Thanks for the link John, got a question though.

If I understand it correctly, your annode load combination,
capacitively coupled through the 0.1 (C21) coupling cap must approach
the 1.5k cathode resistor in value for the circuit to be unity gain,
correct? In looking at V6-A and we remove the connection of R33 and
grid connection going down to R210, I don't need that stuff, we have
the 2.2M (R34), 220k (R31) in series with 220k (R29) all in parallel
with the 100k plate resistor (R30) and we want the resulting value to
approach the 1.5k cathode resistor...right? Assuming my sub will be
around 20k load at the RCA jack, it doesn't appear to be low enough in
value to approach the value of the cathode resistor for unity gain?

Also, I guess you need the high value of series resistance 220k (R29)
so that the feedback/load value combination doesn't load the loudness
pot and input signal too much?

Is it possible to reduce the plate resistor more and thus loading the
dynamic plate resistance (ra) more since we aren't using this as a
'gain' stage?

If anyone else is looking, the schematic is here

http://users.rcn.com/jbyrns/pics/LC-21-23.jpg

Thanks,
Gilbert

In article , Gilbert Bates
wrote:

Thanks for the link John, got a question though.

If I understand it correctly, your annode load combination,
capacitively coupled through the 0.1 (C21) coupling cap must approach
the 1.5k cathode resistor in value for the circuit to be unity gain,
correct? In looking at V6-A and we remove the connection of R33 and
grid connection going down to R210, I don't need that stuff, we have
the 2.2M (R34), 220k (R31) in series with 220k (R29) all in parallel
with the 100k plate resistor (R30) and we want the resulting value to
approach the 1.5k cathode resistor...right? Assuming my sub will be
around 20k load at the RCA jack, it doesn't appear to be low enough in
value to approach the value of the cathode resistor for unity gain?

Also, I guess you need the high value of series resistance 220k (R29)
so that the feedback/load value combination doesn't load the loudness
pot and input signal too much?

Is it possible to reduce the plate resistor more and thus loading the
dynamic plate resistance (ra) more since we aren't using this as a
'gain' stage?

If anyone else is looking, the schematic is here

http://users.rcn.com/jbyrns/pics/LC-21-23.jpg

Hi Gilbert,

Unity gain is not achieved in this circuit by loading down the anode as
you suggest, rather unity gain is achieved by the negative feedback
provided by two 220k resistors, R31 and R29, which form a voltage divider
putting the grid at a "virtual ground". This means that the voltage
across the two resistors must be equal and the voltage at the signal input
end of R29 will be forced to be approximately equal and opposite to the
signal voltage at the anode end of R31, hence unity gain. The purpose of
R29 is not to prevent loading of the volume control, but rather is to
provide the required resistance in the input leg of the feedback network
to insure unity gain. If the grid were driven from a low impedance source
then the feedback resistor R31 would have little effect and the tube would
have its full gain.

The 12AX7 shown in the schematic is a pretty wimpy tube, if you are going
to drive a 20k load, I would use a more muscular tube like a 12AU7 or
better.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

Thanks for the link John, got a question though.

If I understand it correctly, your annode load combination,
capacitively coupled through the 0.1 (C21) coupling cap must approach
the 1.5k cathode resistor in value for the circuit to be unity gain,
correct? In looking at V6-A and we remove the connection of R33 and
grid connection going down to R210, I don't need that stuff, we have
the 2.2M (R34), 220k (R31) in series with 220k (R29) all in parallel
with the 100k plate resistor (R30) and we want the resulting value to
approach the 1.5k cathode resistor...right? Assuming my sub will be
around 20k load at the RCA jack, it doesn't appear to be low enough in
value to approach the value of the cathode resistor for unity gain?

Also, I guess you need the high value of series resistance 220k (R29)
so that the feedback/load value combination doesn't load the loudness
pot and input signal too much?

Is it possible to reduce the plate resistor more and thus loading the
dynamic plate resistance (ra) more since we aren't using this as a
'gain' stage?

If anyone else is looking, the schematic is here

http://users.rcn.com/jbyrns/pics/LC-21-23.jpg

Hi Gilbert,

Unity gain is not achieved in this circuit by loading down the anode as
you suggest, rather unity gain is achieved by the negative feedback
provided by two 220k resistors, R31 and R29, which form a voltage divider
putting the grid at a "virtual ground". This means that the voltage
across the two resistors must be equal and the voltage at the signal input
end of R29 will be forced to be approximately equal and opposite to the
signal voltage at the anode end of R31, hence unity gain. The purpose of
R29 is not to prevent loading of the volume control, but rather is to
provide the required resistance in the input leg of the feedback network
to insure unity gain. If the grid were driven from a low impedance source
then the feedback resistor R31 would have little effect and the tube would
have its full gain.

The 12AX7 shown in the schematic is a pretty wimpy tube, if you are going
to drive a 20k load, I would use a more muscular tube like a 12AU7 or
better.


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

Thanks John,

I played around with the circuit last night and a few things came to
light in my head.

Yes, the two resistors must be equal for unity gain and/or I guess you
could trim them a little for a shift up or down in gain perhaps.

I understand now how I incorrectly stated the loading of R29 on the
volume control but I was still curious about the relative size of R29,
R31. After experimenting, I seem to find that the size of R29 helps to
isolate the input signal from the feedback. By reducing the size of
R29, R31 you *inject* more of the feedback signal back onto the volume
control which reduces or feeds back on the input signal. The 220K size
seems to have a small impact at the volume control but is minimal. I
guess then that R29 (and R31) could approach the maximum size allowed
for the particular tube, 1M for the 12AX7 in this case? Or would
feedback current be reduced to a point to where it would be
impractical?

I'm setting up an SE 2A3 with a single 6SL7 gain stage, yeah I know,
should have used a 6SN7 or paralled the 6SL7, but I think I've worked
out enough of the details to make the single 6SL7 stage work ok and
want to use the other half in a buffered sub-woofer output. While the
6SL7 might be whimpy too, what I plan on doing is to adjust the
feedback loop to give me unity gain with a 20k load attached. It pops
up slightly in gain when the 20k load is removed obviously.

The operating point of the 6SL7 is currently set at a little over a
milliamp with a 100k plate and 1k cathode resistor at about 275V B+.

Does this make sense?

Thanks!!!