Screen-driven pentode in guitar amp?

[_Steve]

Hello!

Im just wondering if anyone knows of any guitar amps/builds that have tried screen drive?

Just for the interest of learning more and mad science etc.

Cheers!
Steve

[sluckey]

I've seen a few amps that connect the LFO tremolo signal to modulate the screen grid to produce the tremolo effect. Selmer Zodiac Twin 30 is one example.

[Merlin]

I doubt there are any examples. Screen drive is rare enough in hi-fi amps, I can't image anyone has tried it in a guitar amp, it would be a lot of work for no clear reward.

[kagliostro]

Your question reminded me of something, but no, this too involves connecting an oscillator signal, even if the tube has more than one screen grid.

It's basically something similar to the circuit Sluckey mentioned, but different anyway.

I'm posting the diagram as a curiosity; the tube in question is a 6BE6.

Franco

[_Steve]

I've seen a few amps that connect the LFO tremolo signal to modulate the screen grid to produce the tremolo effect. Selmer Zodiac Twin 30 is one example.

That is very cool thanks for sharing the link! Years ago I was thinking about trying to make a crude compressor using the same method but never got around to it.

[HotBluePlates]

Im just wondering if anyone knows of any guitar amps/builds that have tried screen drive?

A guitar amp usually uses pentodes (EL84, EL34, some 6BQ5s) or beam power tubes (6V6, 6L6, some 6BQ5s) rather than triodes because we can drive them with a smaller signal.

   - Push-pull 300B triodes can make 20w of output power, but need about 67v peak drive signal to get there.

   - Push-pull 6L6s will make 26 watts of output power with only 22v peak drive signal (Page 4 of the data sheet).

   - Push-pull EL34 are much more sensitive, giving 48 watts of power output with only about 25v peak drive signal if the screen can contribute to output power in a "distributed loading" (aka "ultralinear") connection (Page 2 of the data sheet).  20-26 watts of power output would happen at a proportionately lower drive signal.


But what happens when you drive them by wiggling G2 (the screen) instead of G1 (the control grid)?

   - Look at the EL34's triode connected µ, or "µ_g1-g2":  we see figure of 10.5 or 11  (data sheet Page 1)

   - These figures say essentially the same thing in different ways:  how much more effective is G1 at controlling plate-output than G2?

   - A figure of "11" means you need 11x the drive signal applied to the screen than you used at G1 to wiggle the plate.  The "25v peak drive" used before for the EL34s becomes a requirement to swing the screen 25v x 11 = 275v peak. 

   - You will need two 275v peak outputs to drive your push-pull EL34s.  And I you get that from a tube plate, you will need at least 275v x 3/2 = 413v (and preferably a lot more) for the driver/inverter stage.

   - The challenge of high supply voltage to support a high drive-signal requirement would lead many designers to use a step-up driver-transformer rather than a tube here.  There are weight, cost, space, and bandwidth penalties to pay going down this road.  However, if you wanted perfectly-clean push-pull output, a driver transformer can probably ensure exactly equal-and-opposite output signals.  (But most guitarists aren't looking to play "perfectly clean" as "small amounts of distortion" add harmonics & "excitement" to what are perceived as "clean tones" in a mix)

I've seen a few amps that connect the LFO tremolo signal to modulate the screen grid to produce the tremolo effect. Selmer Zodiac Twin 30 is one example.

I never learned enough about magic eye tubes to have feel for their drive-requirements.

But I notice Selmer took the plate-output of the trem oscillator, which is a "large AC voltage."

Vox tied the trem oscillator cathode to an EF86 cathode in the AC10 (and early-version AC15).  That is a "small AC voltage" but there is also a connection to the EF86 plate via R24, R26 (both 330kΩ_ and C14 (which is about 338kΩ at 10Hz).  The parallel-wiggle-path seems effective at modulating the EF86 plate signal (though I will confess I'm not entirely sure whether that is because of a "signal-input" or simply that the oscillator is changing both plate- & cathode-voltages of the EF86 at the LFO-rate and imposing its modulation).