Figs. 1, 2, 3. In Fig. 4, the soft ferromagnetic core bar 24 is bent into an elongated chain-link form somewhat resembling a large c. with the (conically pointed) Complex.l tones /roma single tone track A simple way of generating and synthesizing component partials of a musical tone or note aachen vgroove l5 in Fig. 2 and attened. This track P may travel to the right, as indicated by the arrow.

For the first partial or fundamental of frequency n, the generator device Q islikethat shown in Figs. 1 and 2. and its generating and exciting coils I5, 25 are connected by leads I5, 2| to sec-` ondary and primary playing circuits 21, 2l. The generator device Qa for the second partial (or nrst overtone) of frequency 2n is shown as having two generating coils I5. I5 on core bars which in this particular instance are Ilimbs of a unitary forked core structure 24. whose single body carries a single exciting coil 2l. 'I'he coils I5, I5 and 25 of this device Qa are also connected to the circuits 21, 25. As shown. the generator device Q is directly opposite a crest of the corrugated wire 25: the nrst or left-hand limb (and coll I5) of the device Qa is directly opposite another such crest; and its second or right-hand limb (and coil I5) is directly opposite a depression of the corrugated wire. If then. a crest of the moving wire 25 passes coil I5 of device Q every i/n of a second. so as to produce electrical impulses of frequency n in tone circuit 21. such a crest will pass one or other of the coils I5, I5 of device every tn of a second. resulting in electrical impulses in the circuit 21 at a frequency 2n, as required for the second partial.

For the third partial (or second overtone) of frequency in, the generator device Q; is shown as consisting of three devices Q like that in Figs. 1 and 2,-with their core bars 24 attached (like limbs) to a common (supporting) member 25 that may be of either soft ferromagnetic or nonmagnetic material. The coils I5, I5, I5 and 25, 25, 2l are connected to the circuits 21 and 2|.-th'ough as here shown the coils 25, 2l, 2l are in series with one another. 'I'he nrst or lefthand limb (and coil I5) of the device Qi is op posite a crest of the corrugated wire 25; the second limb (and coil i 5) is A the pitch or interval of the corrugations to the right of a crest; and the third limb (and coil I5) is 3S the corrugation pitch or interval to the right of a crest: hence electrical impulses of frequency 2n are produced in the tone circuit 21, as required for the third partial. Stating the matter in general terms. it will be seen that the generating coils I5 of each overtone generator device Q3. etc. are spaced along the track P out of phase with one another, with respect to its variations, by I fractions of the corrugation pitchv or interval corresponding to the number of this partial (2nd. 3rd. etc.) ,-or, more properly, to its multiple of the fundamental frequency 1l. Obviously, any

number of such generator devices can be pron puls or alternating currents for all the part'all n, 2n, 3a, etc.,may bein phase with one another. 7- -Itwillalaobeobservedinll'lg.5thatwhilethe limbsofdeviceQrarespacedapartonlyhalft-ba pitchofthe corrugations of the wire 25. thoseof thedeviceQsarespacedadistancelI/Sthispitch.

fromasingletonetrackPisillustratedinFlg 5. wide spacing avoids crowding tbe u limbs together into the pitch of one corrugation, and may be especially desirable for the higher partials, of frequencies 411, 5u, 6n, 7n, 81u, 10u, etc. In practice, of course, the track P may preferably be embodied in a rotor R of any suit,- able type, such as those shown in Figs. 2, 3, 4, or even 6.

Rotors with plural tone-tracks; complex tones; plural manuals or pedals Fig. 6 illustrates a rotor R with a plurality of tone-tracks P of varied permeability concentric with its axis of rotation 3|. These tracks P may be of any suitable construction, several of which have already been explained in connection with Figs. 2, 3, and 4. Such a rotor R can advantageously be used to produce all the partials for one tone or note of an instrument-or even for two corresponding notes of different octaves, if the number of partials necessary to give a pleasing tone quality or timbre does not exceed the number of tracks P that can be satisfactorily accommodated at one side of a carrierrdisc of convenient size. The number of partials required for a good tone depends on both the character of the tone andits fundamental pitch. Four or'flve partials generally suflice for a fair diapason tone; but eight or nine may be used to assure variety or richness of tone, or even more. Violin tones require more partials than diapason tones of the same pitch. But in the higher notes of a keyboardof usual organ range, some two or three or four partlals suffice: i. e.,` the highest note of such a keyboard hasa fundamental frequency of 2093 vibrations per second, and its third partial (second overtone) has a frequency of 6279, which is about the limit of usefulness in music. y The rotor 'R shown in Fig. 6 has six concentric tracks P with one or more generator devices Q for each, so as to produce a complete tone, and may appropriately be distinguished as a full-tone rotor.

Inusing the corrugated hoop or spaced plug constructions of Figs. 2, 3, and 4 for a multi-track rotor R such as shown in Fig. 6, and employing one standard pitch or spacing for the corrugations or plugs of all the tracks P, the tracks P to produce the fundamental and higher partials of frequencies n, 2n, 311, 411, etc. may have radii 1', 2r 3T, 4r, etc., and those for a sub-partial of frequency 1/n may have the radius 1/212 However, it may sometimes be found convenient to make the pitch or spacing of the corrugations or plugs greater for the lower partials of a fulltone-rotor than for the rest, or less for the higher partials, so as to avoid havinglto makethe rotor R excessively large. Accordingly, no attempt has been made in Fig. 6 to indicate the specific construction of the tracks P, or even 'to proportion the radii of these tracks on any consistent basis of corrugation pitch or plug spacing. But an important point illustrated by Fig. 6 is that when the tracks P for all partials of a tone or note are on one rotor (or rotary unit) R, the frequencies of the partials` will always be in the exact ratios fixed by the construction of these tracks, under all circumstances.

Another important advantage of a'multi-track rotor R as shown in Fig. 6 is the compactness of space in which many partials can be generated, as contrasted with forms of rotor in which tracks P for the component partials of the tones are presented peripherally, rather than endwise or laterally. This compactness of many tracks P presented endwise or laterally is not, of course, dependent on the partials of each tone of an instrument being produced by tracks P on one and the same rotor R, but obtains even when the tracks of a rotor are not thus interrelated.

Fig. 6 shows two sets of generator devices Q for the one rotor R, with their secondary coils I6 connected to separate tone circuits 21 which are in turn connected to separate branches of the main output circuit Il that includes amplifier X and sound device S. A variable transformer coupling 290 is interposed in each branch of circuit Il, so that the loudness of the tones from the twosets of devices Q can be varied independently. While the left-hand set of generator devices Q are like those shown in Fig. 1, the right-hand devices Q are different, in that each of them has a pluralityy of primary exciting coils 20, connected to separate circuits 28, 28, which are also separate from the circuit 28 for the left-hand set. In an instrument with a plurality of setsA of playing means or manuals (Whether keyboards or playing pedals), the singly-excited set of devices Q may serve one keyboard manual, while the plurally excited set may serve other manuals,-'e. g., a second keyboard and a pedal. Generally speaking, I use the word manual throughout this specication and my claims in a broad sense, as including tone-control devices of either the hand-played or key. board type or the foot-operated or pedal type. The excitation of the several sets of devices Q through the several circuits 28 may be the same or different, as desired by the player from time to time.

Vibrato eects tained by any suitabley support 36| near the edge of the rotor disc R. By oscillating the arm 36 and its set of generator devices Q around the axis of the shaft 3i, through a convenient angle and at a suitable frequency, a vibra sound effect is produced at the sound device S-sixnilar to that which a violinist obtains by rocking his finger on a. string while bowing it. This is because more crests, plugs, or the like of the tracks P pass the generator devices Q per vsecond when the latter are moving counter to the rotating rotor R, and fewer when the devices Q are moving with the rotor R,as compared with when the devices Q remain stationary. The arm 36 may be oscillatedv by any suitable mechanical vibrator 363, represented in Fig. 5 by a revolving crank with a pitman connection to the outer end of the arm. The vibrator 363 may be thrown into andvout of action as desired by any suitable means (not` shown), such 'as a stop or the like. To simulate the violin vibrato, the arm 36 should make about 8 (complete) vibrations per second,

6 arancia Generator devicesvwith plural coils Generator devices Qwith plural associated generating coils Il, such as that illustrated in Fig. "l, can be used to produce some of the tones 5 of an instrument without corresponding rotors R, by borrowing and 'combining partials from rotors primarily intended for other tones. Like `the generator devices Q in Fig. 6, and for the isame purpose, the generator device Q in Fig. 'l also has plural exciting coils Il. Ihese are shown on the middle ofthe core 24, where their eii'ect is greatest.

-' Organization o! an entire instrument To synthesize a tone of fundamental frequency n and of any desired quality or timbre from the requisite partials n, 2n. 3a, 4u, etes-including the sub-partial tn, if desired,it sumces so to combine these partials in the proper proportions of intensity. By varying the .number and proportions of the partials employed, any desired tone quality can be produced. While this can be done in other ways, such as herelnbeiore indiu cated, I at present prefer to proportion the partials by regulating the activating currents in the exciting coils 2l of the various generator devices Q. One of many possible ways of doing this is shown in Figs. 8 and 9. which illustrate the elec- ,o trical features of an instrument adaptable to play organ music, with the full range and variety of tones and effects available on an organ, besides some eifects of which ordinary types of organs are not capable. Such an instrument may have u any desired number of keyboards or playing manuals (orpedalsh Of course the several manuals or keyboards may be operated separately. orv

coupled together so that playing on one also plays one or more of the others, etc., as is customary 40 in electrical actions of` ilne organs of ordinary weil-known types. I have also devised a verysimple way of doing this, which is illustrated in Pig.

9. A large instrument may preferably have both the keyboard or manual range of ve octaves from C 85.4 vibrations per second to C 2093 and the 45 pedal octave from c 32.7 up to c 65.4 entirely provided for by full-tone rotors R, mahng a total of 73 rotors required to provide C notes at both top and bottom of the six-octave range. In a w smaller instrument. a pedal octave `may be provided by utiiizing the tn tracks P (and others to afford the desired partials) from one of the manuals M. with corresponding reduction in the number of rotors required.

es Borrowing of partials However, such borrowing" of-tones can readily be carried much further. For instance, a fulltone rotor R for a tone oi fundamental frequency n in the lowest octave may be provided with tracks P for frequencies n, 2n, 3u, In, tn. on, tn,

101i; and similar rotors may be provided for corresponding tones two and tour octaves higher. of fundamental frequencies tn and 18m-these three rotors being conveniently designated as u the n, 4u, and 16u-rotors. Instead, then, of

' providing a 2n-rotor R for the intermediate octave-tone of fundamental frequency 2n. the ilrst five component partials of this 2n-tone may be-provided by borrowing from the 2n, 4u, Gn, Bn,

-70 and 101i tracks P of the n-rotor. and its remaining three partials by borrowing from the 121|, 181i, antiv 201i tracks of the 4n-rotor, that produce the third, fourth, and fifth partials of the 4ntone. In like manner, thecomponent partials of 1| the 8ntone may be borrowed from the 4u and 16u-rotors: and the necessary partials for the 32a-tone can be borrowed from the 16a-rotor, since this tone is of very high pitch, and so does not require so many partials as the lower tones. When partials are thus borrowed, plurally-ex- 5 cited generator devices Q with plural generating coils il (such as shown in Fig. 7) are preferably used with the tracks? that are involved, as explained hereinafter in connection with Fig. 9:

i. e.. each generator device should have as many 10 coils of each type as there are partials to be taken from it. It is interesting to remark that in this way six full octaves can be had in an instrument lwith only thirty-six rotors R. l Buch borrowing of partials may be oi' course be 15 resorted to for the purpose of reducing the total number of tone-tracks P required in an instru; ment, without reducing the number of rotors R correspondingly. if at all. Por example, eight partials of relative frequencies n. 2n, 31|, 4u, tn, N Bn, 7n, and Bn can be provided for all tones (that require so many as eight partials) with 73 rotors having (except the highest rotor) no more than four tracks P apiece. corresponding to frequencies n, 311, 5u, 7n; or omitting the 'In-partial, three l. tracks P per rotor, corresponding to frequencies n, 3u, 5u, will provide for seven partials per tone. Making the corresponding hoops 2l of all rotors R or like diameters and corrugation-pitch throughout the instrument, partials of frequencies n, 3n, Bn may be taken from one rotor R; partials 2n, 6u, 101i from another revolving twice as fast; partials 4u, 121i, 20a from another revolving four times as fast; and so on to the rotor R for the top'octave,which, however. l* may require tone tracks P for 32u and 64u, as well as 96u and l60n,but will not need tracks for such uselessly high pitches as 1921i and 2561i. The borrowing of partials is illustrated in Fig. 9, and is further explained in connection therewith, s as well as the use of generator devices Q of the Fig. 7 type.

Borrowing of partials is also practicable when a number of partials are generated from the same track P as described in connection with Fig. 6.-as u will readily be understood from the foregoing explanation. In this case, again, plural generating and exciting coils il, Il are desirable for each generator-device-limb'or member carrying such coils in Fig. 5. se

General organisation electrically 9. only six auch rotors (as for corresponding notes of aix octavos) are shown. ln practice. of course. as many as the full 'i3 rotors represented inlog. l0maybeused,withasmany partlalsper jroturasarei'equireiltogiveagood-toneineacli instance. lnlilgaland'lnvestopsoneach manual are shown, with provision for ilve different proportional combinations oi' the three partials illustrated, giving ilve diiierent tone- 70 qualities for all the notes of each manual. In Practice. thismight suiilce for a small instrument: while on a large one, it might be desired tohave up to fifty stops or more to a manual. with provision for a corresponding variety of tone- 7s qualities in all the notes of each manual.' Provision may in like manner be made for any desired variety of tone-qualities in the pedal notes.

The organization of the instrument in Fig. 8 corresponds essentially with that in Fig. 1, with greater elaboration of the component parts to provide for two manuals or keyboards, for multiplicity of partials and variation of tone by variety of proportional syntheses of the partials, and for loud and soft and tremolo effects, etc. 'Ihere is a sound device S and amplifier X, connected into an output circuit i1 with a branch for each of the manuals M, M; and there is a power circuit 22 with separate branches 221, 22u for the two manuals. In Fig'. 8, a single three-track rotor R is shown in plan', with its generator devices Q; whereas Fig. 9 aiords an edge view of six rotors R with their devices Q,-this number suiicing to show the circuit connections entailed by multiplicity of rotors just as well as if a full complement for six octaves were represented. Although a single set of double-excited generator devices Q (such as shown in Fig. 6) could be used with each rotor R, yet to avoid complication of circuits in the drawings, two separate sets of (three) singlyexcited devices Q are shown,rone set for each of the manuals M, M. In general, the parts and circuits (aside from rotor R, sound device S, and amplifier X) are duplicated for the two manuals M, M in the upper and lower halves of Fig. 8; soA that for the most part, only the lower manual M need be referred to in the description.

For the lower manual M, there is a secondary playing or tone circuit 21 to which the secondary leads I6 of the corresponding generator devices Q are connected, as inFig; 1. There is also a primary playing circuit 28 connected to one side of the lower power circuit branch 221 and to one set of the primary leads 2| of the devices Q, as in Fig. 1. In these primary leads 2| are interposed variable or adjustable calibrating resistances 81, for adjusting and equalizing the exciting currents and the resulting partials; so that each partial of a tone or note may underl all circumstances be limited to a desirable maximum loudness. Such resistances 31 are the more useful because the greater linear velocity of the outer tracks P tends to result in greater intensity of the higher irequency electrical impulses and of,the overtones produced by them; whereas the lower frequency fundamental produced by an inner track P generally needs to exceed (or at least equal) in intensity the strongest overtone(s of a note. While this can be compensated by using lfewer turns (or ner wire) in the'coils I5, 2liI for the higher pitches when an instrument is built, the calibrating resistances 31 afford another means of compensation, that is available if it is subsequently desired to change tone quality, or to correct for changes in lead or circuit resistance due to repairs, etc. Still more important, such resistances 31 maybe useful in assuring a substantially constant standard of loudness for each tone or note of the keyboard, unaffected by concurrent operation of one, two, three, or more keys K at the same time, as explained hereinafter in connectlon with the general volume-control circuit V.

In the playing circuits 21 and 28 are the normally open (and selfopening)tone control or playing switches 26, which are closed by the playing key K when it is depressed, as in Fig. 1.v When rotr tracks P are at all close together, it may be desirable to make and break both the primary and the secondary playing circuits 21 and 28: the primary in order to prevent faint "cross-partials due to slight inductive influence of one rotor track P on devices Q belonging to neighboring tracks, and the secondary to prevent cross-effects due to faint prolongation of a note once sounded,

as a result of residual magnetism in the cores of the devices Q which produced such note.

One means of coupling manuals M, M together so that playing of one will play one (or more) of the others is to provide shunting branch or cross playing circuits 21, 28 from the subordinate manual(s) to the dominant manual(s) with corresponding extra switches 26 operable by the various keys K of the dominant manual(s) In Fig. 8, such shunt branches 21, 28 are shown from the upper manual M to the lower manual M, with extra switches 28 for these branches under the lower key K. Make and break switches 38, :i8 are shownl in these shunt circuits 21, '28, interconnected for operation together by a cam-arm 39 on a shaft N that extends throughout the instrument for actuating corresponding switches 38, 38 in branches 21, 28 of all the upper manual playing circuits. By means such as a stop, tablet, or handle il) associated with. the lower manual M, the shaft N can be operated to couple or uncouple the manuals M, M at will.

The playing circuits 21, 28 for all the rotors R in the instrument are connected to the output circuit I1 and to the power circuit branch 221 through trunk Wires or bus-bars H and I (corresponding to the points h, h and i in Fig. 1) that extend throughout the instrument for this purpose, as will readily be understood from comparison of Figs. 8 and 9. As for the primary leads 2| of the devices Q which are not connected to the primary playing circuit 28 as already described, those for corresponding Itracks P on all the rotors in the instrument are connected to the same one of three trunk Wires or bus-bars J (corresponding to the point y in Fig. 1) that extend throughout the instrument, just like the trunk wires or bus-bars H and I.

Variation of tone quality and of volume In the connections between the trunk-wires or bus-bars J and the corresponding side of the power circuit branch 221, means may conveniently be provided for varying the relative strengths of the exciting currents supplied to the several busbars, and through them to the primary leads 2| of the generator devices Q associated with corresponding tracks P of the various rotors R of the instrument. Thus the relative strengths of the secondary currents generated by these devices Q in coaction with the rotors R will be the same for all the rotors, and the relative intensities of component partials and the resulting tone-qualities will be the same throughout all vthe notes of the (lower) manual M: i. e., all the notes produced -by a given stop of the manual will be diapason,

violin, oboe, etc., as inan ordinary ne organ. Provision may also be made inthe connection between the bus-bars J and the power circuit branch 221 for varying or modulating the aggregate of the exciting currents supplied to the several bus-bars, thus varying or modulating the loudness of all the notes of the manual alike, to produce a soft or loud effect or a tremolo eiect in all, etc.

For the last-mentioned purposes, variable resistance L and T, shown as in series with one another, may be interposed in the connections between power-circuit branch 221 and the several bus-bars J. The resistance L may be adjusted by a loud and soft pedal'for stop,- etc. (not shown). The tremolor'esistanceTmaybeoscillatively varied by any suitable mechanical vibratort(representedin1"igs.8and9byacrankand pitman), which vibrator may be controlled and started or stopped by a tremolo pedal or stop, etc. (not shown). l

For simplicity. the illustration in Fig. 8 includes provision for arly four degrees of loudness of each ofthethreepartialsproducedbythethreetracks P shown on rotor R, and for ilve diiferent denite proportional combinations of these partiels (for each manual M) under the control of five stops Il, l2, 4I. M, Il. thoughobviously any desired number of partials, degrees of loudness. proportional combinations. and stops' can be provided by adding more parts like those shown. Another stop I. is also shown, which will be explained hereinafter. Preferably. the stops ".12, Il, u. Il, Il may be interlocked (as combination pistons of organs often are) to prevent more than one from being operated at the same time. To produce the different strengths of the partials, resistances in the several connections from power circuit branch 221 to the corresponding bus-bars J may be varied, preferably in denite. reproducible steps or gradations. A convenient way is to provide a suitable grouping oi' resistances for each distinct proportional combination of partials, throwing these groups in and out of circuit by the stops 4i, 42, 4I, 44, 4I. As shown in Fig. 8, a general volume-control circuit V is connected across the power circuit 22. and includes the loudness and tremolo resistances L and T and a partial-proportioning resistance-connection W.

all in series with one another. Ihe total re` sistance of the control circuit V should preferably be low as compared with the total resistances between bus-bars I, J (including Duns 2l, Il, 2|, t1, 2t, and 2i) for the devices Q pertaining to each rotor R, in order that the activating current in the coils 20 and the loudness of each note v may be substantially constant regardless of how many keys K aredepressed at any time. The calibrating resistances l1 are helpful in taking care of this. lines or conductors Il are connected to the bus-bars J, and lines or conductors Il. Il, Il, Il are connected to points of the connection W, with suitable amounts of its resistance intervening between them. In the present instance. these amounts are indicated as equal.

Corresponding to the stop tl, there is a 'set of three cross-lines or conductors Il. each connected between a corresponding one of the lines Il and some one of the lines Il, I1, Il, Il. and including a make and break switch Y; corresponding tothe stop ll, there is a similar set of crosslines l! with stop switches Y: and so on for the other stops Il, Il, ,'It Asindicated in Fig. 8', the three switches Y for each set of cross-lines are interconnected, and are operated by the cor- A responding stop Il, 42, I3, Il, II, or II. as the case may be. Operation of each stop Il, 42, u, u. 4l bringsv a different combination of resistances in circuit-.with the bus-bars J pertaining to the'three rotor-tracks P, and so combines the corresponding'partials in diiferent relative proportions, giving a tone of different quality.

l'br example, operation of stop 4i to close its three switches Y gives maximum strength for the partial produced by the inner track P, which maybethefundamentaulessstrengthfor that produced by the intermediate track P, say the second partial; and run 1ere strength for the' other or third partial. Qperation ofstop l2 gives a weaker second'partialthan before, anda third partial of minimum strength: stop II gives second and third partiels both of minimum strength; stop u gives a fundamental lust below maximum, and second and third partiels of like and somewhat less strength: stop Il gives g maximum strength of fundamental and third partial,-and a second partial somewhat weaker.

`While the deiinite stops 4i. 42. 4I. u, 4l are comparable to those of an ordinary organ. Pig. 8 shows the stop as an elective" stop. with provisions enabling theplayer to make at will 1g any desired synthesis of partials, in such numberandrelativestrengthsashemaychoose, determining the quality of the tones of (lower) manual M however he likes. l'br purpose. the three cross-lines or conductors controlled by eachstop-switch Y operated by stopllmayhavemeansforvarylngthe tive'strengths of the currents through One way of doing this is to provide for the three cross-lines" (four) branches (lli 2, l respectively) each connected to o the lines or conductors II. l1. Il. Il, with a selector switch il for each cross-line which the player can operate to connect that cross to any desired one ofits branches. A dead I point l is shown for each switch y, on which `itcanbeplacedwhenitisdesiredtoomitthe corresponding partial. To avoid complicating the drawing, Fig. 8 does not show any stop 4l. etc.. for the upper inanua'l M. il

Rotaryunitsanddve `xanasoumidevieesster thelow andhish" notes, differently adiusted or constructed to produce sounds of widely diiferent pitch with equal correctness. Accordinglyl separate bus-bars cr trunk-wires H are4 shown for the rotors R of the two sections of the unit U. connected to the sep- 00 arate output circuits I1 for the low and high notes.

While rotary units U each comprises rotors R for producing corresponding notes of different octaves are generally to be preferred, there is l some advantage in constituting the rotary units on .a wholly different' principle: i. e.. so as to 4include rotors R. that produce tones whose fundamentals are' of frequencies rather .close to getherf-succesaive notes in the scale. for in- 70 stance. when .uur principle u followed.` the adjacent rotors R can be more easily. adapted (by diirerences in number of corrugations or el -members in their tracks P) to produce currents of the proper relative frequencies when revolvls ing at the same speed; because the, differences of frequency amongst the rotors in one unit U are so much less. In this case, the ratio of the gears 61, 68 may be 1:2 or 1:4, or the like, rather than 1:8; also, the number of rotors R to be included in a unit U is more a matter of arbitrary convenience in design, and may be either more or less than the six shown.

The rotary unit U may be driven from a drive Vshaft 10 through another (parallel) shaft 1| geared to the (upper) shaft 3| through toothed (bevel) gears 12, 13, and also geared to shaft 10 by toothed bearing. For this purpose, there is shown a coupling gear device or idler 15 meshing with gears 16, 11 on the shafts 10 and 1|. The purpose of interconnecting the shafts 16 and 3| in this manner (instead of directly, by

, the same circuit 8| and supplying direct current of convenient moderate voltage (such as 6 v.) to the powercircuit 22 of the instrument.

Aside from those just described, the parts and circuits shown in Fig. 9 correspond generally to those shown in Flg. 8 for thelower manual M, although they,are differently arranged; and the stops 43, 44, 45, 46 and their cross-lines, etc., are omitted to avoid crowding and obscurity in the drawings.4 For the same reason, the circuits for the two lowermost rotors R are omitted, since they would merely` duplicate circuits shown. Some modications of the arrangements illustrated in Fig.- 8 are shown in Fig. 9, and will now be explained.

Special tone effects Certain musical tones (e. g., on the piano) are characterized by delayed initiation of one or as hereinbefore mentioned, so as to enrich the ment, such tones are readily provided for by having separate key-switch(es) 26 for the de- -layed partials, and arranging to close them in due sequence after the other switches 26. In Fig. 9, this is illustrated in connection with the second rotor R from the top, Where the leads i6 oi' the innermost generator device Q (which may be that for the fundamental n) are not connected to the secondary playing circuit I8 for the other two devices Q, but run directly to an additional, separate switch 26. This extra switch 26 is arranged to be closed by key Ksirnultaneously with the switch 26 of the primary playing circuit 28; but the switch 26 for the secondary playing circuit 2.1 is arranged to close a little later,as by placing it at a somewhat lower level, as shown in Fig. 9. As indicative of suitable provision for this, the stationary contacts of the switches t that close iirst are shown as resiliently yielding. Thus the partiais from the outermost and intermediate devices Q (e. g., the iirst and second overtones) begin to sound after the fundamental.

' Partial borrowing circuits Fig. 9 also illustrates the borrowing of partials,

tone of the third rotor R from the top by borrowing from the fourth rotor. As shown, these two rotors R have essentially the same circuits as the third rotor are also just like those of the top rotor. The two innermost devices Q of the fourth rotor, however, are like those shown in Fig. 7, each having two generating coils I5, and also two exciting coils 20.' As shown, the upper coils l5 and 28 of each such device Q are the ones involved in the borrowing. While one side of each of the upper coils 20 for the fourth-rotor devices Q may be connected to the proper busbar J by the same lead-conductor 2| as its companion lower coil 20, the other side of each such upper coil 20 is connected by a separate lead 2| through a calibrating resistance 31 to a part of the primary playing circuit 28 associated with the third rotor R. Similarly, each upper coil I5 oi a fourthrotor generating device Q has its leads I6 connected'to an extension of the secondary playing l circuit 21 associated with the third rotor R. Thus these two inner fourth-rotor devices Q furnish two additional partials for combination with those furnished by the devices Q of the third rotor itself. Assuming a 2:1 ratio for the gears 61, 68, the tracks P of the third rotor R may give frequencies n, Sn, 5u, and the tracks P of the fourth rotor R whence partials are borrowed as just described may give frequencies 2n and Sn. Accordingly, the third-rotor tone will have the ve component partials, n, 21L,-3n, 511, 611., with an obviously richer quality than its own three tracks P could alone afford.

Mechanical organization of an entire instrument Figs. 10, ll and 12 illustrate in a diagrammatic way the general organization and one form of mechanism for an entire six octave instrument adapted to be changed over or converted for playing in different scales and keys of musical intonation. The electrical features of Figs. 8 and 9 are omitted, both to avoid confusion and because the mode ofuconversion illustrated is equally applicable to any type oi instrumentin which the pitch of the tones depends on speed of rotation,- whether the rotors generate electrical impulses or currents, or control light,or what-not. However, rotary units and rotors R similar to those of Figs. 8 and 9 are occasionally referred to in the description, for the sake of specifi explanation of variant points.

Especially in electrical instruments of the general type hereinbefore described, I prefer to group together and interconnect as one rotary unit U all the rotors R for producing corresponding tones in the several actaves covered by the intrument: e. g., all the C rotors R to form an octave unit U, all the C Sharps to form another such unit, etc. This arrangement not only assures correctly related frequencies for 'the notes an octave apart under all circumstances, but also obviates all possibility of discords from cross-induction between generating devices Q belonging to different rotors R; because all the frequencies that can be generated by cross-induction Within such a unit U will be in harmony with one another, as well as with all the .irequencies that `can be generated by each rotor ift and its own devices Q, including all the partials. To prevent the possibility oi `discords from crossinduction between diiierent units U, they should be magnetically shielded from one another, by spacing them well apart or by other suitable means,-indicated in Fig. 10 by septa 84 (as of soft steel or iron) between some units U. To

top rotor, and all the generator devices Q of the cover six octaves, Fig. 10 shows twelve octave umts U, one (for C) with seven full-tone rotors R, and the rest with six such rotors. However, the number of rotors R. required can be reduced by borrowing partials, as already explained. Or by producing the component partials of each note from a single tone-track P as explained in connection with Fig. 5, and providing on one rotor R (such as shown in Fig. 6) a number of tonetracks P adapted to produce fundamentals an octave apart (n, 2n, 4u, 8u, 161i, 321i, and 64u, if

needed), the number of rotors R required can be very greatly reduced: e. g., as an extreme, eleven rotary units" each consisting of one rotor R with six tracks P each. and one rotary unit consisting of a rotor R with seven tracks P, would suillce for a whole instrument. The units U may all be driven from one main or master shaft 13, preferably through (parallel, axially aligned) intermediate drive shafts 1l, one for each unit U, with suitable gearing.

Several types of rotary unit U are illustrated in Fig. 10: e. g., in the unit U at the extreme left, marked as for C tones, a train of seven sections with individual rotor shafts 3| geared together by gears 61, 63 are driven from a shaft 1| through a bevel gear 12 thereon meshing with a bevel gear 13 on one of the shafts 3l;y the next adjacent unit U, marked as for C sharp D flat, has all its six rotors R. on a single shaft 3l. driven from a shaft li by gears 12, 13; the third unit U, marked as for D, comprises two three-rotor sections with their shafts 3| geared together by gears 61, 63, and driven from a shaft 1| by gears 12, 13; the fourth unit U, marked as for D sharp E flat, comprises two trains of three rotor sections with individual shafts 3| intergeared by gears 61, 63, and separately driven from one shaft 1| through gears 12, 13: and each of the remaining units U comprises two three-rotor sections with their shafts 3| separately geared to one shaft 1|. vIn the last two types of unit U, the shaft 1| (with gears 12, 13) interconnects two shafts 3| and is lthus intrinsically part of the unit. and if desired, the shaft 1l may fairly be regarded as a part of each unit U of the other types.

Changes of scale and keu In order to change the relative speeds of vari-y ous tone-producing rotary units U and the musical intonation of the instrument, variable-speed driving means are provided. to drive various rotary umts at various different speeds for each of such units, whereby the notes of musical scales of the various desired intonations are evoked from the units U. as well as speed-selective and correlating means coacting with the driving means to adjust the relative speeds of the units U to various diil'erent concurrent-speed -correlations, severally corresponding to different desired musical scale intonations. With a constant-speed motor 33 to operate the units U, variable-speed driving is accomplished by means of interposed gearings which provide a plurality of driving ratios for each of such units U. and the speedselective and correlating means adjust the several gearings for concurrent operation in various 'suitable correlations of driving ratios. To adapt the particular mechanism illustrated in Fig. 10 for the purpose of playing in different scales and in different keys, provision is made for several possible driving-gear ratios-between the main shaft 13 and the intermediate shaft 1| pertaining to each octave-unit U, in order that each unit U may have a proper rate of rotation. for every scale f the corresponding coupler 1l.

and every key that is to be available. As one means of providing for this, Fig. 10 shows a group of (three or four) gears 11 on each shaft 1|. and corresponding groups of gears 13 onthe shaft 13. As a means of bringing various pairs 'of gears 13, 11 into actual coaction as desired, a coupling-gear device or idler 15 may be provided for each'gearpair 13, 11. This coupler 15 may be rotatably mounted on a shiftable carrier or coupling arm 35, which is shown pivoted at 33 about a (fixed) axis parallel with that of gear 15, and with shafts 13 and 1|. Preferably. the carrier arm 33 is yieldingly biased to carry coupler 15 (upward) out of mesh with the corresponding gears 13, 11, as by a helical compression spring 31. For convenience. the (three or four) gear pairs 13, 11 (and couplers 15) pertaining to each shaft 1| (and its unit U) may be collectively termed a driving-gear set" and designated Z. These sets Z may be individually distinguished by the tone-symbol C, C sharp, D, etc., by whichA each unit U is distinguished in Fig. 10. As one means of throwing into mesh various different couplers 15 in the several gear sets Z, a number of rotatable rods or shafts 33, intended to extend lengthwise of the series of coupling arms 35, are shown in Figs. 1l and 12. In practice, the coupling arms 35 may be over (or under) the corresponding gear-pairs 13, 11, and the rods 33 may overlie (or underlie) the series of coupler arms 35, as indicated in Fig. 11; but to avoid confusion, the coupling arms 33 are omitted from Figs. 10 and 12, and the rods 33 are shown by themselves in Fig. l2. Each rod or shaft 33 is provided with a number of cam lugs or actuator projections 33. one for a coupling arm 35 of every gear set Z: these are shown in Fis. 12 aligned with the corresponding gear pairs 16, 11 of Fig. 10. As shown in Figs. 11 and l2. there are twelve shafts 33, each with twelve actuators 33, yand each provided with an operating stop"crank or handle 33, conveniently located in reach of the player when seated at the keyboards of the instrument.

To guard against damage to the mechanism if a player attempts to set the instrument in different keys at the same time. any suitable interlocking arrangement may be used. In Fig. 13, this is represented by a longitudinally slidable interlock bar 32 extending under all the crankhandles 33 of the rods 33 (shown in Figs. 11 and 12) and having one single notch 33 Vwhich the player can place under any handle. by sliding the bar one way or the other by its own handle 34. On the underside of each crank handle 33 is a projection or lug 33 which ordinarily engages against the top surface of the bar 32 as shown at the left of Fig. 13, thus locking the rod 33 with its actuating arms 33 in the inactive position (shown in Fig. 11 for most of these rods). so that it is impossible to operate the rod 33 to throw in notch 33 is brought under a particular crankhandle 33, as shown at the extreme right of Pig. 12, then by pressing down on this handle 33 the player depresses its lug 33 into notch 33 and turns the rod 33 and its actuating arms 33 clockwise to the active position shown at the extreme right of Fig. 11. When he then releases the bar-handle 34 while holding handle 33 down, a long tension spring 33 pulls bar 32 to the right sumeiently to lock a shoulder on lug 33 under shoulder formed by undercutting one side notch 33, as shown. By the springs 31 acting coupling arms 33, the shafts 33 are all biased turn eounterclockwise; so that when the operasgg But when the Cil tor subsequently pulls bar 82 a little to the left, they automatically return to inactive position.

In Fig. 13 only two of the shafts 88 and crank handles 90 of Fig. 12 are shown, to avoid confusion.

The number of gear sets Z and rods 88 in Fig. 10 is sufficient for laying in the natural scale in the major keys of C, G, D, A, E, B, F, B at, E fiat, A iiat, and D flat, as well as in the equitempered scale; the gear ratios (and tooth-numbers) marked on the various gears 16, 11 in Fig.

10 are adapted for this in an instrument cor-l responding to the illustrative particulars of design now presently to be set forth; and in Fig. 12 eachrod 88 is marked vwith the scale or key into which the instrument will be thrown by operating that rod to throw in the corresponding couplers 15. It ls also to be observed that in the light of the following explanation, other gear sets Z (with proper ratios and teething) and shafts 88 can readily be worked out and substituted or added, so as to permit playing in a great many more keys of the natural scale, or even in scales characterized by very different frequency ratios,such as the old Greek scale, for example. As Fig. 10 is diagrammatic, no attempt is there made to -show the various gears 16, 11, 12, 13, 61, 58 in their true relative diameters. vln practice, I now prefer to make all gears 1E ofeach set Z of nearly the same diameter, and likewise all its gears 11; and this is easily done by adopting suitably different gear-tooth pitches. But of course the diameters of coacting gears 16, 11 must be in the proportions of the numbers marked in Fig. 10.

Illustrative practical gear ratios For purposes of simple illustration, it is convenient to assume an instrument composed entirely of units U of the type illustrated at the extreme left of Pig. 10; to take -such ratios for the gear- pairs 12, 13, and 61, 88, in these units U that the relative rates of rotation of rotors R in each unit are proportional to the numbers 1, 2, 4, 8, 16, 32 which express the frequency relations of tones an octave apart; and to suppose the fundamental tone-tracks P in all rotors R to be made of corrugated wires 25 (such as in Fig. 2) whose corrugations have exactly the same pitch. Taking a fundamental frequency of 440 vibrations per second for the note A, the frequency for middle C -below is 261.625 per sec. 'in the equitempered scale; 264 in the natural scales (keys) of C, F, B flat, and G; and 260.741 in the natural scales (keys) based on E flat, A flat, and D flat. Using "16 corrugations in the Wire 25 forthe 1unis marked (and toothed) 80:80 (Fig. 10) gives 80 16X16.5X\8-0=264 R. P. S.

yforthe middle-C rotor disc R,the correct frequency for C in the natural scales of C, G, F,

and B flat; o1" throwing into mesh the coupler 15 for the C unit gears marked 80:81 gives correct for the natural scales of E flat, A ilat, and D flat; or throwing in the C unit gears marked 110-.111 gives which is virtually the correct frequency of 261.625 for C in the equitempered scale. Similar calculations will demonstrate the substantial correctness of the other gear ratios (and tooth numbers) shown for the various gear pairs 16, 11 in Fig. 10. While the frequencies thus obtained for the equitempered scale are not mathematically exact (except for A 440, the starting' point above), yet the greatest departure for any equitempered scale pitch in the octave of middle C is but 0.052 vibration per second, and the average deviation is only 0.017. As the equitempered scale is in itself only approximately harmonious, these small deviations are unimportant.

For the natural scale, the gear ratios (and tooth numbers) marked in Fig. 10 give mathematically exact frequencies in most cases; but as We get further and further from the middle of the succession 'of fifths, D fiat, A ilat, E flat, B ilat, F, C, G, D, A, E, B, some of the gear ratios (and tooth numbers) in Fig. 10 give frequencies that are only approximately correct: e. g., the gear ratio 126:118 for C sharp gives a frequency of 281.898, as compared with the correct gure 281.918; and the gear ratio 118:112 for D flat gives a frequency 278.143, as against the correct ligure 278.128. However, the deviations from the true pitches are smalL-the average deviation being only 0.027 vibration per second,-and are practically inconsequential. The gear ratios indicated in Fig. 10 are here recapitulated: For C,

for C sharp,

1.0 5 E@ 118 2 6 10o 1,28 112 118 for D,

99 2Q 89 so 81 for D sharp,

lik 2li s4 11s s1 for E,

for F,

for F sharp 1&5! E f'f 132 64 59 84 for G,

for G sharp,

for A,

no si a 90'4848 for A sharp,

for B,

In these ratios, the upper numbers are suitable for the teeth of the gears 16, and the lower numbers for the teeth of the gears 11.

The reason for deviations from exact frequencies in the natural scale is the practical necessity of avoiding the use of gears with the excessively l large numbers of teeth corresponding to the large whole numbers that are required to express accurately the ratios of certain natural scale frequencies. With the extremely simple gearing shown in Fig. 10, this compels resort to approximate ratios, realizable by gear pairs. 16, 11 with moderate numbers of teeth. However, I have devised several simple means of avoiding appreciable inexactitude in such cases.

More accurate gear ratios.' Compound idlers Analysis of natural scale frequency ratios shows that the more complex ones contain the ratio 81:80 or its reciprocal 80:81. For example, the frequency required for C as stated above is 264 in the natural scales of C, G, F, B at. and E fiat, and is 260.741 in the natural scales based on E nat, A fiat, and D nat; and

264 :8 1 260.741 80 Again, `'the frequency for F sharp in the natural scales of G, D, and A is 371.25, and is 375.891 in the natural scales based on E and B; and

One way of introducing the factors 81780 and 80/81 in the ratios between gears 16 and 11 is to replace the simpler coupling idlers 16 in Fig. 10 with coupling gear devices |00 such as shown in Fig. 14 wherever a factor of 81/80 or 80/81 can aid in attaining (substantially) exact ratios -with practicable numbers of teeth on the coacting gears 16 and 11. As shown in Fig. 14, the coupler |00 consists of a compound idler" or gear, with provision for shifting it to one side or the other (in the direction of its axis) relative to the gears 16, 11 with which it is intended to coact. Such a side shift of coupler gear |00 can be very easily effected by shifting its carrier arm 06 along its axis or pivot stud 06, as by cam device(s) |0| on the corresponding `rod-(s) 68, (each) embodying a helical groove |02 in which the arm 05 is always engaged, with a high spot |00 at each end of this groove to take the place of an actuator 00 of Figs. 11 and 12 in depressing the arm 86 when the rod 88 is turned either way.. The gears 16, 11 with which the compound gear |00 is to coact are offset out of line, sidewise, just a little more than their thickness, as shown in Fig. 14. On the periphery of gear |00 there is a mid-zone or band |06 of 80 teeth wide enough to coact with the oil'set gears 16, 11 concurrently, and flanked to either side by (narrower) zones or bands |06, |01 of. 81' teeth each.

supposing, for purposes of explanation, that.

pair in the F sharp set Z at the middle of Fig. 10, it is apparent that when the mid-zone |05 of gear |00 is thrown into mesh with both of them, as in Figs. 14 and 15, this gear |00 functions as a simple idler (just like the idler 16 for these same gears in Fig. 10), and a frequency.

is obtained-for F sharp. which is correct in the natural scales of G. D, and A. In the natural scales based on E and B, however, the proper frequency for F sharp is 375.891, requiring a gear ratio of 729:512. It is impracticable to have gears with so many teeth; but

If, then. the gear |00 is shifted to the right from its position in Fig. 14 before being thrown into mesh, so that while its -tooth mid-zone |05 still engages the 64-tooth driven gear 11, its 81-tooth zone |06 at the left engages the 90- tooth driving gear 16. then there is obtained a frequency the proper value. It is interesting to observe that the gear ratio shown for this F sharp in Fig. l0 is 84:59, giving a frequency of an approximation only .027 vibration per second too low. It may also be remarked that if the gear |00 is shifted to the left from its position in Fig. 14 before being thrown into mesh, so that while its 80-tooth mid-zone |05 engages the 90-tooth drivinglgear 16, its til-tooth zone |01 at the right engages the (i4-tooth driven .gear 11, then a frequency such as 111:110 and noz-111, or 'zomzoiland 2011208, for example.

More accurate year ratios: Plural drive shafts Another way of resolving complex frequency ratlos into simpler gear ratlos is to employ a plurality of main driving shafts revolving at suitably different speeds. As shown in Fig. 16, a driving shaft 10 corresponding to that in Figs. 10. 11, 12 is supplemented with another (parallel) shaft ||0 connected to shaft 10 through suitable gearing. here represented by a chain running around sprockets ||2, ||0 of suitable ratio and tooth numbers. such as 81:80, that are fast on shafts 10, ||0. It is convenient to arrange these sha-tts 10. ||0 at opposite sides of (above and below) the (series of) aligned shaft(s) 1| (corresponding to those in Fig. 10), so that their corresponding gears 16, 16 are at opposite sides of the gears 11 with which they are to coact. A pair of coupling gear devices or idlers 'I5 is provided for each pair of the gears 16 on the shafts 10, H0: in Fig. 16, they are mounted on opposite arms of a unitary rocking carrier H5, which is pivoted (about at its middie) on a fixed axis 86 parallel with the shafts 10, 1I, H0. The carrier H5 is preferably biased to assume the mid-position here shown (in which neither of its idlers 15 is in mesh), as by a springpressed guide-rod-head Ill acting on a suitable flat H8 of the carrier H5, located symmetrically opposite its axis H6, so that rocking of the carrier either way compresses the helical spring IIS. As shown, one group of the operating rods or shafts 88 (such as in Figs. vlland 12) are arranged adjacent one arm of the carrier H5, and another group adjacent its other arm; so that the actuator lugs 89 of the two groups of rods throw the opposite couplers 15 into mesh, one coupler to drive gear 11 from the gear 16 on shaft 10, and the other t drive gear 1l from the gear 16 on shaft III). For clearness, only six of the rods 88 are shown in Fig. 16.

This arrangement affords possibilities of many variations, since corresponding gears 16 on the shafts and H0 may have slightly different numbers of teeth in some cases, besides revolving at slightly different speeds. Moreover, compound couplers I00 such as shown in Fig. 14 may be used with this arrangement, wherever found advantageous.

Flexibility and use of the instrument: The nat- ,ural scale The musical advantages of an instrument that can be changed over for playing in different l scales and in different keys are-very great.

Cil

The equitempered scale in general use today is really but a makeshift or compromise, affording the great advantage of playing on one instrument (without retuning) music written in any key, or in a variety or keys, but at the price of importantsacrifices of harmony and beauty. Exceptingnotes just one or more octaves apart, all of the notes in this scale are out of harmony when sounded together, and some very much so. Hence the chords are generally imperfect, somewhat discordant, in fact. This is because the vibration frequencies of such equitempered notes do not bear to one another exactly the simple numerical relations required for perfect harmony.

On the natural scale, as it is termed, these simple numerical relations forv perfect harmony exvist between al1 notes intended to sound together;

but an instrument with only twelve keys to an octave would not have anywhere near all the notes required-for music written in the desired variety of keys. On the contrary, the playing range of any one instrument would be very limited, unless it were built with a great many more keys than twelve per octave. Such an instrument as this, however, would not. only be excessively complicated and expensive to build, but far too large and cumbersome to be actually played on.. Before theequitempered scale was tend to that scale. Y

,natural scale, all characteristic advantages of that scale are fully realized, including chords that are perfect and beautifully harmonious. An instrument with the simple organization illustrated in Fig. l0 can be set or changed over in a moment for the natural scale in any of the major keys of C, G, D, A, E, B, F, B flat, A flat, and D flat as fundamentals; and without becoming by any means complicated, such an instrument can be built with additional gears and other parts to permit playing in a great many more keys. The scale and key setting or change is made with little more trouble than the setting of an organ stop, or the movement of a handle; the player merely notes the key in which the composition to be played is written (or begins), and operates the proper rod or shaft 88 to set the instrument for-the natural scale in that key.

When a musical composition is partly in one key and partly in another, the excursion into another key may be transient: i. e., lasting only a few measures in which are introduced SharpsV or flats not called for in the signature of the kein-accidentals as they are called. In many cases, such transient excursions will be met by the instrument without resetting or changing over to another key. An inspection of Fig. l0 shows that 12 pitches are provided in each scale, While only 7 would appear to be needed: the extra flve are those required by the adjacent scales in the cycle of fifths, D flat, A flat, E fiat, B flat, F, C, G, D, A, E, B. Transitions into other keys are normally up or down in this cycle of fifths; so that the five extra tones are generally adequate; and the accidentals have the correct pitches, unless, of course, the excursion should go too far. If a transition is made from C upward to G, the required F sharp accidental is provided, and in perfect tune. If it is downward from C to F, the required accidental, B flat, is

available. However, if the instrument is set for the scale of C, the Re or second note of the scale of G will be found to be too -low by the ratio 81/80, and the La or fifth note of the scale of F will be found to be too `high by the same interval. In each case, six of the seven notes of the new scale will be in perfect tune, and one of the less important will be off by a small interval. rhe same holds for any transition from one key to another, up or down, in the seriesof fifths. lf the instrument is set in the key of C, transition up to the scale of D or down to the scale of B fiat will involve two notes slightly out of tune. For most compositions, however, the transitions are only to adjacent keys, and are of short durar tion; so that the slightly faulty pitch of one note of the scale would go unnoticed.

in long compositions, Whole sections may be in different keys, as shown by a change of signature; and in such cases, theinstrument can be changed over from one natural scale key to another between `such sections. Or if a composition involves extensive excursions into far removed and unrelated scales, not provided for by anyl natural scale setting available in a particular instrument, then the instrument need only be thrown into the commonly used equitempered scale, which was devised for just` such purposes. Of course the equitempered scale will commonly be used in playing with other instruments that can only be played in that scale. However, stringed'instruments of the violinclass can be played in the natural scale; and voices naturally Borrowing partiels in the natural scale system thrown into truly harmonious whole-number frequency-relations with one another, and the borrowed partials for each tone are (mostly) brought into similar relation to the proper fundamentals; whereas on the equitempered scale, this is not the case. On the equitempered scale, for example, for the fundamental frequency of C=261.625 vibrations per sec., the tones available to borrow as overtones and their frequencies are C=523.25, 6:78399, C=l0'46.50, E=13l8.51, and G =l567.98. Being octaves of the lower C. the two Cs thus used as second and fourth partials are exact multiples; but whereas the third partial should be 3x261.625=784.875, the borrowed G is 0.88 too low; whereas the fifth partial should be 5x26l.625=l308.125, the borrowed E is 10.385 too high; and whereas the sixth partial should be 1569.75, the borrowed G is 1.77 too low. With my system, on the contrary, when the instrument is in the natural scale, the borrowed third, fifth, and sixth partials are in exact harmony, as well as the second and fourth.

I claim as my invention, and desire to secure by Letters Patent of the United States:I

1. In a musical instrument, the combination of a plurality of rotary units collectively adapted to produce the desired notes of a plurality of octaves, and each provided with a plurality of gears; a drive shaft with gears for coacting with said gears of said rotary units to drive the units, severally, at a variety of alternative speeds; and selective means for bringing into current coaction various combinations of driving and driven gears such that the notes produced with the several resulting concurrent-speed correlations of th:l rotary units conform to diiferent musical sc es.

2. In a musical instrument, the combination of a plurality of rotary units collectively adapted to produce the desired notes of a plurality of octaves, and each provided with a plurality of gears; a drive shaft with gears for coasting with said gears of said rotary units to drive the units, severally, at a variety of alternative speeds; ocupling gear devices for meshing with and operatively interconnecting the pairs of coacting gears of said rotary units and said drive shaft: and selective means for bringing into concurrent mesh as aforesaid various combinations of coupling gear devices such that the notes produced with the several resulting concurrent-speed correlations of the rotary units conform to different musical scales.

8. In a musical instrument, the combination of a plurality of lrotary units collectively adapted to produce-the desired notes yofa plurality oi octaves, and each provided with a plurality of gears; drive shafts running at different speeds andhaving gears for coacting with said gears of said rotary units to drive the units, severally, at a variety of alternative/speeds: coupling gears for meshing with and operatively interconnecting the gears of said rotary units and those oi said drive shafts; and selective means for bringing into concurrent mesh as aforesaid various combinations of coupling gears such.; that -the notes produced with the several resulting concurrent-speed correlations of the rotary units oonform to diiferent musical scales.

4. In a musical instrument, the combination of a plurality of rotary units collectively adapted to produce the desired notes of a plurality of octaves, and each provided withv a plurality oi' gears: a drive shaft with gears for coacting with said gears of said rotary units to drive the units.

severally, at a variety of alternative speeds: coupling gear devices for meshing with and operatively interconnecting the pairs of coacting gears of said rotary umts and said drive shaft, the 10 gears of some of said coacting pairs being out of line with' one another, and the corresponding coupling gear device (s) havng portions with the same toothing for meshing with both or either of the gears of such pair and also portions with different toothing for meshing with one or the other of them; means for shifting the last-mentioned gear device(s) axially to bring the diverse portions into positions for variant meshing with the coacting gears that are out of line as aforesaid; and selective means for bringing into current mesh as aforesaid various combinations of the coupling gear devices such that the notes produced with the several resulting concurrentspeed correlations of the rotary units conform to different musical scales.

5. In a system for generating impulses adapted to produce a complex musical tone in a sound device, the combination of generator devices each having primary and secondary windings in a 3 common field; a circuit, with means for connecting it to said secondary windings. and for amplifying the electrical impulses which it receives from them; rotor means for producing in the elds of said generator devices permeability variations at frequencies corresponding to theV component partials of the tone to be produced; means for supplying to said generator device primaries exciting currents proportioned to the relative intensities of component partials in the complex tone to be produced: and means for modulating said exciting currents during playing while maintaining their relative values unchanged, so as to produce tremolo effects.

6. In a system for generating electrical impulses of musical frequency and for amplifying.

rent from them, and a variable resistance with means for oscillatively varying it so as to produce a tremolo variation of said sounds; and a plurality of parallel branchesl of different resistances connected in series with each of said lexciting windings, with stop-switches in said branches for determining the relative strengths of the exciting currents in the several exciting windings..

. 7. In a musical instrument of the character,

described, for generating electrical impulses of musical frequencies and transmitting them, amplified, to a sound device according to the notes ,to be produced. the combination with playin! circuits for connection to the sound device, each including' a generating winding, exciting circuits including exciting windings associated with said generating windings. for producing magnetic fields affecting them, common paramagnetic lcores for corresponding generating and exciting windings, and means for producing in said iields 'permeability .variations of frequencies corresponding to the notes to be produced by the instrument; `of normally open switches in said playing circuits; normally open switches in said exciting circuits; and a keyboard with playing keys each operative when struck to close the switches in corresponding playing and exciting circuits, whereby cross effects due to cross excitation and residual magnetism are prevented.

8.. In a generator for generating electrical impulses oi' musical frequency for a musical instrument, the combination with a generator device comprising a generating winding and means for producing a magnetic field affecting said winding, and a rotor comprising a paramagnetic ring extending around the axis of rotation of said rotor and travelling .in varying proximity to said generator device as the rotor revolves, so as to produce permeability variations in said magnetic field affecting said winding, of a magnetic inductor adjacent the path of travel of said ring, directly opposite said generator device with respect to said vaxis of rotation of the rotor, for inducfili and a paramagnetic horseshoe core therefor, with means for producing a magnetic field affecting said winding, said horseshoe core being mounted with both its ends closely adjacent the path of travel of said ring as the rotor revolves and spaced apart along said path a distance corresponding to the pitch of the ring serrations, so that when the rotor revolves, the ring completes the magnetic circuit of the core only periodically, as different ring crests concurrently pass the spaced core ends.

10. A musical instrument comprising, in combination, a plurality of rotary tone-produclng units; variable-speed driving means in the instrument for driving various rotary units thereof at various different speeds for each of such units, whereby the notes of musical scales'of various diiferent intonations are evoked from the units; and speed-selective and correlating means in the instrument coacting with said driving means to adjust the relative speeds of said rotary units to various diierent concurrent-speed correlations, severally corresponding to diierent musical scale intonations.

11. A musical instrument as set forth in claim 10 wherein said variable-speed driving means drive the various rotary units at speeds which evoke notes of the equitempered scale intonation, and also at speeds which evoke notes of natural harmonic scales of various different intonations; and said speed-selective and correlating means coact with said driving means to adjust the relative speeds of said rotary units in correspondence to the equitempered scale intonation, and also, e

Images