Dear Fellow Rodents,
Is a Cathode Follower preamp an ideal input for a tube amp? Although with a
theoretical max gain of one and a low output impedance, why do some advocate
that it's best to have as the last stage in your preamp, a cathode follower?
I believe that I understand the theory, but have no empirical knowledge. I
know a fellow who added a CF to his otherwise well sounding preamp to
present the best load to his tube amp. Also, there is an outfit who adds a
CF to all the preamps coming into his shop for upgrading.
Perhaps someone can expand on CF and if they do indeed, present some sonic
benefits. Thanks in advance.
Cordially,
west

Depending on the preamp, I choose either SRPP or Loftin-White CF.

There was a bit of discussion on this in my RIAA preamp thread he
http://geek.scorpiorising.ca/cgi-bin...m=10699406 36

and http://www.tubecad.com has some awesome stuff in their back issues.

Personally, after listening to SRPP and L-W, I'll never go back to a
straight CF again ;-)

--
Gregg
*It's probably useful, even if it can't be SPICE'd*
http://geek.scorpiorising.ca

west wrote:

Dear Fellow Rodents,
Is a Cathode Follower preamp an ideal input for a tube amp? Although with a
theoretical max gain of one and a low output impedance, why do some advocate
that it's best to have as the last stage in your preamp, a cathode follower?
I believe that I understand the theory, but have no empirical knowledge. I
know a fellow who added a CF to his otherwise well sounding preamp to
present the best load to his tube amp. Also, there is an outfit who adds a
CF to all the preamps coming into his shop for upgrading.
Perhaps someone can expand on CF and if they do indeed, present some sonic
benefits. Thanks in advance.
Cordially,
west

I use a CF as the output tubes at a schematic at

http://www.turneraudio.com.au/htmlwe...tubepreamp.htm

The reason for doing so is to isolate the effects of input impedance and
capacitance
of a power amp input circuit, and cable from an other wise high impedance source

if the signal was taken from an anode circuit, or from a 100 k volume pot.
The cathode follower appears after the gain pots, and acts as a "buffer"
between power amp and preamp.
The typical output of a preamp might be 10 k ohms or much higher,
if a CF is not used, but the CF Ro is under 1k ohms,
so there are negligible high frequency losses.

Patrick Turner.

Gregg wrote:

Depending on the preamp, I choose either SRPP or Loftin-White CF.

There was a bit of discussion on this in my RIAA preamp thread he
http://geek.scorpiorising.ca/cgi-bin...m=10699406 36

and http://www.tubecad.com has some awesome stuff in their back issues.

Personally, after listening to SRPP and L-W, I'll never go back to a
straight CF again ;-)

I have always found CF to be the most transparent
of amplifier stages.

From an engineering point of view,
I also like bootstrapped followers, and the White, although
I have reckoned I have never needed to use a White
for an output CF, because the typical signal is less than 0.2vrms,
and a simpler plain CF does the trick.
The use of a CCS DC sink from cathode to a negative voltage
is the my present way to set the follower up, because then the
input grid can be direct connected to a preceding
volume pot, and the output is nearer 0V, which is less stress on the output cap.
The CCS increases the load value seen by the tube, and
the amount of thd is reduced maybe 12 dB down from the already
very low levels with a plain R load.

Patrick Turner.



--
Gregg
*It's probably useful, even if it can't be SPICE'd*
http://geek.scorpiorising.ca

Sonic benefits, They shouldn't color the sound in any way if thats what you
mean be sonic benefits. What they do is lower the output impedance of the
preamp enough so that what ever load they see on the output will have
negliable effect.

"west" wrote in message
...
Dear Fellow Rodents,
Is a Cathode Follower preamp an ideal input for a tube amp? Although with
a
theoretical max gain of one and a low output impedance, why do some
advocate
that it's best to have as the last stage in your preamp, a cathode
follower?
I believe that I understand the theory, but have no empirical knowledge. I
know a fellow who added a CF to his otherwise well sounding preamp to
present the best load to his tube amp. Also, there is an outfit who adds a
CF to all the preamps coming into his shop for upgrading.
Perhaps someone can expand on CF and if they do indeed, present some sonic
benefits. Thanks in advance.
Cordially,
west

Behold, Patrick Turner signalled from keyed 4-1000A filament:



Gregg wrote:

Depending on the preamp, I choose either SRPP or Loftin-White CF.

There was a bit of discussion on this in my RIAA preamp thread he
http://geek.scorpiorising.ca/cgi-bin...m=10699406 36

and http://www.tubecad.com has some awesome stuff in their back issues.

Personally, after listening to SRPP and L-W, I'll never go back to a
straight CF again ;-)

I have always found CF to be the most transparent of amplifier stages.

Exactly. May as well use a 2N5088 :-p

--
Gregg
*It's probably useful, even if it can't be SPICE'd*
http://geek.scorpiorising.ca

The reason for doing so is to isolate the effects of input impedance and
capacitance
of a power amp input circuit, and cable from an other wise high impedance
source

if the signal was taken from an anode circuit, or from a 100 k volume pot.
The cathode follower appears after the gain pots, and acts as a "buffer"
between power amp and preamp.
The typical output of a preamp might be 10 k ohms or much higher,
if a CF is not used, but the CF Ro is under 1k ohms,
so there are negligible high frequency losses.

Patrick Turner.

Viola! This is the explanation that turned on my rather dim light bulb. I
don't know about White or Black's circuits because my valve experience is
short, but I do understand electronics. So Pat, tell me, what's with those
who place 100K level adjust pot on the input of their amp? Besides changing
the level. doesn't this change the input impedance load and is this
something undesirable? What are the sacrifices, if any, in using an input
amp pot for volume control? Thanks as always.
Cordially,
west

Gregg wrote:

Behold, Patrick Turner signalled from keyed 4-1000A filament:

Gregg wrote:

Depending on the preamp, I choose either SRPP or Loftin-White CF.

There was a bit of discussion on this in my RIAA preamp thread he
http://geek.scorpiorising.ca/cgi-bin...m=10699406 36

and http://www.tubecad.com has some awesome stuff in their back issues.

Personally, after listening to SRPP and L-W, I'll never go back to a
straight CF again ;-)

I have always found CF to be the most transparent of amplifier stages.

Exactly. May as well use a 2N5088 :-p

I wouldn't use an emitter follower output stage on a tube pre.

I use enough stages of bootstrapped followers, ( like mu-followers, SRPP )
in my preamps, and they are all pretty transparent as well.
I like the sound, which seems engaging, real, and what I have grown to
expect from a decent tube amp, and better than other tube pres I have tested them against.
The CF at the outputs does no harm, imho.

The White is a PP circuit, and should sound quite blameless.

The measured performance of the bootstrapped follower
gives me 0.1% thd, nearly all 2H, at 10vrms output
and since I only use 0.2vrms max, the thd is negligible.

A normal CF using 1/2 a 6SN7 / 6CG7 might give around
0.07 % at 10v output, and 0.02% at 0.2v output,
and again its mostly 2H, although some 3H is produced
by imd effects of feeding back the 2H with only a moderate amount of NFB.

Using a CCS in the CF tail is enough to reduce the CF thd to less than 0.01%
up to a volt or two, providing the load imposed by the following
power amp is over say 100k.

In one line stage preamp I made, I had bjt CCS
used as loads for the 12AU7 gain tube, which
was then usable with a cap coupled 50k Alps pot,
so 50k was the load seen by the gain tube, then the CF
after the pot had a CCS tail, and the amp was "sweet as"
after I selected the quietest and least microphonic from a
batch of 35 tubes I had accumulated.

Patrick Turner.



--
Gregg
*It's probably useful, even if it can't be SPICE'd*
http://geek.scorpiorising.ca

west wrote:

The reason for doing so is to isolate the effects of input impedance and
capacitance
of a power amp input circuit, and cable from an other wise high impedance
source

if the signal was taken from an anode circuit, or from a 100 k volume pot.
The cathode follower appears after the gain pots, and acts as a "buffer"
between power amp and preamp.
The typical output of a preamp might be 10 k ohms or much higher,
if a CF is not used, but the CF Ro is under 1k ohms,
so there are negligible high frequency losses.

Patrick Turner.

Viola! This is the explanation that turned on my rather dim light bulb. I
don't know about White or Black's circuits because my valve experience is
short, but I do understand electronics. So Pat, tell me, what's with those
who place 100K level adjust pot on the input of their amp? Besides changing
the level. doesn't this change the input impedance load and is this
something undesirable? What are the sacrifices, if any, in using an input
amp pot for volume control? Thanks as always.
Cordially,
west

If you have a 100k pot at the input to your power amp, and
the Ro of the preamp is from a 6SN7 anode circuit, with no loops of FB,
then you have a 10K Ro source connected to a 100k pot,
and the signal from the preamp drops only 10% when the power amp
is connected.

The power amp itself may have only the grid input impedance
connected to the pot wiper, and the gii is many megohms of resistance,
and whatever shunt C and miller C present.
with the 100k pot set at -6dB the pot has 50k to either side,
with say 10k source R.
So the effective R connected to the power amp input is
around 27k, and if the power amp input C totalled 100pF,
then there will be a HF pole at 58 kHz.
( If the pot was a 500k pot, the R would be 126kohms,
and the pole would be at 12.6 kHz.)

But since the power amp may already have a pole at 50 kHz
even with a low Z source connected to its input,
the final pole with the 100k pot will be down around
25 kHz.

But we have not included the cable between pre and power amps,
and this might be 5 metres at 100pF per metre,
so if Ro is 10k, and there is a 500pF cap to ground,
there is a pole at 32 kHz, which combined with the effect
of the poles of the other poles I mentioned,
would bring the response to barely 20 kHz.

I try to make all my amplifier systems, including the preamps,
to have a response out to 68 kHz or better,
which means the phase shift is still pretty good at 20 kHz,
and not the -45 degrees one might see.

The only way to maintain wide bandwidth is to use
a low S source to power a pot, then take the pot output to a low capacitance
impedance changer, a cathode follower, and then
its output signal is low Z, which is unaffected by cables, and power amp input
circuits.

In my ten tube preamp, I often use this to work into
solid state amps with low Ri compared to tube amps,
and maybe I run long cables sometimes, so a CF buffer
makes sense in my case for the intended use.

The White has to be set up rather carefully, because it
can oscillate if the anode R of the top tube is too large.
I find there is little benefit beyound using a simple more passive
CCS tail for a cathode follower, whose thd when working
into a load of even 10k at 1vrms is still quite low,
and the latchup distortion into C loads is generally negligible,
until you get to well above the AF band.

Patrick Turner.

Behold, Patrick Turner signalled from keyed 4-1000A filament:



Gregg wrote:

Behold, Patrick Turner signalled from keyed 4-1000A filament:

Gregg wrote:

Depending on the preamp, I choose either SRPP or Loftin-White CF.

There was a bit of discussion on this in my RIAA preamp thread he
http://geek.scorpiorising.ca/cgi-bin...m=10699406 36

and http://www.tubecad.com has some awesome stuff in their back
issues.

Personally, after listening to SRPP and L-W, I'll never go back to a
straight CF again ;-)

I have always found CF to be the most transparent of amplifier
stages.

Exactly. May as well use a 2N5088 :-p

I wouldn't use an emitter follower output stage on a tube pre.

I was being sarcastic :-p ;-)

--
Gregg
*It's probably useful, even if it can't be SPICE'd*
http://geek.scorpiorising.ca

"west" wrote in message
...
The reason for doing so is to isolate the effects of input impedance and
capacitance
of a power amp input circuit, and cable from an other wise high
impedance
source

if the signal was taken from an anode circuit, or from a 100 k volume
pot.
The cathode follower appears after the gain pots, and acts as a "buffer"
between power amp and preamp.
The typical output of a preamp might be 10 k ohms or much higher,
if a CF is not used, but the CF Ro is under 1k ohms,
so there are negligible high frequency losses.

Patrick Turner.

Viola! This is the explanation that turned on my rather dim light bulb. I
don't know about White or Black's circuits because my valve experience is
short, but I do understand electronics. So Pat, tell me, what's with
those
who place 100K level adjust pot on the input of their amp? Besides
changing
the level. doesn't this change the input impedance load and is this
something undesirable? What are the sacrifices, if any, in using an input
amp pot for volume control? Thanks as always.
Cordially,
west


If the input impedance of the first stage is high enough, say 1meg, the
input impedance of the amp will be almost the same as the resistance of the
pot for just about any setting of the pot.

Jimmy wrote:

"west" wrote in message
...
The reason for doing so is to isolate the effects of input impedance and
capacitance
of a power amp input circuit, and cable from an other wise high
impedance
source

if the signal was taken from an anode circuit, or from a 100 k volume
pot.
The cathode follower appears after the gain pots, and acts as a "buffer"
between power amp and preamp.
The typical output of a preamp might be 10 k ohms or much higher,
if a CF is not used, but the CF Ro is under 1k ohms,
so there are negligible high frequency losses.

Patrick Turner.

Viola! This is the explanation that turned on my rather dim light bulb. I
don't know about White or Black's circuits because my valve experience is
short, but I do understand electronics. So Pat, tell me, what's with
those
who place 100K level adjust pot on the input of their amp? Besides
changing
the level. doesn't this change the input impedance load and is this
something undesirable? What are the sacrifices, if any, in using an input
amp pot for volume control? Thanks as always.
Cordially,
west


If the input impedance of the first stage is high enough, say 1meg, the
input impedance of the amp will be almost the same as the resistance of the
pot for just about any setting of the pot.

If the input to the power amp is direct from the pot, and there is no
"safety" bias R from grid to 0V, say 1M, the
pot value is the input R of the amp.
With IM, the 100k Ri will reduce to 91k.

But if the bias R was 100k, and the pot was 100k
then when gain was max, the Ri would be 50k,
and still able to be powered from the anode circuit
of a medium U triode, but not from a 12AX7.

But the common position of a log volume control is the 12 o'clock
position, where the wiper to 0V R is 10k, with a 100k pot,
so any R between the grid and 0V won't affect the Ri to the pot.

Also, the volume pot has its maximum Ro when set at the -6dB, and that's where
the greatest HF loss occurs, but when the volume is at -20 dB, ( 12 o'clock ),
the Ro has lowered to 10k, so the response is much better.
And sometimes as the pot level is reduced even lower,
there can be a peak in the HF response due to stray C coupling .

With say a 100k DC supply R
The Ro from a 1/2 12AX7 with 100k R is 100k in parallel with Ra of 70 k,
so its 42k.
A pot load of 50k would reduce the output signal by around 6 dB,
and the total load on the 12AX7 would be around 23k,
which will reduce gain and increase thd.
If the power amp has a sensitivity like a Leak amp,
which was 100mV input for full power, a preamp isn't needed,
since mosr CD players have more output than 100 mV,
and also have Ro of less than 1 k.

But a williamson sensitivity is around 2v, and Quad II about 1.4v,
so so a preamp with gain is required, if lowish line level inputs are used .
I have a deletable gain stage in my preamps, and the gain stage
has only around 15 dB of gain, it seems to be enough.
But with more powerful 70 watt amps with 0.5 vrms input voltage needed for
70 watts, the sensitivity is quite OK to not have to have a preamp gain stage.
Sensitivity should include the input voltage per watt.
So in a Leak 20, you got 0.1v for 10 watts.
0.2v will 40 watts, and 0.3v will make 91 watts, and 1v makes 1,000 watts,
Vin/watt is constant, and the amp was set up to be capable.
The signal to noise ratio is better for the insensitive power amp.
About 1 volt / 50 watts is about right, and this is easy to arrange
with medium U triode driver stages instead of the 12AX7/EF86/12AT7 input stages.

I often use a simple 1/2 6CG7 as a preamp to make the volt required,
and the thd is very low, and then the pot set at -20 dB reduces the signal
to 0.1v for 2v ouput, or 0.5 watts into 8 ohms.
This gives 87 dB spl into 90 dB/watt sensitive speakers.
The line gain stage only needs a gain of 20 dB max.
An example of an integrated amp using 6CG7 drivers and a line stage is at

http://www.turneraudio.com.au/htmlwe...0ulabinteg.htm

There is no need for a CF in this schematic, because there
are minimal loses since the preamp tube is so close to the power amp
inputs, and no interconnect cabling.
The SNR is better with this set up rather than placing a pot
ahead of the line stage, because the noise of the line tube is attenuated
with the pot after the gain tube.

Patrick Turner.