Gain control, that really controls gain

[Whiskeyrebel]

Are there any circuits in which the gain control really alters the voltage gain directly? Most circuits I see have a fixed-gain amplifying stage followed by an adjustable voltage divider.

I know that in math, A X 10 X 0.3 = 3A, and 3 x A = 3A. Is there any practical difference in electronics whether you boost a signal by a large amount and then use a small fraction of it, VS boosting it by a small amount and then use all of it?

It does seem like the 'fuzz' control on a Fuzz Face is sort-of a direct control on gain.

[Blue Strat]

There's no difference to the end result. In either case, you have to use a pot.

Using a pot after the gain circuit may limit noise somewhat in the lower "gain" settings since it's allowing only a small portion of the amplified signal to pass.

[RedMan]

You could go with negative feedback but you could run into oscillation problems in a high gain circuit if not careful. Also the higher impedance output seems do more of what we want in that type of circuit. Otherwise we may as well be using SS.

[vibroverbus]

Whisker -

Re:your first question... you can absolutely use a pot to split the plate load and/or cathode resistors and alter the gain characteristics of the stage. Per Mike's comment there's noise and stability problems (and possibly safety) if you have the stage DC current, especially high voltage, running around on long leads through pots. Lots of noise turning the pot if you insert it in the plate load. Which is probably why you seldom see it. In my retarded novice experimentation with this on the plate it's totally impractical (noise and stability) although I am running an amp with variable resistance on the cathode of the PI as a semi-MV setup...

On one hand, it's probably one of many neglected techniques/concepts in our hide-bound rigid tube guitar amp universe, on the other hand, there's reason the ancestor circuits ended up where they are. Sticking a solid state device in there controlled by a pot would eliminate the DC noise problem and give you control, and that's exactly the kind of thing fancy studio tube pre-amp designers do. Of course we dumb guitar players are too cool to have a transistor control our tubes.

Re: Your second point/question with the arithmetic, it's fairly accepted that multiple moderate gain stages create better results than huge steps. The science behind that is that more you drive and hard-clip a stage the more odd-order artifacts dominate, whereas what we're after is tube generated 'soft-bounding' even-order artifacts that happen at more moderate gain points. Which is why guitar amps ALWAYS throw away (in the tone stack and other filters and voltage dividers through the circuit), and/or fail to exploit tons of absolute available gain.

[Pete Cage]

Tapco was doing this with their mixers back in the '70s. The advantage was improved s/n ratio, since the preamp stage wasn't running wide open the whole time.

I'm not sure that the s/n issue is that big of a deal for guitarists, at least in low-gain applications, but the sonic results might be different/interesting/unique.

[Wakarusa]

"direct gain control", to me anyhow, would be directly controlling the tube mu. The other schemes discussed work around attenuating either the input or output signal as opposed to changing how the tube amplifies. Whether you use a pot (with high freq rolloff effects) or solid state (often with capacitance and other impacts), it's still a voltage divider trick.

There are variable mu tubes out there (google "remote-cutoff") but they don't really work the way we'd want 'em to for this application (they reduce mu as the input signal gets larger). There are probably some other tricks you could pull to get sharp cutoff tubes to behave the way we want.. using variable resistances (solid state toys, light dependent resistors, etc) in the plate and cathode circuits, setting up a cascode or SRPP with a secondary input to control amplification, but I'd think the minor improvement in performance/noise (or just the bragging rights of saying "I control gain directly") would be outweighed by the circuit complexity and cost.

[mark norwine]

I have to say.....

Todd,

Not only do you always say it "right", you always say it "well"....

mn

[RedMan]

"direct gain control", to me anyhow, would be directly controlling the tube mu. The other schemes discussed work around attenuating either the input or output signal as opposed to changing how the tube amplifies. Whether you use a pot (with high freq rolloff effects) or solid state (often with capacitance and other impacts), it's still a voltage divider trick.

There are variable mu tubes out there (google "remote-cutoff") but they don't really work the way we'd want 'em to for this application (they reduce mu as the input signal gets larger). There are probably some other tricks you could pull to get sharp cutoff tubes to behave the way we want.. using variable resistances (solid state toys, light dependent resistors, etc) in the plate and cathode circuits, setting up a cascode or SRPP with a secondary input to control amplification, but I'd think the minor improvement in performance/noise (or just the bragging rights of saying "I control gain directly") would be outweighed by the circuit complexity and cost.

Negative feedback isn't attenuating an already clipped signal, therefore the effect is much different than attenuation. It's also the most common form of "gain" control on audio equipment. The other methods you mentioned here may be fun to play with but are impractical, as you have mentioned.

I always thought a cool (but somewhat impractical) way to do something like that would be to have a few different tube sockets that could be switched in/out silently with different (or same) tube types that could provide different amounts of gain/tone.

[Laird_Williams]

Whisker -

Re:your first question... you can absolutely use a pot to split the plate load and/or cathode resistors and alter the gain characteristics of the stage. Per Mike's comment there's noise and stability problems (and possibly safety) if you have the stage DC current, especially high voltage, running around on long leads through pots. Lots of noise turning the pot if you insert it in the plate load. Which is probably why you seldom see it. In my retarded novice experimentation with this on the plate it's totally impractical (noise and stability) although I am running an amp with variable resistance on the cathode of the PI as a semi-MV setup...

On one hand, it's probably one of many neglected techniques/concepts in our hide-bound rigid tube guitar amp universe, on the other hand, there's reason the ancestor circuits ended up where they are. Sticking a solid state device in there controlled by a pot would eliminate the DC noise problem and give you control, and that's exactly the kind of thing fancy studio tube pre-amp designers do. Of course we dumb guitar players are too cool to have a transistor control our tubes.

Re: Your second point/question with the arithmetic, it's fairly accepted that multiple moderate gain stages create better results than huge steps. The science behind that is that more you drive and hard-clip a stage the more odd-order artifacts dominate, whereas what we're after is tube generated 'soft-bounding' even-order artifacts that happen at more moderate gain points. Which is why guitar amps ALWAYS throw away (in the tone stack and other filters and voltage dividers through the circuit), and/or fail to exploit tons of absolute available gain.
Said more bluntly - pass 300-400 volts through a pot that fails and then make contact with your hand to the pot. NO commercial amp manufacturer is going to risk that.