By Reg Williamson
The death of Peter Walker of "Quad" on December 10, 2003, brings to an end an important chapter in the development of high-quality reproduced sound. It was in the immediate post-war years that interest renewed in the design of suitable equipment, taking advantage of the advances in technology, which hitherto were concentrated on the war effort.
After a long debate in the letters columns of the (then) prestigious English magazine Wireless World, the late DTN Williamson published what be came a seminal article on the subject, and he offered a suitable design for a preamplifier and power amplifier.
Thousands were made throughout the world; but it became clear that in many respects, it lacked commercial viability.
INNOVATIONS
Peter Walker (Photo 1) already ran his own small company making public ad dress amplifiers (from 1936), but like so many of us active in audio design, his motivation was the love of music.
He was an accomplished flautist, a hobby he practiced in the local Huntingdon Philharmonic orchestra. In 1941, concentrating now on high-quality amplifier design at a new factory in Huntingdon, he developed a series of tube preamplifiers and power amplifiers, the last of which was the model Quad 22/2 series.
Peter demonstrated an impressive talent for innovative thinking. Negative feedback (nfb) derived from a tertiary winding in the cathode path of the out put tetrodes -- in this case the new KT66s used by Williamson -- helped the power amplifier achieve very low distortion.
Walker also acknowledged a simple fact, seemingly lost today, that you are likely to have imperfections that could be improved by correction in the source of your program material. The preamplifier was unique in providing a variable slope low-pass steep cut filter, with three chosen cutoff frequencies along with variable slope bass and treble boost/cut filters. He remained faithful to this principle right into the transistor era.
With the transition to transistors, the company's first effort was not without its problems--as indeed, many other designers found out. Transistors, it was soon discovered, did not have the inherent robustness and tolerance of tubes under overload conditions, particularly if there were any appreciable reactive components in the load. It was this latter characteristic of some loudspeakers that produced conditions where the output devices could be required, albeit momentarily, to simultaneously exceed both the rated maximum current and voltage. The result was usually instant demise! In collaboration with Peter Baxandall, another prolific worker and friend of many years, Walker examined thoroughly and solved the problems. The concept of the Safe Operating Area or (SOAR) of a power transistor's operating characteristics is now well established. The Quad 303 was--and still is--a highly successful design, and the Quad model 33 preamplifier, incorporating most of the features of its tubed predecessor, still attracts enthusiasts.
Walker's major contribution to the de sign of high-quality transistor power amplifiers was the extraordinary concept of current dumping.
Early transistor amplifiers were often based on the first Lin quasi complementary design and used Class B working for efficiency, but insufficient attention was given to minimizing the phenomenon that became known as crossover distortion. This is caused by the tiny hiatus that occurs when both output transistors are simultaneously in a non-conducting mode.
Since high levels of overall nfb are essential, at this point, there is no gain--therefore, no nfb. So, the amplifier produces minute but particularly unpleasant levels of high odd-order harmonics, giving rise to the so-called "transistor" sound. While circuit design techniques existed to minimize it, other designers tried to eliminate it altogether by using the highly inefficient Class-A mode.
Walker developed what many consider the ideal solution. He designed an amplifier that operated in Class-A at low levels, but--should higher power be required--relatively low-cost, high power transistors would "dump" the extra current into the load. It was an extraordinarily ingenious answer to the problem, and the transition from one mode to the other was seamless and in audible. The first of these current dumping amplifiers was the Quad 405, and, to this day, if you could look on the back of many monitor loudspeakers in the studios of a broadcast company, you are more likely than not to see a dust-covered Quad 405.
MAJOR ACHIEVEMENT
However, Peter Walker's most important contribution was in the field of loudspeaker design. It had long been recognized that the electrostatic principle was ideal for minimal distortion and linear frequency response; indeed, tweeters already existed for use with conventional middle and bass systems.
PHOTO 1: Peter Walker.
Walker set himself the task of producing the world's first full-range electro static loudspeaker, and, in collaboration with DTN Williamson, published a sequence of articles in May, June, and August 1955 of Wireless World, discussing the many possibilities and most of the design criteria. Even today, for anyone interested in the subject, it is required reading.
The Quad prototype first saw the light of day in 1956. Many stood in line to hear this wonder at an audio show organized by the (now defunct) British Sound Recording Association. Even now, I can vividly recall the impression it made on me when I had the opportunity to hear it locally, and can remember the recording: Rita Streich singing the Lost Rose of Summer from Flotow's "Martha." Although it was in mono, the sheer clarity and transparency of sound bowled me over, almost as though the singer was somewhere just through the diaphragm.
Walker reminded me later of another pioneer's famous comment (Paul Voigt--when he suggested that it should sound as though you are listening to the real thing through an open window), which he regarded as one of the basic tenets of his own design philosophy. The production models came out a year later and many are still in use today. However, stereo was soon to arrive and the original ESL57 (as it be came known) showed some deficiencies when used as a stereo pair.
Walker was already thinking ahead and in 1963 began work on a revolutionary concept, still using the electro static principle. Only a hint of what was to come came from a paper presented to the AES in 1979. By this time, Peter and I were firm friends, after he gave me valuable advice on the design and construction of capacitor microphones, my particular preoccupation at the be ginning of the '60s. Every time I found myself near the factory at Huntingdon, a visit was a must--and after I was invited into the secret room behind his office to view progress, we'd have a leisurely lunch at a nearby hotel. It was invariably accompanied by a long discussion on the steady deterioration in equipment review standards, a distaste we both shared.
The new speaker appeared in 1981 and was an immediate success. For a full description of the principles on which it is based, I can only refer you to my article in Speaker Builder 1/82. Suffice to say, it was based on the concept that the ideal sound radiator was a sphere.
Since it was physically impossible to make a speaker of this shape, Walker made the flat plastic diaphragm operate as a sphere by delaying the sound to arrive sequentially to each of a series of six concentric rings. This produced the illusion of sound coming from a pulsating sphere about 30cm behind the speaker (which incidentally, like its predecessor, was a bipolar radiator so the effect was the same, front or back-no cabinets were involved, so no structural resonances).
The actual plastic film radiator was lighter than the air surrounding it. The ESL63 is still the benchmark standard to which all other designs are inevitably compared.
Peter Walker belonged to that generation of audio designers who firmly believed, as I do, that--provided the natural laws of physics are faithfully observed--it is usually possible to predict with almost 100% accuracy how a particular design will behave. This was vividly illustrated to me when on a visit to the factory with my wife, who unusually (and I do not intend being sexist) has a fascination with hi-fi (a tautological corruption I hate). We had just taken delivery of a pair of ESL63s. I casually mentioned her interest to Peter, and an invitation came with alacrity! She regards this as one of the most interesting days of her life, not so much seeing the work going on, but the obvious respect and affection in which Peter was held by his workforce--all were on first name terms.
The final test of a production ESL63 consisted of being in close proximity opposite a reference model with a B & K high-grade microphone positioned approximately between them. Each was fed a square wave (if my memory is correct, around 800Hz). The phase was reversed on one, so it followed that if each performed to specification, it should be possible to position the microphone equidistant from both speakers, and the output should then be nil.
As indeed, it was--just noise.
I very unwisely asked whether it was desirable to listen to some music. Peter looked at me in genuine amazement and exclaimed "Whatever for” Oh, dear me, no. No point. We leave our customers to do that. . . ." He became increasingly irritated by the outpourings of an emerging breed of equipment reviewer, whose talents were confined to writing entertainingly--but invariably with no engineering background whatsoever.
For many years, he refused to allow reviews of Quad products, particularly after an incident in which a reviewer insisted that all [sic] amplifiers sounded different and that tube amplifiers always differed in some degree from transistor amplifiers. He made the serious error of citing Quad products. Peter, like me, does not accept this premise, if the amplifier is designed properly in the first place. But his Quad products had been held to be defective in some way, resulting in a deluge of letters to the factory.
Peter set up a test panel to take place over two days, and invited a number of Golden Ears to take part. He invited the original complainant to suggest suitable speakers. He chose a pair of very costly Japanese models but subsequently declined to take part because "The relays switching the speakers from amplifier to amplifier were not gold plated." No comment is needed, I think.
Three amplifiers were used: the original tube Quad 22; the first transistor model 303; and the later 405. The chosen panel was invited to suggest for test material, and I provided a master tape of one of my own recordings. The switching between amplifiers was random, literally on the toss of a coin.
Individually and independently, each member of the panel was invited to decide which amplifier was in use, and at the end of the two days, a statistician examined the results. His conclusion was that the toss of a coin would have produced the same results. Another respected engineer, James Moir, also handled the whole expensive exercise independently. The reviewer--now discredited--departed the audio scene and started to write for the emerging computer magazines!
Peter developed a number of fascinating tests to support his thesis, one of which consisted of passing a signal-- any signal, pure tones as well as complex programs--through one of his power amplifiers, and, after attenuation, into a simple difference amplifier.
He fed the original signal into the other input via some passive reactive net works to simulate what happened in any amplifier, such as minute phase shifts. When summed and at the same level, cancellation would occur if identical and without any added artifacts. He could always demonstrate that the result was, inevitably, just noise!
HONORS
Peter was responsible for the succinct description of the ideal amplifier as "A Straight Wire with Gain." It was towards the end of his highly creative career that his work began to be appreciated. The AES finally recognized this with an award after a short, and I suggest slightly tarnished, episode in which that au gust body had some misgivings, since its rules required a recipient to have a degree. Redemption came when he was ap pointed a Fellow in 1980 and given the Society's Silver medal in 1989. No such problems attended the award of an OBE and the Queen's Award for Industry to the Quad Company.
It was while I was part-time teaching at the University of Keele that I discovered my colleagues in the Electronics and Music departments shared my admiration for this remarkable man, so nomination for an honorary doctorate presented no problem. The then Chancellor, Sir Claus Moser, himself a music lover and owner of a complete Quad system, made Peter's award.
Peter was a family man, with two children, and his son Ross succeeded him as CEO of the company after Peter's retirement. Through the difficult days of a small company striving to make a name, his wife Peggy was loyally supportive. She died a few years ago from cancer, but Peter subsequently re married a former school friend. Sadly, fate can be quite vindictive at times and he found himself bereaved once more after a distressingly short time.
Peter's final years were marred by serious ill health, the consequences of a progressive lung disease, and he be came dependent upon others for care until the end. The family no longer owns his company, and the jury is still out on whether the new owners will maintain the high standards on which he insisted. Sadly, the world of commercial hi-fi is now highly competitive and the label is slapped on almost anything that makes a noise. For example, I can well imagine his robust reaction to the suggestion that connecting cables have a characteristic.
So, as to the future, I would like to be more optimistic, but it is difficult. With recent publication of a book about the company, Quad--The Closest Approach (Available from Old Colony Sound Lab, PO Box 876, Peterborough, NH, 03458-0876, 603-924-9464, sales@ audioXpress.com, #BKIA1), there is some encouraging evidence that a genuine attempt is being made to maintain the prestige Peter Walker enjoyed in his lifetime, accompanied by a rightful tribute to his exceptional legacy.
------------------
Also see:
Quad ESL 63 Loudspeaker (Jun. 1985)
Quad (Acoustical Mfg. Co., Ltd.) Model 405 Power Amplifier (Apr. 1979)
Quad Component System -- One-Brand Systems (Dec. 1983)
The QUAD Manchester Competitions (Apr. 1989)
BOOK REVIEW--Producing in the Home Studio with Pro Tools.