Motorboating in Valve Amps

As a child in the late 50's and 60's I spent many a happy minute on the boating lake at Peasholme Park.

I remember the sound the little kiddies motor boats made: put, put, put, they went.

When I began my AV career in 1974 I heard that sound again from a valve amp.

A cap had gone faulty allowing the amp to pass a lower frequency than it was designed for - it was obviously on the supply rails too (the supply IS part of the signal "path" after all), and that was feeding back to the input. Also, this amp had negative feedback (I know hi-fi lovers prefer the extra distortion that zero negative feedback amps give them...) and at such a low frequency where the output transformer phase was nearing reversal, the feedback at this low frequency was in fact more positive than negative, adding further to this phenomenon.

Recently I became aware of an abundance of affordable valve amps (OK, sometimes I'm behind the times being a solid-state man...), quite a few being of PRC origin. I managed to find a schematic on the web, and shock! horror! I find they're now DC coupled!

Well, I thought, Motorboating on a stick!

But I read from everywhere I look "heaped praise" from the press. Just shows how dumb (some of) the press is about anything remotely technical - preferring the manufacturer's spin?

DC input coupling is WORSE than that faulty cap I told you about above. Somehow these amps get away with it being fed by the bandwidth limited signal of a "digital" source, but you just need a trigger - a nice bit of real HF like you get from vinyl and ..... put, put, put, put....

In fact, you don't need to put the stylus on the record - the normal hiss that is due to the high analogue gain of a phono stage is enough.

Just one capacitor (per channel) added to the design of these amps in the input just before the first grid bias resistor and bingo! You'd have a stable amp. But no, some idiot got it about that "all capacitors are sh*t" and the lemmings followed.

The big headache I have over this? People who buy such amps keep blaming our phono stages

dvv wrote: Alternating between an AC and a DC coupled on the same device, one can hear a minute difference in the subjective speed and clarity...

This is my whole point about capacitor coupling: "one can hear a minute difference"

I believe it is possible to engineer-out that minute difference and the cost of the components to do so is similar to the alternative of DC coupling with servo's. But, as you know about me all too well, I like to be different as well as being openly honest (as far as I dare).

Capacitors are not perfect - I doubt anything we, mankind, have invented is. Electrolytics cannot handle high frequencies - the frequencies needed for faithful reproduction of analogue high frequencies, not so much for us to hear, but for amp stability to prevent interaction of distorted highs with the frequencies we do hear. High frequency electrolytics don't do the complex waveforms of music too well - they are after all designed for switching purposes - they distort a musical signal and give it a "signature" which is also known as "brightness".

In came the bypass capacitor - originally simply a liberal smattering of 0.1uF's as per JLH, but soon a better solution was provided by the high speed digital processing industry, with the understanding of impedance curves - capacitor coupling can now follow that trend and leave much less of a "signature" on the sound - so much less as to be hardly noticeable, if at all noticeable.

The cost? In PCB "real-estate" is not that different, but it's distributed around the board instead of being a circuit in its own right like the servo. In money terms? Hardly any difference either. So has one got any benefit over the other? Maybe just one - removal of the risk of damaging DC voltages under fault conditions.

There is however one major advantage of AC coupling from my point of view, and that is the property of being able to use a single rail supply. Is that a cost reason? It may have a bearing, but no. All audio circuits rely on a thing called "ground" to sink and source dynamic current. Apart from a few isolated cases of valve and J-FET circuits, the vast number of DC coupled solid state designs rely on a transformer centre-tap to both source and sink current. The only way it can do this is to change its rail voltages in relation to this centre tap - think of it as a resistive split. In a resistive split the ground would fluctuate up and down. Now "fix" that ground, and therefore the ends are moving up and down. All ground currents flow in and out of the centre-tap - they are not all going in the same direction at the same time. Clearly to overcome this (if it can be overcome) a very low secondary winding impedance is required, and this gives rise to the practice of using massive transformers - a sledgehammer to crack a nut - and the preoccupation of bigger and bigger power supplies that "improve the sound".

This single fact has led the bigger-is-better/heavier is better mentality of the hi-fi industry to the point that most people who don't understand the real reasons, feel that a 1 inch thick front panel must improve their listening. Crazy isn't it?

DC coupling is hailed as being the fix-all to the extent that many simply won't try any other solution. But worse than that, those who only half understand electronics but yet manage to become manufacturers held in high esteem, change AC coupling to DC coupling in such as valve amp inputs without recognising the consequences.

dvv wrote: Well, divided we stand now...

Never! But we can agree on peaceful coexistence?

dvv wrote: ...somehow I've always regarded Britain as an isle of sane tradition...

It was until we squandered our inheritance...