Tuesday, January 14, 2014

Building a ca.1928 Dual Triode Class-A Audio Amplifier

This is another recap of a previous project.

Basically the task came to me to clean up an old Victrola Radiola audio cabinet from 1928. The phonograph worked fine so it was left be, but the radiola (the AM radio) was toast. I figured I might at best get a few tubes to play around with or just a couple of inductors. It turns out nearly all of the tubes were good, and the PSU worked fine, but the tuning coils and caps were wrecked. I spent a bit over a week going through various experiments with the old tubes and modern tech to see just what all I could do with the parts I found. Here are the data files and the videos of progress:
Power Supply and Output Coupling:
RF amp, tuning, and AF amp "final" sections:

Simplified schematic:
Modified schematic showing rebuild as dual triode Audio Amp:
Videos detailing the various incarnations and progress:

Final Result:
The trickiest part, which is not on the last schematic shown, was finding a good negative feedback scheme (NFB) to try to balance the frequency response (enhance bass response by impeding treble's gain) and cut down on some of the hum. Unfortunately not all of the hum can be removed without remaking the entire power supply section, because these triodes are directly heated cathode tubes (filament valves) powered by AC, so since the filament IS the cathode the AC on the filament is imposed on the output by design. One has to remember that back in 1928 having sound come out of a box without a little man inside it was AMAZING, and having hum was of no consequence.
NFB on thermionic amplifiers simply involves taking some of the antiphase output from the output transformer and feeding it in to either the grid or the cathode (most often the cathode) of first amplifier stage. This is as simple as a resistor and a capacitor in series between the antiphase leg of the output or impedance matching transformer and the pre-amp triode's filament. I can't remember the values I used, but that info is of little consequence really.
What all this really shows is that triode class-A amplifiers are SUPER SIMPLE if you can find the right parts. The real unspoken hero here is the interstage transformer that takes the output of the first triode and couples it to the grid of the second triode. This is a job of both precision impedance matching and voltage transformation. There are no modern equivalents to get off a shelf. Interstage transformers were abandoned in the 30s for the most part when capacitive coupling was chosen as its successor. Using capacitive coupling is far cheaper, but it does require the engineer to figure out a proper value of anode resistor. The resistor forms a voltage divider with the tube, think of it this way; the tube becomes a variable pull-down resistor. The capacitor then couples off the dV (change in voltage) seen at the voltage divider output node. This scheme may be much cheaper to implement but there is a marked reduction in gain to having a series transformer directly couple the output. Such is the way of technology progression though.


No comments:

Post a Comment