A.J. van den Hul: Going Dutch Page 3
Three things are very important in cartridge design. I would say first to have very low tracking force, so that you can retrieve on the next play what you retrieved the previous time. Second, to have very high channel separation. And third, to have enough output.
Atkinson: When it comes to channel separation, many papers have been published saying that 20dB is enough.
van den Hul: It is not enough! I will tell you why. You will never have 20dB of acoustical channel separation in a room, so you may ask yourself why is it worth putting a lot of effort towards getting good electrical channel separation? But electrical channel separation involves adding distortion from one channel to the other. If you look at the cartridge crosstalk signal when you run your B&K test record, you see shifted signals, very strange-shaped signals, a lot of distortion. If you have an amplifier with a distortion level 20dB below the original signal, you will never accept this amplifier. Never. But as soon as it's a cartridge, everyone is very forgiving.
Crosstalk is very interesting. Fundamentally, it is an elliptical movement of the cantilever. If I have signal just in the left channel, the motion is in a line. But when a little bit of that movement appears in the right channel, it opens the line into an ellipse somewhat. The higher the frequency, the more the ellipse turns into a circle, so the higher you can keep that circle movement above 20kHz, the better the channel separation will be. The lower the distortion in one channel, the nicer the sounds coming only from the violin section in the left channel, the crosstalk not appearing as a very harsh cello from the right section. This cross-intermodulation of the violins over the cellos always makes recorded concert halls sound bright, because too much high-end distortion from violins comes from the other corner of the concert hall. It sounds like there is a bathroom-tile covered wall behind the cellos.
My idea is that a cartridge should have a minimum 35 to 40dB channel separation—I have a cartridge at home working with 43dB separation. That is less than 1% of the movement of one channel coming into the other.
Atkinson: What made you venture into cartridge design and manufacture?
van den Hul: The cartridge was the result of having done some work on tips, and some work on winding coils for repair purposes. I thought, why not make the whole unit myself? Some parts are made in Switzerland, but the whole unit is put together in Holland and I rebuild it completely. That's the only way how to do things well, to be honest. The man who buys a van den Hul cartridge knows that I've checked the whole unit.
Atkinson: After developing the stylus profile, you introduced a range of cables. What are your feelings on cable design? It seems that every week, someone, somewhere, introduces a different kind of cable construction, or conducting material, or dielectric. The lack of consensus suggests that there is still a lot to be learned.
van den Hul: Firstly, I would say that designing a cable is somewhat more complicated than designing an amplifier, because circuitry is known: you can make a mixture of part of circuit A, part of circuit B, and part of circuit C which can be even better than A or B or C on their own. Cable design, however, is an art. It should be a mixture of chemistry, mechanical engineering, physics, and the joy of music. If you are missing one of these four, I think you just do your job half.
Secondly, I do not think so much about the supposed super-quality of mono-crystal or large-crystal or any typical item of a cable. I think these are used partly for promotional reasons. I do care very much, however, about how a cable is made and how pure are the ingredients. This is very vital. Typically, it's not how the soup is cooked but the ingredients you have to buy before you can even make the good soup. I think cable manufacturers should pay more attention to the quality of the conductor. The quality of American copper is not very high, a lot of impurities are involved, which means that there can still be a lot of inherent failures in a wire where an over-exaggerated attention has been given to the design or layout. Maybe this sounds somewhat accusing—someone else may have a different opinion. I will not accuse anyone for doing his work, but I think I should say what I think about it.
You don't need to have three or four different windings made from different materials like the Japanese, or several different-diameter wires, one for the low frequencies, one for the mid and one for the high, like some companies in the United States. That I do not believe, because no-one tells the wire that it is not allowed to transport low frequencies through the thin wire, or high frequencies through the thick wire. It just assumes that that will happen. But that is the same as putting lemonade in a house's water system and expecting the lemonade to find its own way to the hot tap but not the cold. It doesn't happen: as soon as the information is mixed together, you need filters or maybe devices non-invented up to now to split them again.
Large-crystal copper is nice, but you should not bend any large-crystal wire because if you bend the large crystals, you open the structure, oxidation takes place because there is no protection against it, and, when it is pushed back again, there is now a barrier inside. But a wire is bending all the time. You put it here, you take it out, you put it there, you take it out, you solder it, it is all the time bending. And as soon as it is bent, it is finished with. So a wire intended for consumers should not be so critical that it should not be bent. My theory is that you should have a type of structure inside the wire that matches the crystals. The initials MC in my wire do not fundamentally mean Mono-Crystal, they mean much more Matched-Crystal. The crystals are matched so that when you bend the wire, they do not open immediately, oxidize, close again, and introduce an oxide barrier.
Very careful production is extremely important. Wires, like cars, degrade after a certain time, thanks to chemical processes. A good cable should have perfect protection against chemical alterations, but I'm afraid that a lot of cables do not have this protection. A Japanese-manufactured cable—I won't say which one but the first initial is "H"—features oxygen-free copper wire. Say that you took this morning a shower: you're fresh this morning; certainly you are somewhat fresh this afternoon; you may even be fresh tomorrow; but you will certainly not be fresh the day after tomorrow. Similarly, if you produce cable that you call oxygen-free, it's certainly oxygen-free at the moment that it is made, but definitely not so in three or four months. In a year it will certainly be oxidized, and we all agree that copper oxide is not the most perfect conductor.
So you must protect your conductor wire against oxidation. All my copper wires, for example, have at minimum a very dense coating of silver. Any cable manufacturer working without a good protection against oxidation is working with a cable he hopes next year to sell again to the same client because by then it sounds awful.
I'm especially aware of the effect of deterioration because I used to live in Delft. There is so much rubbish coming from the chemical plants close to Rotterdam, sulphur and nitrogen oxides and mercaptans, that the air in Delft is extremely polluted. The zinc on our roof had to be changed every four years, my car rusted very quickly, my Apple computer had a lot of contact problems—I would clean the slots but in several weeks they were green again.
Consider a wire: every day it warms up, forcing air out; in the evening, the air comes in again. The cable works as a thermal pump, pumping clean air out and impurities in which become chemically bound. The wire should not, therefore, be uncoated, uncovered. You protect yourself, why not protect a wire? In my cables I have first a migration-free jacket, so that nothing can be transported through the jacket. Then there's a double braiding—a woven braiding with a very dense silver coating, inside an aluminum braiding—covered with a Mylar film, so that even when the outer jacket is broken, the Mylar film will protect the conductor. Then there's a very high-density polyethylene dielectric, which also means there will be no penetration, with artificial fibers inside so that there is no humidity transportation. This high-density polyethylene is so dense around the wire that the wire is really forced together and the thermal pump effect does not exist so much. The high-density polyethylene also copes with soldering: the solder runs inside and closes the wire for ever.
Atkinson: Is the choice of dielectric material important?
van den Hul: The dielectric material is critical in that it should not have any energy storage or microphony. Polyethylene is good stuff to work with, certainly, but I prefer to work with Teflon. Teflon, however, is rather stiff so the flexibility of the wire is low; it can have some microphony; and it is expensive, costing about 40–50 guilders/kilo compared with polyethylene at about 2 guilders/kilo. For flexible consumer cables, therefore, I have to do things somewhat differently.