US2924784A - Electronic musical instrument - Google Patents

Description

Feb. 9, 1960 R. H. PETERSON 2,924,784 I I ELECTRONIC MUSICAL INSTRUMENT 4 Sheets-Sheet 1 Filed July 18, 1956 llx iaini lilla EgLc/zaxd A! peiarsod Feb. 9, 1960 R. H. PETERSON ELECTRONIC MUSICAL INSTRUMENT 4 Sheets-Sheet 2 Filed July 18, 1956 5505mm! J. geiersdrz y' "a fiiivrfieq Feb. 9, 1960 R. H. PETERSON ELECTRONIC MUSICAL INSTRUMENT Filed July 18, 1956 4 Sheets-Sheet 3 Inz/mi Ede/card A/Feixemz @Zz'bfmg/ Feb. 9, 1960 R. H. PETERSON ELECTRONIC MUSICAL INSTRUMENT 4 Sheets-Sheet 4 Filed July 18, 1956 b A UN T COUPL NG rzz/zrzz'or 5 Elcfzard AK Pi'ersazz quirement must be met.

United States Patent 2,924,784 ELECTRONIC MUSICAL INSTRUMENT Richard H. Peterson, Oaklawn, Ill. Application July 18, 1956, Serial No. 598,582 4 Claims. (Cl. 331-49) My invention relates to electronic organs and analogous musical instruments, of the type in which the original tone is generated electronically, as by means of electrical oscillators, with or without additional distorting devices for changing the wave shape of the signal generated by the oscillators. For use in'such instruments the oscillator, ideally, must have certain characteristics not generally found in ordinary oscillators.

Specifically, the frequency of the generated signal must be exceptionally stable over long periods of time, substantially independent of aging of the vacuum discharge devices or transistors, and unaffected by variations in the potential of the power supply, over the range of voltage variations to which ordinary commercial power circuits are liable.

Further, in cases where it is desired to have the oscillator normally inoperative, and to key or play the instrument by applying activating potential to selected combinations of oscillators, a much more diflicult re- To secure-effective control of the rates of attack and decay when a note is played, it is essential that the amplitude of the signals .generated should correspond roughly to theactivating voltage delivered to the oscillator, and that over substantially the entire range of amplitudes there must be no material change in frequency. It is essential to secure this constant pitch over any desired period of attack or decay because even a small change in pitch, especially during decay, produces a chirp that is most objectionable from the esthetic point of View. Many percussion instruments are characterized by long decay periods up to several seconds at perfectly constant pitch and the extended period of the decay makes this'requirement critical, when percussion effects are desired.

In designing oscillators for these exacting requirements of electronic musical instruments, two different stability citeria must be carefully considered. First, the primary elements of the tuned circuits, namely, the inductance and capacitance, must remain stable, not only with regard to aging, temperature, and humidity, but especially in regard to instantaneous changes as a result of the normal functioning of the instrument itself with varying excitation applied to the oscillator circuit, and changes in the currents flowing in the circuit.

Secondly, the Q of the circuit must be as high as possible to minimize the frequency shift due to the exciting impulse itself. This means that the reactance of the tuned circuit elements must be high compared with their resistance. This is vitally necessary because .any resistance in the tuned circuit, whether it is the inherent characteristic of the inductor or capacitor, or Whether reflected into the tuned circuit by the exciting circuit, will cause phase shifts of thereactive currents, which will cause a frequency shift away from the theoretical frequency during functioning.

In every-tuned circuit, there is a mathematically precise theoretical frequency depending on the values of the component parts of the tuned circuit proper. This freqnency never obtains in practice because it could only exist momentarily with no excitation taking place. The presence of the exciting circuit introduces disturbing factors which can never be eliminated but can be reduced to a minimum by designing the circuit with as a high a Q as possible. This variation from theoretical frequency will ordinarily be a maximum when the excitation is maximum, and may decrease materially before the applied plate voltage gets low enough so that oscillation ceases.

With the air wound coils of which I have knowledge, gradual decay down to the voltage where oscillation ceases appears to be accompanied by a lowering in pitch or frequency too slight to be detected by an untrained ear and too slight to be olfensive in any event. Ferromagnetic. material for a minor part of the magnetic circuit has a tendency to cause the frequency to rise during the decay. During decay, this counteracts the eflect of the elimination of the exciting circuit (which is in shunt with part of the inductor Winding), as a factor tending to make the actual frequency higher than the mathematical, theoretical frequency of oscillation without excitation. A practical limit is reached, where the net tendency to rise during decay becomes excessive. With ferro-magnetic metal cores which are also of conductive material, the amount of ferro-magnetic material that can be tolerated, while maintaining adequate stability, is very low. With the ceramic, insulating, ferro-magnetic material disclosed herein, it is possible to use much greater amounts of ferro-magnetic material and much smaller physical dimensions and still maintain adequate stability during decay. I believe this to be because hysteresis, and the variations in the permeability and the incremental permeability of the ceramic material, as a function of flux density, differ substantially from the corresponding characteristics of metallic ferro-magnetic materials. Even with the ceramic material, it is still necessary to have maximum flux densities much lower than those customary in the ordinary uses of such material. This limits the operation to a relatively narrow range within which the permeability is relatively constant and the incremental permeability has a negative coefiicient.

A special combination of selected constituents and proportioning of the parts has been found to accomplish:

(1) minimum pitch fluctuation with varying flux density;

(2) minimum pitch fluctuation due to ohmic and/0r magnetic losses;

(3) a residual fluctuation with variation in flux density that tends to counteract the variation due to the phase differences between the currents in the tank circuit, the plate circuit, and the grid circuit.

(4) small, compact units having small, compact magnetic fields, such that complete tone generator assemblies can operate without magnetic interference in space volumes amounting to only a small fraction of the volumes previously required.

It is significant that the space requirements are not merely the actual volumes occupied by the physical units themselves. They include also the volumes of the spaces surrounding each individual inductor, in which spaces the magnetic fields of the inductors may come into such proximity that the fields of two adjacent inductors occupying adjacent positions in space, tend to overlap and affect each other. When this happens, the signal from each inductor will be transferred into the other inductor.

If this contamination were followed by mere amplification without further alteration of wave shape, the final result might still be tolerable. But when the contaminated signal is subsequently subjected to amplitude distortion of its wave shape, the distorting instrumentaliasaansa r ties generate not only overtones of the contaminating signal, but oscillations having a frequency equal to the sum of the frequencies of the two signals and overtones of that oscillation, and other oscillations having a frequency equal to the difference between the frequencies of the main signal and the contaminating signal, and overtones of that oscillation. Not infrequently three different oscillators may be functioning next to each other and if there is such leakage followed by such distortion, the number of possible combinations builds up to a point such that the product is likely to be noise rather than music.

Even where the degree of contamination is relatively slight, the subsequent attempts to atler the wave shape and produce tones of great esthetic beauty, as described in U.S. Patent 2,649,006, is certain to generate so many of these irrelevant or ghost frequencies, as to render the music esthetically unacceptable.

In the accompanying drawings:

Figure 1 is a schematic wiring diagram of an electronic organ;

Figure 2 is a schematic wiring diagram of an oscillator employing a vacuum tube;

Figure 3 is an elevation of the separated parts of an inductance unit;

Figure 4 is a top plan view of an inductance unit; Figure 5 is a section as on line 5--5 of Figure 4; Figure 6 is a diagram of flux distribution;

Figure 7 is a perspective of a chassis for producing each of two difierent semitones in each of six octaves;

liigure 8 is a plan view of the chassis on a reduced sca e;

Figure 9 is an end view of the same chassis;

Figure 10 is a schematic wiring diagram of a transistor oscillator; and

Figure 11 is a side elevation, partly in section, of a modified inductor.

In the embodiment selected to illustrate the invention, and referring first to Figure 1, there are indicated three tone-generating units 10, 12, and. 14. Each of these units may be according to Figure 2 or Figure 10. Each oscillator delivers a wave of dependably constant frequency and substantially perfect sine curve shape. The complete organ includes as many oscillators as there are notes within the gamut of the instrument. The signals from the oscillators pass through distorting devices 16 of which six are illustrated for each oscillator. Distorted signals of similar tone quality are collected on the buses 18. Stops 20 controlled by the player connect any selected assortment of buses to the amplifier 22, with the resultant music delivered by the loud speaker 24. These distorting units are fully described in Patent 2,649,006 August 18, 1953 and per se form no part of the present invention.

The playing keys 26 are used for normal playing to activate selected combinations of oscillators. The playing keys 28 activate the oscillators with the same attack characteristics but with prolonged decay characteristics to produce percussion effects. This is accomplished by capacitors 30 protected by resistances 32 and separated from the oscillators by rectifiers 34 so that keys 26 can operate the oscillators without charging the condensers but when keys 28 are used the capacitors 30 charge when the note starts and the charge is available to prolong the decay of the note when the key is opened.

Referring to Figure 2 the conductor 36 receives positive potential from the power source 38 when the player closes either key 26 or key 28. The connection to the tube 46 is through a time-delay circuit comprising the resistor 40 and grounded capacitor 42 to the plate 44 of the vacuum tube 46. The cathode 48 is kept continuously active so that the oscillation is initiated by supplying potential to plate 44 and controlled during the function of the oscillator by the grid 50. The cathode and grid are connected across only 30 percent of the winding 52 4 of the inductor 54. Thus the conductor 56 connects the cathode to an intermediate point in the winding and the grid is connected through the protecting capacitor 58 to the junction point 60, which is equivalent to the upper end of the coil as illustrated in Figure 2. The theoretical frequency of the tuned loop comprising the inductance 54 and the capacitor 62 is a function of dimensions of those elements only, but that frequency never exists in actuality because the exciting current has to supply an increment of energy during each oscillation to replace that dissipated in the loop. The dissipation of energy takes place due to the ohmic resistance of the winding 52; and to hysteresis in the core of the inductance if it has a core. There are also dielectric losses in the capacitor 62, magnetic leakage losses, and resistance losses in the connectors.

The intermittent supply of energy by the exciting circuit alters the frequency and this alteration varies with the intensity of excitaiton. This variation cannot be avoided but it can be minimized by designing the circuit with maximum Q.

Referring now to Figures 3, 4, 5, and 6 each mductance comprises a cylindrical core 64 and an annular winding 66. The length of the core and the outer diameter of the winding have a ratio approaching umty, say between 0.7 and 1.4.

A suitable structural frame is provided including the guide tube 68, in which the core may slide freely, and spaced end plates 70 and 72 for the coil. The tube 68 extends beyond the plates in one direction and the opposite plate carries a plurailty of connector terminals 74, 76, and 78.

Figure 6 is a diagram of the putative approxlmate distribution of the lines of magnetic flux when the energy storage in the inductance is at its peak. It will be noted that the outer envelope of the lines of force is approximately spherical and that the radial dimension of the cross sectional area in which the lines of force in air are located can be very small at the equator because the other dimension is the entire periphery of the sphere. Accordingly, the lines of force at the point indicated at in Figure 6 will all lie very close together in a thin annular band and if there were another inductance beside the one illustrated, the spacing between them needed to prevent deleterious intermingling of the lines of force of the two oscillators would be a minimum. For instance, the point 82 at the right side of Figure 6 represents a degree of separation that would be efiiective.

Means are provided for shifting the axial position of the core 64 with respect to the coil. This provides a very fine and precise adjustment of the inductance. I have illustrated a cap 84 telescoped over the tube 68 and provided with small prongs 86 which embed themselves slightly in the material of the tube. The core 64 carries an integrally assembled non-magnetic threaded extension 88 provided with a screw driver slot 90, and the cap 84 has tangs 92 engaging the threads of the adjustment screw 88.

Referring now to Figures 7, 8, and 9, the geometrically compact configuration of the field indicated in Figure 6, in which field the pulsating energy is stored, makes it possible to locate a plurality of inductances much closer to each other than was heretofore possible. The central channel 94 of the chassis is of non-magnetic metal such as sheet aluminum. Twin panels 96 and 98 of fiber board are aflixed to the edges of the channel legs. The panel 98 is apertured to receive six inductances of which four are in spaced relation along a straight line extending down the middle of the panel. The largest inductance 100, which produces the lowest note, is at the upper end of the panel 98 in Figure 7. The next inductance, musically, is the fourth inductance geometrically, at 102, and is part of a loop tuned to produce a frequency twice that of the inductance 100. The inductance 104 is adjacent the inductance 100 and produces a note two octaves higher than the inductance 100. The next inductance 106 is at the extreme remote end of the panel 98 and offset toward the channel 94. It produces a frequency four octaves above that of inductance 100. The fifth octave is produced by inductance 108 located on the center line between inductance 102 and inductance 104. Finally the sixth octave is produced by inductance 110 which is at the extreme end of the panel, almost but not quite abreast of inductance 106 but set out to the outer corner of the panel remote from the channel 94. As most clearly indicated in Figure 8, panel 96 has identical arrangement except that the largest inductance for producing the lowest note is at the end of: panel 96 farthest from the inductance of panel 98. Also, when a plurality of chasses according to Figure 7, are set side by side, each inductor 100 is nearest the inductor 110 of the adjacent chassis.

It is customary to have each panel produce the same note of the musical scale in each of six different octaves. For instance. panel 98 might produce the note C and panel 96 might produce the note C sharp. It will be apparent that twelve panels assembled on six channels can be used to produce 72 consecutive semitones. No inductor is adjacent another inductor of frequency closer to it than two full octaves, but the sixth octave is so small that it can find room beside the fourth octave.

Along the web of the channel 94 are provided sockets 108 receiving twin vacuum tubes 110. Each such tube is provided with two plates and two grids and constitutes the tube 46 for two oscillating circuits, one of which will be on panel 96 and the other on panel 93.

A convenient and geometrically compact arrangement for the resistors and capacitors indicated schematically in Figure 2 is a pair of fiber board panels 112 and 114 provided with terminal strips 116 each carrying a plurality of terminals 118 for establishing the connections called for in Figure 2 with the various resistors and capacitors indicated generally at 120 in Figure 7.

By arranging the panels or Wings 112 and 114 in the diagonal relationship shown, a compact arrangement of the complete assembly within the confines of a rectangular parallelepipedon of minimum volume is obtained.

Referring now to Figure 10 I have indicated schematically a transistor 122 with its emitter 124 connected to conductor 56 through an adjustable resistor 126. The base 128 is connected to the capacitor 58 and the collector 128 is connected to the time-delay circuit through a resistor 130. The gridbias resistor 132 of Figure 2 is omitted. A resistor 133 is connected between the collector and the base and a resistor 134 is connected between the base and ground. The three resistances 138, 133, and 134 constitute a three-element network which keeps the potentials of the collector and the base within pre-determined limits.

All other parts of the oscillating circuit illustrated in Figure 10 perform the same function as in Figure 2 but the absolute values need to be changed because the transistors illustrated operate with the power source 38 in opposite polarity and at voltages between about 6 and about 18 volts.

Audio frequency oscillators are commonly made with a plate circuit including from 10 to 30% of the tank circuit, the rest of the tank circuit becoming part of the grid circuit. An oscillator according to the invention employs an inductor of much less inductance and an abnormally large capacitor, such that the circuit does not oscillate well with such a disposition of the tap connected to the cathode. Specifically, the inductor is so small and the capacitor so large that oscillation achieves maximum intensity and stability with about 70% of the tank circuit in the plate circuit. Beginning with about 10% of the tank circuit in the plate circuit the response of the oscillator and its stability both improve gradually up to optimum values over a short plateau located at about 70% and then decrease abruptly. The small inductor and large capacitor constitute an abnormal and relatively inefficient combination, electrically speaking, but the high stability of the frequency of the entire combination transcends other considerations.

Referring now to Figure 11 I have indicated an inductor in which the ceramic core 138 has a cap 142 of the same ceramic ferro-magnetic material. The cap may have a peripheral lip 144 extending down around the outside of the coil 146, which coil may occupy the entire enclosed space except for the winding tube 148. A duplicate cap 144 results in reducing the air path to the relatively short gap at 148. The air path is not only short in axial dimensions but of quite large cross section and if made up in the same size as in Figure 6 the air gaps at 148 and through the tube 68 would present only ten or twenty percent as much resistance to the flux as the long air gap in Figure 6. Because of the high permeability of the core 138 and the caps 142 the air gap of Figure 11 may still represent considerably more than percent of the total resistance to flux, but the necessary total flux to obtained with a structure having dimensions less than half and volume less than A; of what would be required to get the same frequency with an inductor according to Figure 6. For very low notes where percussion effects are not desired this secures adequate stability with a great decrease in size.

Others may readily adapt the invention for use under various conditions of service by of the novel features disclosed For instance, with the transistor the channel 94 is omitted.

As at present advised with respect to the apparent scope of my invention I desire to claim the following subject matter.

oscillator of Figure 10, structurally superfluous and may be each transistor having an emitter a collector and a base; each tank circuit comprising an inductor and a capacitor connected into a closed loop; said inductor having a wound coil of many turns; said exciting circuit having a connection from said base to one point in said quency down to cessation; celving and assembling into a composite signal, the signals from a plurality of simultaneously oscillating oscillators.

2. A combination according to claim 1 including means for securing small precision pitch adjustments of the pitch of each oscillator by shifting said ferro-magnetic core axially with respect to said coil.

3. A combination according to claim 1 in which said exciting circuit includes about 70% of said inductor coil.

4. A combination according to claim 1 which six inductor coils are mounted in a common plane; there being a single panel for supporting said coils as a unitary assembly; each of said coils being part of one of said oscillators; each oscillator being electrically separate from all the others to the extent that it can oscillate independently of all the others; the frequencies of five of said oscillators being aliquot portions of the frequency of the oscillator of highest frequency; the aliquot portions being V2, 34, /8, A the coil for the lowest frequency, identified for convenience as the first octave, being adjacent one end of said panel; the coils for octaves l, 3, 5

and 2 being arranged in a longitudinal line in the order stated; the coils for octaves 4 and 6 being spaced apart transversely on opposite sides of said line and beyond the coil for octave 2.

References Cited in the file of this patent UNITED STATES PATENTS 2,051,012 Schaper Aug. 11, 1936 2,216,513 Hammond Oct. 1, 1940 2,223,539 Baker Dec. 3, 1940 2,291,787 Beanland et al. Aug. 4, 1942 2,555,039 Bissonette May 29, 1951 2,588,082 Brown et al. Mar. 4, 1952 2,603,774 Gusdorf et a1 July 15, 1952 2,630,560 Earl et a1. Mar. 3, 1953 2,728,054 Schoenberg Dec. 20, 1955 2,790,906 Hammond Apr. 30, 1957 min...

UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Pat ent No. 2,924 7841 February 9. 1960 Richard H. Peterson It is hereby certified that error a of the above numbered patent re Patent should rea ppears in the printed specification quiring correction and that the said Letters d as corrected below.

Column 2, line 19 after "circuit" insert a comma column 5 line 14 for the numeral "10" read 100 Signed and sealed this 26th day of July 1960 (SEAL) fittest:

KARL H.- AXLINE ROBERT C. WATSON .ttesting Officer Commissioner of Patents UNITED STATES PATENT OFFICE v CERTIFICATE OF CORRECTION Patent No. 2,924,784 February 9, 1960 Richard H. Peterson It is hereby certified that error appears in the printed specification of the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.

Column 2, line 19, after "circuit" insert a comma; column 5, line ll for the numeral "10" read lOO Signed and sealed this 26th day of July 1960.

(SEAL) ittesti KARL H.- AXLINE 7 ROBERT C. WATSON tttesting Ofi'icer Commissioner of Patents

Images