Cap Size!!!

[emj12]

22uf interstage coulping cap in 12ax7 preamp..... bad idea???

[R.G.]

Somewhere between possibly and probably. 

Diminishing returns bites you hard here. If the 12AX7's grid leak resistor is the common 1M, then you hit full audio range (that is, down to 20Hz) at

C = 1/(2*pi*20*1M) = 0.007uF.  Anything bigger than that just makes  you have bass response below 20Hz where you can't hear it.

And the bigger that cap, the slower the stage recovers from any bias shift caused by overdriving the input into positive grid conduction.

[emj12]

Somewhere between possibly and probably. 

Diminishing returns bites you hard here. If the 12AX7's grid leak resistor is the common 1M, then you hit full audio range (that is, down to 20Hz) at

C = 1/(2*pi*20*1M) = 0.007uF.  Anything bigger than that just makes  you have bass response below 20Hz where you can't hear it.

And the bigger that cap, the slower the stage recovers from any bias shift caused by overdriving the input into positive grid conduction.

would this slow recovery result in distortion?

Also this cap will be going into a 1M gain pot

[R.G.]

When you overdrive a tube grid, the grid conducts. This charge gets dumped onto the coupling cap, and an any recovery is purely by the grid leak resistor discharging the cap back to zero. The grid acts like a rectifier, charging the cap.

The voltage on the cap turns the tube more off until it bleeds away, making for what is usually termed "farting" or "gating". The tube quits doing anything until the voltage drains back down to where the tube can amplify.

[PRR]

22uFd is _probably_ an electrolytic.

Electrolytics leak DC. Not too bad in low-impedance circuits, but more than enough to overwhelm the normally extra-teeny current around a 12AX7 grid. The 12AX7 is mis-biased, moving it into a distortion zone. This is not even a useful distortion effect because electrolytic leakage is very unpredictable over temperature, short time, and long-time.

WHY do you use a 22uFd for a grid coupling cap? As R.G. says, 0.01uFd is ample, and usually cheap.

[emj12]

22uFd is _probably_ an electrolytic.

Electrolytics leak DC. Not too bad in low-impedance circuits, but more than enough to overwhelm the normally extra-teeny current around a 12AX7 grid. The 12AX7 is mis-biased, moving it into a distortion zone. This is not even a useful distortion effect because electrolytic leakage is very unpredictable over temperature, short time, and long-time.

WHY do you use a 22uFd for a grid coupling cap? As R.G. says, 0.01uFd is ample, and usually cheap.

Basically because I have spares and nothing around a 10uf other than thwe 22uf solen caps. I what at least 10uf so i will not encounter any frequency attenuation.

[PRR]

> I what at least 10uf so i will not encounter any frequency attenuation.

As a wise man said: "0.007uF.  Anything bigger than that just makes  you have bass response below 20Hz where you can't hear it."

0.007uFd caps are odd; we use 0.01u (17Hz).

22uFd into typical 1Meg grid resistor give -3dB response to 0.0072Hz! This is equal to one vibration every 2 minutes! 10,000 times (13 octaves!) lower than guitar!

Additionally it takes a minute to charge-up, and a minute to recover from overdrive. "slow recovery", "farting".

A handful of 0.1u and 0.01u caps is good to have, and not expensive.

[R O Tiree]

emj - look at that formula again...

f=1/(2*pi*R*C)

Solve for "C" and you get:

C=1/(2*pi*R*f)

f = 20 Hz (limit of human hearing and about half of the lowest freq you can get on a 4-string bass guitar)
R=1M

C=0.007µF (nearest common value 0.01µF, which will lower the response, as PRR says, to about 17 Hz, although you could go with 0.0068µF, perhaps? That's as close as it gets.)

So, that's 2 1/2 wise men saying, "0.01µF is enough and nasty, farty things will happen if you go 1000 or 2200 times higher than that".

Hmmm... do I count as even 1/2 of a wise man, yet? Maybe "2 1/4 wise men", then

[emj12]

I see... thanks for your help guys.