P.S. Actually, I'm no purist and I do see the usefulness of transistors and their merits---in those places where they have considerable strengths, and even surpass tubes. And it really does depend on the song and the context. There's a place for a squeezebox, too, but it may not be in the solo section of an electric blues guitar trio. Similarly, with some stuff, like speed metal
, for instance, the notes go by so fast that not only does it not matter what one plays
through it doesn't even matter what one
plays! As one Jazz great (I forget who) once said, "There's no such thing as a wrong note if you play fast enough." Probably something similar applies to transistors in the signal chain. There's no such thing as a 'bad' transistor if it's buried far enough down into the mix.
(Smiley means that was the punchline of a joke. If you don't understand jokes or sarcasm, it's probably best to keep that to yourself.) And no matter how careful one is with the signal chain into the recording console, the listener is going to put transistors into the playback circuit at some point anyway. Actually, it's critical
where the transistor shows up: in terms of sustain, the first impedance seen at the input from the pup coils on a guitar is the worst possible place for a transistor. The high grid impedance of a tube is good for sustain. (10 Megohms is good, infinite is better, every milli-microampere the input takes is going to rob the strings of motion. If you really want to get a "feel" for this, go ride a bike with a generator-powered headlamp and turn the lamp on and off while you pedal the bike. Whenever you turn on the lamp, ... oooff! It gets a lot harder to pedal.) Other places where transistors fall short of tubes are when they are used as the first impedance seen by the transducer in a mic or other electromechanical device, the last
or next-to-last driver for a set of speakers, or the very last driver for a low-power device (like a reverb spring, cable driver, or splitter). In places other than these, all things being otherwise equal, a transistor will escape notice so long as the signal doesn't drive it anywhere near saturation or too close to the noise floor. FETs at an input improve things from an impedance perspective, but nowhere near to the extent that a well-designed tube circuit can. Electrons can leak through a tenth of a millimeter of silicon (a semi-conductor) much more easily than they can a quarter-inch of vacuum or clear glass (oxide of silicon, a non-conductor). And FETs just can't completely get rid of the amplification of the thermal motion in a solid. Now, having said all that, I have to admit that the first gig I ever played went through a transistor bass preamp I bult from a circuit in Popular Electronics with a then unheard of +/- 15db cut/boost in treble and bass going into a home-made monster of a tube amp. So the bass pups played into a transistor load. But that was a situation where the loss of sustain of the bass strings from the transistor load was miniscule compared to the +15db boost in the bottom end that I was able to send to that pair of Ham radio output tubes (with plates glowing red hot). These were
driven (into class C operation!) by a pair of 6L6s, to give you an idea of the power that went into the single 15-inch speaker connected to it. The output transformer alone (between the 807 output tubes and the speaker) was about as big as an ATX power supply! It must have weighed 15 pounds! (This was my first ever gig, so it was certainly the first with this guitarist. The guy that frequently hired him for his private parties said after the gig that his guests were all commenting that they had been dancing all night because it was the first time they'd ever heard the bass player.) So, to answer the burning question, yes, I do use transistors, too, when they're the right tool for the job. But if someone gave me $6000 to spend on a high-fidelity stereo sound system preamp and power amp, would I get an audio system containing a signal-path transistor? No way.
--thndrsn