SS Vs Valves

Graham, I like valves ... and I like SS. They are both ok in my book if the designer and builder knows what he/she is doing. Like anything in life the attention to detail and knowledge of the designer comes through when you hook up a pair of speakers or headphones.

Hi AK,

Last question first: No. But let me qualify that: everything has a lifetime - for example: the plastic bags you get at the supermarket checkout take many generations to decompose. Likewise, a transistor is encapsulated in plastic. As for the "element" (the transistor junction), that is etched in silicon - "written in stone" - and if used within its limits will last nearly as long as stone - whatever that is. The only other "life-threatening" thing is its wires: they will tend to corode before the "stone" withers away - possibly hundreds of years in the right conditions...

As humans if we don't understand something we experiment. I see most things as an experiment - hybrid amps included.

The property of a valve is its inherent linearity, which doesn't mean all audio valve circuits will be linear by default - it still needs the skill of the design engineer.

The valves I'm used to (which mainly covers them all - but a few wierd ones) have voltage supplies in the region 100 - 300 volts DC (the HT rail). When biased into class A in say a preamp, half the rail is across the valve and the other half is across its anode resistor - these items connected in series (there will usually be another resistor in series with its cathode to the negative rail but I'm trying to keep this explanation simple).

The output signal is produced by the anode "swinging" a facsimile of the signal that appears on its input (the grid). It is like two pieces of elastic taking it in turns to pull one way then another (in fact, on paper it's up and down) thus producing a sine wave (or in real life a complex musical wave).

Because the output of a preamp is in the order of a volt, the two "elastics" hardly move: taking a 100V HT, one has 50 volts of "stretch" and so does the other. 1 in 50 is not perfectly linear and so the distortion is going to be 1 in 50 or 2%. But we have a "tool" called negative feedback which further linearises things and brings down the distortion within acceptable limits.

That is what I mean by inherently linear.

The transistor isn't as inherently linear because it operates on a much lower supply voltage. Otherwise, for the purpose of this explanation, it works just the same as a valve. The "elastics" are "shorter" so the distortion is higher. And that's why you won't see many single transistor preamps like you see single valve preamps. Good (properly designed) transistor stages use a few transistors taking it "a bit at a time" and using the same "tool" of negative feedback, and in that way can be made as linear as a valve.

I will add here (before somebody pulls me up) that I have left out the "non-linear" base-collector relationship simply because most SS designers (me included) know how to linearise that on "auto-pilot"!

Eventually, when I get sufficient time I will post some simple circuits (valve and transistor) and explain how they work.

The danger is that through not knowing (unfortunately the word for that is ignorance) people can be easily swayed into one belief or another just like the way vigilante groups establish themselves. The two camps will always have their "extremists" or "fundamentalists" who try to coerse their followers into dirtying the opposite camp. Unfortunately that has happened here in high-fidelity circles, and we all waste precious time arguing about it. What we should be doing instead is learning.