Theory behind selecting plate resistor?

[j-po]

I've read that the plate resistor should be about twice the tube's internal resistance and there seems to be an awful much 100k's in amps. Dumble being more of an exception with 220k/150k. Got something to do with linearity something I guess. But then again, since when did we care about linearity in guitar amps. Anyone got an explanation?

[Ronsonic]

There are a few ways to approach this. There's the pages of calculation method, boring and no fun. There's the "Let's try what others have done, introduce some variables then see what it sounds like." This is the usual thing and far more entertaining. Or, take a graphic approach to calculating what's going on using the published curves with load lines.

If you google triode load lines, you're pretty sure to find a description of how to do this voodoo. For the most part we all end up using 12AX7 with something between 82K and 390K on the plate and 820 - 4.7K on the cathode.

[benoit]

I just stumbled across this article. A little ways in there's quite a lot of discussion of plate resistors.


http://www.tubecad.com/articles_2003/Gr ... index.html

[El_Martin]

Hi there!

One stage is not enough - it's the chain of stages making the sound...
A series of two/three (not to mention the nu higain amps) triode stages with or without attenuation shapes the sound. Some rely heavily on powerstage distortion and some don't.

I like to have two gain stages with different Rks. Asymmetry is the key to additional overtones. Good for cleans and lite crunch w. powerstage distortion. Not so good, if following stages also generate overtones of the overtones

When introducing some more gain stages, you have to keep an eye/ear on the generated overtones. Sound tends to be raspy. That's why the multi stage heavy distortion amps usually need snubber Cs across the Ras. They keep the generated overtones under control.

To get the best tonal results, most designers of multi channel amps end up with at least two chains of gain stages: one for the clean and one for the distorted tones. ->less compromise, more tubes

Ciao
Martin

[wscrane]

The rule of thumb is that the plate resistor is twice the rp of the tube, and that the downstream load is twice that. The whole idea is to keep noise down by using the smallest resistors you can get away with and still have a gain that is a reasonable fraction of the tube’s mu.

The main reason for using a different value is higher gain. A lot of amps use 150k, 220k, or 330k. The first stage of a Kalamazoo Reverb 12 uses 470k. I recall seeing 680k too but don’t remember what amp it was. The cost of doing this is that the noise goes up faster than the gain – partly because the tube is noisier with lower current, and partly because the resistor noise goes up as the square root of its value.

There are other reasons for using unusual values: maybe you want a little lower rp for driving a tone control, maybe your supply voltage is particularly high or low, maybe you’re using the stage to achieve a particular shelving effect, et cetera.

Morgan Jones book Valve Amplifiers is an interesting reference on choosing parts for gain stages, among other things.

[M Fowler]

I will attach a plate load resistor chart from my book.

Mark

[j-po]

Wow! Good viewpoints guys.