Quite some time ago, in the July 1994 issue to be precise, I wrote an editorial entitled "Valve Amplifiers Are Dead & Buried". The gist of the editorial was that valve amplifiers were far too costly and poor in performance, relative to even run-of-the-mill solid-state amplifiers. I went so far as to make the statement that "SILICON CHIP will never publish a design for hifi valve amplifier unless it is of academic interest only. In fact, let's be even more absolute and just say NEVER".

So why are we now publishing a design for a valve preamplifier? Well, as they say in politics, never say never! I still believe that valve amplifiers are far too expensive and that their performance is mediocre compared to very cheap solid-state designs. In fact, our high quality amplifier designs published in the intervening years since 1994 have continued to widen the gap.

That has not discouraged readers and kitset suppliers from periodically suggesting that we do a valve amplifier of some sort or other. In fact, only a month ago, one of the kitset suppliers suggested that we do a 60W valve guitar amplifier with its own speaker, etc. When they did the sums for the likely kit price (over $1000), they quickly back-pedalled.

So why do a valve preamp? Again, there have been a number of suggestions from readers and a number of circuits have been published elsewhere, all of which by the way, we have regarded as jokes. There has even been a PC motherboard with an on-board valve preamplifier for the sound section. Again, what a joke. A bad joke at that.

But having cast such aspersions, we were then more or less obliged to show we could do better. And we have. The triode preamp circuit featured in this issue is considerably better than anything we have seen published elsewhere, either recently or in the past.

The good performance comes about because of three factors, two of which were not available in the days when valves ruled. First, we have run the valve heaters from pure DC. This was sometimes done years ago but it was difficult. Now it is easy, using a 3-terminal regulator.

Second, the critical grid resistors are metal film types which have very low noise. Thirdly, and most important, our circuit has a substantial degree of negative feedback to greatly improve distortion and frequency response. It turns a very average performance into something we regard as acceptable (for valve technology, that is).

Mind you, some valve fans will turn up their noses precisely because we have used negative feedback in the circuit. Perhaps we can reassure them: the amount of applied negative feedback is still nowhere near as much as is commonly used in op amp circuits and the circuit still displays "soft clipping" when driven hard. But does it have "warm sound"? Probably not, because it does not have distortion levels of more than 1% unless it is driven to very high levels. Build it and see for yourself.

So there you are. We have changed our stance (slightly) and published a valve preamplifier. It is still a long way from publishing a high-quality valve stereo amplifier which would cost lots of money for fairly average performance and not much power.

And let us not mince words. While this mono preamp will probably sell quite well, to people wanting to satisfy their curiosity about valve circuits, its performance is still well below what can be achieved with a common low-noise op amp IC such as the LM833 which costs just a few dollars.