US2919620A - Automatic tuning device for polyphonic instruments - Google Patents

Description

Jan. 5, 1960 R. H. DORF AUTOMATIC TUNING DEVICE FOR POLYPHONIC INSTRUMENTS Filed March 25, 1954 AMPLIFIER LOW- FRE ENGY LO -P 55 FILTER MIXING DETECTOR HIGH FRE UENCY AMP 1F ER a E E L 3 INVENTOR a iaz'chdrdjifl yf 2 ATTORN EY United States Patent AUTOMATIC TUNING DEVICE FOR POLYPHONIC INSTRUMENTS Richard H. Dorf, New York, N.Y. Application March 23, 1954, Serial No. 418,180

5 Claims. (Cl. 84--454) This invention relates to tuning devices and more part1cul arly to a device for tuning polyphonic electronic musical Instruments. It is particularly adapted to the tuning of musical instruments of the type employing frequency dividers or locked oscillator chains. This includes, for example, electronic organs. My improved tuner can be used to set the temperament octave in any polyphonic instrument including not only that of individual oscillator organs but also pipe and reed organs and pianos.

improved tuner consists of a single small case contaming an electronic amplifier and detector, a motor rotated,stroboscopic wheel and a group of lamps. A window on the front panel is observed by the tuner and through it can be seen twelve concentric bands of rotating dots or transparent lines. The pitch of each note is adjusted until the band associated with that note appears to stand still. The tuner requires a power source of 100 to 120 volts with 60 cycle volt alternated current, although models maybe built for other voltage and frequencies.

One of the principal objects of the invention is to furnish a reference standard of good accuracy for the frequency to which a musical instrument of the polyphonic type may be tuned. Whilethe invention will be used for tuning the so called tempered scale, it may also, with appropriate modification of the stroboscopic disc, be adapted to tune to the scale of just intonation or any other scale in common use.

A further purpose of the invention is to enable unskilled personnel to tune a polyphonic instrument in a minimum of time. In the past stroboscopic discs have been used to tune musical instruments, but the arrangement and set up of the tuning mechanism has been such that it was necessary to use separate discs for each note and to rotate the discs at different speeds. It has been found that single notes cannot be tuned adequately or accurately with a single constant-speed disc. The conceptive idea of my invention lies in the use of two tones simultaneously and measurement of the frequency interval between them rather than in the measurement of the single tones.

Another feature of my invention is that I reduce the number of indicia on the single constant-speed disc so that they may be more readily distinguished and the number of indicia in the various stroboscopic rows may only range from 105 to 220 still allowing measurement of frequencies or frequency intervals in a range from 880 to 1661 cycles per second with a maximum error at any frequency of .0238 percent of the frequency of the musical note involved. I

My invention consists of four principal divisions. First, a means of converting the sound of the musical instrument to audio frequency electrical current or voltage if it is not already in that form. Second, a means of amplifying said audio frequency electrical current or voltage and of extracting from two such currents or voltages representing two musical tones played simultaneously, a single audio voltage representing by its frequency the arithmetic difference between the frequencies 'of said two musical tones. Third, one or more electrical lights, preferably but not necessarily of the gaseous type, such as neon, excited by said difference frequency voltage. Fourth, a stroboscopic disc rotated by a motor whose speed is controlled by the frequency of the primary power source, usually 60 cycles per second, said disc containing 12 or more concentric bands of marks or dots so located and spaced that each band of marks will appear to stand motionless when the rotational speed of the disc and the frequency of the exciting voltage for the lights are in predetermined relationship.

The sound vibrations of the instrument to be tuned may be converted to electrical impulses by any input means commonly used for that purpose, such as a microphone, or, as in the case of electronic musical instruments, the input means of my tuner may comprise Wires connected directly to the device since the vibration tones already exist in such instruments in the form of electrical impulses and no translation is required.

In the accompanying drawings Fig. 1 is a block diagram showing one arrangement of electronic circuits and associated devices for use in an apparatus embodying my invention.

Fig. 2 is a sectional view of the stroboscopic disc used in my invention. It will be understood that while I have shown a disc having linear apertures any type of strobo scopic discs may be used.

Fig; 1 illustrates the arrangement of the electronic circuitry required for the practice of my invention. It is the purpose of this circuitry to cause the diiference in frequency between two simultaneously sounded musical tones to flash one or more stroboscope-illumination lamps at a rate equal to two times the said frequency dif ference. Since the circuit illustrated in Fig. 1 is composed of elements of entirely standard types, each of which may in practice be designed as to configurations and component values in a number of different ways to accomplish the same results, no schematic diagram is necessary to make this part of my invention clear. Suitable configurations and component values may be designed by any person of normal skill in the art, using standard methods and circuits.

As hereinafter explained in detail, two musical tones are sounded simultaneously on the musical instrument to be tuned. These tones are, in the embodiment illustrated, of frequencies between 880 and about 1661 cycles per second, constituting the notes A to G-sharp in the upper range of the normal organ console.

Said tones are applied to the input of high-frequency amplifier 1, either by means of a microphone, or, if the instrument to be tuned is electronic, through direct connection of the audio-frequency output of said instrument to the input of said high-frequency amplifier 1.

High-frequency amplifier 1 is simply an audio-frequency voltage amplifier of standard design, the gain and frequency response of which is designed by ordinary methods to be adequate for the purpose. It may consist of a single stage but more probably will consist of two stages, perhaps a pentode followed by a triode, especially if a microphone is to be used. It is the purpose of the high-frequency amplifier to amplify the original two tones to a voltage level sufiicient to apply to the detector 2.

The function of detector 2 is to cause the original two musical tones to intermodulate or beat against each other so as to produce a beat frequency which is equal to the arithmetic difference between the frequencies of the two original tones. The detector is essentially a nonlinear mixer. It may be of any common type, for instance a simple diode rectifier such as is used in amplitude-modulation receivers for the detection of the audiofrequency intelligence carried on the radio-frequency signal. It may also be an amplifier stage designed with a deliberately nonlinear transfer characteristic, or a simple nonlinear resistor such as a Thyrite, or any other of the many circuits known in the art for this purpose.

The detector, as is true of any nonlinear element, causes the two tones passing through it to intermodulate. This intermodulation produces two major additional frequencies, one equal to the dilference between the two originals and the other equal to the sum of the two originals. It is the difference frequency which is desired. Therefore low-pass filter 3 is provided to substantially attenuate all frequencies emerging from the detector except for the said difference frequency. Since the difference frequencies in the embodiment discussed lie below 100 cycles per second while the original frequencies lie above 800 cycles per second, any of a large number of standard filters well known in the art will suffice to attenuate the original frequencies and the sum beat without affecting the difference frequencies to any appreciable extent.

At the output of the low-pass filter 3, therefore, there appears only a signal of a frequency equal to the arithmetic difference in cycles per second between the frequencies of the original two tones.

This difference frequency is amplified by low-frequency amplifier 4, which is a standard audio amplifier stage or stages, so designed as to amplify frequencies between 50 and at least 110 cycles per second. The gain and output of the low-frequency amplifier are so designed as to yield sufficient output to light lamps 5, 6 and 7.

Lamps 5, 6 and 7 are lamps of any type whose light output is not of substantially longer duration than the electrical signal which causes it. Typical lamps suitable for the purpose are neons and other lamps which operate by ionization of gas, since such lamps do not have the thermal delay of filamentary types and flashing on and oif of such lamps coincides in time with application of electric power.

I have in one embodiment used neon lamps of the type which contain two electrodes. It is characteristic of such lamps that when they are excited by an alternatingcurrent signal each electrode lights only on an alternation of one specific polarity. Thus, when a frequency of, for example, 70 cycles per second is applied to such a lamp, a separate flash of light is produced for each alternation of the exciting voltage, so that the number of light flashes per second is double the number of cycles per second of the exciting current and would be, in this illustration, 140 flashes per second.

Lamps 5, 6 and 7 are, in this embodiment of my invention, excited by alternating current. The coupling between the output of low-frequency amplifier 4 and the lamps is made by means of a transformer or a capacitor 4:: to avoid the transmission of DC. to the lamps.

As many lights are provided as may be desirable to illuminate the rotating stroboscopic disc described below. Such lighting may be from the front, side, or rear, as may appear most advantageous for operating and manufacturing purposes, without affecting the principle of operation.

The stroboscopic disc and the method of carrying on the tuning operation are the most essential features of my invention. The stroboscopic disc has always been recognized as a very desirable standard of frequency because it can be rotated by an inexpensive motor. However, previous workers have been forced to employ several discs rotating at several speeds to attain the required accuracy of frequency, because there is no simple relationship between the frequencies within one octave of the tempered scale.

The present invention takes a different approach to the tuning problem in order to simplify the tuning device and make it inexpensive. Instead of setting up frequency standards for the notes of the musical scale, it measures the difference or beat frequencies between the notes.

To do this, a stroboscopic disc such as that illustrated in Fig. 2 is provided. The disc 8 is attached directly to the shaft of an inexpensive synchronous motor (not shown) of the type used to operate electric household clocks; the speed of this motor is 60 revolutions per minute or one revolution per second in the embodiment used for illustration. The speed is rigidly controlled by the frequency of the power line and is subject to negligible error on account of mechanical loading since the loading imposed by the disc is negligible. Since almost all power companies control the frequency of the power lines to a very small fraction of one percent, the speed of the disc can be assumed to be one cycle per second within negligible tolerances.

If there isone mark at some point on the disc and the lights which illuminate the disc are excited by a current of a frequency of 0.5 cycle per second, the light will flash once per second, since it flashes on each alternation of the exciting current. The mark on the disc will then appear at the same point of revolution each time the light flashes. If there are 15 equally spaced marks on the disc and the light is excited by a frequency of 7.5 cycles per second, the marks will appear to stand still. The same will hold true when ever the light is excited by a frequency equal to one-half the number of equally spaced marks on the disc; and where there are 15 marks (and flashes) or more, persistence of vision will render the flicker invisible and the disc will appear to be steadily illuminated, with the marks standing still. The number of different corresponding lamp excitation frequencies and equally spaced dots is unlimited except by the practical question of mark spacing for a given disc size. The minimum increment between one excitation frequency and the next is one-half cycle per second, corresponding to 1 dot. It is thus apparent that the device as so far described may, with preparation of a suitable stroboscopic disc, be used to measure frequencies in increments of one-half cycle per second from a minimum of 7.5 cycles to a maximum limited by space on the disc and the speed of response of the light to its excitation.

The system as so far described might be used directly to measure musical tone frequencies to the nearest halfcycle. In a disc of reasonable size, however, it is impractical to have a sufficient number of dots or marks to provide good musical accuracy. For example, to indicate directly the frequency of C above middle C at a frequency of about 523.3 cycles per second, 1046 dots would be necessary, and an error of almost 0.3 cycle would occur. Both the number of dots and the error at such a frequency are excessive. The situation becomes worse, naturally, as higher frequencies are indicated, with respect to number of dots, and worse with respect to percentage inaccuracy as the frequency is lower.

In operating the present invention, the first step is to set the frequency of All0 cycles. This is done by playing the note A-llO cycles per second on the musical instrument to be tuned. This note is applied to the input of the electronic circuitry illustrated in Fig. 1. Since only a single frequency is present, it will pass through each of the stages l-4 without frequency change, since there is no second frequency with which it can beat in detector 2. It will excite the lamps 5, 6 and 7, causing them to flash at twice the frequency of or 220 flashes per second. A band 8A of 220 equally spaced dots is provided on the stroboscope disc of Fig. 2. When the frequency of the A on the musical instrument is adjusted the 220-dot band on the disc will appear to stand still when the exact frequency of 110 cycles per second is reached.

Once the A-110 cycles is tuned as described above, the successive higher-frequency As (220, 440, and 880 cycles per second) are tuned by the zero-beat method by ear in the manner well known and commonly practiced in the art.

The musical tone A at 880 cycles per second is now in tune. Next, notes A-880 and Ali-932.328 are played. These are fed into the amplifier, detected, and operate the lights. A band 8L of 105 dots on the disc is made to appear to stand still by tuning the musical tone Ail. There is then a difference of 105/2 or 52.5 cycles per second between the A and the Ail. Since the A was set at 880, the All must now be at 932.5 cycles per second, with an error of 0.172 cycle from its theoretical value of 932.328. Such an error is negligible and cannot be heard at this frequency. In fact, mathematical examination of the simplest kind will indicate that, as succeeding notes are tuned by playing the last tuned note and the note one semitone above, the maximum resolution of the stroboscopic disc is 0.5 cycle per second, which means that the maximum error at any frequency is 0.25 cycle. Each such error is multiplied by two in the next upper octave and by two again for each succeeding octave, but at no times does it constitute an audible error. In lower octaves, of course, the error in terms of cycles per second is divided by the inverse ratios of the frequencies of the notes themselves, in the same manner. (Such other octaves are tuned by car by the simple zero-beat method, my invention being used only to octave set temperament.)

In the manner just described I have succeeded in tuning one octave around and above 1,000 cycles per second with maximum error of 0.25 cycle. The remaining notes may be tuned by beating octaves with those already tuned in a well known manner. Electronic musical instruments of the frequency divider type will, of course, have been completely and accurately tuned by the tuning of the single octave provided for. Tuning this octave, between 880 and about 1161 cycles per second, requires a quantity of 12 concentric bands of dots or marks or other indicia 8A, 8B, 8C, 8D, 8E, 8F, 8G, 8H, SI, SI, SK, and SL, respectively, on the stroboscopic disc, ranging from 105 marks in the smallestdiameter band to 220 in the outer band.

By way of information and illustration I am setting forth a table of disc information in which the headings convey the following information:

Band refers to concentric bands of marks ordinarily numbered beginning with the innermost.

Tune denotes the note to be tuned.

Sound indicates the notes to be sounded simultaneously.

Ideal Beat and Ideal Light Frequency are those which would give ideal-frequency (Zero-error) tuning.

Marks indicates number of marks in the band.

Result Freq. is the frequency resulting from the tuning.

Error Freq. is the error in cycles per second.

Error, Percent is the percentage error.

Separ. is the number of degrees by which the marks parts are all well known in the art, but the combination and organization of the parts is claimed.

It will be understood, of course, that while specific examples have been given of the notes and frequencies to be measured as well as the type of circuit and parts which may be used, considerable variation of such particulars is possible Without changing the conceptive idea of my invention, which is the use of means to translate sound into audio frequency current and to measure by means of an electric circuit and a stroboscopic disc a frequency interval between respective notes.

I claim:

1. A device for tuning polyphonic instruments comprising input receiving means responsive simultaneously to first and second vibrations representing adjacent notes of a musical scale, said input means being operative to deliver first and second electrical signals having corresponding frequencies, mixing means including low-pass filter means and connected to said input means and operable to mix said signals to derive an actual difference frequency, illuminating means connected to said mixing and filter means and responsive to said difference frequency and operable to provide flashes of light synchronized with the alternations of said difference frequency, and stroboscopic means adjacent said illuminating means and having indicia representative of at least one desired difierence frequency between adjacent notes of said scale and said stroboscopic means being illuminated by said light flashes to indicate coincidence between said actual difference frequency and a corresponding desired dilference frequency.

2. In a device as set forth in claim 1, said stroboscopic means comprising a movable stroboscopic member; and drive means for moving said member at a constant rate, said member having a plurality of endless rows of light varying indicia, the number of such indicia in each row representing one of said desired difference frequencies between adjacent notes on said musical scale.

3. In a device as set forth in claim 2, the number of indicia in each of said plurality of rows being mutually related as integral multiples of the following numbers: 105, 111, 116, 125, 132, 140, 14-8, 157, 166, 176, 186.

4. In a device as set forth in claim 2, at least one additional row of indicia representative of the frequency of a note of said musical scale.

5. In a device as set forth in claim 1, said first and second sound vibration both lying within a frequency range between 880 and 1760 cycles per second.

References Cited in the file of this patent UNITED STATES PATENTS are separated on the disc for a reentrant band of equally 890,803 Severy et al June 16, 1908 spaced marks. 1,916,782 Crossley July 4, 1933 Ideal Ideal Result Ideal Error Error, Band Tune Sound Beat Fght Marks Freq. Freq. Freq. Percent Separ.

req.

# AA# 52. 328 104. 050 932. 500 932.328 0. 172 .0187 a. 429

0 BO 50. 153 110. 300 1010. 000 1040. 153 0. 153 153 s. 103

o# OC# 02. 730 125. 400 1108. 000 1108. 731 0. 231 0209 2. 880

D# DD# 70.008 140. 010 1244. 500 1244. 508 .008 00004 2. 571

E D#E 74. 010 148. 020 148 1318. 500 1318. 510 010 00000 2. 432

F E-F 78. 413 150. 820 157 1397. 000 1390. 913 087 .0002 2. 293

F# FF# 82. 978 105. 950 100 1480. 000 1479. 978 .022 0015 2. 109

G Fit-G 87. 982 175. 904 1508. 000 1507. 982 .018 0010 2. 045

G# GG# 93. 219 186. 438 1001. 000 1001. 219 0. 219 .0132 1. 935

It will be appreciated that I have provided a method 70 2,066,581 Severy Jan, 5, 1937 and a device whereby an electro optical instrument tunes 2,070,867 Severy Feb, 16, 1937 a musical instrument by measuring the frequencies of 2,086,892 Barton July 13, 1937 the intervals between notes in one or more stated ,191,203 Place et a1. Feb. 20, 1940 octaves. 2,221,523 Railsback Nov. 12, 1940 The individual sections of the circuits and the other 75 2,542,540 Kunz Feb. 20, 1951

Images