THE HISTORY OF THE DEVELOPMENT OF BEVERIDGE SPEAKERS

by
Rick Beveridge

My father, Harold Beveridge, was a highly skilled engineer with a deep love for music. He studied electrical engineering at McGill University in Canada. From 1947 to 1954, he worked at Raytheon, in Waltham, Massachusetts. He was involved with designing vacuum tubes and the early commercial development of Radar.

During those years, he always bought season tickets for himself and my mother to the Boston Philharmonic and often attended the Boston Pops. He almost never missed a concert of either. When I was old enough, about six, he would take me when my mother couldn't join him. I enjoyed many concerts with him. He enjoyed them so much that it became his passion to reproduce that sound at home.

By 1951, he had already decided that the inherent limitations of a dynamic speaker were so severe that, even with tremendous gains in technology, they would never produce a satisfactory transient response. Several people had been experimenting with electrostatic designs and had written about their work. My Father read their reports and decided to look into the potential he saw in electrostatic devices.

His first effort was a single-sided electrostatic transducer, or panel. It was made from a rectangular piece of 1/4 inch gray slate. The plate, which was 12 inches by 16 inches in size, had over one thousand 3/16 inch holes drilled through it. The side away from the Mylar membrane was painted with a conductive silver paint, spreading the charge over the entire surface.

My Father would put the transducer on a table with the membrane side up, sprinkle table salt on it and photograph the patterns which would form at different frequencies, amplitudes, and membrane tensions. He would also put a piece of white paper on top of the membrane and sprinkle it with iron filings so that he could observe the electro-magnetic field patterns.

I was often invited to watch. It was fascinating and good fun for both of us. My Father spent quite a bit of time explaining to me how different wave forms behaved and interacted. By then, I was probably seven or eight years old. He explained concepts such as sine waves, resonant oscillations and nulls, and intermodulation distortion. He studied every detail of these fledgling transducers, and through watching him I too learned a great deal both about sound and about experimental methodology.

By 1953, we were listening to music on this electrostatic transducer. It was placed vertically in a frame and sounded pretty good. We heard highs that regular speakers couldn't come close to producing. In fact, that was part of the problem. Many speakers of this era sounded like the public address system in the movie version of M.A.S.H. My Father's "speaker", in contrast, accentuated the highs and diminished the lows. So, compared to the standards of the day, it sounded far too "bright" or "tinny".

Also, the back wave interfered, producing peaks and nulls at certain frequencies. At least a half-dozen different forms of baffles and reflectors were tried in an effort to diminish these effects and, if possible, use the back wave to augment the sound. None were very successful.

Another problem arose when the speaker was driven at too high an amplitude; the membrane would get too close to the electrode and stick to it, stopping the sound. At that point, we would have to turn off the power and wait ten or twenty seconds until the polarizing voltage had dropped enough to let the membrane "peel" itself away from the electrode. Then we could turn the equipment back on again, this time being careful not to drive it quite so hard.

Reflective Dispersal

By 1957, stereo was coming out, and my father decided to build two new speakers. He had also decided to move to a two-sided transducer, increasing both accuracy and sound pressure level (SPL). This meant, however, that he would now need four electrodes instead of just the one.

Drilling all those holes in a sheet of slate was definitely out of the question. In an attempt to replicate certain electrical properties of the slate, he experimented with mixing several types of carbon powder and barium titanate into Epoxy resin. After some experimentation, he came up with a working mixture and we cast four electrodes.

By 1958, I was thirteen and helping to make these new electrodes. While my father was back east on a business trip, I followed his recipe and cast four new electrodes. There were many factors we hadn't learned about yet, and for reasons we learned about much later, these electrodes were much more conductive than the ones my father had made before. He did, however, make them into transducers and found that they had some very interesting properties. The "sticking" of the membrane to the electrode no longer occurred and we could drive each transducer much harder than before.

We used a pair of 1/4 inch glass plates to mold these electrodes. Each plate was a little over one by two feet. We put a thin coating of grease onto one side of each plate, then spread a sheet of aluminum foil onto one of the plates. The plates were placed with their greased sides (one with aluminum foil) facing each other. The plates were held 1/4 inch apart by three lightly-greased wooden sticks. This was all clamped together and stood upright with the open side up. The epoxy mix was poured into the cavity, then allowed to cure.

This produced a one by two foot, 1/4 inch slab of epoxy composite with aluminum foil on one side and a very flat surface on the other. Using the table saw, we cut a large number of parallel 1/8 inch slots, leaving about 3/16 inch of material between adjacent slots. We then poured a non-conductive epoxy rim around the entire border, finishing the electrode. These electrodes made fairly good transducers.

To build the new speakers, My Father made two cabinets. Each one measured about 36 inches wide by 24 inches deep by 30 inches tall and had a rectangular opening in the top. A transducer measuring 12 by 24 inches was placed horizontally below this opening, with a grill cloth above it.

To disperse the higher frequencies into the room, my father created a pair of acoustic "reflectors". He made a wooden bowl, about one foot tall and eighteen inches in diameter. The sides dropped, following a parabolic curve, to a 6 inch diameter base. The bowl was then cut in half vertically, forming a pair of half-bowl reflectors. The reflectors were fastened in place on the top of the cabinets, at the back side of the transducer openings, with their curved sides facing towards the room.

The reflectors protruded out and over part of the transducer, causing some of the high frequency sound beaming up from the transducers to be reflected out into the room. This brought the ratio between the highs and lows to a more acceptable level and did, in fact, sound excellent.

The back (bottom) wave from the transducer, or at least the lower frequencies from it, came out the bottom of the cabinet, which was open. By going to larger transducers, of a design which produced better bass, and by only reflecting part of the highs into the room, my father had greatly improved the frequency balance. The "tinny" sound, or "over-brightness", was mostly gone.

Two of these speakers were built by 1959, giving us a very good sounding stereo sound system. The cabinets looked good, but took up a substantial amount of floor space. Also, the artificial flowers my mother had put in each "bowl" for decoration had to be removed after the first time that a cleaning lady watered them, drowning the transducers!

The Acoustic Lens

In 1965, my father was more excited than I have ever seen him about anything, before or since. He had conceived of the acoustic lens and the full-range cylindrical wave front. These innovations were, and continue to be, the unique hallmark of Beveridge systems. Providing comfortable levels of crystal-clear sound throughout the room, they offer a unique audio experience for the listener.

The acoustic lens solved several problems and solved them beautifully. Because the entire lens is driven by a single, continuous driver, the sound retains complete phase coherency. All frequencies of interest, including the "beamy" upper frequency ranges, are formed into a six foot tall, 180 degree, vertical, cylindrical wavefront. Everywhere the lows go, the highs go, too. He had achieved truly uniform dispersion over the entire frequency range!

In addition, my father made an important decision about cabinet design. He could find no way to allow the back wave into the room without seriously adulterating the sound in one way or another. So, from then on, he never allowed it into the room at all.

My Father's lens and cabinet combination accomplishes a tremendous amount. The lens throat, combined with the "empty" space in the cabinet, creates a Helmholtz resonator in the 40 to 50 Hertz range. This works very well, extending the bass capability of the speaker.

So, we built two cabinets, each three feet wide, two feet deep, and six feet tall. Each cabinet held a continuous six-foot lens and transducer combination. We had also managed to cast twelve new electrodes in a new mold that my father had made. Those made six new transducers, enough for the first pair.

We later called this pair our model 1's. Only one pair was ever made. Eight years later, when we decided to go into production with this concept, we honored this first pair by calling our first production speakers our model 2's. The pair of 1's was sold to a friend in Santa Barbara who had a huge house overlooking the ocean. The pictures taken of our first model 2's (the white lacquered ones shown in our first brochure) were taken at his house.

What my father had created was truly novel: a high performance, phase coherent, full range, full height, full width, cylindrical wave front, line source speaker with incredible transient response. A series of patents attests to the novelty of his designs.

The new design had the added benefit that it was a true line source, as opposed to a point source. Consequently, sound pressure level drops off as one over the distance from the speaker to the listener. In contrast, a conventional cone speaker is a point source and the sound pressure level drops off as 1 over the distance squared.

Because these new speakers were true line sources with a 180 degree dispersion patterns, their placement in our home was novel. Our living room at that time was about twenty feet wide by fifty feet long. There was a large circular fireplace on the centerline and about twenty feet from one end. Each speaker was placed with its back to a long wall, facing the fireplace and the other speaker. The highs, mid-range, and lows were dispersed uniformly throughout the entire room. There were no more "bright" and "dull" spots.

They manner in which these new speaker introduced sound into the room was astonishing. As one moved about the room, the relative change in the volume from each speaker was minor. We found that we could set the volume for an acceptable listening level, go right up to either speaker, put one ear right up against it, and still hear the other speaker with the other ear.

Commercial Development

In 1972, Akio Morita, the founder of Sony, spent several days at our house in Santa Barbara. He was very impressed with the sound made by my father's speakers. He wanted to purchase all of the rights to build them and made my father an offer to do so. Unfortunately, this offer precluded any further involvement by my father and was unacceptable to him because of that. My father had some other ideas and wanted to remain involved with development.

I suggested to my father that I could help and that we could build speakers ourselves. He was retired by then and really didn't want to start a manufacturing company, but my offer of a year's free labor was an significant enticement and we got started.

During that year, Don McFarland, a close friend and talented designer, helped us reshape the cabinet. Because the cabinets were so large, we couldn't employ the usual brute-force, high density particle-board style of cabinet design: the result would be impossibly heavy.

So we tested a thin-walled cabinet at every possible frequency, making notes of all the resonant locations. These places were strategically reinforced, adding little to the overall weight. The result was a remarkably light cabinet which was remarkably sound-dead. It was, however, not completely sound dead; some listeners believe that the speakers sound better because of this fact.

My Father and I spent an intense five weeks brain-storming and developing the tooling to manufacture the lenses. Several modifications were made in the electrode molding and coating and transducer assembly processes. A new amplifier was also designed, using conventional tubes instead of the radio transmitter tubes used in the model 1's. The result of this effort was the Model 2.

The systems performed beautifully, as far as acoustic accuracy. The major thrust in those days, however, was for the highest possible sound pressure, with real fidelity taking a back seat. Consequently, some speaker designs were beginning to reach some pretty amazing sound pressure levels. One very highly respected reviewer even refused to audition ours until we made them louder! Hence the birth of the model 2SW.

The addition of a subwoofer allowed us to redesign the lens and open up somewhat the throat of the lens: a more constricted throat reinforces the bass response. With the line source pumping out a few more decibels, and the subwoofer pumping out more, we actually did get higher total sound pressure level. The model 2SW then enjoyed some very favorable reviews. It was even called, by one of the most critical reviewers, "a prime candidate for the world's best speaker." One reviewer kept the speakers, using them as the standard against which he measured all other speakers for nearly ten years.

They were, however, expensive to build. We had been selling the Model 2's for $2,000, wholesale to dealers, for about two years before we found out that they were actually costing us about $2,500 a pair to build! So, we simply doubled our price and kept on going.

For the Model 2SW's, we were paying a little over $500 per pair just to have the cabinets built. The Model 2's used three two foot transducers, and also three two-foot-tall lenses per speaker. Both systems were very labor intensive to produce.

The Model 3 was my Father's answer to trying to cut the cost of production. The two-foot diameter cylindrical cabinets were intended to be much less expensive to build. It was also hoped that by building three foot long transducers and lenses, we could seriously reduce the production costs. Eliminating a built-in amplifier also reduced costs and increased our customer base.

Unfortunately, it turned out that the costs did not actually go down. The increased transducer length meant that everything in the tooling had to be that much bigger. With the increased difficulty in handling, it actually took about twice as much labor to build a three-foot transducer as a two-foot one. The same was true of the lenses.

In addition, the round tubes we used to build the cabinets took up a tremendous of space to store and were in many ways more time-consuming to build than we had thought they would be. I personally believe that we didn't save a dime on any of those round cabinets.

It was in 1980, during the development of the Model 3's, that I left the company. I was never again involved in any of the companies that followed. What I know from there on, I have gotten second hand.

After I left, my father designed an all-new type of transducer. The electrodes were made from circuit-board material. They measured one foot by three feet and performed very well. They required less labor to make, but they didn't hold up very well over time. They would burn up the Mylar much more readily because they lacked some of the electrical characteristics of the cast epoxy-composite type of electrodes. These circuit board type transducers were used in Models 5's, and 6's, which also had cylindrical cabinets, but smaller diameter, about eighteen inches, and were sold without their own integral power amplifiers.

The Model 5 was five feet tall, with a single three-foot transducer and lens above. It only had about a 130-degree dispersion angle, and had a subwoofer in the base. Model 6 was a return to the full-height, six-foot line source, although not with a full width 180-degree dispersion, but again more like 130 degrees, and with a separate subwoofer cabinet about eighteen inches tall underneath it.

The Model 2 was the last speaker my father engineered solely to satisfy his own personal criteria. That is:

  • It was a full range electrostatic system: no crossovers,
  • It was full height: at least 3/4 of the distance from floor to ceiling,
  • It was full Width: 180 degree dispersion starting essentially at the wall,

For its simplicity of concept and clarity of sound, it is the personal favorite of some, including my younger brother Ross. The step from the Model 2 to 2SW exchanged the simplicity of a single source for increased volume and bass presence. While the Model 2 is perhaps the finest for solo guitar music, the Model 2SW is better for demanding pieces such as pipe organ music. The models following the 2 and 2SW, while excellent, represented compromises that met with varying degrees of acceptance in the world.