Richard Dunn Interview

Audio Arts Magazine 1992

Q: What does NVA mean?
A: NVA stands for Nene Valley Audio, we still use it but most people know us by our initials. The river Nene runs through the town of Peterborough in North Cambridgeshire and that is where the company started in 1982.

Q: What are the main ideas you hold in designing your products?
A: The basic concept is not to use anything that is not needed. There is no such thing as a good component. All components are bad, but some are less bad than others. There is only one good thing and that is nothing. The way I design my products is to use the simplest circuit possible, and make the most of it, and make a good product that the customer can buy at a reasonable price. We never planned to make a universal or unconditionally stable amplifier. Anybody can design these you just put capacitors everywhere, but the capacitors you put in the circuit will limit the bandwidth and contaminate the sound. What we want is an amplifier with clean sound and wide bandwidth, that is why we take every possible capacitor and all filtering circuits out. In other words these amplifiers are not unconditionally stable and they still perform at 100kHz. They are not compatible with all kinds and makes of loudspeaker cable. For example if you bi or tri wire your loudspeakers the capacitance of the cable run will double or triple and the resistance and inductance will lower by one half or two thirds. A wide band amplifier will go into an unstable condition under this situation, so we must set the conditions under which the amplifiers will work comfortably, although this may not be conventional. As far as amplifiers are concerned, convention stresses stability, and the more stable the better. Designers usually put filters at the input and output. The frequency range beyond 30-40kHz will then not be able to get past. An amplifier based on these kind of ideas sounds dull, and lack dynamics and clarity. Another problem started several years ago when Litz style cables were introduced. Litz cables are two coated (insulated) wires, one positive the other negative, woven together, which ends up having very high capacitance. This creates a virtual short circuit load on the amplifier at very high frequencies.

Q: But they must use it for a reason?
A: Sure. Litz wire or bi and tri wiring connected to the output of an amplifier acts as a slight high frequency boost within the audio frequency range. The higher the frequency the lower the load impedance. The sound becomes brighter when the impedance at high frequencies lowers. Some people prefer the sound to be brighter because they feel it is more live, more dynamic when it is brighter. But they do not realise it is like someone getting sick, and they try to heal him, he may get better but the medicine always has side effects and creates other problems, it is always better to not have the illness in the first place. You use Litz, or bi or tri wire because you suppose your amplifier is sick because it sounds dull, and you need these things to heal it or compensate it. It is my idea to just design a healthy amplifier in the first place, which does not need any form of compensation! However an amplifier that operates well this way is more difficult to design. The location of every component in relation to others will affect the stability of the amplifier. Yet if you can do it well, and them use some high quality copper or even better silver wire as loudspeaker cable, the amplifier will definitely sound better.

Q: Are you saying that you demand that speaker cables of normal impedance are used with your amplifiers?
A: Definitely. It is because high frequencies of 150 or more kHz going through an NVA power amp stage will be amplified. High capacitance cable connected to the output of my amplifiers makes the load very low, sometimes down to even 0.1 or 0.2ohm at RF. At these frequencies non filtered amplifiers, when facing that kind of impedance, will become unstable, which results in oscillation. When this happens, the power from the amplifier will be drawn to amplifying this, and it will sound terrible, get hot, and could be damaged.

Q: Will the high frequency oscillation hurt the loudspeaker?
A: If the loudspeaker is correctly designed it wont get damaged, but you need to consider another problem - cables with high capacitance like Litz are only suitable for those amplifiers that are filtered and cannot be driven unstable at high frequencies. So if you insist on this cable you are limited to only using filtered amplifiers and those extra components interfere with the sound so you will never hear good amplifiers!!

Q: Now here is a very basic question - why would components affect the sound?
A: ALL components damage sound quality (music). It is like when you go to a concert. The closer you are to the stage, where the music happens, the less the chair backs and heads there are in front of you. All components have the same effect of corrupting information.

Q: Do you think it is appropriate to compare mechanical things to electronic components.
A: Of course it is. Mechanical structures and electronic circuits are analogous. In the mechanical world all objects have their own resonant frequency. A table has its own sound when you knock it, so has everything. This is timbre, a combination of fundamental frequency and harmonic pattern. Even the planets rotate around the sun at their own frequencies. The human body has a biological clock which runs at specific frequencies. Sound is just an oscillation that is high enough to hear but too low to see! Music operates on these oscillations. In the path of music, the more resonances (timbre oscillations) there are the more the music gets corrupted. In my amplifiers the cables that connect the output stages on the PCB to the speaker terminals are uninsulated tin plated copper. The reason why it is not coated is to keep all resonant structures to a minimum and a dielectric is a resonant/absorbent structure. Do you know how this will effect the sound? It acts as a storage medium, add resonance and will colour the sound. Even dielectrics on the DC voltage rail wires will have a small effect on the sound. The best dielectric is air (no dielectric) so we choose to use bare wire.

Q: Not even Teflon?
A: The thin cable at the amplifiers inputs is coated in Teflon for protection sake. This is because these are thin wires which could be damaged or shorted. However if you can do without it, just do without it.

Q: Conventional amplifier designers must have their reasons, mustn't they?
A: Maybe they just want to follow convention. You know, there is a broadcast university in the UK called The Open University. They teach electronics and they use the conventional electronic circuit structures for their text materials. All those that take the course are taught to make stability their priority and there is no mention of sound quality!! So it is rare to find a designer or electronic technician who will think that their test equipment is wrong when what they hear contradicts. They just depend on and trust their oscilloscopes and distortion analysers. If you were to judge NVA amplifiers by these criteria you would say they are not very good, but if you listen it is a completely different story. Test equipment readings have virtually nothing to do with the musical ability of an amplifier.

Q: Why do you think they have nothing to do with music?
A: I have designed amplifiers for very flat frequency response and low distortion, but I do not like the music they create. I design my amps to enjoy music it is that simple! The only valid test equipment is the human ear. We go to a concert with ears not distortion analysers.

Q: The owner of the company Swiss Physics is a physicist. He told me that he uses test equipment to test components and get the kind of sound he wants.
A: All test equipment manufacturers use the best components they can to manufacture their products and so do I. In fact I probably use better components, so how can they measure me, if anything I should measure them. It makes no sense, if test equipment was so wonderful and they used wonderful components what is to stop me using the same. If you wanted to design a better cow would you give the job to a cow?

Q: So what kind of test equipment do you use to test your amplifiers?
A: Ears, or I should say ears and heart. I use the emotional experience not intellectual reasoning. To most of us music is an emotional experience, you apply your subjective feelings to the music you listen too at a concert, not objective intelligence. Objective criteria are applied in manufacture to create reliability and product standards, they cannot tell you what kind of music is good music. The best test equipment for that is the little hairs on the back of the neck!!

Q: Have you ever thought about the possibility that someday somebody like the designer with Swiss Physics, who know how to get the sound he wants by testing components, will come to you showing you a chart and declaring how he knows how an amplifier will sound by simply looking at the chart?
A: Anything is possible in this world, but for me to give that statement credibility beyond the realms of good bull**** would be extremely unlikely. There are too many variables. For example, lets talk about components. You look at a resistor and you say it is resistor, WRONG! This kind of knowledge is not enough, because what you see is just the major function of each component. A resistor is also a very small capacitor and if wire-wound is also an inductor. Even a circuit board acts as a resistor, capacitor and inductor in a very unpredictable manner. Also components interact unpredictably when in close proximity to each other. Amplifier design is a discipline of infinite possibility, what we know today is but a very limited portion. A majority of things that affect the music in amplifier design we haven't yet invented a word for, let alone be able to measure.

Q: In other words there are no absolute standards in the world, is that what you mean?
A: Exactly. You know the human being is a strange creature. We are talking now. I talk to you, you talk back to me, and there is something going backwards and forwards between us, communication of information, this is reality. When we go to a concert, the singer or orchestra on the stage cares about how the audience feels, this is an interaction in reality. You go to a concert because you want this kind of interaction or reality. The communication of the note structure as it goes up and down, or moves from major to minor key, when they are singing or playing is communicating emotion and creating an emotional reaction in you the listener. And my voice when I am talking to you, and the inflections I am putting in my voice, are communicating to you things about me that are not in the words. That is what all forms of sound are, the music of life.....and the problem I believe with Hi Fi is we have lost the ability to perceive this reality or to compare what we are hearing with the reality and communication of real life. All the time we are walking around or working we are listening to reality. Yet when we go home and switch the Hi Fi on, we also switch into another part of the brain in order to actualise and intellectualise the experience.....we sit there thinking "wow, what a beautiful holographic image; you know that high frequency is really clean, and the bass is really deep" do you think this at a concert? NO you have an emotional experience! You are just there, and you perceive it.

Q: But you still have to use test equipment when you design your amplifiers, don't you?
A: The only reason I have to use test equipment is to make sure the amplifier is safe.

Q: So if you were to design a complicated amplifier you would use more test equipment to make it safer.
A: I don't want to offend you, but this to me is a nonsense question, because you are looking at the problem from a completely different angle. I don't want to design complicated amplifiers. If complicated amplifiers really did the job, then I would design them. It is function-before-fact as opposed to fact-before-function. May I ask why would somebody want to design a complicated amplifier.

Q: I don't know, probably because they want it to be, for example, unconditionally stable.
A: So we are manufacturing product for the public, so they can sit there and look at the test results and say: wow, I have got a wonderful amplifier, and not listen to music.

Q: Is musicality and simplicity synonymous?
A: In Hi FI terms, musicality is an expression that is being hacked around by journalists and pundits. Musicality to my mind is where somebody is changing the reproduction of music to the way they want to perceive it. It is a bit like the valve (tube) versus transistor debate. You know valves tend to make things sound "nice", a bit like looking at the world through rose coloured glasses, but is that reality? It is not my way of doing things.

Q: Let me put it this way - is an amplifier being musical the result of it being simple?
A: It can be. Again one of the reasons why valve amps are making such a strong come back is that they are very simple. In a valve amp you have got four gain devices (valves) which in a transistor amp you would need anything up to twenty transistors to do the same thing, each of those, in the signal path, is affecting the sound. So in terms of simplicity valves are simply better, but there are other things about valves and the way you have to use them, that I do not like, which is the reason I use transistors. So, everything has to be compromised to some extent as we are way away yet from state of art. Obviously, if I could design an amplifier to use only one transistor, that did not need other components around it, then I would do it, but it doesn't work. What I am talking about is doing the minimum within your capability, necessary to the circuit, to make it reproduce music.

Q: So if you could design a more simple amp you would do so?
A: Of course, yes. I am always trying to take components out of the circuit. That and using the best components I can cost in. We have designed really top end amps - the Statement Range - where thing that are not cost effective in the Black Box Range can be designed in , but simplicity is still the by-word.

Q: You know, I think that you have had experiences in designing complicated amps, but you probably failed, so you changed your policy, am I right?
A: I have been working with amplification for numerous different purposes. It all started as a hobby which goes back to the 1960's. I have designed complicated amplifiers, for reasons that require complicated amplifiers. The previous company that I owned was Tresham Audio (sold to Tannoy in 1982). This was a professional audio company. We produced the first non-application-note FET amplifier in the world. It was known as the SR402. Now that was a 200w professional studio or PA amplifier capable of going into bridged mono mode and producing over 1000w into a 4 ohm load. Now, in that situation you are designing an amplifier to be used in very rugged conditions, especially on stage at live concerts. It has to survive being thrown around by roadies, drive unpredictable loads in a very unstable RF rich environment. It could even have to survive getting soaked with rain at outdoor concerts. All this abuse makes the design a completely different compromise to Hi Fi design. In Hi Fi, hopefully the customer doesn't pour water over the amp while it is running, overload it, drive parallel loudspeaker loads, use hundreds of feet of connecting cable. In which case, I can design an amp that is a lot more simple. You only apply complication when complication is required and necessary, for the amp just to survive! My customer, hopefully, is a lover of music, not the person who just wants to talk watts and frequency response, and says - hey, I've got a 200w amp yours is only 100w, so mine is better than yours. Specmanship is self defeating and is open to corruption and bull**** its the music that matters.

Q: You said you use this tinned copper wire with no dielectric outside it and yet you are not afraid of it being oxidised?
A: That is why it is tinned.

Q: What's the difference?
A: If copper is tinned it doesn't oxidise. Non tinned copper will oxidise and will change the sound of the cable similar to adding a dielectric.

Q: Why don't you make an amp totally hard wired with uninsulated cable?
A: Purely for production purposes. We actually did produce a completely hard-wired amp, but we found it impossible in production because of the amount of time involved and the possibilities of error, because you have wires climbing over wires, it ends up more like a birds nest than an amplifier. It is a good solution for one off designs as printed circuit boards affect the sound, but compromise told me that I must use PCB's, so I use the best quality I can cost in and avoid all the damaging dielectric effects of printing and solder resist that cover most other manufacturers boards.

Q: How much better is it when you use hard wiring?
A: It is very difficult to quantify.

Q: You know Mark Levenson uses Teflon circuit boards. To me, although they are so much more costly, I cannot find it to be worth it, as it doesn't sound that much better to reflect the additional cost.
A: Everything has to be compromised. There are times when you only make new improvement of say 5%, yet you have to increase your cost by 20% to do it. You find it very hard to justify. It is all a question of which market you are producing your products for. With the Black Box Range we aim for a niche market of people who don't have a lot of money but are enthusiast for music, whose Hi Fi is one of their major enjoyments in life, and I am trying to produce the best sound that they can afford. With the Statement Range we aim for the Mark Levenson or High End type customer and for this we can cost in far more costly improvements. Most people, even though they may be in it, don't seem to understand the Hi Fi market. To draw an analogy, most people are happy to own a Kodak or Polaroid camera to take holiday snaps. Conversely most people are happy to have a mini or rack system at home, just to make a noise as background music, you know to use at Christmas or for parties etc. I don't think they are stupid people, it just means music is not as important to them as it is to me. Other people buy flash separates or systems, all flashing lights and controls, it impresses the neighbours but it doesn't sound any better than a portable radio (the Japanese are masters of perceived value and the American marketing concept of "don't sell the meat sell the sizzle" just reinforces it). The equivalent in the photographic market is the camera with automatic focus, in fact automatic everything (by trying to not do something wrong, you do nothing right!). In both the Hi Fi and Photographic markets you have got enthusiastic people who wish to get as close to reality as they can, but they only represent a small proportion of the market. What they buy is like a Leica, on the outside it doesn't look impressive, the money is spent on the lens, on the shot system, and if you notice it is a simple system but with nice quality. So we are looking for a small portion of market, we are looking for people whose hobby in life is to enjoy music. I get mad when I hear people say: `Why can't you get everyone to buy your amplifiers?' You cannot, the same as you cannot get people to buy a Leica, different things appeal to different people. I've got my own ideas, but I am no so conceited to think I have all the answers to everybody's needs. If somebody would like to have an amplifier to impress their neighbours, or to compete with their hi-fi friend by saying: `Look at this frequency chart, look at that distortion analysis', if they want to do that, that's what they want to do, but to me it is stupid. I drive an Alfa Romeo Spider because I love to drive, I like to feel the response of the car on the road, it is my joy. If driving was of no interest to me , I would own something Japanese. That's their market, that is what they want, not what I want. I have only seventy/eighty years on this planet and I want to experience life to its best. It it is music, I want it to give me joy, if it is a car, I want it to give me joy. That is just me.

Q: You said NVA amplifiers can operate at the frequencies of 100 or 200 kHz...
A: There is virtually no filtration in the amplifier, the amplifier is wide open. Everything that is fed to it, will come out.

Q: Is it good then to cut the output band higher than 20 kHz, in other words, in order to keep the amplifier in a safe condition, do you think we should cut the frequency range?
A: I still think we shouldn't try and do it. One of the major problems of CDs is the fact it chops off everything so quickly. I think the musical information at these frequencies is very small, it is so small that test equipment has problems reading it, but there is information, that information is part of the music. We might not hear it in technical terms, because my hearing starts going down above 15 kHz, but if it is not there, it really affects how you perceive music.

Q: But it is really not there with CDs, right?
A: Right.

Q: If I am right, then how can you reproduce true music when you reproduce information on CDs?
A: You are looking the other way to the way we are looking at it. It is not so much a question of if it is good to reproduce these frequencies. It is what you have to do to an amplifier to stop reproducing those frequencies that causes harm. You see, most amplifiers will amplify wide bandwidth if they are just left the way they are, but because the way the circuit is being designed most amplifiers and most of components they use, and from the fact that they want to see nice clean charts, they don't leave it that way - they put compensation in, they filter the amplifier, and it is the filtration that damages the sound within the range we can hear. It is not a question of there not being music at those levels. I think that is why in a way analogue is more open emotionally and is more appealing than CD, even though you have the problems of noise, analogue still scores emotionally over digital.

Q: Is it true to say that if I have an amplifier that is full of capacitors around the circuit, and if I take them off, I can make the sound as good as I want?
A: It depends on the circuit design and also the printed circuit board layout, because you will probably have to relay the circuit, repositioning components relative to other components. Because basically what happens when someone is designing an amplifier, they look at their oscilloscope, they see instability, and they say: `Ok, I need compensation', instead of saying: `OK, I've got problems here, let's try to solve the cause of the problem instead of curing it'. And that takes a lot of care in the layout of the boards and selection of components. It is more care than most people are willing to take, especially when you are in a highly commercial situation, like the Japanese, when every year they have to produce different amplifiers. You know, you've got to spend time with circuits. You've got to listen to every single component change, every change in the circuit board layout. I mean, crazy things happen, I couldn't believe when I first discovered it - when you lay out a circuit board with 90 degree turn on the signal path, and then you lay out the circuit board with gental curves on the signal path, the differences are audible. Now, if you tell an electronic engineer that it is audible, they will laugh at you, they will say that is stupid. You know, electrons turn corners, they don't flow off the end, it is not like turning off a tap. You cannot measure it, but it has an effect. That is why I say that lots of things we are looking at and trying to achieve, we cannot measure. We do not have either the terminology to identify them, or the measurement equipment to measure them. The only way we can find it is through experience and listening.

Q: So, can you tell me what are the most important things that you look at when designing an amplifier?
A: The prominent point, the No 1 guideline for me is simplicity. After simplicity, I look at what is required for the amplifier to actually `live' in the real world. Now, living in the real world is having to drive a speaker, and amplifier has to live with a loudspeaker as a partner, so I have to design an amplifier to drive loudspeakers. Now, I cannot, I am not responsible for what loudspeakers people design. If I were to design an amplifier that could drive any loudspeaker, then I would design an amplifier that was highly compensated and band-limited, etc. So again, you have a situation that this is not a universal amplifier. There are situations that will make this amplifier not necessarily be distressed or damaged, but will make the amplifier not sound as good as it should. Now, it normally works out OK if my No 1 principle has been applied, a loudspeaker should be simple because again, all forms of cabinets, all forms of crossovers, all forms of capacitors, all forms of drive units, or resistors, they are all filtering, all holding information, all doing things. My own mind tells me, if you can design the HF drive and low frequency roll off, if you can design a bass driver with no crossover network, if you can design a cabinet that will work properly without any form of damping, that is a good loudspeaker. So if you mix a simple speaker with a simple amplifier, you've got it. If you mix a complicated speaker with a multi component crossover network that is making phase changes all over the place, and highly damped cabinet, OK, maybe you will end up with the situation that the only thing you want to listen to is the human voice, the way BBC design their speakers, you get the voice sounding right. OK, if that is what you want, but it kills the music, it kills emotion, kills the surge, kills the separation. So, that is the load I want to drive, I mean, there is another principle, I have to look at in what I am doing. Now, in order to drive the load, I need a good current capability from the amplifier, the current capability of the amplifier is more important than the voltage swing. To produce a 100 Watt amplifier, to me, is unnecessary. If I can produce a 50 Watt amplifier that will drive the majority of loads, in which case I need a good current capability. Now my principle of design dictates that I go for the largest V/A ratings in the power supply I can. Toroidal transformers are a must. Frame transformers tend to create higher impedance in the power supply, but no matter what V/A the power supply is, it is constricted by the ability of the circuit to pass current quickly and react to the signal. You want to try to get as many constrictions out the way of the flow of current as you can. Similarly, you want to get constrictions out of the way of the flow of signal in the amplifier, same as the power supply, try to get it as clean as possible. In the AP35 we use a 150 V/A toroid. We have listened to numerous different toroids and they all sound different. We now use an Avel Lindberg., which we consider to give us the sound we want. Mark Levenson also uses their transformers. They are also very compact for their V/A rating because they have found a way of winding the primary outside the secondary, which is actually more difficult. A majority of toroids secondaries are would outside the primary, for safety sake, because when the primary is outside the secondary, if you damage the transformer, you've got mains there, if the insulation goes, so that is why most people wind the secondary to the outside. But this design of transformer is better, you'll lower impedance if you wind primary outside the secondary. So we are very pleased with that, that is a question of selection. Again on power supply, we go against the trend of high capacitance in power supply smoothing. We don't like large-smoothing capacitors, because again, what is a smoothing capacitor doing in the power supply? Out of the bridge rectifier you get pulsed DC. Now what you need the capacitors for is: No 1 to fill in the pulse, to act as a battery, and No 2 is to remove the last bit of ripple out of the voltage rail. Now again, if you design it with test equipment. You look at the voltage rail, you see the ripple, and you increase the size of the capacitor, until the ripple completely disappears. You say: `OK, that's it, I've got good power supplies'. In my opinion wrong. Because you, have not listened to it. What I do is I listen to the power supply when I am listening in the normal position, away from the speaker, and I increase the level of capacitance until I can no longer hear the ripple. And that's a much lower level of capacitance than not being able to measure it. All that's important to me, is to stop it from being audible. The reason I do this is also cost saving, because I end up with 6,800 mfd, which is quite small, compared with 10,000 or 15,000 mfd that other people use. A capacitor is not a capacitor, a capacitor is and inductor, a resistor, and a capacitor. The higher you go in capacitance, the more the winding, the higher the inductance. A capacitor, if it is in series, blocks low frequencies and if in parallel - blocks high frequencies, where as an inductor, is the opposite. So there you have two complete opposing principles high capacitance you've got to increase the inductance. So big capacitors increase the impedance of the power supply rails, and what that does to music is it stops the amplifier responding to quick transients within the music. Again you look at any capacitor, any electrolytic capacitor in the power supply, looking at what happens at the extreme high frequencies, up towards RF, where these capacitors are not doing their job properly. You get a lot of induced high frequency mush, as well as the audio signal and induced noise from the transformer on the power supply rails. So we are great believers in putting higher quality capacitors in parallel with lower quality electrolytic capacitors in order to get the high frequencies field correct, especially with an amplifier with is a very wide bandwidth like ours. So we are putting a 10 mfd capacitor, and a 0.68 mfd capacitor in parallel in the power supply to progressively filter out higher frequencies, they are poly-propylene and paper devices. We use a high current bridge rectifier, there are horrible things, they influence the sound, but we are forced to use them. Batteries are cleaner, but they are also quite high impedance, far higher impedance than an AC power supply. So you still have problems there in the way it responds to the music, the way it allows attack. All amplifiers designs are compromises. Another thing that is very important to me is the volume control. A very important principle people apply to the VR, is that the VR must be quiet in its operation. This is because people tend to believe that if a VR is making noise, then there must be something wrong with it. There have been volume control potentiometers on the market for a long time, that are superior technically to the normal plastic film volume control known as Cermet type, but people reject them because they are noisy when used. We tried one and realized how good they were. Now we've two manufacturers, one French, the other English. The one in this model is made in England. We are going to use the French one in future because the French one is quieter than the English one. But still customers who bought our amplifiers are complaining that there is something wrong with the volume control because they are making noise. Please, do not worry about it, when the music is happening, you won't hear it. But when there is no music, and you turn it up and down, you will hear noise. So again, it is compromise. It is more important for us that we get volumes controls that sound good, and this one does sound very good. Again, coming to the preamplifier stage, because of the principle of simplicity, we go for passive. We don't believe in applying line amplification to the amplifier, we just like to see the signal coming in, being controlled by a high quality potentiometer. But you cannot do this with every amplifier, you have to design your amplifier to function that way. Because you think about it, what is a tone control? A tone control is a capacitance. In the input of an amplifier you always have capacitance. We try to get the capacitance in the input of our amplifiers as low as we can. Some amplifiers, again, for filtration reasons, make it higher. If you are going to use passive, in this case, it is a 100 kOhm linear potentiometer, that is blagged to simulate a 10k log law, by having a high quality 10k resistor put across it, (it is impossible to get accurate log law with cermet track). You have to have low capacitance coupling in the input to the power amp, otherwise you end up having the high frequencies being attenuated as you lower the volume control. So you have to design the power amp input stage to go with a passive. Which is the reason why some people do not like passives, because they've tried it with their amplifier and they think it is dull. If they tried it flat out, it wouldn't sound dull because at low levels it is acting as a tone control. All the signal processing we need to do, such as the RIAA amplifier, DAC, we do externally in a separate box. We also produce DAC and phono amps. Again, cost verses quality, people say: `Why do you put the switch on the back?' Now, that is an extremely good switch. What is important is the quality of its contacts. You can get really flash looking push-button switches, but so what, what is important is the contact. This switch has a very good sound, extremely good contracts, and a very sensible price. It is my recommendation to the users, to always leave the amplifier on, and as a way to help them keep it in mind we put the mains switch on the back... so that they have to make a conscious effort when they want to turn the amp off.

Q: Why don't you use screws?
A: It all started because when I first put the circuit together, I was listening to the amplifier in early bread boarded state, and I was very happy with the circuit. I designed a case, I put it in the case, and it was not as good. I could never understand this, so I started to investigate. The first thing I came up with was that it was steel that was the major problem. You see, a case of steel has a very strong effect on the sound. My theory on this is that there is a lot of magnetic field, that is not only coming off the transformer, but it comes off all AC circuits, comes off the capacitors, comes off the transistors. Now, with a steel case, it cannot get out of the amplifier, just bounces around inside, that turns the sound muddy and loses information. Our next case was made of aluminum, a good change, but I still was not entirely happy. I still could not understand why the original sounded better. Then I had an idea, to actually form the case, almost like a carved shell in one whole lump, that was better again. Then we made a case with no electrical connections between the panels, that worked better. I started to think: `Why is it better?', and the only thing I could think of was the screws. So the next stage was to think: `If it is the screws, then it could also be the electrical connections between the panels, because when you break the whole amplifier, you have no case'. So we designed a glued case, where the actual panels have no electrical connections between them at all, the case itself is anodized, the glues are electrical insulators - so it created a case, that electrically wasn't a case. There is a disadvantage to it, as you cannot put an amplifier of this type near to a phono stage - you know, the field is getting out very easily, creating induced noise, but it made such a sonic difference. The next stage was to look inside the amplifier, I said, OK, how can we remove more screws. We then came to the conclusion that we could actually glue the capacitors down, we could glue the circuit board down, we even found a thermally conductive glue, to glue the output devices down, so we don't need any screws there. And then the amplifier sounded better again. It gives you more of the information of the music, gives you more insight into the music, lets you see how the musicians want to play with each other. My theory is that the whole planet creates electro-magnetical field, which is the poles. Whenever you use a screwdriver to do up a screw, that screwdriver becomes magnetic. If you use it as much to undo, as to do up, it demagnatizes. So I told myself, something is going on here. I also have a theory about twisted cables. They create a magnetic field, and I don't like that. Our NVA cable, when we tried to twist it, the sound became worse. And as soon as you twist the cable, the cable becomes directional. If you don't twist the cable, it is no longer directional. We couldn't glue the transformer because it was too heavy, so we had to bolt it. Basically, it was the process of trying to create a case that has no sonic signature.

Q: But isn't it easy to get interference when the case cannot block it out?
A: This I agree. You see everything is compromises. If you allow everything to get out, then everything can get in. We suffer from it, but we suffer from it mostly in analogue. The thing that gets most affected by it is the phono stage. Whenever we got the the Heathrow show in London, we always had a room, which was overlooking the forcourt of the hotel. Whenever a taxi came, you could hear him talk to base. We have a dealer who has a taxi office right next to his shop. He cannot use my amplifiers with a disc stage on them, because they've got an AM transmitter. Whenever they switch on the AM transmitter, there will be a big `bang' coming out of the speaker. So it is a great disadvantage, but this is a good compromise to my mind, because it make the music better. There are very few people who live next to transmitters, or who live next to taxi offices. If I've got customers that have that kind of problem, I'll help them out.

Q: Why do I not see a protection circuit within the amplifier?
A: Basically for purist reasons, there is no such thing as a protection circuit that doesn't influence the sound. We've tried numerous types, and we have come to the conclusion, that none of them work. Our customers are people with, hopefully, a modicum of knowledge and common sense. And we say in our instruction book: `If you want to switch speaker cables, please always switch the amplifier off first.' If they will do that, fine. If, take me as an example, I never do it, because I know how to do it - as long as you do not touch the positive to negative, you've got no problem at all. There is nothing else that can harm the amplifier. If you used a fuse, you can either put a fuse in the power supply rail, or a fuse in the speaker lines. The problem with a fuse is, you can always hear it. Get yourself a fuse, put it down in the circuit, pass current through it, and as you start increasing the current, you can actually watch the fuse wire flex, it flexes as it heats up. As it heats up, it increases the resistance and when the heat gets to a point, the fuse breaks. You can actually hear it in any amplifier that has a fuse in it, whether it will be in the power supply or in the output stage, or even in the mains input. The least you can hear it is on the mains, but you can still hear it. A fuse has a compression effect. You'll lose dynamics within the music. So we have come to the conclusion, even though some people say it is dangerous, we will not put a fuse in the amplifier. The reason we feel happy about it, is because it is totally sealed case. We write on the back: `Do not remove the covers, no user serviceable parts inside'. There are only two connections to the mains, and they are very well connected. If for any reason they should break off, wherever they should land is also electrically insulated, so no power can get into the case, the case is quite safe.