The AM Forum

I do a lot of weird stuff, not all of it Ham related and one of the recent projects was repairing an Amplifier Research 100-watt broadband amplifier. The spec sheet says its broadband and flat from 100 kHz to 100 MHz and tolerant of gross output matches. This amplifier is about thirty years old and no longer supported so that?s why I am working on it.
No documentation and had to draw up a basic schematic of the amplifier and this is the output stage in a ?simplified? version, the real thing has additional decoupling and unlike the drawing you are seeing it uses twelve 8122 tubes all in parallel.
The one Ohm resistor shunting to ground is what throws me, figure this amplifier is dissipating huge amounts of power to get the 100 watts of output and has no tuned circuits so its efficiency is marginal at best. Maybe five or ten percent?
Just thought I would pop it up here for others to see how to build maybe the most inefficient amplifier possible but it is broadband.
 

   I'm wondering what that amplifier was originally used for. Probably some sort of industrial use, like drying wood, maybe. I believe there was at least one industrial wood drying machine that used 810 triodes.
   That actual amount of energy dissipated in that 1 ohm resistor would depend, of course, on the reactance of both the capacitor and the inductor in series with the resistor, with the reactance of the inductor decreasing as frequency decreases and the reactance of the capacitor increasing as the frequency decreases.
   The schematic shows no circuitry for applying a regulated DC bias to the cathodes. The inductor from the cathodes to ground would need to have a very high reactance throughout the operating frequency range and relatively high resistance at DC. Since the bias will be whatever is dropped across the DC resistance of the inductor, the bias and thus the operating point would, I'd think, be all over the place.
   That 8122 tube would seem to be a good tube for amateur amps, at 400 watts of plate dissipation, but they don't seem to be very common these days.

I don?t know what this is, but one look at the curves for an 8122 shows that a grounded grid and screen voltages of several hundred volts will render put this in a saturated state pulling over one ampere per tube even at 600volts on the plates.
See curve for 400volts on the screen at zero grid voltage in image.
Am I missing something?