Real ESLs

During the ten years from 1925 to 1935 there were quite a few patents issued and technical articles written about electrostatic loudspeakers. These include patents granted to Frederick W. Lee (U.S. Pat.: 1,622,039) in 1925 and Walter Hahnemann (U.S. Pat.: 1,674,683) in 1926. Both of these are referred to by Williamson and Walker in researching their own project. Despite some ingenious designs during this period, the electrostatic speaker made little, if any, practical progress. The major reason for this was undoubtedly the lack of materials with suitable physical properties. Substances such as gelatin, japan varnish, gold size, silk gutta percha and rubber are mentioned in early patents! Lacking a suitable insulator, early designers were also forced to rely solely on the air gap and diaphragm for protection against electrical breakdown. Clearly, this limited the voltages that could be used and the audio power resulting therefrom.

So, lacking suitable materials (e.g. Mylartm, E.I. Du Pont, March 22, 1949 (U.S. Patent 2,465,319) for diaphragms and so on, early researchers in this field became very clever (or cunning). One of the first electrostatic speaker patents, issued to Colin Kyle (U.S. 1,644,387) used a thin, flexible diaphragm, fly wire, and mosquito netting! The conductive coating used thin metal leaf, insulated from the stator by placing it under thin layer of gelatine. This particular unit was studied by Walker when researching the ESL '63.

P.E. Edelman, around 1931, produced and electrostatic speaker using gold foil, japan varnish, tensioned springs and expandable corners. He was granted six patents in all, (U.S. Patents: 1,759,809 / 1,776,112 / 1,759,810 / 1,759,811 / 1,767,656 and 1,767,657) in the field of electrostatic loudspeakers.

A couple of significant developments by Kellogg described in U.S. Patent: 1,983,377 and G.B. Patent: 346,646 were referenced by Williamson and Walker when researching the Quad Electrostatic. Kellogg's U.S. Patent was not granted until December 4, 1934, but the filing date is September 17, 1929! This paper has significant sections on segmenting panels in an electrostatic speaker such that it presents a simpler load to the amplifier. It describes both single-ended and push-pull speaker concepts. The connection of panels through a resistor and and inductor was used to great effect by a number of designers after World War II, and the Quad folks were well acquainted with Kellogg's work. Indeed, the designs Kellogg propounded in G.B. Pat: 346,646 was a very heavy influence on Walker in his later design of the ESL '63 as he noted in a lecture in September, 1979. The popular magazine "Wireless World" often contained significant articles on speakers, and one such article on May 29, 1929 seems to have caught Williamson's or Walker's eye. Walker would have been a young lad of 13 years at the time. This article by Hans Vogt, entitled (what else?) "The Vogt Electrostatic Loudspeaker" seems to have made a lasting impression on one or both of the Quad designers, since they (much later) refer to both of Vogt's Patents (G.B. 387,546 and G.B. 372,649 ) [Sheet 2 ], which data from around May 12, 1932.

After World War II and until around 1965 significant effort was expended on electrostatic speaker design on both sides of the Atlantic. Arthur Janszen in the United States developed his famous high frequency units (U.S. Patent 2,631,196 [Sheet 2 ], March 10, 1953. His second patent (U.S. Patent 2,896,025 July 21, 1959) describes a unique (for the time) method of manufacturing stators from wound wires, instead of perforated metal. A great enthusiasm of the period was to combine the Janszen electrostatic tweeter with Acoustic Research’s first bookshelf loudspeaker, the AR-1. Janszen eventually developed and built the full range KLH-9. This full range speaker, was purely electrostatic, but perhaps not as capable in the bass as it could have been.

Meantime, Peter Walker (Acoustical Mfg. Ltd.) and David Williamson (Ferranti Ltd., Edinburgh) were designing and building the very first Quads. The Quad was a 'horizontal' design compared with Janszen's 6 foot high line source. The centre strip of the middle Quad (treble) panel was 11/2inches wide with a cutoff frequency of 7 000 Hz. The two mid-range strips placed either side of this strip in the same panel, but segmented (after Kellogg) from the central strip electrically. The bass panels were placed symmetrically on either side of the treble panel. The treble panel was charged to 1 500 Volts and the bass panels were charged to 6 000 Volts.

Around 1963, according to legend, Peter Walker set out, on his own this time, to improve on the original Quad ESL. His efforts as we know, led to the speaker we now call the ESL '63 (U.S. Patent 3,773,984 ). The unusual segmentation features of this speaker can trace their line of development directly to Kellogg's original G.B. Patent 346,646, and a comparison of the two patents reveals interesting similarities. The original AES lecture [Page 2 , Page 3 ] given by Peter Walker in September 1979, clearly refers to Kellogg's work in connection with the ESL '63.