Oddball Mullard mullard amp circuit

[octal]

Came across this crazy amp circuit today. The ECL80 oscillates at 2Mhz and I guess the output tube grids rectify the signal to provide negative bias. (Note the lack of cathode resistors or a bias supply for the output tubes.) It doesn't seem like they're actually doing any Class D/ PWM stuff in this circuit. Just a really odd way to generate bias. Thought you fellers might find it interesting.


(Since this is an "AC/DC" circuit, the schematic is UNSAFE as drawn and is presented for THEORETICAL study only. An isolation transformer is mandatory to even prototype something like this. ) If AC/DC circuits are against the rules of the board, mods, please take it down and I'll redraw it with an isolation transformer.

[PRR]

> The ECL80 oscillates at 2Mhz and I guess the output tube grids rectify the signal to provide negative bias.

Almost.

The triode is used like in a radio's local oscillator. All simple oscillators start high-gain, build-up "to infinity", but hit something which limits their swings. Most commonly, they try to swing their grid positive, but that's too difficult, so the large AC swing is pushed negative. (In amps we say "grid blocking".)

I can't find oscillator conditions for this triode, and the common low-output LO in a radio usually only goes -10V to -15V. I assume they jazzed-up the coupling to get the -30V these power bottles need.

The negative voltage can be pictured across the 47K (a *very* common grid-leak value). It's nasty, so they take it out with resistors and capacitors to smooth it. These caps may be quite small because they filter the 2MC.

Aside from -30V for output stage, they take -3V for the second pentode because the common cathode must be grounded for good oscillation.

The second pentode is a simple phase-inverter, gain of -R7/R6 (-1).

The NFB return is wacked. They cudda flipped L6, moved R1 over under R3 C5, gotten a grounded input pot and constant NFB at all volume settings.

I smell the death of Tube TV here. When TV was young they couldn't make the tubes fast enough. Transistors came in slowly, so the tube lines kept running flat-out. But then the demand for tubes dried-up as transistors took over. So the TV Tube department is looking at lay-offs and wondering "how can we stir-up a few sales?" Transistor radios had already come in, so they had heaps of tube-radio oscillator coils.

Aside from "do not try this at home!", note that it is scaled for 200-250V power lines.

[octal]

PRR, I knew you wouldn't be able to resist giving your take on this circuit.  It seems like doing things the hard way, doesn't it? I like your "using up obsolete TV tubes and radio oscillator coils" theory.  But if the oscillator stops for any reason, your output tubes are going to be red plating! (Which I guess would lead to more sales of obsolete sweep tubes.) It's engineering-marketing genius!

[kagliostro]

It's engineering-marketing genius!

Yes may be 

K