<<     TUBE CAD JOURNAL     >>

October 

Page 10

Copyright © 2003 John Broskie    All Rights Reserved

If so, rush to the patent office, but first come up with a new name as everyone knows that you can judge an amplifier by its title. For example, when the phase splitter that was used in countless Audio Research amplifiers was called the “cross-coupled phase splitter,” no one cared and it was considered somewhat dead sounding, but now that it is known as the “Van Scoyoc phase splitter” it is much more interesting and better sounding as well.

But which name? “Mega-path” or “Ultra-Mega-Path” or “Megacirc” or “Ultratron,” yes that’s it: the Ultratron amplifier; single-ended-class-A glory by the wave of a tongue.

What a joy it is to live in an age of miracles and magic, unburdened by logic and unfettered by common sense. Fifty years ago, people were not so lucky. Back then, most people did not believe in astrology and UFOs. In fact, tube circuitry was still taught in universities and libraries were filled with solid, well researched pages devoted to the vacuum tube and its functioning. In those bleak days, many of the brightest, most capable scientists and engineers devoted themselves exclusively to understanding how a tube worked in a circuit; and we all know what party poopers those kinds of people are.

Imagine if a time capsule had a reverse gear and that we could send the pearls of today’s better understanding of the vacuum tube back to the fifties. Here’s the scene: the oppressively stolid electrical engineer, the one with Buddy Holly glasses and white socks, finds what looks like a thermos on his desk. Opening it, he finds a single sheet of paper upon which is written the following:

 

300B*, 2A3, 45, 76, 211, 845, carbon resistors and oil capacitors, single-ended and push pull transformers, circlotron with triodes, tube rectifiers, octal tubes, connect the power supply capacitor to the output tube’s cathode.

* actually anything at all marked Western Electric including lamps and doorstops

Of course, so dense a dollop of wisdom might be too much for any one human being to absorb without passing out (I felt a bit queasy just typing the list). It would be as if Mozart were presented with a performance of John Cage’s famous composition 4'33 (of Silence). Really, it’s truly amazing that so much has come from so little. Or maybe… confusing our list of received truths for someone’s old shopping list, the engineer would hurl it in the garbage. Alas, we will never know.

Wait one sarcastic, solid-state-loving, I’m-not-as-trendy-as-thou minute: how do we know that these two amplifiers do in fact function identically?

Well, one could derive all the formulas for gain, Zo, PSRR, bandwidth, and distortion for both amplifiers and then compare the formulas (this is my preferred path, as it can be done on the back of an envelope, while eating cereal). Or one could actually build two amplifiers and test both for gain, Zo, PSRR, bandwidth, and distortion. Or, lastly, one could model both amplifiers in SPICE and compare results, which is what we will do next, as it requires the least amount of work (it’s hard to format the back of an envelope for the internet).

On the next page, we see the totem-pole amplifier with the center-shifted reference point modeled in B2 Spice A/D. This amplifier uses 6AS7 tubes and a 32-ohm load resistance, which implies that four 6AS7 type triodes would actually be used per phase leg.

The secret to using SPICE successfully is to realize that you are not playing a game of The Sims: you do not want as much detail and reality mapping as you can get. Instead, you only want to get the same results that reality would give; hence the SPICE op-amp models that hold only a few transistors, not the several dozen that the real op-amp holds. A SPICE model that held a perfect 1:1 representation of reality would as useful and cumbersome as a map of New York City likewise rendered in a scale of 1:1.