The Plate that talks
The plate that Talks
Electrostatic speakers generate sound in a way that has thrilled many hi-fi fans; here's how they do it.
LONG AWAITED and much discussed, electrostatic speakers are here -apparently to stay. Although available commercial models cover only the midrange and high frequencies of the audio spectrum, we may expect that sooner or later fullrange electrostatic reproducers will be on the market. Development work, here and abroad, is now under way on such devices. Recent models are, of course, hi-fi news -although their operation is based on tried-and-true electronic principles. If you take apart an electrostatic speaker, you have what is, essentially, an overgrown capacitor. One plate of this capacitor is stationary and can't move: the other is suspended elastically so that it can move. Feed an a.c. signal across these two plates and the movable one will vibrate, creating sound waves. That, in a nutshell, is how the electrostatic speaker works.
Two-way speaker system employing electrostatic tweeter (Pickering) which is used with conventional cone woofer (Bozak].
But, as in all things electronic, there's more to this story than the insides of a nutshell. The basic principle of electrostatic speakers was known years ago, but audio engineers chose the electromagnetic principle as the big hook on which to hang their speaker designs. There were many good reasons for such a choice. Among them were:
- cost of construction and operation
- over-all smoothness of response
- wide frequency band coverage
- output level; efficiency.
Many experts still feel that these advantages have not been seriously challenged by the electrostatic speaker and never will be. Others point to the highly rated performance of electrostatics at the high frequencies, emphasizing the importance of good reproduction in this area for the ultimate perception of clean overtones that characterizes real hi-fi.The Moving Plate
In preparation for hi-fi-dom's reception of electrostatic speakers, it's a good idea to understand just how these units operate. As stated above, the basic action is capacitive, depending on the buildup of an electrostatic field between two conductive plates. When the charges on the two plates are of the same polarity (both plus or both negative), the static field (or tension, or force) between the two plates tends to push them apart. When the charges are of opposite polarity (one positive and the other negative), the two plates are "attracted" and the force pulls inward between them. Varying the charge-as an audio signal voltage can-varies the force and creates a "push-pull" effect across the two plates.
Curved surface of Pickering "Isophase" electrostatic tweeter disperses highs uniformly into room.
To make a loudspeaker of two such plates, or panels, three things must be done. First, fix one panel so that it can't move. Second, suspend the other in an elastic mounting so that it can move a very short distance with respect to the first plate. Finally, feed across the two panels a signal voltage that represents speech or music (see figure 1).
The movable panel will be pushed and pulled with respect to the fixed panel with a force that is proportional to the signal voltage. The resultant vibrations will move the surrounding air to create sound. This arrangement makes possible a reproducer diaphragm that moves as an integral unit. The force of any point on the surface equals that at every other point, thus eliminating "breakup"-long a bugaboo of conventional speakers. Breakup occurs when some parts of the speaker diaphragm do not move in step with the voice coil.
At middle and high frequencies, for example, the flexible cone of a conventional speaker is pushed at one area only, near its attachment to the voice coil. Some distortion may result. In the better speakers, this difficulty has been controlled so that distortion is fairly low. Well-designed horn tweeters reduce it to a minimum. But, it appears that the electrostatic speaker has no such problem at all. This would indicate a high degree of purity in the propagation of high frequencies.
There were two big obstacles that stood in the way of electrostatic speakers clearing the high-frequency hurdles. One was the spacing of the two plates. The diaphragm. or moving plate, must be very close to the fixed plate for any usable electrostatic force to be developed. The closer together the two plates, the stronger the sound available. But the plates can't get close enough to touch each other during the movement of the diaphragm; this would, naturally, short out the audio signal and defeat the purpose of the whole setup.
In the past, experimenters attempted to solve this problem by stretching a thin metal foil tightly in front of a plate. Another method was to mount the foil on a layer of rubber. Neither method worked very well. The foil could not be made both lightweight and strong enough to do the job. Thus, the spacing problem became, effectively, a problem of finding and using the right material for the diaphragm.
Four panels, at various angles for dispersal of highs, characterize the Janszen electrostatic tweeter, a push-pull unit handling 1000 cpr and up.
The second problem involved a basic law of physics: the mechanical force between two charges varies inversely as the square of the distance between them. For example, if two charges attract each other with a certain force when they are one inch apart, the same two charges will attract each other with only one-fourth that force when they are two inches apart. This square-law relationship between distance and force means that when the diaphragm moves well away from the fixed plate the resultant force will not be in strict proportion to the applied signal voltage. Harmonic and intermodulation distortion will result: the speaker will not be linear in response.
The "Push-Pull" Speaker
To solve the distance-force problem, both American manufacturers of electrostatic speakers (Pickering and Janszen) use a push-pull system which is said to be completely linear (see figure 2). The force applied to the diaphragm is proportional to the applied signal voltage no matter where the diaphragm is at any time during its vibrations.
The push-pull system uses two fixed electrodes; the moving diaphragm is suspended between them. A fixed d.c. voltage (the "polarizing" or "bias" voltage) is applied to the speaker by means of the transformer center-tap, creating an initial electrostatic force between the diaphragm and each electrode. With no a.c. signal, the force is balanced; the diaphragm is pulled equally in both directions and so stays in the middle. No sound is generated.
When a signal is applied to the transformer, the voltage varies. In effect, the polarizing d.c. voltage is "modulated" by the a.c.signal. This action pulls the diaphragm first toward one electrode, then toward the other. Resultant vibrations generate the sound.
Diaphragm and Suspension
The problem of choosing a suitable material for the diaphragm is solved in both the Pickering and Janszen units by the use of a very thin, lightweight, strong plastic which is made into an electrode by depositing on it a microscopic Iayer of metal. The suspension of the diaphragm, however, is quite different in each make.
In the Pickering speaker (see figure 2), the diaphragm is not stretched and held at its edges. Instead the "inert diaphragm" method is used, with the plastic supported by a number of small spring elements across its surface, between it and the two outer electrodes. This entire "sandwich" (consisting of the outer electrodes and the diaphragm) is curved in the horizontal plane, which helps in the even distribution of highs throughout the listening area. In addition, the outer electrodes have numerous openings to allow both the front and the back waves to emerge from the unit.
Two models of the Pickering speaker are currently sold: the larger (2' by 3') is designed to cover the frequency range from 400 cps up. The smaller (12" by 18") handles the range from 1000 cps up. Obviously, the lower the frequency range handled by the electrostatic method, the larger--and costlier-must be the unit.
Both units are free-standing and selfcontained. They can be placed anywhere near the woofer and its enclosure. Each speaker is used with an adapter unit which serves as a crossover network, supplies the polarizing voltage, and matches the speaker to the amplifier. The rig can be used with any standard hi-fi amplifier. A typical hookup is shown in the photo above of the Pickering.
The Janszen speaker consists of a series of four flat panels, each about six inches square. The plastic diaphragm is stretched across the frame of each panel. On each side of the diaphragm, in each panel, is a network of parallel wires which forms the outer electrodes of the push-pull circuit. This unit is intented to cover the frequency range from 1000 cps and up. It is furnished with a high-pass filter for crossover, and a polarizing voltage supply. Like the Pickering, it is free-standing and self-contained. (For further inlormation on these units, contact Pickering and Co., Inc., Oceanside, N. Y., and Janszen Laboratory, Inc., 69 Harvey St., Cambridge, Mass.)
In addition to the push-pull speakers, there are a few small, single-ended electrostatics such as the Isophon (Arnold Ceramics, Inc., 1 East 57 St., New York, N. Y.), and the Lorenz (Kingdom Products, Ltd., 23 Park Place. New York 7, N. Y.).
Examples of single-ended electrostatic tweeters are the Arnhold "lsophon" (above) and the Lorenz SKL-100. Both units are small, fairly inexpensive, and simple to connect. lsophon handles range from 7000 cycles and up: Lorenz, 5000 cycles and up.
Designed to handle only the very upper range (the Isophon takes off at 7000 cps, the Lorenz at 5000 cps), these tweeters serve as "top" reproducers in three-way systems or as high-frequency aids to widerange cone speakers where relatively high crossover points can be used. The distortion inherent in their single-ended construction is minimized by permitting the diaphragm to move only thousandths of an inch, or less. At that, such microscopic motion is enough to produce sizable sound levels at high frequencies.
How Do They Sound?
Most listeners agree that properly connected and matched to suitable woofers, electrostatic speakers make excellent mid-range and tweeter units. Of course, in the last analysis every listener must decide for himself. Listen, if you can, to "A-B" tests in which the highs are played first through a horn or cone tweeter and then through an electrostatic unit.
Advocates of the electrostatic devices urge you to listen for two main things. First, the over-all response is said to be very smooth with no audible peaks and no "break-up." Transient response is claimed to be excellent. Intermodulation and harmonic distortion have been measured at less than 1% at full output.
Second, the point-source effect, or "sound coming from a hole in a box" is eliminated. The wide radiating surface spreads the sound over a large area so that it "just seems to be there." While the highs may not come at you with the impact of a pile-driver, they are all present and perhaps in a more natural-sounding way.
What About Bass?
Practical use of electrostatic speakers, at least so far, involves a two-way speaker system in which the bass is handled by a cone woofer while the mid-range and highs are sounded by the electrostatic unit. The crossover point depends on the particular electrostatic speaker used; some take over at about 500 cps, others at 1000 cps. In any case, the electrostatic reproducer must be 'fairly well balanced with whatever woofer is used. Power ratings and relative efficiencies must be watched lest the sound output is unbalanced in favor of highs or lows.
One of the big problems faced in developing electrostatic speakers for the full audio range, including bass, is the spacing of the electrodes. As frequency is lowered, the diaphragm has to move farther to maintain balanced sound output.. At 100cps, for example, the diaphragm must move four times as far as it does at 20(cps. But the two outer plates must remain very closely spaced for optimum/operation. This spacing gives the suspended diaphragm very little room in which to vibrate. At the middle and high frequencies, where t h e motion is in thousandths of an inch, it is no problem. But the limits are reached when the diaphragm is forced into the low bass regions.
To an extent, this difficulty is overcome by increasing the size of the diaphragm. The bigger the diaphragm area, the betted is its bass performance, and the less it had to move to put out the bass notes. Here though, lies another problem involving ultimate size and cost and general practicability of a bass-reproducing electrostatic speaker. To Dat, this problem has been solved in theory only. Reports and rumors tell of experimental models both here and abroad. What will actually be produced remains to be seen.
Popular Electronics, September 1956