by Jacob Turner [With the assistance of Douglas Elliott, Koss Corp., Milwaukee, WI.]
SINCE THE LATTER half of the nineteenth century (circa 1871) the reproduction of sound through electrostatic transducers has stirred the creative vision of professional engineers and idle dreamers alike. It is an interesting fact of history that no other single device in the audio equipment hope chest has enjoyed such an extensive and prolonged courtship between engineer and audiophile as the electrostatic transducer. Early attempts to embody this means of sound reproduction were only marginally successful partially because of inadequate design but primarily because of the lack of suitable materials and processes.
What is the glamour of the electrostatic principle that gained it such extended, devoted attention? Why has the electrostatic transducer remained the standard of excellence by which other acoustic devices are so often measured? The answer to these questions lies in at least three areas, which will be discussed in the following order:
1. Some peculiarities of the hearing process;
2. The nature of the acoustic medium, air, and
3. The operational features of electrostatic acoustic devices as related to the above and to dynamic acoustic transducers.
The recent increase of activity in the highly elusive area of psycho acoustics promises to contribute significantly to a more profound grasp of the complexities of man's perception of his sound environment. Several recent studies have been carried out concerning the sensitivity of normal adults' ears to different levels of harmonic distortion. The results suggest that relatively high levels of harmonic distortion (odd and even order) are imperceptible in the presence of normal musical program, while quite small changes in amplitude and phasing are readily ascertained. Amplitude changes were described as altering the tonal quality of the program, while phase displacement between two major frequency bands, e.g. bass and treble, of no more than 5° degraded the clarity and definition of musical transients and upset the homogeneity of the stereo image.
Other studies have pointed out that the inherent transient nature of musical and speech sounds dictates a high level of transient fidelity as a prerequisite of high quality acoustical transducers.
The significance of these observations insofar as electrostatic transducers are concerned will be pursued a little later. (See Bibliography. -Ed.) Another vital link in the chain is air, which has the following major characteristics of behavior that are germane to our topic.
Air is highly compressible, that is, the amount of air pressure (number of air molecules) in a given space can be increased or decreased beyond its normal condition. Air, then, can be said to be like a spring, a means for storing energy; a compliance.
Air also has weight or mass. Ten pounds of air are just as heavy as ten pounds of potatoes. Air is, therefore, like an inertance, which opposes an action or force; an inductance.
Air can also be placed in a condition of motion or vibration, so that when randomly excited, air molecules consume power by generating heat. Air can then be said to be like a resistance.
The combination of these particular properties of air can be labeled its acoustic impedance or radiation resistance. This acoustic impedance is normally very low, although at high audio frequencies it is considerably greater than at low audio frequencies.
In order to insure that the transfer of diaphragm or cone motion to air motion occurs with the greatest efficiency, it is necessary that the total mechanical impedance of the device be as close to the acoustic impedance of the body of air it is exciting over as much of the audio range as possible.
To relate the preceding discussion to the topic of the electrostatic transducers, it will be necessary to outline the operational features of the push-pull electrostatic device. The previous points will be related at the same time to the operational features of the dynamic transducer.
As illustrated in Fig. 1, the electrostatic transducer is composed of a thin membrane (diaphragm) made of Mylar that is stretched and contained between two acoustically open plates.
The two plates are connected to either end of a coupling transformer which provides the high voltage audio signal.
The diaphragm is connected to a high voltage, low current bias supply, which provides an electrostatic charge that becomes trapped in the diaphragm. In recent years a method called "electretification" has been developed whereby the bias charge is permanently embedded in the diaphragm material, so that the diaphragm is self-energized without an external source of bias voltage. The net result is the same in both cases. The two plates provide an electric field that is the voltage equivalent of the audio signal. In the presence of an audio signal, the electric field exerts forces on the electrostatic charge that is trapped in the diaphragm.
These forces are transferred to the diaphragm, causing the diaphragm to move in synchronization with these forces.
[left] Fig. 1--Electrostatic operating principle. [right] Fig. 2--Dynamic
driver operating principle.
By contrast, Fig. 2 will illustrate make-up of a dynamic driver, which consists of a frame housing a magnet, and a voice coil attached to the apex of a cone which is suspended at its edge by a flexible cloth or other material. The voice coil is positioned in the magnetic field of the magnet structure, and is set into motion in synchronization with an audio signal that causes current flow through the coil. As the coil is set into motion by this signal, it in turn sets the cone into motion.
Although both units achieve air excitation through diaphragm or cone motion, the manner in which this is done involves radically different techniques and results. The electrostatic device employs the use of a moving member for all its operating frequencies that is usually only 0.0004 in. thick and weighs only as much as a body of air 7 mm thick whose boundaries are equal to those of the moving diaphragm. The electric field, which acts to make the diaphragm move, exerts its actuating force uniformly over essentially the entire area of the diaphragm.
A diaphragm of such extreme lightness, in combination with an actuating force that is uniformly distributed over the entire surface of the diaphragm, results in a transducer whose transient response closely duplicates the electrical input.
The net result is a diaphragm motion that is a very good replica of the electrical forces acting upon it, with all sections of the diaphragm surface moving with highly accurate phase and amplitude linearity throughout its entire range of travel, at all frequencies within its area of operation.
The forces acting to move the dynamic transducer's cone, however, produce different results. The application of the driving force only to the apex of the cone necessitates a sufficiently stiff cone to prevent buckling and deformation of the cone structure.
Such a stiff cone normally has considerable mass, which degrades its efficiency, its transient response capabilities, and its high frequency performance. In addition, the forces applied at the apex do not act uniformly over the surface of the cone, causing the cone to "break-up" into an infinite variety of vibrational modes, only one of which is truly representative of the original signal. This mode of operation produces amplitude and phase non linearity often of considerable magnitude, and these tend to increase as the cone is driven to greater excursions.
Obviously the discussion of dynamic driver operation relates quite strongly to the previous discussion concerning the unusual sensitivity of the human ear to the problems of transient response, amplitude linearity, and phase linearity. The basic conclusion is that an electrostatic unit behaves with better composure in all of the above areas.
The second major area of distinction involving electrostatic transducers deals with the considerable problem of coupling to the air with reasonable efficiency over the entire audio band.
The electrostatic unit, because of its extremely low mass diaphragm and the uniform distribution of the driving forces over the entire diaphragm surface, is inherently a unit with low mechanical impedance at all frequencies. As such, the coupling problem at low frequencies (where the problem is greatest) for electrostatic units is considerably less than for dynamic units, which are encumbered by a high mechanical impedance. The result of these conditions is that the electrostatic unit performs quite well down to its frequency limits and within its maximum excursibility with equal fidelity at all drive levels.
The dynamic cone unit, because of its poorer coupling, must be driven harder to produce satisfactory excitation of the air at low frequencies, and usually encounters a number of problems involving cone break-up, non-linear motion of the voice coil due to loss of magnetic coupling in the gap, suspension non-linearities, etc. In all fairness, it must be said that the performance level of today's highly popular dynamic acoustic transducer is incredibly good given the economic and operational constraints of that type of unit.
On the other hand, the superiority of the electrostatic principle has been demonstrated by the great acceptance of the increasing number of electrostatic headphones which have already emerged in the market. In addition, of course, several electrostatic loudspeaker, products are highly regarded by the audiophile community. Koss Corp. will introduce early in 1974 a new line of electrostatic speakers, offering wide-range performance and small size, to complement the ESP-9 and ESP-6 electrostatic headphones already in the line.
We feel that this could be the year in which the electrostatic dream will finally be fulfilled.
Bibliography:
- Robert Carver, Results of an Informal Test Project on the Audibility of Amplifier Distortion, Stereo Review, May, 1973.
- Erik R. Madsen & Villy Hansen, Threshold of Phase Detection by Hearing, paper presented at the 44th Audio Engineering Society Convention, Rotterdam, Feb. 20-22, 1973.
- R. J. Matthys, Telstar-Shaped Electrostatic Speakers, Audio, May, June, 1964.
- Harry F. Olson, The Psychology of Sound Reproduction, AUDIO, June, 1972.
- Transient Response as it Refers to Musicology, Audio & Video News, Jan., 1973.
(adapted from Audio magazine, May 1974)
Also see:
Electrostatic Loudspeakers (Mar. 1971)
Koss electrostatic speaker system (May 1977)
How to Add WOOFER to an Electrostatic Speaker (Mar. 1970)
= = = =