US3015979A - Electronic musical instrument - Google Patents

Description

Jan. 9, 1 962 M. DAVIS ELECTRONIC MUSICAL INSTRUMENT Filed Nov. 29, 1957 VOLUME CONT/m1.

50 c 5 46 4 6 E's'u JOFT 8 Sheets-Sheet 1 Fi .l

INVENTOR Jan. 9, 1962 M. DAVIS ELECTRONIC MUSICAL INSTRUMENT 8 Sheets-Sheet 3 Filed NOV. 29, 1957 INVENTOR Mer m mlsmm g oh Jan. 9, 1.962 M. DAVIS ELECTRONIC MUSICAL INSTRUMENT 8 Sheets-Sheet 4 Filed Nov. 29, 1957 IN VENTOR ATTORNEY L i J. .1... N *N N Jan. 9, 1962 M. DAVIS ELECTRONIC MUSICAL INSTRUMENT 8 Sheets-Sheet 5 Filed Nov. 29, 1957 INVENTOR v Mar/m [Jaws 8 Sheets-Sheet 6 Shh! M. DAVIS ELECTRONIC MUSICAL INSTRUMENT IN VEN TOR Merl/h Dav/'5 ATTJ 2E Y Jan. 9, 1962 Filed Nov. 29, 1957 Jan. 9, 1962 M. DAVIS ELECTRONIC MUSICAL INSTRUMENT 8 Sheets-Sheet 7 Filed NOV. 29, 1957 MN 0w m Q ou $86: mks w 1 Q 1 b u m m m n w mg g m m w m INVENTOR Mer/lh Dav/'5 BY ATTORNEY Jan. 9, 1962 M. DAVIS. ELECTRONIC MUSICAL INSTRUMENT 8 Sheets-Sheet 8 Filed Nov. 29, 1957 INVENTOR Mr/in Lbv/s xwkSbku wziubzw Mm Patented Jan. 9, 1962 3,015,979 ELECTRGNEC MUSJ UCAL RNSTRUMENT Meriin Davis, 5424 31st St. NW., Washington, D.C. Filed Nov. 29, 1957, Ser. No. 699,779 7 Claims. (Cl. 84-118) This invention relates to musical instruments or more particularly to an electrical musical instrument for the production of complex sounds and having a fixed keyboard.

Particular musical instruments having several keyboards and a large number of keys per octave have heretofore been experimentally capable of rendering chords substantially in the perfect interval relationship of the just tempered scale. Such instruments are obviously exceedingly complex and diificult of manipulation. Non-fretted instruments, such as the violin are capable of such rendition; however, only in the hands of the most talented musicians. The average musician is therefore denied the pleasure and satisfaction of performing with perfect temperment.

Accordingly, it is an object of the invention to provide a musical instrument with a fixed keyboard which requires a minimum of skill to play but which is capable of rendering chords in the perfect interval relationship of the just-tempered scale.

It is a further object of the invention to provide a musical instrument on which the performer is given a wide range of self expression with respect to musical parameters.

It is also an object of the invention to provide improved means for controlling these musical parameters.

It is also an object of the invention to provide an improved electronic fundamental and harmonic frequency generator.

Musical characteristics fall generally into two classes; those which are predetermined and fixed and those, in contrast, which are variable in nature. The first class comprises the pitches of the melody and harmonic accompaniment of a musical composition. These are fixed characteristics and are represented by the position on the staff of the notes on the written score. In the latter class, in general, are: (a) those qualities which represent the distinguishing features among various instruments, ([2) timing or rhythm, volume accent, and other less salient features.

These attributes of the latter class, except for rhythm, may be widely varied in form and nature without changing the basic structure of the composition or generating discordancy. They constitute the adornments and embellishments of the fixed basic musical structure. Rhythm, although essentially a basic part of a composition, may not be completely predetermined, if the performer is to be afforded the satisfaction of full personal expression.

The manipulation of these variable tonal parameters constitutes the essence of pleasure and satisfaction to be derived by the performer and lends character and distinction to the rendition.

Accordingly the instrument of this invention provides means whereby the first noted or fixed characteristics of a musical composition are automatically generated in response to arbitrary rhythmic manipulation of any one or.

more of a plurality of parameter control keys. The parameter control keys are arbitrarily selected and therefore the variable musical characteristics of the last noted class are determined at will by the performer.

The invention also provides, in combination with the above noted features, improved parameter control devices operating in conjunction with an improved cathode-ray tube harmonic generator. Possible usages in other than the musical field may be mentioned; (a) harmonic generator, (b) function identification, (c) transient function conversion, (d) tri-coordinate keying, (e) forward-reverse saw-tooth generator. The invention, however, will best be understood from the following description with references to the accompanying drawings in which:

FIG. 1 is a schematic view illustrating the general principles involved in the operation of the musical instrument.

FIG. 2 is a block diagram expanding the principles illustrated in FIG. 1.

FIG. 3 is a schematic view showing in more detail the elements of the block diagram of FIG. 2.

FIG. 4 is a chart illustrating the inequalities between equal and perfect tempered scales, and including an arrangement of the pitch compass of the instrument into overlapping groups for simplification in selection mechanisms.

FIG. 5 is a schematic showing a potentiometer circuit for primary pitch activation, an accompanying pitch commutation system, and a program register tape with details of construction and fabrication.

FIG. 5a is an enlarged fragmentary perspective view showing structural details of parts of FIG. 5.

FIG. 6 illustrates schematically the details of a harmonic generating system responsive to fundamental fre-. quencies and means for control of the harmonics by preselected mixing and attenuation by manual selection.

FIG. 6a is a fragmentary sketch showing an alternative harmonic attenuation method.

FIG. 6b is a detail of harmonic attenuating light filters and their density chart.

FIG. 7 is a schematic view of an alternative secondary emission set of harmonic patterns employing push-pull operation.

FIG. 8 is a schematic view of curve shaping means for an attack and decay envelope of an automatic attack and decay system which employs manual selection of these parameters.

FIG. 9 schematically illustrates the shape of the electrical signals resulting from depression and release of the push buttons of the manual selection means of the attack and decay system.

FIG. 10 schematically illustrates the action of variable density light patterns of the attack and decay system used for shaping the electrical signals which limit the crest height of the sound waves during the advent and terminaa tion of a sounded note or notes.

FIG. 11 is a circuit employing transistors for the purpose of producing a signal with an adjustable linear rise time to a limited manually held stabilized condition, followed by an adjustable linear decline, resulting from the closure and subsequent reopening of a switch in this circuit.

FIG. 12 illustrates attack and decay circuitry designed to reshape a linear signal function to any desired curvature for ultimate control of volume.

GENERAL PRINCIPLES OF OPERATION Reference is now made to FIG. 1 which illustrates generally the combination of basic components of the system for semi-automatic selection of the fundamental pitches of a musical composition, and arbitrary selection and control of the numerous parameters of the composition.

The system comprises a tone generating component, generally indicated by the reference numeral 1, a perforated tape 2, which is punched in accordance with the musical score pitches of a composition, and a series of contact brushes 3 adapted to sense the perforations and complete the tone generator circuits, respectively.

Tape 2 is advanced in response to partial depression of either one of a pair of spring retracted keys or buttons 4. Upon partial depression of a key 4, a ratchet mechanism is operated. This steps the tape a distance equal to the linear spacing between the holes of the tape. Consequently, one or more of the brushes 3, depending on the transverse location of the hole or holes of the tape, will complete a respective tone generator circuit. Upon further depression of the key, an associated switch 6 will be closed, thereby controlling the volume of sound, loud or soft, emanating from loudspeaker 7.

As long as a key 4 is held depressed, the note or notes will be sustained. Upon release of the key and redepression thereof, or, depression of the other key 4, tape 2 will be advanced another step to establish contacts of an other tone generating circuit or circuits. Thus, the rhythm and timing of the notes of the composition are arbitrarily determined by the performer as well as a musical parameter, that of volume.

It is understood that the general concept of the invention illustrated in FIG. 1 is not restricted to any particular tone generating system, to any one type of commutating sequentially operated register, not is it the intent that the system of the invention be limited to but one controllable musical parameter, i.e. volume intensity selection. Each key may control several musical characteristics in addition to motivating the tape and, as explained later, a number of such keys may be arranged in checkerboard fashion as cross coordinates so that variations in selection may be made along X or Y axes and also depth, representing three distinct parameters.

THE TONE GENERATING SYSTEM A tone generating system, generally indicated by the reference numeral 1 of FIG. 1 may employ an improved cathode-ray tube harmonic generator as shown at 10, FIG. 3, utilizing the properties of secondary electron emission.

Prior disclosures of the kind have described tone generating systems, including a cathode-ray tube harmonic generator, having a group of wave-shaped metal anodes which operate by straight conduction.

In the present and previous disclosures the generator has a group of wave-shaped anode targets. The anodes are insulated from one another and each anode target is made up of in-line, wave-shaped patterns in harmonic multiple relationship to each other, i.e., 1, 2, 3, etc. Each of these groups of interceptor plate patterns can be simultaneously scanned by the beam at a sweep rate corresponding to a desired fundamental frequency. By so doing, a series of harmonically related electrical pulsations are available in the respective interceptor circuitry responsive to fundamental-frequency sweep circuit oscillators. Alternatively, the anodes can be scanned at more than one fundamental frequency sweep-rate on a timesharing basis. This is done for the purpose of producing chords of desired harmonic content with a single beam cathode-ray tube.

With reference to FIGS. 3, 6, 6a and 7, the preferred form of tone generating system, generally indicated by the reference numeral 1 of FIG. 1 is comprised of: a number of relaxation-type oscillators 11, equal in number to the notes desired in any chord to be played by the instrument, and their associated frequency control circuit network 12. (NOTE: These relaxation-type oscillators 11, otherwise known as saw-tooth or sweep generators, depend, for the most part, on the employment of the charging potential across a capacitor upon application of a source of potential through a resistor and, in the case of the saw-tooth type, means for its rapid periodic discharge. Triangular oscillators are also applicable as sweep generators for this purpose): a time-sharing electronic switch 13 connected thereto and driven at an inaudibly high frequency by a pulse generator 14. Sweep circuits 15 are joined by the common output lead T6 of the switch 13 and the input leads 17 to a single set of deflection plates 18, of the previously mentioned harmonic generating tube 10.

The aforementioned sweep circuits 15 comprise those means, currently employed in practice, for centering and limiting the sweep travel of the beam in both directions of its excursion.

The cathode-ray tube 10 of FIG. 3 contains the usual electrodes supplied for producing a ribbon-shaped electron beam; a heater 19, a cathode 20, control grid 21, accelerator 22, and focusing elements 23. Accompanying circuitry is conventional and is not shown. The beam 33 impinges in the form of a line 24 on the surface of a plurality of improved output anodes or targets 25 on which wave-shaped patterns 26 are delineated. The electron image 24 is moved back and forth across these patterns 26 by the varying potential applied to the deflection plates 18.

Further details regarding the design features and mode of operation of these anode targets, along with their output circuitry and accompanying secondary emission collector 27 will be mentioned later. Suffice to say at this point, that electrical pulsations derived from the several output targets 25 form harmonic multiples of the fundamental sweep frequencies which may be separately attenuated, or synthesized, by the series connected photo-sensitive elements 28. The high frequency pulsations resulting from the action of the time-sharing chord switch 13 tend to be smoothed out by an integrating circuit 29' which follows it.

The resulting signal finally passes through the volume control amplifier 30 to the loudspeaker 7. Before commencing the detailed discussion of the improved harmonic generating tube, the operation of the electronic switch 13 will be described. Several types of such switching means are commercially available. The purpose of the electronic switch 13 as previously mentioned is to permit chord rendition with but a single-beam cathode-ray harmonic-generating tube. The time-sharing principle may be explained by the following illustrative example:

Suppose separate output circuits and oscillographs were connected to each of the anode targets 25. If the patterns were of sinusoidal form and one of the sweep oscillators 11 would be generating a saw-tooth wave form 31 at an audio frequency equivalent to that of the lower note of the first chord of the melody line FIG. 1 (i.e., G=39l.995 cycles/second of the equally tempered scale, based on a standard pitch A=440 cycles/second), then, with the oscillograph sweep rate adjusted for the observation of audio frequencies, sinusoidal traces would appear on each of the oscillographs screens. The first trace would be of the same frequency as the saw-tooth sweep oscillator. The second, double this frequency. The third, triple the saw-tooth, or fundamental frequency, and so on. The last might be say 20 or more times that of the first. If now the sweep rates of the several oscillographs were materially increased in speed to repetition rates far be yond that of audible frequencies, for the purpose of increasing the resolution of the traces on the fluorescent oscillograph screens, there would then appear a series of fine dots separated by spaces in place of what previously appeared to the eye as a continuous trace.

Suppose now a second oscillator were set into operation along with the first to produce a second saw-tooth sweep and both were fed to the deflection plates through the time sharting electronic switch 13. Let us presume that the frequency of the second oscillator were represented by saw-tooth wave 32. It might, for example, be that of the higher note of the first chord in the aforementioned melody (i.e., E=659.255 cycles/second of the equally tempered scale with standard pitch A=440 cycles/second). Now with the low resolution oscillograph setting, two sinusoidal traces would appear on each oscillograph screen and with high resolution these would appear dotted with dots of one trace falling vertically below or above spaces of the first. On the first oscillograph the two traces would have the same frequencies as that of the chord. The second oscillograph would show traces each.

of which were double the frequency of the first, or a chord an octave higher, in musical terms, and so on.

Like the eye, the ear does not perceive the breaks in the continuity observed at high resolution and the integration amplifier 29 further smoothes out the breaks. Let us further presume that the separate oscillographs were replaced by a single ins-trainer connected to all of the anode targets 25 of the harmonic generating tube it) through circuit elements designed to prevent inter-circuit current robbing. Under these conditions, if the same sawto-oth fundamental frequencies 31 and 32 remain operating, two traces will appear on the oscillograph corresponding in frequency with that of the saw-tooth sweep oscillators but differing in wave shape from the original sinusoidal form. If now attenuating means are inselted in each target cir cuit before it is connected to the common oscillograph, the shape of the trace may be changed at will in any manner desired, provided sufiicient harmonics are available, i.e., a sufiicient number of anode targets are used. If the oscillograph were replaced by a loud speaker the sound produced by the harmonic generator may then be varied at will to create sustained sounds of any description desired, such as those produced by conventional musical instruments or new tone creations, provided however sufficient harmonics are present in the higher frequencies to produce the brilliance required.

The following more detailed description of the improved harmonic generating tube and accompanying circuitry is made with reference to FIGS. 2, 3, 6, 6a and 7. The primary electrons emanating from the gun 21 (FIG. 3) are first accelerated and focused by elements 22 and 23 respectively into the form of a flat, ribbon-shaped beam 33 which is later ostensively bent, as it were, back and forth during its passage between deflection plates 18 for the purpose of scanning a series of special anode targets 25 as heretofore described.

As illustrated in FIG. 6, each of these targets 25 are supported and joined together by insulating rods 25a with a slight overlapping of the targets as shown. Each target 25 comprises a back plate of rectangular form made from a good conducting metal, say aluminum, to which output conducting leads 34 are attached. The front surface of these metal back-plates is coated with a thin film of some insulating material which has a high rate of secondary emission when struck by the primary electrons in the beam. If the plates are aluminum, this may be the natural oxide which is always present on this metal when it has been exposed to the air. This is represented by the portion of the surface indicated as 35 in FIGS. 6 and 6a. On the front of these targets, facing the electron beam 33, wave shaped patterns 26 are imprinted or otherwise formed on the aforementioned film 35. This may be done with some material such as ordinary black printers ink and subsequently baked in a hydrogen oven. This provides an area which has a low secondary emission ratio in contrast with the complementary portion coated with the aluminum oxide. If the printed patterns are of sinusoidal form, as represented, each of the targets are oriented such that the path of the electron beam image 24, which crosses them, is in the same direction as the usual time axis associated with such wave forms.

In operation, the primary electron beam 33, at any instant of its travel, will impinge within narrow boundaries, as represented by the image line 24, across the entire surface of each target which is exposed to the beam. It should be noted that the insulating supports 25a, which are unexposed, will therefore be unaffected by the beam and produce no extraneous secondary emission. Secondary electrons 36 will be ejected along the entire length of the electron image 34 where the primary beam 33 hits the several targets. These will be drawn away by the relatively higher potential collector 27. Because of the difference between the emission ratio of the aluminum oxide versus the carbon surface films on the targets, the total emission will depend upon the relative areas of these films within the exposure zone. The number of secondary electrons knocked out of the target film can exceed the number of incident or primary electrons and these, under the influence of an accelerating potential, result in a current flow in excess of that which would have been obtained in the original tube construction, which depended only upon the collection of the primary electrons.

If the beam were to remain stationary, the charge on the surface of the targets would reach a state of equilibrium and no signal would result but the constant movement of the beam does not permit such equalization between the collector and the targets and a varying charge is obtained on the front surface insulations of each target which results in a how of current from the metal backplates of targets 25 through their respective photo-sensitive attenuators 28, thence through the common lead wire 37 and load resistor 35% to ground by way of the DC. blocking capacitor 39. The remaining circuit is then completed by way of the collector 27. (The secondary electrons would be ejected from the front target surface to the collector as indicated by dotted lines 36.)

Fabrication of the individual patterns, as outlined, is similar to that employed in the construction of the monoscope type of tube employed for transmitting television test patterns. (See R.C.A. Rev., vol. 2, pp. 414, April 20, 1938, and its reference to British Patent 465,715.) In the present usage, however, a plurality of targets are provided with patterns designed for the generation of a plurality of repetitive or cyclic signal waveforms in harmonic order which may be individually controlled in amplitude, whereas the signal derived from the monoscope tube corresponds with a graphic pattern intended for reproduction as previously mentioned.

The composite pulsating current flow through the load resistor 38 and large capacitor 39 provides the signal voltage which is supplied via capacitor 40 to the integrating amplifier 29, amplifier and volume control 30 and loudspeaker 7, in the order mentioned (see FIG. 3).

An alternative target construction is shown in FIG. 7 for push-pull circuit operation. Here, the wave pattern targets 41, marked A, are etched or cut away to form the wave-shaped outlines, while targets 42, marked B, are made rectangular in shape. In this construction, all of the targets of both A and B construction may be made of aluminum with an insulating oxide coating or other high secondary emission ratio dielectric material facing the electron beam 33. These coatings are marked 41a and 42a for the wave-shaped and rectangular-shaped targets respectively while the corresponding target backplate conducting portions are marked 41b and 42b respectively. As in the first described targets, the wave forms are related harmonically, i.e. 1, 2, 3, 4, n. In place of one output conductor lead 34, for each harmonic generating tanget anode 25 as employed in the form shown in FIG. 6, each harmonic generator consists of one target of type 41 and one of type 42 each provided with separate conducting leads 43 and 44 with separate attenuating means represented by the dotted lines. The electron beam image 24 as it scans a set of these targets, comprised of a wave-shaped target 41 directly facing the beam and one of rectangular form 42 partly hidden from the beam by the front target 41, recognize, at any instant, complementary ordinates of the wave form, so that the output signal pulses from each are out of phase. For this reason, after individual attenuation, the outputs from all of the rectangular shaped targets 42 can be paralleled and connected to one push-pull amplifier tube while the other targets (wave shaped) may be similarily connected to the other tube of a push-pull input amplifier stage. In this construction, a maximum of secondary electrons may be collected, since all of the target front faces are highly emis-sive.

CONTROL OF TONE GENERATING SYSTEM Reference is made first to FIG. 3. The harmonic generating tube 10, including the output circuit has been described. A general reference was made to the employment of separate attenuating elements 28. These may be in the form of photo-sensitive elements of the small crystal type. The purpose of these attenuators is to control the relative amplitude of the several target signal-s, representative of the harmonics of one or more selected fundamental frequencies. The degree of attenuation imposed by these photo-sensitive elements 28 on the harmonic signal pulses is determined in turn by the relative transparency of a series of selectable groups of light filters 45 interposed between a light source 46 and the said photosensitive elements 23. As shown, the image of the elongated filament lamp 46 is selectively focused so as to cover one group or set of the several groups on the composite filters 45 to subsequently fall on the sensitive surface of the several attenuators 28, there existing a separate filter in each group of the series of filters 45 corresponding with each such attenuator element. The optical parts 47 provided for this purpose are attached to the end of a movable lever 48, better shown in FIG. 6. The optical elements, consist of a spherical lens 47a (in the simplest form) backed up by a mirror 47b. When properly positioned, this combination reverses'and focuses the light from lamp 46 on to a selected group composite filter 45. Selected positioning is accomplished by the actuation of the push-buttons 4 of manual 49. Each optics actuating lever 48 is linked to a fore and aft bar 50 (Y axis of co-ordinates) restrained to move downward in a parallel motion. Movement is accomplished for each bar by depression of any one of a row of pushbuttons 4 on it. Each bar 50 is therefore representative of various harmonic amplitude compositions and therefore different tone qualities.

FIG. 6b is a representation of one of the series or groups of light filters of several on composite filter 45 with a chart beside it, such as might be obtained with a densitometer, showing the relative densities corresponding to sections 45a, 45b, 45c, 45d in a group for the fundamental F and the 2nd, 3rd and 4th harmonics respectively, as illustrated. FIG. 6a shows an alternative method for filter selection in which each filter group 45 are separate units and have a direct mechanical linkage with a corresponding movable bar In this case the light beam from lamp 46 would remain undefiected.

This is one phase of push-button control for several musical parameters. The push buttons 4 of the simplified form of instrument shown in FIG. 1 control the degree of volume and is only a one-parameter push-button control system. Description will later include the control of other parameters with the same push-button manual.

ATTACK AND DECAY CONTROL The distinguishing characteristics attributable to sustained sounds, from whatever source, are determined by the number and relative intensity of their harmonic components. The manner of rise (attack) and fall (decay) however of separately emitted instrumental sounds, or musical notes, also marks the individuality of the various musical instruments to a high degree.

A semi-automatic system for the control of a second musical parameter, the attack and decay, which employs the same manual used for the harmonic tone blending, will now be described. The manual serves still other functions to be mentioned later.

The attack and decay system herein described, apart from the method of control, employs an electronic circuitry and certain optical elements.

For the purpose of this disclosure, transient volume control is understood to refer to the built up and falling off in amplitude of the sound oscillations during the advent and termination of each note or chord. More precisely, and for present usage, the expression is intended to refer not only to the rise and decay time subsequent and following a sustaining action, if any, but to trated in FIG. 8. Here an imaginary line 51 has been drawn which envelops the peaks of the sound oscilla-- tions, shown in the form of pulsations 52 emanating from the loudspeaker 7. The manner of rise and decay of these oscillations is indicated by (a) the form of curvature and (b) duration (time axis in the direction of the arrow) of the envelope at the beginning and end of the diagram. The intermediate portion, where the height of the pulses are constant, would correspond with the sustained period of a sounded note.

Refer again to FIGS. 2 and 3 and also FIGS. 8, 9, l0 and 12 for details regarding attack and decay control. It has been shown how each row of push-buttons 4 which are arranged along the Y axis (front to back) of the manual 49 control a different tone quality. In similar fashion, rows of these same push buttons 4 which lie along the X axis (left to right) on the manual 49 may be arranged to control different attack and decay conditions through the medium of the crosswise bars 53.

In FIG. 3 each such bar 53 is shown in the capacity of transferring a positive potential (+B B or B etc.) 54 and a like or unlike, negative potential (-B B or B etc.) 55 to certain electronic circuits which are thereby activated to initially produce a signal with a linear rise when the push button is struck, then a flattened top signal (constant potential), while the push-button is held down and finally signal with linear fall in potential upon release of the push button. The period of rise and fall is completely independent of the finger action. The resulting signal is shown in block diagram 56. The full line graph 57 indicates a condition which might apply to a certain one of the crosswise bars 53 and their respective activating push buttons. The dotted lines might indicate the action with the selection of other push-button groups. Following these electronic circuits 56 are the elements, represented by block 58, which change the curvature of the attack and decay signal outline, as represented by the full and dotted lines 58a, 58b, etc. The arrows drawn to these blocks mean they are selectable. This final output signal controls the volume circuitry 30 and hence the loudspeaker 7.

FIG. 8 gives further details regarding the attack and decay system particularly the contour shaping electro-optical devices employed. At the top left hand corner is shown a fragment of the manual bars 50 and 53 and two of their many push buttons 4. Upon initial depression of a push button 4 the companion tone-quality control bar 50 moves down throughout its length to set up a corresponding optical element 47 (FIGS. 3 and 6). Upon further depression, the bottom of the button comes into mechanical contact with the crosswise bar 53 and moves it downward along its length against the urge of resilient member 59. These bars 53 actuate the multiple switches necessary for the operation of the attack and decay sys tem.

Essentially, the switches 60 (FIG. 8), when depressed by means of the push button 4, initiate the flow of electrical current at some constant preset rate of change by appropriate circuitry 56 dependent upon the cross bar 53 selected. Each such bar being associated with differing attack and decay systems. The current, so initiated, flows through the movable spring restrained coils of a galvanometer-like device 62, or the equivalent such as an electromagnetic speaker coil, giving it essentially a constant angular velocity (uniform linear velocity in the case of a speaker coil). By means of a mirror 63 attached to the galvano-meter-like device 62 the image of an elongated filament lamp 64 may be focused and deflected in such a manner that it passes across a continuously varying light filter 65 at a uniform rate so that the light passing through it and picked up by the photo cell 66 (marked X) translates this action into a varying potential which follows the selected density changes of the light filter. It is only necessary to apply this varying potential to a tube control grid of the amplifying tube in the gain or volume control stage of amplifier 30 preceding the loudspeaker 7, FIG. 3, to cause the sound oscillations to vary in amplitude rise with the selected light filter 65 employed. In reverse order, these optical elements may be supplanted by a similar decay set of elements.

In operation, switch 60a in its normally closed position provides a negative cutoif potential to the grid of the input tube of the integrating amplifier from potentiometer 55, thereby rendering it inoperative. With de pression of push button 4, and consequent movement of cross bar 53, the double-pole single-throw switch 6011 applies a positive potential from potentiometer 54 on the input tube grid of the integrating amplifier 56a, thereby causing the capacitor, as normally placed in such an amplifier, i.e., between the plate and grid, to charge at a substantially linear rate for a period of time t determined by the potentials supplied by the limiting circuits 56b. So long as the switch remains in the depressed position the output potential from the limiting circuit remains fixed but upon release to the original position the grid-plate capacitor of the integrating amplifier discharges at a rate depending upon the setting of the potentiometer 55. Trapezoidal amplifier 56c transforms the linear rise and fall in potential to a linear rise and fall of current which activates the galvanometer winding. A transistorized circuit which performs all of the functions characterized by block diagram 56 representing rise and fall time circuitry, will be described later.

As the galvanometer mirror 63 swings in one direction, in response to a rise in current, against its restraining spring, the S.P.D.T. switch 60b connects in the proper photo cell X or Y and accompanying amplifier (attack A or decay B) which has a controlling influence on the speaker volume control 30 in a manner such as to follow the shaped and timed photo-cell signals.

As illustrated in FIG. 10, the attack and decay filters which shape the signal along predetermined curvatures, as indicated by densitometer charts shown above them 91, functions as the result of the filament images 67 sweeping across them back and forth together and always in the same direction. The images start in the more dense portion with only the attack photo cell active. The image passe-s across both filters in the direction of lesser filter density and with the decay photo cell still inoperative. This results in a rise in volume to a maximum and it will remain so as long as the push-button is depressed. Upon release of the push-button 4 the images return at their prescribed rate to the initial positions on their respective filters. During this latter stage the decay photo cell is made operative by switch 601) while the attack photo cell is shut off. As a result, the active decayphoto-cell receives less and less light as the images move across the decay filter in the direction of increased density. Consequently the loudness diminishes in conformity with the combined filter density changes and the listeners aural response, until the sound is essentially silenced. Graphical charts FIG. are shown to clarify these various stages of operation. These are drawn above the attack and decay filter density curves.

In FIG. 9 charts are drawn which show the circuitry performance corresponding with the initial (1) and final (2) positions of the push button 4. Diagrams A and B will be discussed later. C shows the action of the potentiometers 54 and 55. D shows the output current from circuits 56 which activate the galvanometer '62. E shows the photo cell amplifier combined outputs which control in part the rise and fall in volume of each note as it is sounded. It should be noted that a different linear time rate of rise and fall circuitry 56 is required for each cross wise bar 53, although only one galvanometer is required. A plurality of filter sets 65 is, required and these are selectable. It should also be noted that the rate of rise and fall at any instant and the duration of such rise and fall of any note or chord is independent of the finger action and is selective without dependence on other parameter controls.

Use of a transistorized circuit as an alternative method for the first stage of attack and decay control 56, FIG. 3 will now be described. FIG. 11 shows such a transistorized circuit 56d operable by a normally open S.P.S.T. switch S for producing a signal voltage which will rise at an adjustable linear rate to a limiting value after the switch is closed, remain sustained at this value as long as the switch is held closed and then be lowered at an independently adjustable rate after the switch is opened again.

This is accomplished in two stages. The first stage employs one transistor V of the P-N-P type connected with a common emitter and an adjustable capacitance reactance negative feed-back between the collector and base, analogous to the plate to grid feed-back as employed in tube circuit amplifiers of the integrating type (Miller) providing long time constants. The common emitter circuit has a phase reversal between base and collector. To obtain a quick return for the saw-tooth (rapid flyback (using oscillograph terminology) at the end of the wave) waveform required in this stage, a crystal diode D is inserted in series with capacitor C in the feed-back loop. A high leak resistor is connected between the common lead joining the cathode terminal of the diode D and capacitor C and the common emitter of this stage. Closure of the switch S in the potentiometer circuit, which includes the battery B with positive side common, initiates current through the load resistor R provided by the collector battery B (collector negative). By adjusting C the rise time can be lengthened or shortened as desired. If this period were so adjusted to extend from a given time 2 until a later time and the switch could be closed for this exact period, the wave shape of the signal across R i.e. e would appear as shown at the top of the sheet (Diagram A). In practice the switch S might be held closed for a period from say t until t as indicated in Diagram B of the negative bias battery potential applied to the base of transistor V (E). In this case the wave form provided by the first stage would appear as in the Diagram C.

The second stage is made up of two transistors V of the P-N-P type and V of the N-P-N type. The potential across the load resistor R of the first stage is coupled to the input transistor V of the second stage by a large electrolytic capacitor C with the negative side attached to the base of transistor V and the positive side to the negative side of load resistor R The second stage is equivalent in tube circuitry characteristics to a cathode follower with a constant-current load. This circuit follows the input wave form followed by a linear fall, provided only that the first stage wave terminates in a fall that is faster than the one to be generated. The composite result is an output potential across capacitor C C; which appear as shown in Diagram D. The adjustable capacitors C C are charged through a low impedance circuit and allowed to discharge through a high impedance circuit. The input transistor V to the second stage is forced to conduct by a pulse at the base that must be as large as the output triangle and of duration sufficient for the charging of the capacitors C C In operation, if a saw-tooth wave for example, is applied to the base of the second stage input transistor V the same saw-tooth wave form is reproduced in the second stage output except that the return to the quiescent level is a linear fall at a rate determined by the constant current and the magnitude of the adjustable charging capacitor C C The shift from P-N-P to N-P-N type of transistor is necessary to provide the correct current flow directions of the capacitor charge and discharge. A constant potential is applied to the base of the transistor 1 1 V by means of the potential divider, made up of equal resistors R and R I An ammeter M is shown connected in series with the load resistor R and if not too heavily damped and of low ohmic resistance will follow substantially in the manner of the signal Diagram D, so that by adjustment of C and C C the needle of the meter will rise rapidly and return slowly or vice versa or in any other manner desired. With a minute mirror cemented to the meter it would serve as the electro-optical movement (galvanometer 62), previously mentioned in the first part of the attack and decay control description.

A transistorized circuit 56d has just been described which may serve as an alternative construction for the first stage 56 of the attack and decay system. This circuit supplies a controlled linearized rise and fall signal input 57 to a subsequent electro-optical signal-shaping stage 58, as described, before the final signal is applied to a grid of an amplifier tube in the volume control stage of the amplifier 30 which operates the loudspeaker 7.

An alternative signal shaping stage method 58a, employing passive elements, will now be described. No galvanometer 62 is employed in this case and the linearized output potential 57 derived from the first stage 56 serves to actuate the second stage shaping circuits 58d directly (see FIG. 12).

The circuits to be described are modifications of a biased-diode square-law detector circuit which is illustrated in FIG. 34, page 24, N.A.C.A. Report 1209- Development of Turbulence Measuring Equipment by Leslie S. G. Kovasznay, 1954, The Johns Hopkins University. (See also Electronic Analog Computers, Korn and Ko-rnpage 290.) Although this circuit had been designed for the square-law conversion of the input signal, it may be changed in component values to reshape the input signal to any concave-upward form.

Considering one side only of the push-pull circuit, as illustrated, in the above reference. (A modified form of this circuit is shown by reference numeral 68 in FIG. 12.) It will be seen that the circuit consists of a load resistor 69, FIG. 12 in series with increasingly biased parallel branches, each branch consisting of a rectifier 70 and a series connected variable resistor 71. The bias voltage prevents the rectifier from conducting before the signal overcomes the bias voltage. Thus with increasing instantaneous voltage, more and more stages of rectifiers 70 are conducting. The curvature (concave upward only) of the rectified current through the load resistor 69 vs. input voltage e is controlled by the relative values of the series variable resistors 71 (photo-sensitive elements in the modification of FIG. 12).

FIG. 12 also shows a modification of this principle in circuit 72 designed to provide a concave-downward characteristic output. For this purpose, a series of fixed resistors 73 are provided which are in turn in series with the output load resistor 74. Between each of the said series of fixed resistors 73 biased diode bleeder circuits are provided which eventually terminate ahead of the load resistor at one of the input terminals 75. Each of these branch circuits are made up of a variable resistor 76, which may be in the form of a photo sensitive element and a diode or rectifying element 77. A floating power supply 78 with a potentiometer 79' is employed in lieu of separate bias batteries between the rectifier stages as illustrated in reference relating to the square-law circuit. (The bias for the concave-upward circuit modification 68 FIG. 12 is also shown provided by the voltage drop through resistors 80 which are supplied with current fiom a separate floating power supply 81.) As in the concave-upward circuit 68 the bias voltage between rectifier stages again prevents the rectifiers 77 from conducting before the signal overcomes the bias voltage and with increasing instantaneous voltage e more. and more stages of rectifiers are again conducting, but the increasing current through these branches is by-passed away from 12 the load resistor 74 so that the current through it is increased or reduced for attack or decay respectively at a diminishing rate as opposed to an accelerated rate prevailing for the concave-upward circuit 68.

By combining these two types of circuits, 68 and 72, in series connection, any shape of signal may be developed including that with a change of inflection.

To facilitate rapid and varied curvature shaping by these circuits without materially multiplying their nurnber, the variable series resistors in the rectifier branches may be replaced by light sensitive elements or with photo-diodes and their relative resistances changed by the use or" a selectable series groups 82 of varying density or variable area apertures 83 through which light may be passed in varying amounts to the sensitive surfaces of these ele ments as indicated by the arrow 84. For example, in the case of the concave-upward circuit 68, if the apertures 83 are substantially identical, then the resistances of the photo-sensitive elements 71 will also be nearly equal and the curvature will approach that of the squared instantaneous. values of the signal.

It is to be understood that the more rectifying stages in the circuits the more nearly the characteristics approach a smooth curve.

Arrangements may be provided for eliecting either concave-upward or concave-downward attack and decay characteristics by routing the linear rising and falling signals, 57a and 57b respectively, a derived from the preceding attack and decay stage 56 separately and alternatively by means of an electronic switch 85 to selectable circuits of either type, concave-upward 68 or concavedownward 72. In this case, the electronic switching may be triggered with pulses which are separately responsive to the attack and decay periods. These may be produced by making and breaking a direct current inductive load circuit 86, or moving a coil through a magnetic field, simultaneous with the depression and release, respectively, of a push button 4 and companion switches; switch S of the transistorized circuit 56 of FIG. 11, and switch 87 in the circuit 86. If an electronic switch, such as a flip-flop circuit, is employed the trigger pulses so obtained may be converted to like polarity by methods known to the art. Also buffer stages, not shown, will be required.

As shown in FIG. 12, the output signals are indicated by diagrams 88 and 89 for attack and decay when concave-upward and concave-downward circuits 68 and 72 respectively are selected.

All output terminals from the four circuits may be paralleled, since only one circuit at a time is active of any two in use. The signal is applied to a grid of a tube in the volume control stage of the amplifier 30 which operates the loudspeaker 7 of FIG. 3.

It is understood that where a linear rate or attack and decay only is desired the duration control circuit 56 may be used and the transient curvature rate control systems 58 would then be by-passed.

VOLUME CONTROL Finger-tip volume control Two control parameters have been described which employ the same manual but with selective components along the X and Y coordinate axes respectively. A third parameter controllable in depth (Z axis) will now be described. This is finger-tip volume control.

Displacement of any push-button 4 (FIG. 8) beyond that necessary to make the second contact on switch 60 will continue to make such connection by means of a sliding contact on a circular arc, but additionally, after an initial contact arc movement will actuate a variable volume control resistor 600.

It is significant to note that all of the push buttons are on one manual and provide means for initiating the notes of the music and also the control of the various .important tonal parameters, thereby providing facilities which would be manually impossible to duplicate in performance with the numerous keyboard keys and separate control stops normally employed in conventional organ-type instruments.

OTHER CONTROLS AFFECTING VOLUME (A) Tremolo cntr0l.- Reiteration of a note, chord or the harmonic components of such a note or chord with great rapidity, which is to say, a modulation in volume to produce a trembling or quivering effect is easily adapted to this system by interposition of suitable movable screens in the light path of either the attack and decay electrooptical system or quality-control optical system. The use of vibrating reeds with shutter or light-filter ti s and other methods have been described in patent references 1, 2 and 3 for this purpose.

(B) Over-all volume controL-ln addition to fingertip volume control, overall increase or decrease in volume is assisted by means of spring return foot-pedal controls, such as indicated by 109, FIG. 3.

(C) Rhythm accent c0ntr0l.A method was described and illustrated by FIG. 6, Nos. 280 and 281 of patent reference 3 for a pulse-generator-timed accent of the sounded notes, as an aid in acquiring proper rhythm. This was done by intensification of the cathode-ray harmonic generating tube beam and had the effect of periodically augmenting the signal strength, and consequently the loudspeaker volume, at intervals to be synchronized with the required beats of the measure. I

In FIGS. 3 and 5 methods are illustrated for automatic accent of the first note in each measure of a composition by code registration on a tape 2 by use of an independent set of commutating brushes 101 leading to suitable attenuators within the volume control portion of the loudspeaker 7 amplifier 30.

In the consideration of the various controls of volume, it is not to be implied that a linear relationship exists for equal steps of loudness, and so the attenuating elements must be tailored accordingly.

TRANSIENT PITCH CONTROL (A) Glide.In order to permit gliding of a tone from one note to the next, but so rapidly that the intermediate notes are not defined, sliding members 102 FIG. 3 are conveniently placed on the same manual 49 with that of the push buttons 4. Springs 103 return the member automatically to central position after displacement. The sliding member internal mechanism 102a acts as the moving contact of a variable resistor 12a in the fundamental tone generating initiator 12 (FIG. 3). Interlinkage with the cross bars are provided. The subject will be taken up again when the pitch activating potentiometer 12 is described in more detail.

(B) Pitch fringe.-A slight frequency modulation about the true center pitch may be accomplished by rotary capacitors 104 FIG. 3 in each fundamental pitch sweep circuits 11 of the harmonic tone generating cathode-ray tube 10. Speed may be controllable by motor 105 and a foot pedal 105 (FIG. 3).

FUNDAMENTAL SCALE INITIATION Under the subject of tone generation, the use of selectable charging potentials from the potentiometer resistance network 12 to activate one or more relaxation type fundamental sweep oscillators of the harmonic generating tube was briefly mentioned (FIG. 3). This unit 12 will now be examined more minutely in connection with the proposed use of both the usual fixed keyboard equally tempered scale and that of perfect or consonantly tuned intervals, usually referred to as the just tempered scale. The adaptation of perfect scale intervals has been described, patent references 1, 2 and 3. The use of semiautomatic selection of notes, as here proposed, lends itself particularly to such scale use, since accidentals (i.e., single notes introduced apart from the prevailing key signature) can, with preselection, be executed with rapidity.

SCALE FORMATION A brief review of the principles involved in scale formation will be undertaken with the aid of the charts shown in FIG. 4. Beneath the frequency scale of FIG. 4 there are shown portions of four musical scales. Two of these represent scales graduated with perfect intervals, While the lowest scale in the group is that of equal temperment. The disparagement between the equal and perfect intervals is apparent by comparison. The perfect scale in the key of C major differs from that of A major, the scale below it, in that, although the intervals for each are of like geometrical proportions, such proportioning is based on a different basic frequency, i.e., ground tone or tonic, in this case C and A respectively.

In FIGS. 3 and 5 by maintaining these same geometrical proportions for the contact on both a scale (12 h) potentiometer resistance and a series connected keysignature potentiometer resistance (12c) a combination of settings, pitch and key signature, may be obtained such that the potentials which are applied in the activation of fundamental tone generation permit rendition of perfect intervals, without the necessity for a multiplicity of extended keyboards to take care of the divergence of pitch requirements exhibited by such just tempered scale formations.

Although means for scaler transposition or modulation have been described in prior art, its use in combination with a proposed mode of semi-automatic pitch selection, about to be described in detail, would permit automatic key-signature pre-programming so as to eliminate the need for key-signature change stops. Accidentals would therefore no longer present a problem to tax the dexterity of the operator on rapid passages and perfect intonation could be achieved. The potentiometer system, shown in block form 12 FIG. 3, is more realistically illustrated in FIG. 5. Here the potentiometer resistance 12 is shown as consisting of a plurality of adjustable series resistors connected at one end to a source of potential while the other end, and the remaining battery terminal, are grounded.

Starting at the top of the sketch, the first or standard manually regulatable resistor 12d is used to set the pitch level to conform with accompanying instruments. The standard concert pitch of 440 cycles per second was used for convenient illustration (FIG. 4).

The second resistor, in order, 12a (FIG. 5) is adjustable about a fixed return center and provides the gliding pitch control, previously mentioned. Details are shown in FIG. 3. A forward movement of any one of several sliding members on the manual increases the pitch of the note sounded, while a reverse motion decreases the pitch. Springs 103 FIG. 3 return each slider 102 to their original position automatically. These sliders and accompanying rails are interlinked with adjacent pushbutton cross bars to permit simultaneous finger-tip volume control and a tone quality corresponding with the position of the push button which was used last.

The third series-potentiometer from the top, 12c, is provided with connections, as previously mentioned, for shifting the pitch level to that of a required key note base or key signature setting. Semi-automatic selection is provided and will be mentioned again later in the description under pitch commutation.

The next series or range resistor potentiometer, 122, is also arranged for semi-automatic selection and is employed for reducing the number of commutation brushes. This is accomplished by grouping the pitches within the compass of the normal keyboard into double octave ranges which overlap by one octave as shown in FIG. 4. This permits an octave stretch from any point within the pitch compass.

The last series resistor potentiometer, 12b, is provided 15 with alternate taps for initiating scale pitches in either just or equal temperment and is semi-automatically commutated as were the two previously described series potentiometer resistors.

SEMI-AUTOMATIC PITCH SELECTION Reviewing the basic ideas presented in FIG. 1, it will be noted that subject matter has been covered involving details of improved versions of the tone generator and the push-button manual 49, including the control of several tonal parameters.

It now remains to elaborate more fully regarding the construction of the semi-automatic pitch programming and selection system. Such a system is represented in FIG. 3 by the commutating brushes 3 and the programmed tape 2. In operation this tape is moved one step at a time in response to nudges given by the ratchet drive 5, manually operated by the same push buttons of the manual 49 which serve as tone parameter control selectors.

The directly connected ratchet may be replaced by an electromagnetically operated ratchet. Such a ratchet 1W7 is shown in FIGS. 3 and 8. In FIG. 3 the electrical connections are indicated as leading to the manual 49. In FIG. 8 a fragment of the push buttons 4 of the manual show a means for energizing the ratchet motor 1197 each time any one of the push buttons is depressed. Conducting strips 50a on either side of each quality control bar 50 serve to complete a circuit between the ratchet motor 107 and a suitable power supply 10% by means of a metallic coating on the bottom of each push-button body. In FIG. 9 Diagram A indicates the timing of the ratchet motor current pulse in relation to the attack and decay generated signals previously discussed. (The ratchet 107 may be arranged to operate either on depression or release of the push buttons.)

The perforated tape 2 shown in FIG. 3 may be replaced by one made up of composite layers FIGS. 3, 5, and a. FIG. 3 shows the general arrangement of the tape transport mechanism including a drive pulley 1W3, take-up pulley 109, feed pulley 110 and tape rolls 111, including a spare. Tape rolls may be self winding in the manner of window blinds. A slack portion is shown which permits r-apid incremental shifting. The tape may be sprocket driven (not shown).

A section of composite tape is shown in FIG. 5a. This fragmentary sketch also shows parts of the brushes. Although not shown, it is understood that a cam mechanism is included in the commutative brush system, well known in the art, which will lift the brushes away from the tape when in motion and, with the aid of individual brush spring tension retainers will compress the opposed brushes firmly against the tape during the period when it is stationary. This is for the purpose of insuring good electrical contact. Since, however, these contact brushes are part of a high impedance circuit, the contact resistance is not overly critical. The tape, as shown, is made up of a paper or other insulating inner layer 2a which is coated on each side with a thin film of some plastic material, e.g., kryoline 2b after thin strips of metal 20 or other conducting medium are imbedded in place. When electrical contact is desired the plastic coating is melted away (120) by the use of an electrically heated wire coil so as to expose the conductor beneath. The top brushes 3 are shown in contact with such a strip 20 to complete a circuit existing between them. The lower brush is shown as held against the plastic coating, since no hole has been melted away at this point, as illustrated.

Turning now to FIG. 5, a portion of the tape 2 is here shown along with the pitch activating potentiometer 12 and the completion of this circuit with brushes 112 through the burnt-out openings are illustrated. These .are shown as black dots. The underlying conducting .strips are shown as dotted lines. The chord selection commutation brushes 113 are also indicated at the bottom portion of the tape and a set of rhythm-accent contacts '16 101 previously discussed in part under auxiliary volume control methods.

PREPARATION OF THE PROGRAM TAPE The tapes may be prefabricated or prepared in place on the instrument from blanks. The latter process should be an aid to composition, since the music can be heard during the preparation of the tape and later executed with the regular push-button manual and with the addition of appropriate control features.

For the latter purpose, auxiliary equipment would be supplied in the form of a special keyboard 114 (FIG. 3) with conventional type keys 115. One function of these keys would be to operate switches interconnecting the scales section 12b of the pitch activating potentiometer 12 with the sweep oscillators 11, in lieu of the commutative selector brushes 112 and 113. The latter would be temporarily disconnected for the purpose of preparing the register tape 2. The keyboard 114 consists of two octaves, and as shown in FIG. 4, this permits the playing of any note with one hand throughout the complete composition when provided with overlapping range-selector stop 116 having operable switches, which temporarily are used for cutting in range selector potentiometer resistor 12!: to different ranges. Other stops 117 are connected to the key signature potentiometer resistance 120.

The keyboard serves a second purpose, for, coincident with the rendition of say an original composition, the performer by actuation of these same key and stops completes appropriate electrical heater element circuits 118 which shrink away the plastic coating on the tape to expose the conducting portions underneath as illustrated in FIG. 5a. In addition, each time a key is struck the ratchet motor 1tl7 connected thereto will move the tape one step forward in preparation for burning the next series of register holes. Note that at this time the semiautomatic manual is to be disconnected from the ratchet motor with the switch 119. When using this keyboard,

each push button 4- of the semi-automatic manual 49 is latched into a selected position of quality control. One type of such latching mechanism is shown in FIG. 9 Nos. 410 and 414, patent reference 3.

What I claim is:

1. In a musical instrument; means for generating fundamental pitch frequencies, selection means including a record having recordings in successive areas corresponding to given pitch frequencies, means for sensing said record, transport means for advancing said record with respect to said sensing means, a keyboard, drive means for said transport means operable upon each depression of any one of the keys of said keyboard to advance said record a single step of movement to bring a recorded area of said record into registration with said sensing means, means operable in response to each key depression to initiate operation of the frequency generating means corresponding to the recorded area sensed by said sensing means, a plurality of preset tone quality control means for said frequency generating means, and means operable upon depression of given keys respectively of said keyboard to effect operation of said tone quality control means.

2. A musical instrument according to claim 1; said keyboard comprising a group of keys for each tone quality control means and each tone quality control means being operable by depression of any key of the associated group.

3. A musical instrument according to claim 1; including volume control means for said fundamental pitch frequency generating means, and means responsive to depression of any key of said keyboard to regulate said volume control means for operation of the selected frequency generating means at a volume level proportional to the pressure applied to the depressed key.

4. A musical instrument according to claim 1; wherein said keyboard comprises a plurality of groups of keys, and including a present attack and decay means for said pitch frequency generating means associated with each key group and operable by any one key of said group independently of the depression and release time thereof.

5. In a musical instrument; means for generating fundamental pitch frequencies and harmonics thereof; a plurality of wave-shaped patterns of low secondary emission rate dielectric material in harmonic multiple relationship, a complementary pattern of a relatively higher secondary emission ratio dielectric material for each wave-shaped pattern, a conducting sheet backing each pair of wave-shaped and complementary patterns, means for insulating each pair of patterns one from the other to form a series of harmonic targets, means for constantly sweeping said targets with an electron beam, means for controlling the rate of sweep of said beam, and collector means for attracting secondary electrons emitted from said targets and raising the potential of said secondary electrons thereby causing a flow thereof between said collector and the respective target conductor backings; selection means including a record having recordings in succesive areas corresponding to given rates of sweep of said electron beam, means for sensing said record, transport means for advancing said record with respect to said sensing means, a keyboard, drive means for said transport means operable upon each depression of any one of the keys of said keyboard to advanc said record a single step of movement to bring a recorded area of said record into registration with said sensing means, said sensing means thereupon being operable to regulate operation of said sweep control means to cause said electron beam to sweep said targets at a rate corresponding to the recording sensed, and means operable in response to each key depression to initiate operation of said frequency generating means.

6. In a musical instrument; means for generating fundamental pitch frequencies, selection means including a record having recordings in successive areas corresponding to given pitch frequencies, means for sensing said record, transport means for advancing said record with respect to said sensing means, a keyboard, drive means for said transport means operable upon each depression of any one of the keys of said keyboard to advance said record a single step of movement to bring a recorded area of said record into registration with said sensing means, means operable in response to each key depression to initiate operation of the frequency generating means corresponding to the recorded area sensed by said sensing means, and a plurality of parameter control means each operable to modify the generated pitch frequencies and each operable in response to manipulation of any one of a plurality of said keys of said keyboard.

7. The invention according to claim 6; wherein said means for generating fundamental pitch frequencies includes an electronic time-sharing switch and a single beam cathode-ray secondary emission harmonic generator.

References Cited in the file of this patent UNITED STATES PATENTS 2,069,441 Headrick Feb. 2, 1937 2,171,936 Kucher Sept. 5, 1939 2,438,709 Labin et al Mar. 30, 1948 2,441,296 Snyder et al May 11, 1948 2,478,867 Hanert Aug. 9, 1949 12,496,633 Llewellyn Feb. 7, 1950 2,497,331 Swedien Feb. 14, 1950 2,500,821 Hanert Mar. 14, 1950 2,516,886 Labin et a1 Aug. 1, 1950 2,528,187 Sziklai et a1 Oct. 31, 1950 2,541,051 Hanert Feb. 13, 1951 2,562,670 Koehl July 31, 1951 2,601,265 Davis June 24, 1952 2,734,100 Kendall Feb. 7, 1956 2,747,130 Goldberg et a1 May 22, 1956 2,801,563 Koehl Aug. 6, 1957 2,842,021 Oncley July 8, 1958 2,855,816 Olson et a1 Oct. 14, 1958 FOREIGN PATENTS 1,178,926 France Dec. 15, 1958

Images