Archives - 1956

Archives Main | 1956

 

Audio Craft Magazine

June 1956
MUSIC FROM ELECTRONS By Robert Moog

ON the Tuesday evening of January 31, 1928, a capacity crowd filled the New York Metropolitan Opera House. The occasion was nothing as conventional as the performance of an opera, however. The audience braved the winter evening to witness what was then a remarkable novelty. Leon Theremin, a Russian physicist, was to perform on a musical instrument which he had developed and named after himself. The instrument was made of radio components and was operated by electricity. Most important, however, was the fact that Professor Theremin would play his instrument without touching it.

The printed programs which the audience held set the proper mood: Music from the Ether, produced by a pair of hands reaching out from a raging inferno. Rudolf Wurlirzer introduced Professor Theremin and explained to the audience that Theremin would play his instrument by varying the position of his hands in the space around it. The Professor then proceeded to play. Although Theremin was a creditable musician, the audience was undoubtedly more impressed by the mysterious undulations of his hands than by the musical quality of his Performance.

To show that the Theremin was a worthy musical instrument as well as an intriguing novelty required the devoted efforts of several fine musicians. Fortunately, a small body of artists extended the frontier of theremin music so that all but the most conservative critics granted the theremin a place in the realm of "serious" music. The best known of these pioneers are Lucy Bigelow Rosen, Clara Rockmore, and Elena Moneak, who gave many public concerts in the 1930's, and Jeno von Takacs, who composed works for the theremin.

As a commercial venture, the manufacture of theremins was not a notable
success. Shortly after Theremin's first concert the Radio Corporation of America bought his patent and began production of the instruments. Introduced at a time when few people were willing to invest in anything, the RCA theremins did not sell very well. After making only a few hundred instruments RCA discontinued production. With no instruments being produced, theremin music seemed to be on its way to oblivion. Only recently has there been a promising renewal of interest.
Two years ago we completed the design of a theremin which is played in the same manner as the RCA instrument, but which utilizes more modern circuits and
components. In addition, it incorporates some features which were not present in the original. This instrument, the Model 350, will be described and compared to
other musical instruments.
The Model 351 Theremin is an electronic device played by movement of the performer's hands in the space stir rounding it. Pitch of the tone is determined by the distance between the performer's right hand and the pitch antenna, which is a long slender rod. Volume of the tone is determined by the distance between the performer's left hand and the volume antenna, a flat metal plate. Two switches on the front panel enable the performer to select the tone quality or timbre suitable to the music being performed. The entire instrument, except for the amplifier, is housed in a wooden cabinet 20 in. long 11 in. deep, and 6 in. high. It weighs less than 20 lbs.

Pitch Control

Musical instruments are traditionally grouped in categories according to the way in which the tone is produced: string, brass, wood wind, and percussion
The Theremin belongs in none of these categories, since its tone is produced by
electronic circuits. This in itself is not enough to characterize it, however; we must also state that the pitch is controlled directly by the position of the performer's hands. If we were to divide musical instruments according to the way in which the pitch is controlled we would have three categories:

1) Instruments which have a separate key or valve for every note (a piano or clarinet, for instance).

2) Instruments for which pitch is controlled partly by keys, and partly by the player himself (trumpet, French horn).


3) Instruments which have no keys, and for which pitch is controlled by the position of the performer's hands (stringed instruments, theremin). This list is arranged in order of increasing flexibility in pitch control. On the piano, only those pitches for which there are keys can be produced. With the theremin, however, any pitch can be produced. There are several reasons why this is desirable. First, it is often desired to go from one note to another in a glissando, or glide. Obviously, this cannot be done smoothly on a piano. Second, in many systems of harmony the intervals between notes are different from those of the traditional tempered scale. Even in playing classical music, departure from the tempered scale is often made. For instance, a good violinist may differentiate between G-flat and F-sharp but, on the piano, which is tuned to the tempered scale, G-flat and F-sharp are the same note.

So much for the musical aspect of pitch control for the theremin. How is this continuous pitch control achieved? The pitch circuit takes advantage of the fact that the hand is a conductor of electricity. Its connection to the rest of

the body effectively grounds it. Therefore, the hand can be used as a grounded
plate of a capacitor. If the hand is moved with relation to another electrical conductor, we have a variable capacitor. It is this variable hand capacitance which is used to control the pitch of the theremin.

Of course, the hand capacitance is very small only a few micromicrofarads. It could not be used to tune an audio oscillator directly. A special type of pitch generator, called a beat frequency oscillator, is used in place of a conventional audio oscillator. The beat-frequency oscillator in the Model 351 theremin consists of two radio-frequency oscillators operating at frequencies very close together. The oscillator outputs are fed into a mixer circuit which effectively subtracts one frequency from the other. If the difference or beat frequency lies in the audio range, the mixer will deliver an audio output. For instance, if one RF oscillator is operating at 200 Kc and the other is operating at 599 Kc, the output of the mixer will be 1,000 cps. A small percentage change in frequency in one of the oscillators will result in a proportionally larger change in the audio output. Thus, with one of the RF oscillators operating at 200 Kc, the entire audio spectrum can be covered by changing the other oscillator frequency only 10%.

The beat-frequency oscillator which generates the pitch is composed of V1, V2, the triode section of V3, and their associated components. The oscillator coil (T1) of the variable oscillator V1 is designed to effect a relatively large change in the frequency of oscillation for a small change in capacitance of the pitch antenna caused by variation of hand capacitance. The fixed oscillator V2 is identical with the variable oscillator, except for the absence of a pitch antenna. The RF signals from the two oscillators are fed through mixing transformer T3 into a mixer, which is the triode section of V3. The output of V3 is passed through and RC filter composed of C9, Rio, and C11, which removes the RF components and allows only the audio signal to pass.

At that point the signal has very little harmonic content. This might appear at first to be desirable. Musicians often call a pleasing tone a "pure" tone, but it is usually far from pure in the sense of being free from overtones. For instance, the fundamental component of a violin tone is only a small part of the total. The remainder of the tone consists of harmonics, or overtones, whose frequencies are integral multiples of the fundamental frequency. These harmonics are nor perceived by the listener as discrete tones but instead give the fundamental tone an ear-pleasing timbre.

Harmonic Generation

In all conventional musical instruments, the tone source, be it a string, a reed, or the player's lips, generates all the harmonics of the fundamental tone. Before the sound is released into the air, it is transmitted through the body of the instrument, which attenuates some harmonics and allows others to pass. Thus it is the body of the instrument which determines, for the most part, its timbre. In the flute, harmonics are attenuated sharply, giving the tone a mellow quality. All harmonics are allowed to pass in the violin, although some are attenuated slightly. That is why the violin has a rich pleasing tone. The oboe body is highly resonant, reinforcing a narrow band of harmonics and attenuating the rest. The result is a sharp, nasal quality which makes the oboe tone easy to identify.
In the Model 350, the audio signs passed through special circuits which introduce the desired harmonics, and then through attenuating filters which are electrical analogs of the body transmission responses of conventional musical instruments. The signal, as it enters is shown in Fig. 5A. V4 and its associated circuitry act as a clipper, so that signal emerges as a square wave. A square wave contains only odd harmonics, however. If you have ever heard a square-wave test on an amplifier, you know that a square-wave tone is hollow and woody, like that of a clarinet. While the tone may be pleasing, it is too distinctive to be the basis for more than one timbre. The next circuit, consisting of the left section of V5 and its associated components, forms a wide pulse from the square-wave input. This pulse which contains the fundamental tone and all its harmonics, is fed into two filters. One is an RC filter which gives the signal a string-like quality. The other is a resonant filter involving a phase-shift amplifier which gives the signal a sharp, horn-like quality. The outputs of these two filters, together with the output of a filter fed by square waves, and a signal taken directly from the beat-frequency oscillator, are all connected to switch S3. The performer can select the timbre which he desires simply by setting this switch.

In addition to being able to select one of four timbres, the performer can also select one of three overtones. Note that the mixing transformer T3 has two secondaries. The upper secondary is broadly tuned to the fundamental frequency of the RF oscillators, and feeds the triode section of V3. The lower secondary is tuned by one of four capacitors which can be selected by switch S2. These capacitors are adjusted so that they tune the lower secondary to one of the harmonics of the RF oscillators. The lower secondary feeds its own mixing circuit composed of a diode (pin 1 of V3) and the RC filter R32, R33, C17, and C18. The output of this mixing circuit is the audio harmonic corresponding to the RF harmonic to which the lower secondary of the mixing transformer has been tuned. For instance, if the fixed RF oscillator is operating at 200 Kc and the variable RF oscillator is operating at 199 Kc, the output of the triode section of V3 will be 1 Kc. If the lower secondary of T3 is tuned (by one of the condensers connected to S2) to a frequency of 6oo Kc, it will transmit the third harmonics of the fixed and variable RF oscillators, which are 6oo and 597 Kc respectively. When these are mixed, the resultant will be 3 Kc, which is the third harmonic of the 1Kc audio fundamental. With S2 the performer can select either the fundamental, second harmonic (octave), third harmonic (quint), or fourth harmonic (superoctave).
Once the proper harmonics have been added, the signal can be amplified and fed into a loudspeaker. The Model 355 contains one stage of amplification, which is part of the volume-control circuit.

Volume Control

Hand capacitance is used to control the volume also. A variable RF oscillator(the triode section of VS), similar in design to the variable-pitch oscillator, is constructed so that changes in capacitance between the left hand and the volume antenna change the frequency of oscillation. A signal from this oscillator is fed into a narrow-band RF amplifier (the pentode section of V8) so that, as the frequency of oscillation changes, the amplitude of the output of the RF amplifier also changes. This output is rectified, and the resultant potential is applied to the control grid of the control amplifier V6a. When this bias is made negative, the gain of V6a decreases; when the bias is made positive, the gain of V6a increases. By varying the capacitance of the volume antenna with the left hand, the performer is able to make the tone loud or soft, or to silence it altogether. From the control amplifier, the signal is fed into a cathode follower, and then to an external amplifier and loudspeaker.
The power supply is conventional in design. A glow voltage regulator tube is used to supply constant plate voltage to the oscillators, so that variations on line voltage do not cause variations in pitch or volume while the performer is playing.
Theoretically, any type of music can be played on the theremin. Both pitch
and volume-control circuits respond instantly to changes in hand capacitance. Melodies can be played as fast as the hands can move. In practice, however the theremin has proved itself best adapted to slow melodies which give the hands time to locate themselves accurately on each note. Slow melodies in addition, lend themselves to the use of vibrato. The thereminist's greatest asset is his ability to impart a beautiful vibrato to the tone merely by moving his right hand back and forth through a small distance.

In this age of network broadcasting, tape and disc records, and high fidelity sound, music has enriched the lives of millions of people. The theremin was the first electronic instrument that generated, rather than reproduced, music. As such it has pioneered the field of electronic musical instruments. Because of his knowledge of electronics and appreciation of music, the high fidelity enthusiast will find the study of electronic musical instruments an interesting and worthwhile pursuit.