Vacuum Tube Stuff

But first, can you identify the tube over there on the right? Doing so will put you in a rare category, and there will be no hints, as the markings came off when I cleaned it with Formula Pro409. It was made by Western Electric, though, to give you a head start, and it was popular in pulse-modulated service, later used by some amateur radio operators as a grounded grid linear amplifier. I have two of these guys, plus, more importantly, their sockets (also shown). Al knows the answer to this. Who's Al? Ah, that's another hint.

Anyway, generally, what kind of tubes do we like?

Triodes! (Where possible.) Why? Simple--they're simple. Easy to control. Linear in their operation. Elementary to power. A plate voltage, a grid bias, a cathode voltage (or not), and thou. And what is the most common triode in use in low-level preamplifier circuits in my guitar amplifiers and amplitude-modulated ham radio transmitters? The 12AX7, of course. A dual triode, actually, and aside from octal tube equivalents, perhaps the best selection available for a trouble-free design.

What, design a microphone preamp for an AM transmitter? Using an Astatic D-104 crystal microphone? How about a 270K plate resistor, a 2.7K cathode resistor bypassed by a 25 microfarad polarized electrolytic capacitor (at 25 wvdc), and a 10 megohm grid resistor? Decouple the plate resistor with a 47K and a 20 microfarad polarized electrolytic capacitor (at 450 wvdc), couple the output to the next stage with a 0.05 microfarad coupling capacitor, and you're there. What size grid resistor should you use in the next stage? Well, it's time for our first ROT!

Rule Of Thumb #1--the value of the coupling capacitor multiplied by the value of the next stage grid resistor should equal roughly 25000. Let's see--0.05 times, oh, 500K equals, wow, 25000! So, we're talking a half-megohm grid resistor in the next stage. Where do I get this? I forget, but it works. Now what? Well, if you make that 500K a potentiometer, now you've got a volume control. Nice, huh?

Is this too easy? Hang in there, we'll get to the good stuff. Tubes, you know--the Rosetta Stone of modern technology.

Hold on, now. We're tossing these plate and cathode resistor values around a little too lightly, here. What B+ voltage are we talking about? What gain do we want the stage to provide? These issues will largely determine the PROPER values for plate and cathode resistor values. (We'll get to the grid resistor value in a moment.) I've recently found an excellent resource in Dan Torres' new book on fixing and modifying tube guitar amplifiers. It's available at many mail-order places, one being J. K. Lutherie, who advertises on the back page of Vintage Guitar Magazine. I strongly recommend this book (Dave Funk's, too).

What other elements should you be concerned with for best sound? Stiff power supplies, which means choke input filters and large capacitor values, usually. Except for some guitar amps, which depend on power supply "sag" for their characteristic sound. They should employ tube rectifiers, unlike some of the better vacuum tube audiophile designs. Solid-state rectifiers should be used with care, however, as turn-on can take out cold capacitor banks if no voltage has been applied in a significant period of time. Inrush protection is essential, although a choke input can mitigate this condition somewhat. But choke inputs are rare in consumer equipment, so careful with that silicon stack, Eugene.

Back to our circuit. Now what? Well, next is the second stage. We already have a half-meg volume control, so let's use a, oh, 220K plate resistor (with a similar 47K and 20 microfarad decoupling circuit of his own), and a 2.2K cathode resistor, which we'll leave unbypassed, for a little bit of negative feedback here. Another 0.05 coupling capacitor, and we're off into the third stage. What should we use here? A tube, sure, but what kind? My Valiant uses a 12AU7, as does the Ranger. The fact is, E.F. Johnson used to parallel up both halves of a 12AU7 to get more capability in the Ranger, plus a lower plate load impedance, but that may not be necessary. A 6C4, which is actually one-half of a 12AU7, works for most AB1 modulator output stages.

These are all triodes, by the way. We're going to use either a phase inverter or an interstage driver transformer to drive the modulator tubes at this point, so we need to know where we're going here. My Valiant preamp drives a pair of 6146B's in triode-connection through its stock interstage driver transformer, if that's any help. I think there's a bit of negative feedback in there to straighten out of that scrotumless piece of iron, as well. Oops, another ROT!

Rule Of Thumb #2--Negative feedback sucks in an audiophile design, but straightens out crappy hardware in an amateur radio audio design. Why is feedback bad? Phase smear. Different frequencies propagate at different speeds through the feedback-equipped stage, so thay all arrive at times different than that of the original waveform. THD (Total Harmonic Distortion) may be down, but headaches are up! Listening fatigue, as the brain tries to reestablish the subtle interrelationships between all the subject frequencies gets real old, real fast. So, no feedback is optimal, right?

Right! Except, ham gear is generally not optimal, so we tame the savage beast with some hair of the dog that bit you, as we used to say in the music business. But there are more things to consider--phase smear, transient response deterioration, phase inversion from stage to stage, and so on. This calls for test equipment. Get an oscilloscope and an audio generator--while it isn't absolutely necessary, it's nice stuff to have and helps a lot.

More as it comes.

To get back to the WA3FLE homepage, click here.