i am looking for a circuit diagram for a tube direct box. something with an
input impedance of at least 10Mohm. thanks.

Scott Dorsey wrote:

Go to the RCA Radiotron Handbook and look up "cathode follower." Put a
transformer on the output to balance it. The input Z is entirely due to
the leak resistor, so if you use a 10M leak resistor, you have a 10M input.

Very clean and transparent if it's biased right. The transformer will
be pretty much all of the coloration and it'll be damn near impossible to
overload.

How do you calculate the output impedance of a cathode follower? I
think I know how to measure it, but I'd rather do the math. It can't
be higher than the cathode resistor, right?

Also, I know any cathode follower will have a gain less than unity, but
what can you typically expect from a 12AX7 or 12AY7? I know a 12AY7
has lower gain than a 12AX7 in a typical plate-loaded circuit, but is
there a substantial difference in the signal level you'll get out of
them in a cathode follower circuit? Seem like a good idea to use both
halves of a 12AY7 in parallel and thereby use a lower-ratio output
transformer.

Sorry, I dunno where Johnny put the Raditron. So I ask and I ask.

ulysses

Justin Ulysses Morse wrote:
Scott Dorsey wrote:

Go to the RCA Radiotron Handbook and look up "cathode follower." Put a
transformer on the output to balance it. The input Z is entirely due to
the leak resistor, so if you use a 10M leak resistor, you have a 10M input.

Very clean and transparent if it's biased right. The transformer will
be pretty much all of the coloration and it'll be damn near impossible to
overload.

How do you calculate the output impedance of a cathode follower? I
think I know how to measure it, but I'd rather do the math. It can't
be higher than the cathode resistor, right?

The math is in the Radiotron Handbook. For a triode with a single
cathode resistor, Rk, the output Z is Rk/1+gmRk. For a pentode you
fudge the gm for whatever the operating point you have the screen set
at. I don't remember the other configurations but they are all in
the book.

Also, I know any cathode follower will have a gain less than unity, but
what can you typically expect from a 12AX7 or 12AY7? I know a 12AY7
has lower gain than a 12AX7 in a typical plate-loaded circuit, but is
there a substantial difference in the signal level you'll get out of
them in a cathode follower circuit? Seem like a good idea to use both
halves of a 12AY7 in parallel and thereby use a lower-ratio output
transformer.

You can get as low an output impedance as you want, as long as you don't
want much voltage swing. From a 12AT7 you should be able to drive a
50-ohm load up to -10 dB without any linearity issues. Look at the output
stage of the Harman Kardon Citation II for a nice example of a well
designed follower output stage.

Sorry, I dunno where Johnny put the Raditron. So I ask and I ask.

You need it! It has everything you need to design tube or fet circuits.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On Tue, 02 Sep 2003 18:21:01 -0500, Justin Ulysses Morse
wrote:

How do you calculate the output impedance of a cathode follower? I
think I know how to measure it, but I'd rather do the math. It can't
be higher than the cathode resistor, right?

The active device's output impedance is the reciprocal of its
transconductance. Circuit resistances all contribute as you'd
expect 'em to.

Also, I know any cathode follower will have a gain less than unity, but
what can you typically expect from a 12AX7 or 12AY7? I know a 12AY7
has lower gain than a 12AX7 in a typical plate-loaded circuit, but is
there a substantial difference in the signal level you'll get out of
them in a cathode follower circuit? Seem like a good idea to use both
halves of a 12AY7 in parallel and thereby use a lower-ratio output
transformer.

Lightly loaded high-mu devices have gains of about mu/mu+1, IOW,
pretty darned close to unity.

A good way to think about loading is that loading should not be
heavier for a cathode follower than for common cathode use. The
device operates on the same load-line, same plate curves, etc.


Chris Hornbeck