Balanced Audio Technology VK-P10 phono preamplifier Victor Khomenko Interview
Jonathan Scull: Victor, why balanced circuits, if you please?
Victor Khomenko: I remember reading an article that said balanced topology is inherently much more complicated than single-ended. They felt it was a cumbersome way of building circuits. And they provided proof in the form of a very ugly circuit representing "balanced" configuration.
Now, of course, a balanced circuit can be ugly. But that doesn't mean that every balanced circuit has to be like that. On the contrary, when properly implemented, a balanced circuit is inherently one of the most beautiful creations in electronics. I remember the first time I learned about the nature of differential balanced circuits from one of my tutors. He showed me that nothing in electronics was more elegant than a simple differential pair of transistors.
Look at the circuitry used in some classic 1950s McIntosh components, for example. If you take some of their monoblock designs from that period and cut off the input connector, you're left with, essentially, a balanced circuit.
Scull: The question begs to be asked: Why didn't they use a balanced input?
Khomenko: Sure, because at that time the marketplace hadn't accepted balanced as the right way of doing things. So they provided a single-ended RCA connector, and the signal was converted to balanced internally.
Scull: This happened often?
Khomenko: Exactly. And I can understand why. There are many reasons why I believe balanced is better inherently from an electrical perspective—as well as, how shall I say, from the philosophical point of view, in the sense of being more symmetrical. So balanced naturally becomes the choice of many designers.
Once you've accepted this, it's a shame not to use the full potential of the balanced topology. But what you often find, as with the McIntosh, is an RCA single-ended connector attached to one input of a differential pair, the other side grounded. In this way the distinction between balanced and single-ended becomes very blurred. It's almost nonexistent in many designs.
Scull: There seems to be a misconception regarding balanced equipment and its suitability for use in single-ended systems.
Khomenko: All our equipment—and I mean all our equipment—is fully compatible with other single-ended components. Which is to say, you can take the VK-5 or VK-5i and connect any mix of single-ended and balanced sources to it, and you won't lose much performance as a result.
Scull: You provide the best-sounding XLR-to-RCA adapters I've ever heard...
Khomenko: Well, even with the adapters, we feel there is some degradation when going from balanced to single-ended configuration. But you understand that,in all our models, the signal processing is always differential balanced.
Scull: And other benefits accrue from that?
Khomenko: Yes. One big advantage is in the way balanced circuitry interacts with the power supply. When you look at the interaction between a single-ended circuit and its power supply, it's always a current-on-demand situation. The power supply has to struggle to keep up with the instantaneous demand.
Scull: Especially when heavily regulated?
Khomenko: Yes, exactly.
Scull: You don't regulate your power supplies at all, just filter them heavily?
Khomenko: Correct, because when you're talking about regulation, you're talking about feedback. That can result in delay and slew-rate limitations. Balanced circuitry, by virtue of its symmetrical in-and-out of phase operation, does not have this problem. Basically, the power supply maintains DC to the power rail, and this is traditionally accomplished with regulators. But in our experience, the value of the DC rail in and of itself does not determine the sound quality to any great degree.
Scull: And what about the AC side, so to speak?
Khomenko: That relates to how the power supply absorbs and provides instantaneous current changes. And that's a much more difficult aspect of design because of feedback implications and AC fluctuations.
Scull: So the power supply has an easier time of it when it's symmetrical?
Khomenko: Yes. Once you remove this major demand from the power supply by doing circuits in differential topology, the power supply starts breathing easier. It becomes unshackled. It runs freely. It becomes what it really wants to be—a circuit by itself.
Scull: Victor, what was the design brief behind the VK-P10 phono stage?
Khomenko: First of all, it had to be extremely versatile. There's a switch-selectable configuration for any type of cartridge out there. And I'm talking about an output voltage range anywhere from well under 0.1 all the way up to 4 or more millivolts.
Scull: And you even included step-up transformers in your design.
Khomenko: Right; it's an additional way of providing flexibility to the user. We're not saying that you must use the transformers in every case. One can try and see what works best for one's cartridge and system. It's another tool for the owner to use.
Scull: Victor, how about some specifics of the circuit? I suppose it's not very typical?
Khomenko: Of course, nothing we introduce is just another me-too product. [laughs] We wanted our phono stage to be different from essentially everything else on the market. And there are several areas in which we address this. First—and I would say it's our trademark—we don't use any cathode followers or buffers in the signal path. All circuits are typically plate-loaded triodes throughout.
Scull: And I suppose that means no feedback?
Khomenko: Right, the phono stage doesn't have any global feedback.
Scull: There are three amplification stages in the VK-P10?
Khomenko: Yes. There are many ways of implementing high-gain circuitry, of course, but the one we take is a three-stage design: very crisp, very simple, with no buffers. And, of course, we rely on totally passive RIAA equalization. Except it has a twist—we call it "Flying Passive RIAA." To our knowledge it's a unique topology, and we have a patent application in process.
Scull: I gather you won't be giving too many details of it during the interview!
Khomenko: [laughs] Right! I guess it's enough to say that it allows us to simplify the RIAA equalization circuit tremendously. That impacts cost to some degree, and makes the unit more manufacturable, so to speak.
Scull: Poking around inside, I see you use a pair of toroidal transformers...are those Plitrons?
Khomenko: That's right.
Scull: And you use Russian 6922s, which are also known as 6DJ8s?
Khomenko: Yes. It was also called the 6H23 in Russia. It's a wonderful low-impedance tube that was designed for cascode circuit applications. In fact, we don't use cascodes; I don't like the sound.
Scull: And then there's a pair of 6SN7s...
Khomenko: Yes. If you look at the basic configuration of a differential gain stage, it always requires some type of current source. This can be supplied by a simple resistor, a solid-state device, or vacuum tubes. Most of the time I try to use vacuum tube-based current sources.
Scull: Why is that?
Khomenko: Oh, maybe for the sense of purity. When you build a "classic" tube phono stage, it naturally pops into your mind to use a vacuum-tube current source. My own inclination is always to try to use tubes, unless I run into practical or other difficulties. Out of the three stages in the VK-P10, two use resistors as a current source and one uses tubes. It's a tradeoff. You want to make the product practical and have it come in at a particular price point.
Scull: Thanks for the lowdown, Victor!
Khomenko: My pleasure...