Hi Pat !
I don't expected less from you

"Patrick Turner" wrote in message
...


Yves Monmagnon wrote:

Was building my fisrt tubed RIAA preamp.
Before optimizing the working point of the first stage, I spend
considerable
amount of time
until I realize than an MM cartridge specified to give 5mV rms at 1000Hz
will produce about 20dB
more level at 20000Hz due to the RIAA groove preemphasis and so can
deliver
up to 140mV pp.

I have never ever seen such a high level from any MM cartridge.

If
the first stage gain is 36dB (and it is !), the level at the stage
output is
near 9v pp. It must be able to
deliver this level with lo thd. But the RIAA deemphasis pulls down by
near
40dB at such
frequency, removing most harmonics before they reach the second stage,
nevertheless, sound is horribly
dark, even if frequency response, ploted using white noise at -40dB and
spectral analysis, still
looks very good (well within the dB).

There is a simple answer.
Having a gain of 63 sounds like you have a 12AX7 as the input tube.
If you use a mu follower, the gain tube sees a very high
load resistance, maybe a megohm, and you can easily get the thd down from
2% at 9 volts output, to 0.1%, and the follower part has even lower thd.

See my circuit at
http://www.turneraudio.com.au/htmlwe...tubepreamp.htm
Sure I've already seen, like as much other and helped me a more than a lot

The input stage has low thd, and the gain tube sees the same load value
across
the band.

I've a PC900 (4HA5) CCS loaded at 15 mA to obtain those 36dB.
B+ is 200v
Funny tube, lo noise, no microphony but large specs variations.
After having 'centered' the plate voltage at about 130v (by adjusting the
cathode resistor)
things works much better.(Last night at 11 pm when everybody sleeped in
house, confirmed this morning at higer level)
Found too that cathode resisitor does not need to be bypassed, since CCS
voids any current change, thus no voltage change and no degenerative effect
!!

The gain is 39 dB.

Note that the second stage just has to
sustain half this
level, since we just need it be clean at say +3dBv, that is about 4v pp!
I've never read anything about that could be called 'The Passive RIAA
Filtered Preamp Paradox' ?
Am I wrong ?

The RIAA will allow the low levels of bass through,
but attenuate 20 kHz about 39 dB below the 27 Hz level.
The attenuation of 1 kHz entering the filter from the first stage
is about 18 dB, so don't worry, be happy,
use a normal second stage like I have, and as we say,
she'll be apples mate.

The 5 mV of quoted output signal of an MM cart is a high level of output,
and this would be confirmed when looking at a music signal from a cart on
a CRO.

I have never seen peaks above 50mV.
The circuit of mine copes with anything you throw at it.
You might be worried if you were to try to use low voltage fets or
transistors,
but not with tubes.

Patrick Turner.

To be continued latter, with schematic, time to work now !
Cheers, Yves

Found too that cathode resisitor does not need to be bypassed,
since CCS
voids any current change, thus no voltage change and no
degenerative effect
!!

Also true of mu-follower.

But consider effect on output impedance of the stage.

cheers, Ian

Hi Ian

"Ian Iveson" wrote in message
...

Found too that cathode resisitor does not need to be bypassed,
since CCS
voids any current change, thus no voltage change and no
degenerative effect
!!

Also true of mu-follower.
Right, IMHO, mu-follower is a CC loaded by a CCS, true ?

But consider effect on output impedance of the stage.
Right again, anyway any load means current change throught it, so load is
obviusly no longer an infinite impedance.
And thus in real word, plate load line is never perfectly horizontal, and
gain always lower than mu.

I'm a bit psycho maniac looking at cathode bypass capacitors who have to be
so perfect that I'm very
happy when I can remove it
I remember of the nightmare of self resonating caps in RF circuits !

Cheers, Yves

cheers, Ian

Ian Iveson wrote:

Found too that cathode resisitor does not need to be bypassed,
since CCS
voids any current change, thus no voltage change and no
degenerative effect
!!

Also true of mu-follower.

But consider effect on output impedance of the stage.

In my revised 10 tube preamp, I have adopted the use of bootstrappeed
followers,
which are between a Mu follower, and SRPP.
The Mu follower has a large value R between top and bottom tubes, so the
gain
of the tube on the bottom approaches the mu for the tube.
The SRPP, has a small value R between the tubes, and PP action happens,
its extent depending on the value of the R between the tubes.
A bootstrapped follower is intermediate between the former two ways,
and in my case with a 12AT7, I have 4.7 between the tubes, and RO
from the top tube cathode is a low 2.0 k, even with the bottom tube Rk
unbypassed.

So the Ro is nice an low for a bootrapped follower,
and even when Rk of both tubes is eqaul, Ro taken from the top tube
cathode
is lower than Ra even with an unbypassed bottom Rk.

Your simulations should have told you all this.

Patrick Turner.




cheers, Ian

"Patrick Turner" wrote

Why have you written all that stuff?

Patrick. If you take either a mu-follower, or a common cathode
stage with a CC anode load, and unbypass the cathode resistor at the
bottom, you will increase the output resistance considerably, and
possibly very significantly.

Your comment about simulation is not out of character I notice.

cheers, Ian

Ian Iveson wrote:

Found too that cathode resisitor does not need to be bypassed,
since CCS
voids any current change, thus no voltage change and no
degenerative effect
!!

Also true of mu-follower.

But consider effect on output impedance of the stage.

In my revised 10 tube preamp, I have adopted the use of
bootstrappeed
followers,
which are between a Mu follower, and SRPP.
The Mu follower has a large value R between top and bottom tubes,
so the
gain
of the tube on the bottom approaches the mu for the tube.
The SRPP, has a small value R between the tubes, and PP action
happens,
its extent depending on the value of the R between the tubes.
A bootstrapped follower is intermediate between the former two
ways,
and in my case with a 12AT7, I have 4.7 between the tubes, and RO
from the top tube cathode is a low 2.0 k, even with the bottom
tube Rk
unbypassed.

So the Ro is nice an low for a bootrapped follower,
and even when Rk of both tubes is eqaul, Ro taken from the top
tube
cathode
is lower than Ra even with an unbypassed bottom Rk.

Your simulations should have told you all this.

Patrick Turner.




cheers, Ian

"Ian Iveson" wrote in message
...
Patrick. If you take either a mu-follower, or a common cathode
stage with a CC anode load, and unbypass the cathode resistor at the
bottom, you will increase the output resistance considerably, and
possibly very significantly.

How can it if there's no AC voltage across the resistor?

Tim

--
In the immortal words of Ned Flanders: "No foot longs!"
Website @ http://webpages.charter.net/dawill/tmoranwms

"Ian Iveson" wrote in message
...
Two arguments, take your pick. Easiest for Yves' simple CC with CCS
at anode.

I'm most familiar (such that I am) with a mu-follower.. which I presume
could be done with a CCS in place of the resistor between the two tubes.

In the case where the load is placed on the top tube's cathode, above
the R/CCS, the load develops a voltage across the resistor, causing
a stronger voltage at the bottom tube's plate, which is coupled to
top's grid, which acts as a CF, bootstrapping out the load impedance.
This is my understanding.

If you vary the load, you vary the current robbed from the CCS and
therefore the current arriving at the cathode. This results in
series nfb from the unbypassed resistor.

But the CCS (top tube you mean?) ensures a CC into the bottom tube, by
supplying additional current to the load, if loaded down.

Naturally there will be a bit of voltage change on the cathode resistor,
but nothing is perfect.

The signal cannot go
up the CCS so for a low output impedance Ra and Rk must be kept low.

Ah, but it does, for what else is that coupling cap doing?

Rk is effectively multiplied...can't remember the formula and too
tired to work it out. You know it anyway.

Uhm... notes are downstairs. But yes I have it.

The problem is not so bad for the mu-follower, because a returning
signal does not see a CCS in either direction. But still the bottom
Ra is part of the path to ground, and the bottom valve is
responsible for the gain, so the resistor makes a significant
difference.

Are you referring to the CC w/ CCS load mentioned earlier? Or a mu stage?
Your post didn't specify too well :-\

Tim

--
In the immortal words of Ned Flanders: "No foot longs!"
Website @ http://webpages.charter.net/dawill/tmoranwms

Tim, I think my explanations work for CC with CCS at anode. I
stated that is what my explanations were about.

If you took output from the bottom anode of a mu-follower, then the
same
argument applies because that anode sees the CCS *above* it. In
these
cases there is not constant current through the bottom cathode
resistor...constant anode current does not necessarily result in
constant cathode current because some is robbed by the load. Such a
configuration has a high output resistance anyway, and an unbypassed
Rk increases it even further. To find the output resistance you
just have to consider ra, since rl is infinite.

r'a = ra + (mu +1).rk

where r'a is anode resistance in case of unbypassed rk, and ra is
anode resistance with zero or bypassed rk

That is, unbypassing adds (mu + 1).rk to the output resistance.
That could easily be 100k or more.

That is why I drew Yves' attention to the output resistance of *his*
stage. It wasn't about a mu-follower then...he said CCS loaded but
had he meant mu-follower he would have said so, presumably. I
responded to his remark that rk sees a constant current, which is
not true when a signal and a finite output load is present.

Hence Patrick's remarks to me had nothing to do with what I said.

The mu-follower is not so straightforward. Lots of things going on.
Only ones I have in practice have mosfets on the top and transformer
loads, and behave in some surprising ways (to me...dunno what goes
on in the tiny mind of a mosfet), and struggle if I take out the
cathode bypass caps.

Your arguments seem perfectly correct to me though.

But wait...bootstrapping increases the output resistance of a
cathode follower, right? If the mu-follower has no signal at its
bottom grid, then the bottom valve's ra is more or less constant,
and forms the bootstrapped portion of the top cathode follower's rk.
If you increase ra of the bottom valve, then you increase the
series-derived feedback to the top valve's grid, and hence the
output resistance of the cathode follower. So it seems that the
bottom ra is still significant. And it would still be raised by
(mu+1).rk.

Now you should be kind to me because I don't know what bootstrapping
means really. Every explanation I find makes reference to mechanics
but the analogies are always nonsensical. I can't pull myself up by
my own bootstraps, it doesn't work.

And there are other issues to consider...such as hum and noise
rejection.

There must be some reason why nearly all the mu-followers I see have
bypass caps?

cheers, Ian

"Tim Williams" wrote in message
...
"Ian Iveson" wrote in message
...
Two arguments, take your pick. Easiest for Yves' simple CC with
CCS
at anode.

I'm most familiar (such that I am) with a mu-follower.. which I
presume
could be done with a CCS in place of the resistor between the two
tubes.

In the case where the load is placed on the top tube's cathode,
above
the R/CCS, the load develops a voltage across the resistor,
causing
a stronger voltage at the bottom tube's plate, which is coupled to
top's grid, which acts as a CF, bootstrapping out the load
impedance.
This is my understanding.

If you vary the load, you vary the current robbed from the CCS
and
therefore the current arriving at the cathode. This results in
series nfb from the unbypassed resistor.

But the CCS (top tube you mean?) ensures a CC into the bottom
tube, by
supplying additional current to the load, if loaded down.

Naturally there will be a bit of voltage change on the cathode
resistor,
but nothing is perfect.

The signal cannot go
up the CCS so for a low output impedance Ra and Rk must be kept
low.

Ah, but it does, for what else is that coupling cap doing?

Rk is effectively multiplied...can't remember the formula and
too
tired to work it out. You know it anyway.

Uhm... notes are downstairs. But yes I have it.

The problem is not so bad for the mu-follower, because a
returning
signal does not see a CCS in either direction. But still the
bottom
Ra is part of the path to ground, and the bottom valve is
responsible for the gain, so the resistor makes a significant
difference.

Are you referring to the CC w/ CCS load mentioned earlier? Or a
mu stage?
Your post didn't specify too well :-\

Tim

--
In the immortal words of Ned Flanders: "No foot longs!"
Website @ http://webpages.charter.net/dawill/tmoranwms

Maybe it's time to post this link again :
http://laplaza.org/%7Efvuotto/mustage.html

Ronald .


"Ian Iveson" schreef in bericht
...
Tim, I think my explanations work for CC with CCS at anode. I
stated that is what my explanations were about.

If you took output from the bottom anode of a mu-follower, then the
same
argument applies because that anode sees the CCS *above* it. In
these
cases there is not constant current through the bottom cathode
resistor...constant anode current does not necessarily result in
constant cathode current because some is robbed by the load. Such a
configuration has a high output resistance anyway, and an unbypassed
Rk increases it even further. To find the output resistance you
just have to consider ra, since rl is infinite.

r'a = ra + (mu +1).rk

where r'a is anode resistance in case of unbypassed rk, and ra is
anode resistance with zero or bypassed rk

That is, unbypassing adds (mu + 1).rk to the output resistance.
That could easily be 100k or more.

That is why I drew Yves' attention to the output resistance of *his*
stage. It wasn't about a mu-follower then...he said CCS loaded but
had he meant mu-follower he would have said so, presumably. I
responded to his remark that rk sees a constant current, which is
not true when a signal and a finite output load is present.

Hence Patrick's remarks to me had nothing to do with what I said.

The mu-follower is not so straightforward. Lots of things going on.
Only ones I have in practice have mosfets on the top and transformer
loads, and behave in some surprising ways (to me...dunno what goes
on in the tiny mind of a mosfet), and struggle if I take out the
cathode bypass caps.

Your arguments seem perfectly correct to me though.

But wait...bootstrapping increases the output resistance of a
cathode follower, right? If the mu-follower has no signal at its
bottom grid, then the bottom valve's ra is more or less constant,
and forms the bootstrapped portion of the top cathode follower's rk.
If you increase ra of the bottom valve, then you increase the
series-derived feedback to the top valve's grid, and hence the
output resistance of the cathode follower. So it seems that the
bottom ra is still significant. And it would still be raised by
(mu+1).rk.

Now you should be kind to me because I don't know what bootstrapping
means really. Every explanation I find makes reference to mechanics
but the analogies are always nonsensical. I can't pull myself up by
my own bootstraps, it doesn't work.

And there are other issues to consider...such as hum and noise
rejection.

There must be some reason why nearly all the mu-followers I see have
bypass caps?

cheers, Ian

"Tim Williams" wrote in message
...
"Ian Iveson" wrote in message
...
Two arguments, take your pick. Easiest for Yves' simple CC with
CCS
at anode.

I'm most familiar (such that I am) with a mu-follower.. which I
presume
could be done with a CCS in place of the resistor between the two
tubes.

In the case where the load is placed on the top tube's cathode,
above
the R/CCS, the load develops a voltage across the resistor,
causing
a stronger voltage at the bottom tube's plate, which is coupled to
top's grid, which acts as a CF, bootstrapping out the load
impedance.
This is my understanding.

If you vary the load, you vary the current robbed from the CCS
and
therefore the current arriving at the cathode. This results in
series nfb from the unbypassed resistor.

But the CCS (top tube you mean?) ensures a CC into the bottom
tube, by
supplying additional current to the load, if loaded down.

Naturally there will be a bit of voltage change on the cathode
resistor,
but nothing is perfect.

The signal cannot go
up the CCS so for a low output impedance Ra and Rk must be kept
low.

Ah, but it does, for what else is that coupling cap doing?

Rk is effectively multiplied...can't remember the formula and
too
tired to work it out. You know it anyway.

Uhm... notes are downstairs. But yes I have it.

The problem is not so bad for the mu-follower, because a
returning
signal does not see a CCS in either direction. But still the
bottom
Ra is part of the path to ground, and the bottom valve is
responsible for the gain, so the resistor makes a significant
difference.

Are you referring to the CC w/ CCS load mentioned earlier? Or a
mu stage?
Your post didn't specify too well :-\

Tim

--
In the immortal words of Ned Flanders: "No foot longs!"
Website @ http://webpages.charter.net/dawill/tmoranwms

"Ronald" wrote

Maybe it's time to post this link again ...

Thanks, Ronald.

It is, according to Mr Kimmel, how I have described it: the output
impedance is partly dependent upon ra of the bottom valve, which is
reduced if its rk is bypassed.

Also important for hum and noise rejection according to Morgan
Jones. Does Alan say that too? Lemme see...er..

cheers, Ian

Ian Iveson wrote:

Tim, I think my explanations work for CC with CCS at anode. I
stated that is what my explanations were about.

If you took output from the bottom anode of a mu-follower, then the
same
argument applies because that anode sees the CCS *above* it. In
these
cases there is not constant current through the bottom cathode
resistor...constant anode current does not necessarily result in
constant cathode current because some is robbed by the load. Such a
configuration has a high output resistance anyway, and an unbypassed
Rk increases it even further. To find the output resistance you
just have to consider ra, since rl is infinite.

r'a = ra + (mu +1).rk

where r'a is anode resistance in case of unbypassed rk, and ra is
anode resistance with zero or bypassed rk

That is, unbypassing adds (mu + 1).rk to the output resistance.
That could easily be 100k or more.

That is why I drew Yves' attention to the output resistance of *his*
stage. It wasn't about a mu-follower then...he said CCS loaded but
had he meant mu-follower he would have said so, presumably. I
responded to his remark that rk sees a constant current, which is
not true when a signal and a finite output load is present.

Hence Patrick's remarks to me had nothing to do with what I said.

But Ian, you are so long winded, vague, pedantic, and roundabout in all
your posts,
its impossible to know WTF you are saying, quite often.
If people appear to not understand you, spell things out more clearly,
if you wouldn't mind!

Thus time what you have said above abour effective output resistance
is about right.
The above formula for Ra' is correct, if the tube has a CCS DC supply,
but if it is supplied via a resistor, RL, or a tube acting as an active
current
source, then Ra' has to include RL or the active current sourse value in
parallel,
to get the total output resistance.
Readers in the group should now turn to their copy of RDH4
to confirm what I am saying, and extend their knowledge, so I don't have
to keep
describing it ad infinitum.
The point of output from seriesed topology we are talking about
is critical to getting a low output resistance.



The mu-follower is not so straightforward. Lots of things going on.

No, is simple, and it is best regarded as a bootstrapped follower.
unless the R between top and bottom tubes is such a high value of
dynamic
impedance, such as a j-fet set up as a very near constant current
source,
ie, it acts as a resistor equal to a megohm or more.



Only ones I have in practice have mosfets on the top and transformer
loads, and behave in some surprising ways (to me...dunno what goes
on in the tiny mind of a mosfet), and struggle if I take out the
cathode bypass caps.

?????????



Your arguments seem perfectly correct to me though.

But wait...bootstrapping increases the output resistance of a
cathode follower, right?

Not much.

If the mu-follower has no signal at its
bottom grid, then the bottom valve's ra is more or less constant,
and forms the bootstrapped portion of the top cathode follower's rk.
If you increase ra of the bottom valve, then you increase the
series-derived feedback to the top valve's grid, and hence the
output resistance of the cathode follower. So it seems that the
bottom ra is still significant. And it would still be raised by
(mu+1).rk.

But where is the input signal if there is no signal at the top tube's
grid? what are we talking about? a mu follower, or a cathode follower
loaded by the bottom tube? and where is the output from?
where is your schematic?
Readers would be totally confused by what you have posted.



Now you should be kind to me because I don't know what bootstrapping
means really.

I will for one, But just remember, the penny drops about bootstrapping
with most audio engineers, sooner or later.
It's a general principal, used in many SS amps to boost the gain of the
voltage amp stage, and used in McIntosh tube amps for the same reason.

Every explanation I find makes reference to mechanics
but the analogies are always nonsensical. I can't pull myself up by
my own bootstraps, it doesn't work.

We know this to, regardless of how much we weigh.
Its an absurd name for a circuit concept you are having difficulty
visualizing mentally.
Its also like a form of positive voltage feedback.
But once someone says "bootstrapping", many of us know exactly what is
meant
electronically, and it is different to the usual positive FB.



And there are other issues to consider...such as hum and noise
rejection.

There must be some reason why nearly all the mu-followers I see have
bypass caps?

Just habit. Noise input in a phono stage is more likely,
since the unbypassed Rk generates noise.
Noise from the heaters could get in because of the capacitance between
cathode and heater.

There are quite a few stalwarts who HATE capacitors of any sort,
especially electrolytic low voltage types used for byapassing,
and the latest craze is teflon caps, which are the only coupling/bypass
caps which
can be relunctantly tolerated.
I think this is BS, but they are entitled to their BS, ok.

I got rid of my bypass caps in my preamp recently, and its still
remarkably quiet, but the phono stage now has a 2SK369 j-fet input
instead of a tube, to get the noise down about 15 dB lower than any tube
I might
use, except perhaps a 6C45pi, which probably is noiser,
but nobody here has ever posted any measurements about its use in an
actual
circuit.
I say this because its pertinent to to the subject, RIAA preamps.

The following stages after a tubed RIAA input stage
usually will contribute negligible noise to the signal.
The bootstrapped, or mu follower is used to take advantage
of the low Ro, and low thd, and high gain, of such a stage.
This means, for example, that the output of such a stage can be taken to
a sound card
via a 5 metre long cable, for transfering LP to CD, without any losses,
or need for extra buffers.

Patrick Turner.



cheers, Ian

"Tim Williams" wrote in message
...
"Ian Iveson" wrote in message
...
Two arguments, take your pick. Easiest for Yves' simple CC with
CCS
at anode.

I'm most familiar (such that I am) with a mu-follower.. which I
presume
could be done with a CCS in place of the resistor between the two
tubes.

In the case where the load is placed on the top tube's cathode,
above
the R/CCS, the load develops a voltage across the resistor,
causing
a stronger voltage at the bottom tube's plate, which is coupled to
top's grid, which acts as a CF, bootstrapping out the load
impedance.
This is my understanding.

If you vary the load, you vary the current robbed from the CCS
and
therefore the current arriving at the cathode. This results in
series nfb from the unbypassed resistor.

But the CCS (top tube you mean?) ensures a CC into the bottom
tube, by
supplying additional current to the load, if loaded down.

Naturally there will be a bit of voltage change on the cathode
resistor,
but nothing is perfect.

The signal cannot go
up the CCS so for a low output impedance Ra and Rk must be kept
low.

Ah, but it does, for what else is that coupling cap doing?

Rk is effectively multiplied...can't remember the formula and
too
tired to work it out. You know it anyway.

Uhm... notes are downstairs. But yes I have it.

The problem is not so bad for the mu-follower, because a
returning
signal does not see a CCS in either direction. But still the
bottom
Ra is part of the path to ground, and the bottom valve is
responsible for the gain, so the resistor makes a significant
difference.

Are you referring to the CC w/ CCS load mentioned earlier? Or a
mu stage?
Your post didn't specify too well :-\

Tim

--
In the immortal words of Ned Flanders: "No foot longs!"
Website @ http://webpages.charter.net/dawill/tmoranwms