Wednesday, October 3, 2012

Paraphrase Circuit: Brilliant Audio Engineering Solution?


As a product post World War II era English frugality and common sense, did the paraphrase circuit withstood the test of time as a brilliant audio engineering solution?

By: Ringo Bones

Maybe I’m just paraphrasing from Ken Kessler’s book Quad: The Closest Approach whenever I hear a fully restored Quad II amplifier and praising its sound quality made possible by an unseemly circuit design of the famed amplifier’s input section called the paraphrase circuit, or am I just looking this venerable vacuum tube based design that’s now over 70 years old with rose-tinted glasses. Anyway, what is it about Peter J. Walker’s famed paraphrase circuit that had pundits calling it a brilliant audio engineering solution cobbled up in the austere fiscal climes of post-World War II Britain that later become an iconic fixture in the Golden Age of Stereo?

A typical Peter J. Walker designed Quad II amplifier, whether original, a 1996 era Diamond Jubilee Edition or a 2004 era Classic, is a class-A push-pull monophonic amp with only two active stages. The input stage contains two EF86 miniature low noise pentode tubes, each of which is capacitor-coupled to its own output tetrode tube, but only one which – designated as tube No.1 – gets the original input signal. From there, in addition to driving its corresponding output tube, input tube No.1 drives input tube No.2 with a portion of its reverse-phase output, attenuated by some 6dB. Thus the output of the second EF86 is equal to but out of phase with the output of the first EF86 tube, and this balance is maintained through the output transformer’s primary winding, where the full music wave is put back together. This is the working description of a paraphrase circuit, used to be quite popular, arguable because it accomplished phase inversion and voltage gain in one stroke – and so allowed the designer to keep both distortion and parts cost to a minimum. By the way, the KT66 output tubes of the Quad II amplifier was later used by the late, great Dr. Jim Marshall in designing his first Marshall electric guitar amplifier – the JTM 45 Blues Breaker.

But to “audio purists” - what they see as the greatest weakness of the paraphrase circuit is that the signals at the grids of the two output tubes – and thus the two halves of the music waveform at the output transformer – are not perfectly in time and in phase with one another, given that the signal on the grid of input tube No. 2 has already been amplified, whereas the signal on the grid of input tube No.1 has not. Of course, one has to remain open to the possibility that this very technical anomaly of the paraphrase circuit resulted in some form of another a “musically consonant form of distortion” that most converts to Peter J. Walker’s Quad II amplifier just happens to find pleasing. When paired with another of Peter J. Walker’s famed designs – the Quad ESL-63 electrostatic loudspeaker – almost all hi-fi enthusiasts on the planet claim it to be the most perfect hi-fi rig money can buy.

Believe it or not, the only other power amplifier utilizing the benefits of the paraphrase circuit was designed by Kevin Hayes in his VAC PA160 mono amp which was marketed around the mid 1990s – a time when vacuum tube circuit revival in the hi-fi world was steadily gathering steam. For all intents and purposes, the VAC PA160 can be seen as a more powerful version of the 12-watt Quad II amp that uses KT88 tubes that can provide 160 watts in pentode mode, 152 watts in Ultra-Linear mode. Though quite different from the paraphrase circuit used in the Quad II, Kevin Hayes’ version uses a 6SN7 octal-based dual triode configured in a floating paraphrase phase inverter. In this arrangement, the input signal is applied to the grid of a triode acting as a common cathode amplifier, which provides one phase of the stage’s output signal. This output is resistively coupled to a common point – which for convenience, I shall call “point x”. The other triode’s plate, also resistively coupled to point x, provides the other phase of the stage’s output. A coupling capacitor from point x drives the grid of the second triode. In actuality, this amounts to a stage that provides one output phase directly and couples that output to another stage, which has 100% inverting feedback, to provide the other output phase. From a circuit purist’s standpoint – i.e. fans of single-ended triode amps – this scheme has the glaring disadvantage that the second generated phase goes through one more tube than the first generated phase. Further, the output impedance of the two phases is likely to be quite different. But given they resulted in a good sounding design, does this mean that the paraphrase circuit – born out of postwar English frugality and common sense truly had survived the test of time?