Reply
 
Thread Tools Display Modes
  #1   Report Post  
Posted to rec.audio.pro
[email protected][_2_] seanbroderick20003@yahoo.com[_2_] is offline
external usenet poster
 
Posts: 26
Default Hi Z Buffer Distortion

What are the dominant distortion mechanisms when one tries to build a
unity-gain (or greater than unity gain) buffer to buffer a signal with
a source impedance of, say, 100 megohms, at audio frequencies? Do
vacuum tubes suffer from the capacitance-modulation effects that FETs
do?

Thanks,

Sean B
  #2   Report Post  
Posted to rec.audio.pro
[email protected] makolber@yahoo.com is offline
external usenet poster
 
Posts: 614
Default Hi Z Buffer Distortion

On Apr 6, 1:18*pm, "
wrote:
What are the dominant distortion mechanisms when one tries to build a
unity-gain (or greater than unity gain) buffer to buffer a signal with
a source impedance of, say, 100 megohms, at audio frequencies? *Do
vacuum tubes suffer from the capacitance-modulation effects that FETs
do?

Thanks,

Sean B


is this for a condensor mic buffer?

Mark
  #3   Report Post  
Posted to rec.audio.pro
Chris Hornbeck Chris Hornbeck is offline
external usenet poster
 
Posts: 1,744
Default Hi Z Buffer Distortion

On Mon, 6 Apr 2009 16:12:24 -0400, "Soundhaspriority"
wrote:

I do not have expertise in this area, but there is probably wide agreement
that vacuum tube solutions have higher distortion than FETs.


This is widely expressed, but fundamentally incorrect. An
appropriately used vacuum valve can give linearity with
transfer exponents less than 1.05, and exteremely low
distortion, and at high signal levels, and monotonically.

Many years of deliberate abuse have given vacuum valves
a reputation as distortion generators, but it comes from
the abuse, and not as a fundamental of the device.

For example, a classic type 12AX7 made back in the good
old days (when men were men and women liked it that way)
by real craftswomen, can deliver +20dBu outputs with less
than -40dBu total distortion products (before any feedback).
No single semiconductor can deliver even remotely similar
performance in linearity. Type 5687 in larger devices and
type 211 in really big devices are also very linear.

Vacuum valves have an additional advantage for use from
capacitive sources because signal voltages don't modulate
their input capacitance, ehich is purely parasitic. Many
hi-Z sources (the OP didn't specify) tend to the capacitive.


I seem to always be defending some old technology or another,
but, hey, audio's old tech. I can live with that.

Much thanks, as always,
Chris Hornbeck
  #4   Report Post  
Posted to rec.audio.pro
Chris Hornbeck Chris Hornbeck is offline
external usenet poster
 
Posts: 1,744
Default Hi Z Buffer Distortion

On Mon, 6 Apr 2009 21:23:50 -0400, "Soundhaspriority"
wrote:

It would seem that, among commercial mike designs, the following is
generally true (and I will be very interested in the exceptions!)

1. Solid state is used by the top makers, DPA, Schoeps, Neumann and
Sennheiser, for designs that are admired, more or less, for neutrality and
low noise. Some or most of the designs may no longer use the single FET
impedance converter, which you have by implication impugned.
2. Tube mikes are noisy, colored, and favored for the human voice..

I am sure that other members of this forum can point out interesting
exceptions. There would be little point, however, in saying that a high-spec
measurement mike can be built better with a vacuum tube impedance converter,
unless there is a product to point to.


Well, if you want to get all practical, and ever'thing like that,
well, yeah. Doesn't mean I have to stop listening to the Beach Boys!

Or something like that...


Said by someone who may own one of the last working specimens of a Tektronix
555 oscilloscope, which has around 129 tubes and draws 1250 watts. At least,
it worked the last time I checked


My (inherited) 503 is still used for some loudspeaker measurements -
excellent differential input sensitivity and "common mode rejection".
Still nothing better in my low-dollar world (and I have the complete
battery of Sound Technology gear from the 1970's in mostly working
condition. Keeping stuff from the 1950's and 1960's going is actually
easier than the 70's stuff right up to the point of death).

Also still useful, and sometimes used, is an old HP 120B that can do
10mV/cM differential - good enough for most loudspeaker work, immune,
like all vacuum valve stuff, from over-voltage issues, and probably
free these days.


Within the audio range, a buffer stage and a sound card's conversion
to computer manipulation is a huge step beyond squinting at an old
scope's screen, but above the audio range important things still
happen, and a real 'scope is still needed to try (!) to see what
they are.


Much thanks, as always,
Chris Hornbeck
  #5   Report Post  
Posted to rec.audio.pro
Chris Hornbeck Chris Hornbeck is offline
external usenet poster
 
Posts: 1,744
Default Hi Z Buffer Distortion

On Mon, 6 Apr 2009 23:08:09 -0400, "Soundhaspriority"
wrote:

Chris, the Beach Boys are obsolete, having been replaced by superior forms
of music. I suspect you probably listen to Simon & Garfunkel as well.
Please, get a hold of yourself and stop the slide toward deviancy. You can
go slow. Start by listening to some of these three or four hours a day:

http://knowledgerush.com/kr/encyclop...amous_rappers/


I work with a young man who makes great rap music (folks willing
to sign into MySpace (I think - not personally willing) can hear
him as Chuck Sweet. But I'm too old to put it into the same
category as the Beach Boys. Chuck probably wouldn't either;
I'll ask him tomorrow.

We'll know in forty years. (Except, I'll be dead - maybe that's the
difference...)


Much thanks, as always,
Chris Hornbeck


  #6   Report Post  
Posted to rec.audio.pro
Paul Stamler[_2_] Paul Stamler[_2_] is offline
external usenet poster
 
Posts: 178
Default Hi Z Buffer Distortion

"Soundhaspriority" wrote in message
...

wrote in message
...
What are the dominant distortion mechanisms when one tries to build a
unity-gain (or greater than unity gain) buffer to buffer a signal with
a source impedance of, say, 100 megohms, at audio frequencies? Do
vacuum tubes suffer from the capacitance-modulation effects that FETs
do?

Thanks,

Sean B


I do not have expertise in this area, but there is probably wide agreement
that vacuum tube solutions have higher distortion than FETs.


Not necessarily, and particularly not when high source impedances are
involved. As the OP notes, there are major problems with
capacitance-modulation in FETs, which can lead to remarkably high levels of
distortion if the source impedance is above a few tens of kilohms. Vacuum
tubes typically have extremely low levels of capacitance-modulation, which
makes them good candidates to buffer very high impedances. A
*properly-designed* vacuum tube circuit can have very low levels of
distortion; 0.01% harmonic distortion at audio frequencies isn't
particularly hard to achieve, depending on the level you want to put out.
Tubes aren't very good at driving low impedances; if the load on the buffer
is going to be lower-impedance, than a good solid-state driver circuit
following the buffer is in order.

So the first two questions for the OP a what exactly do you expect to be
buffering? And what will the buffer have to drive, and at what level?

Also, no active device exists that has zero capacitance, because such a
device would have the theoretically impossible ability to measure without
disturbance. This is forbidden by physics.


True but irrelevant. The issue isn't capacitance by itself, but capacitance
which is modulated by signal level. That adds distortion to the signal at
high frequencies when the impedance is high, sometimes plenty of it (like
2-3% THD, and presumably comparable levels of high-frequency IMD).

Peace,
Paul


  #7   Report Post  
Posted to rec.audio.pro
Eeyore Eeyore is offline
external usenet poster
 
Posts: 8,474
Default Hi Z Buffer Distortion



" wrote:

What are the dominant distortion mechanisms when one tries to build a
unity-gain (or greater than unity gain) buffer to buffer a signal with
a source impedance of, say, 100 megohms, at audio frequencies? Do
vacuum tubes suffer from the capacitance-modulation effects that FETs
do?


You might get a good technical answer in sci.electronics.design.

Graham

  #8   Report Post  
Posted to rec.audio.pro
Eeyore Eeyore is offline
external usenet poster
 
Posts: 8,474
Default Hi Z Buffer Distortion



Soundhaspriority wrote:

Correction, Should read: Vacuum tubes disappeared from instrumentation
with the advent of the FET back in the late 60's.


I can still remember my Tek 545 scopes ! And the fan at the back.

Graham

  #9   Report Post  
Posted to rec.audio.pro
Eeyore Eeyore is offline
external usenet poster
 
Posts: 8,474
Default Hi Z Buffer Distortion



Chris Hornbeck wrote:

On Mon, 6 Apr 2009 16:12:24 -0400, "Soundhaspriority"
wrote:

I do not have expertise in this area, but there is probably wide agreement
that vacuum tube solutions have higher distortion than FETs.


This is widely expressed, but fundamentally incorrect. An
appropriately used vacuum valve can give linearity with
transfer exponents less than 1.05, and exteremely low
distortion, and at high signal levels, and monotonically.


If used as a buffer stage ( cathode follower ) the linearity would be quite
good. Not sure how it'd fare noise wise though. Ultra low noise and THD solid
state circuitry could follow it.

Graham

  #11   Report Post  
Posted to rec.audio.pro
Eeyore Eeyore is offline
external usenet poster
 
Posts: 8,474
Default Hi Z Buffer Distortion



Paul Stamler wrote:

"Soundhaspriority" wrote in message

Also, no active device exists that has zero capacitance, because such a
device would have the theoretically impossible ability to measure without
disturbance. This is forbidden by physics.


True but irrelevant. The issue isn't capacitance by itself, but capacitance
which is modulated by signal level. That adds distortion to the signal at
high frequencies when the impedance is high, sometimes plenty of it (like
2-3% THD, and presumably comparable levels of high-frequency IMD).


Believe it or not I've seen this effect in mic amps designed to handle high
signal level with the now unobtainium bipolar Rohm 2SB737 input device. Being
large geometry to get thermal noise down, Rbb was only about 4 ohms ! Ccb was
also large and acted like a varicap diode and you could see the distortion
increasing over about 10kHz.

Graham

  #12   Report Post  
Posted to rec.audio.pro
[email protected] jwvm@umich.edu is offline
external usenet poster
 
Posts: 24
Default Hi Z Buffer Distortion

On Apr 7, 6:45*am, Eeyore
wrote:
Soundhaspriority wrote:
Correction, *Should read: *Vacuum tubes disappeared from instrumentation
with the advent of the FET back in the late 60's.


I can still remember my Tek 545 scopes ! And the fan at the back.

Graham


That was a great scope and very rugged! :-)
  #14   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default Hi Z Buffer Distortion

1. The distortion issues with high-Z input followers and gain stages are all
the same as with conventional follower and gain stage circuits, PLUS a whole
raft of other problems which are all related to parasitic capacitance.

2. The input capacitance of both tubes and FETs is modulated by signal but
in different ways. For an introduction to how the effect works for tubes,
look at the data sheet for any reactance modulator tube.

3. One good solution to deal with capacitances that change with modulation is
to reduce the modulation. Look at the B&K 2615 as an example of an
extremely-low-distortion tube design. Notice what is done to sacrifice
noise floor for low distortion.

4. Even the worst input stage has far less distortion than even the best
output transformer.

5. With any competently-designed microphone today, the distortion due to
mechanical and acoustical issues will be far, far higher than the
distortion due to any of the electronics.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."
  #15   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default Hi Z Buffer Distortion

In article ,
Eeyore wrote:


Soundhaspriority wrote:

Correction, Should read: Vacuum tubes disappeared from instrumentation
with the advent of the FET back in the late 60's.


I can still remember my Tek 545 scopes ! And the fan at the back.


I have one on my bench right now and it's just fine, even though the cal
guys are always making fun of it and trying to give me an Agilent digital
scope to replace it.

The reason tubes disappeared from instrumentation had more to do with
stability than anything else... the 545 really needs to warm up for an
hour before it's reasonably stable and you can't use anything in it for
quantitative measurement directly.

A hell of a lot of work went into making tube circuits stable at all over
long time periods. Korn and Korn's book on analogue computer circuits is
fascinating to read to look at some of the stuff they did. A lot of it is
still applicable today, too.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."


  #16   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default Hi Z Buffer Distortion

In article 49e22ecd.76189031@localhost, Don Pearce wrote:

There should be no capacitance modulation in a unity gain buffer since
the voltage between gate and source does not change with the input
signal.


This is true, but most microphones do not have the front end operating in
unity gain. Many older designs in fact use a huge amount of gain on the
front end, driving a step-down transformer.

All active devices have inherent non-linearities, but the degree to
which they appear in a complete circuit depends inversely on the
degree of feedback that corrects them. That in turn depends on the
amount of open loop gain before feedback. Generally that is much
higher for the FET than the tube, so the linearity of the FET circuit
is likely to be better.


Well, the thing is that although the gain of the FET is probably higher
so that you can put more feedback on it, the straightline section of the
curve isn't as flat as with, say, a frame grid tube. So you pays your
money and you takes you chance. (Once you get out of the straightline
section of the curves, all bets are off too).

Of much greater concern is the fact that the source impedance is of
the order of 100 megohms. Even the odd pF or so of capacitance will
kill the top end of the audio - a cable connection is certainly out of
the question.


Yes, and you can have all kinds of parasitic issues too... with 1G to 10G
input impedances, if you have wires shaking around even that can create
modulation. Poor wiring layout and PC board layout are big issues.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."
  #17   Report Post  
Posted to rec.audio.pro
Don Pearce[_3_] Don Pearce[_3_] is offline
external usenet poster
 
Posts: 2,417
Default Hi Z Buffer Distortion

On 7 Apr 2009 11:29:17 -0400, (Scott Dorsey) wrote:

In article 49e22ecd.76189031@localhost, Don Pearce wrote:

There should be no capacitance modulation in a unity gain buffer since
the voltage between gate and source does not change with the input
signal.


This is true, but most microphones do not have the front end operating in
unity gain. Many older designs in fact use a huge amount of gain on the
front end, driving a step-down transformer.

For a microphone preamp sure - the source is grounded and capacitance
modulation will happen. But not in the OP's unity gain buffer.

All active devices have inherent non-linearities, but the degree to
which they appear in a complete circuit depends inversely on the
degree of feedback that corrects them. That in turn depends on the
amount of open loop gain before feedback. Generally that is much
higher for the FET than the tube, so the linearity of the FET circuit
is likely to be better.


Well, the thing is that although the gain of the FET is probably higher
so that you can put more feedback on it, the straightline section of the
curve isn't as flat as with, say, a frame grid tube. So you pays your
money and you takes you chance. (Once you get out of the straightline
section of the curves, all bets are off too).

Of much greater concern is the fact that the source impedance is of
the order of 100 megohms. Even the odd pF or so of capacitance will
kill the top end of the audio - a cable connection is certainly out of
the question.


Yes, and you can have all kinds of parasitic issues too... with 1G to 10G
input impedances, if you have wires shaking around even that can create
modulation. Poor wiring layout and PC board layout are big issues.
--scott


Quite so.

d
  #20   Report Post  
Posted to rec.audio.pro
[email protected][_2_] seanbroderick20003@yahoo.com[_2_] is offline
external usenet poster
 
Posts: 26
Default Hi Z Buffer Distortion

Thanks for all the responses!

Yes it is a condenser capsule I am buffering. One with around 80 pF
of capacitance. I am messing around with a mic design.

I would like to drive loads as low as 1500 ohms, in order for a user
to be able to use mic preamps with input transformers. But my
personal philosophy is that the mic preamp should be near the
performer. It seems ridiculous to knock the impedance down with a
transformer in the mic and then up in the preamp; to have 2
transformers in the path that aren't really necessary if you just
place the preamp close to the performer. What would be a good tube
type and bias current to drive 1500 ohms to about +14 dBu on the very
highest peaks or so?

I'd like to point out that there is a small voltage change between
grid and cathode, or gate and source, in any real-world follower that
has a load attached and signal on it's input. If there were no change
in voltage difference there would be no output signal, since the
device's open-loop gain is finite. I don't know in practice if this
small a wiggle is big enough to produce capacitance-modulation related
distortion with JFETs (Mosfets are too noisy for this app) or tubes.

I also wouldn't say that Jfets always have higher Gm than tubes.
There are Jfets that have a Gm of only a couple thousand microMhos,
and tubes with 5000 or more. Gm is quite dependent on bias current as
well.

Scott, what is the capacitance modulation behavior for a typical tube
like a 12AU7? I've never seen this mentioned on a tube data sheet,
but then I've only looked at data sheets for the most common types.

One more question: grid to plate, grid to cathode, and cathode to
plate are all obvious terms for a tube's capacitances, but what do the
terms "input" capacitance and "output" capacitance really mean? I've
never found a proper definition for these terms.

Check out page 3 for capacitance versus voltage behavior for a high Gm
Jfet:

http://www.datasheetcatalog.org/data...shiba/1027.pdf

Thanks!

Sean B



  #21   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default Hi Z Buffer Distortion

wrote:

Scott, what is the capacitance modulation behavior for a typical tube
like a 12AU7? I've never seen this mentioned on a tube data sheet,
but then I've only looked at data sheets for the most common types.


You'll have to measure it. The HB-3 has numbers for some devices, but
not too many.

The 12AU7 is a crappy microphone tube for a lot of different reasons,
including microphonics.

One more question: grid to plate, grid to cathode, and cathode to
plate are all obvious terms for a tube's capacitances, but what do the
terms "input" capacitance and "output" capacitance really mean? I've
never found a proper definition for these terms.


Input capacitance is the capacitance seen at the input of the circuit...
that it, it's the grid-to-cathode capacitance, plus grid-to-plate
capacitance, plus and minus some stuff on the outside.

Likewise output capacitance is the capacitance seen at the output of
the circuit.... it's plate to cathode capacitance plus and minus some stuff.

There is a discussion of a lot of this stuff in Terman's book _Radio
Engineering_ as well as in the Radiotron Handbook.

Check out page 3 for capacitance versus voltage behavior for a high Gm
Jfet:

http://www.datasheetcatalog.org/data...shiba/1027.pdf

This is a lot wider variation than you will ever see with a tube, which
is unfortunate for people designing reactance tube modulators for FM.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
  #22   Report Post  
Posted to rec.audio.pro
Paul Stamler[_2_] Paul Stamler[_2_] is offline
external usenet poster
 
Posts: 178
Default Hi Z Buffer Distortion

Hey guys -- we still don't know what the original poster wants to buffer.
Everyone's assuming it's a condenser mic capsule -- but he didn't say that,
and we don't know it.

Peace,
Paul


  #23   Report Post  
Posted to rec.audio.pro
Paul Stamler[_2_] Paul Stamler[_2_] is offline
external usenet poster
 
Posts: 178
Default Hi Z Buffer Distortion

wrote in message
...

Yes it is a condenser capsule I am buffering. One with around 80 pF
of capacitance. I am messing around with a mic design.


I stand corrected -- now we know what's going on.

I would like to drive loads as low as 1500 ohms, in order for a user
to be able to use mic preamps with input transformers. But my
personal philosophy is that the mic preamp should be near the
performer. It seems ridiculous to knock the impedance down with a
transformer in the mic and then up in the preamp; to have 2
transformers in the path that aren't really necessary if you just
place the preamp close to the performer. What would be a good tube
type and bias current to drive 1500 ohms to about +14 dBu on the very
highest peaks or so?


The thing is, tubes don't really like that kind of impedance. The maximum
current you'd be swinging is about +/- 3.7 mA, so in theory something biased
at about 5mA would work -- maybe a 6SN7. But it wouldn't really be low
distortion.

My suggestion: use a tube to buffer the capsule, then something like a
Schoeps solid-state circuit to create a balanced output that'll happily
drive 1500 ohms.

Peace,
Paul

I'd like to point out that there is a small voltage change between
grid and cathode, or gate and source, in any real-world follower that
has a load attached and signal on it's input. If there were no change
in voltage difference there would be no output signal, since the
device's open-loop gain is finite. I don't know in practice if this
small a wiggle is big enough to produce capacitance-modulation related
distortion with JFETs (Mosfets are too noisy for this app) or tubes.

I also wouldn't say that Jfets always have higher Gm than tubes.
There are Jfets that have a Gm of only a couple thousand microMhos,
and tubes with 5000 or more. Gm is quite dependent on bias current as
well.

Scott, what is the capacitance modulation behavior for a typical tube
like a 12AU7? I've never seen this mentioned on a tube data sheet,
but then I've only looked at data sheets for the most common types.

One more question: grid to plate, grid to cathode, and cathode to
plate are all obvious terms for a tube's capacitances, but what do the
terms "input" capacitance and "output" capacitance really mean? I've
never found a proper definition for these terms.

Check out page 3 for capacitance versus voltage behavior for a high Gm
Jfet:

http://www.datasheetcatalog.org/data...shiba/1027.pdf

Thanks!

Sean B



  #24   Report Post  
Posted to rec.audio.pro
Don Pearce[_3_] Don Pearce[_3_] is offline
external usenet poster
 
Posts: 2,417
Default Hi Z Buffer Distortion

On Tue, 07 Apr 2009 18:45:02 GMT, "Paul Stamler"
wrote:

Hey guys -- we still don't know what the original poster wants to buffer.
Everyone's assuming it's a condenser mic capsule -- but he didn't say that,
and we don't know it.

We know it isn't a condenser mic capsule - he said it had a source
impedance of 100 megohms. Having said that, I have no idea what that
might be.

d
  #25   Report Post  
Posted to rec.audio.pro
[email protected][_2_] seanbroderick20003@yahoo.com[_2_] is offline
external usenet poster
 
Posts: 26
Default Hi Z Buffer Distortion


The 12AU7 is a crappy microphone tube for a lot of different reasons,
including microphonics. *


What are the other reasons?

I often like the sound of low-mu tubes, and the 12AU7 has a decent
current drive capability for a noval tube. Low capacitance, too. I
don't think a tube like a 6SN7 will fit in this mic. Also, a lower mu
gives you more headroom at the grid, if you do want to use a plate-
loaded configuration. Especially in a mic, where you're inclined to
use lower plate voltages for noise reasons.

Sean B


  #26   Report Post  
Posted to rec.audio.pro
[email protected] jwvm@umich.edu is offline
external usenet poster
 
Posts: 24
Default Hi Z Buffer Distortion

On Apr 7, 10:23*am, (GregS) wrote:
In article , wrote:
On Apr 7, 6:45=A0am, Eeyore
wrote:
Soundhaspriority wrote:
Correction, =A0Should read: =A0Vacuum tubes disappeared from instrument=

ation
with the advent of the FET back in the late 60's.


I can still remember my Tek 545 scopes ! And the fan at the back.


Graham


That was a great scope and very rugged! :-)


I had to fix a 565 recently and several plug ins. I hate the multi
component stuff. transistors, Nuvistors, and tubes.
I had to put in a nos crt in the scope. Funny thing, the rear
connector had corosion around a couple pins only, and
ate the pins alltogether. Couple weak tubes and was ready to go.
Troubleshooting the HV was the main problem, and was an internal
crt short.

greg


Parts, especially tubes, will wear out. Given the age of the hardware,
some things will go bad. These scopes are antiques. However, the
switches seemed to be almost indestructable compared to what followed
in the later scope series.
  #28   Report Post  
Posted to rec.audio.pro
[email protected] makolber@yahoo.com is offline
external usenet poster
 
Posts: 614
Default Hi Z Buffer Distortion

On Apr 7, 1:31*pm, "
wrote:
Thanks for all the responses!

Yes it is a condenser capsule I am buffering. *One with around 80 pF
of capacitance. *I am messing around with a mic design.

I would like to drive loads as low as 1500 ohms, in order for a user
to be able to use mic preamps with input transformers. *But my
personal philosophy is that the mic preamp should be near the
performer. *It seems ridiculous to knock the impedance down with a
transformer in the mic and then up in the preamp; to have 2
transformers in the path that aren't really necessary if you just
place the preamp close to the performer. *What would be a good tube
type and bias current to drive 1500 ohms to about +14 dBu on the very
highest peaks or so?

I'd like to point out that there is a small voltage change between
grid and cathode, or gate and source, in any real-world follower that
has a load attached and signal on it's input. *If there were no change
in voltage difference there would be no output signal, since the
device's open-loop gain is finite. *I don't know in practice if this
small a wiggle is big enough to produce capacitance-modulation related
distortion with JFETs (Mosfets are too noisy for this app) or tubes.

I also wouldn't say that Jfets always have higher Gm than tubes.
There are Jfets that have a Gm of only a couple thousand microMhos,
and tubes with 5000 or more. *Gm is quite dependent on bias current as
well.

Scott, what is the capacitance modulation behavior for a typical tube
like a 12AU7? *I've never seen this mentioned on a tube data sheet,
but then I've only looked at data sheets for the most common types.

One more question: grid to plate, grid to cathode, and cathode to
plate are all obvious terms for a tube's capacitances, but what do the
terms "input" capacitance and "output" capacitance really mean? *I've
never found a proper definition for these terms.

Check out page 3 for capacitance versus voltage behavior for a high Gm
Jfet:

http://www.datasheetcatalog.org/data...shiba/1027.pdf

Thanks!

Sean B


Sean,
the reason i asked about the condensor mic is that the condensor mic
ITSELF suffers from a (very small) distortion due to capacitance
modulation... so if you connect it to a device that also has some
capacitance modulation and you get the polarities correct, you might
even get some first order cancellation.

But this is all theory, in reality these effects are very very small
and unless you are dealing with some really loud sounds and/or very
critical instrument type application, I would say you shuold probably
have bigger fish to fry than the distortion caused by capacitance
modulation from a singal of a few mV.

Mark

  #29   Report Post  
Posted to rec.audio.pro
[email protected] makolber@yahoo.com is offline
external usenet poster
 
Posts: 614
Default Hi Z Buffer Distortion

On Apr 7, 2:06*pm, (Scott Dorsey) wrote:
wrote:

Scott, what is the capacitance modulation behavior for a typical tube
like a 12AU7? *I've never seen this mentioned on a tube data sheet,
but then I've only looked at data sheets for the most common types.


You'll have to measure it. *The HB-3 has numbers for some devices, but
not too many.

The 12AU7 is a crappy microphone tube for a lot of different reasons,
including microphonics. *

One more question: grid to plate, grid to cathode, and cathode to
plate are all obvious terms for a tube's capacitances, but what do the
terms "input" capacitance and "output" capacitance really mean? *I've
never found a proper definition for these terms.


Input capacitance is the capacitance seen at the input of the circuit...
that it, it's the grid-to-cathode capacitance, plus grid-to-plate
capacitance, plus and minus some stuff on the outside.

Likewise output capacitance is the capacitance seen at the output of
the circuit.... it's plate to cathode capacitance plus and minus some stuff.

There is a discussion of a lot of this stuff in Terman's book _Radio
Engineering_ as well as in the Radiotron Handbook.

Check out page 3 for capacitance versus voltage behavior for a high Gm
Jfet:


http://www.datasheetcatalog.org/data...shiba/1027.pdf


This is a lot wider variation than you will ever see with a tube, which
is unfortunate for people designing reactance tube modulators for FM.
--scott
--
"C'est un Nagra. *C'est suisse, et tres, tres precis."


the input capacitance of tube (triode) amp CIRCUIT will be dominated
by the grid to plate cap which is magnified by the Miller effect. So
the effective value of CIRCUIT input capacitnace is not simply
determined by the capacitance between the elements in the tube alone
but the amount of gain is also a factor.

Mark





  #30   Report Post  
Posted to rec.audio.pro
Mike Rivers Mike Rivers is offline
external usenet poster
 
Posts: 8,744
Default Hi Z Buffer Distortion

On Apr 7, 2:58 pm, (Don Pearce) wrote:

We know it isn't a condenser mic capsule - he said it had a source
impedance of 100 megohms. Having said that, I have no idea what that
might be.


Probably a piezoelectric instrument pickup. He must have heard the
term "buffer" applied to interfacing a high impedance source with the
real world, and didn't want to use the old, tired, "Hi Z guitar input"
terminology that we'd all understand.



  #32   Report Post  
Posted to rec.audio.pro
Paul Stamler[_2_] Paul Stamler[_2_] is offline
external usenet poster
 
Posts: 178
Default Hi Z Buffer Distortion


wrote in message
...

The 12AU7 is a crappy microphone tube for a lot of different reasons,
including microphonics.


What are the other reasons?

It's less linear than the 6SN7 family, for one. And microphonics in this
application are a deal-breaker.

I often like the sound of low-mu tubes, and the 12AU7 has a decent
current drive capability for a noval tube. Low capacitance, too. I
don't think a tube like a 6SN7 will fit in this mic.

The 6CG7/6FQ7 is a 6SN7 in a noval envelope. It's taller than a 12AU7, but
the same diameter. Very, very linear, and if you bias it up to a high enough
current, low-noise as well.

Another alternative, popular among tube mic manufacturers, is the EF86,
which I gather is back in production, or the somewhat similar 5879. These
are pentodes, but were popular in microphones when triode-wired. Don't know
the linearity specs, but they're quiet and very non-microphonic, and they're
both about the size of a 12AU7.

Peace,
Paul


  #33   Report Post  
Posted to rec.audio.pro
[email protected][_2_] seanbroderick20003@yahoo.com[_2_] is offline
external usenet poster
 
Posts: 26
Default Hi Z Buffer Distortion

On Apr 8, 1:29*am, "Paul Stamler" wrote:


The 6CG7/6FQ7 is a 6SN7 in a noval envelope. It's taller than a 12AU7, but
the same diameter. Very, very linear, and if you bias it up to a high enough
current, low-noise as well.


Thanks, Paul! Have you ever heard the type 5687 triode? The curves
look similar to the 6CG7.
  #34   Report Post  
Posted to rec.audio.pro
Paul Stamler[_2_] Paul Stamler[_2_] is offline
external usenet poster
 
Posts: 178
Default Hi Z Buffer Distortion

wrote in message
...
On Apr 8, 1:29 am, "Paul Stamler" wrote:

The 6CG7/6FQ7 is a 6SN7 in a noval envelope. It's taller than a 12AU7, but
the same diameter. Very, very linear, and if you bias it up to a high
enough
current, low-noise as well.


Thanks, Paul! Have you ever heard the type 5687 triode? The curves
look similar to the 6CG7.

I haven't myself, but it has a good reputation; I gather each section
behaves approximately like two 6CG7 (=6SN7) sections in parallel, making it
a pretty powerful tube for a small-signal triode. However, IIRC it runs
quite hot, making it perhaps a poor choice for a microphone. And the input
capacitance is on the high side, again IIRC.

Peace,
Paul


  #35   Report Post  
Posted to rec.audio.pro
Chris Hornbeck Chris Hornbeck is offline
external usenet poster
 
Posts: 1,744
Default Hi Z Buffer Distortion

On Sun, 12 Apr 2009 07:25:12 GMT, "Paul Stamler"
wrote:

wrote in message
Thanks, Paul! Have you ever heard the type 5687 triode? The curves
look similar to the 6CG7.

I haven't myself, but it has a good reputation; I gather each section
behaves approximately like two 6CG7 (=6SN7) sections in parallel, making it
a pretty powerful tube for a small-signal triode. However, IIRC it runs
quite hot, making it perhaps a poor choice for a microphone. And the input
capacitance is on the high side, again IIRC.


The 5687's are a dual, with huge cathodes and twice the heater
power of the already hefty 6SN7/6CG7/6FQ7 family. The big cathodes
and other good geometry gives them very sharp low-current knees
and excellent linearity, maybe the best of the 3 to 10 mA/V bottles
(and!, they're at the top end of that gm group). And when the
military dumped their supply a few years ago, they were a stone
bargain.

But as you say, they're not a first choice for stuffing into an
existing mic body. How important is gm to this gig? Valves with
1.5 mA/V (with 1.0 mA plate current) and exceptional linearity are
still available in a 7-pin as type 6AV6. Its triode is a half of
a 12AX7, probably the most linear device (into a suitably high
impedance, but also into a somewhat reactive load) that *can* be
made, and "NOS" devices shouldn't cost much of anything. (Selection
for microphonics will still be a project for *any* valve for this
gig, natch).

Random thoughts,
Much thanks, as always,
Chris Hornbeck


  #36   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default Hi Z Buffer Distortion

Chris Hornbeck wrote:
But as you say, they're not a first choice for stuffing into an
existing mic body. How important is gm to this gig? Valves with
1.5 mA/V (with 1.0 mA plate current) and exceptional linearity are
still available in a 7-pin as type 6AV6. Its triode is a half of
a 12AX7, probably the most linear device (into a suitably high
impedance, but also into a somewhat reactive load) that *can* be
made, and "NOS" devices shouldn't cost much of anything. (Selection
for microphonics will still be a project for *any* valve for this
gig, natch).


6AV6 and the ever-popular 6AU6 pentode are both used in AA5 table radios
and therefore are quite plentiful. The bad news is that there are a lot
of noisy ones out there so you may have to do some selection.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
  #37   Report Post  
Posted to rec.audio.pro
Paul Stamler[_2_] Paul Stamler[_2_] is offline
external usenet poster
 
Posts: 178
Default Hi Z Buffer Distortion

"Chris Hornbeck" wrote in message
...

But as you say, they're not a first choice for stuffing into an
existing mic body. How important is gm to this gig? Valves with
1.5 mA/V (with 1.0 mA plate current) and exceptional linearity are
still available in a 7-pin as type 6AV6. Its triode is a half of
a 12AX7, probably the most linear device (into a suitably high
impedance, but also into a somewhat reactive load) that *can* be
made, and "NOS" devices shouldn't cost much of anything.


Naah; the 12AX7 isn't anything like the most linear tube out there. A 6SL7
can beat it, and so can a 5751, but if you want real clean behavior, a good
6SN7 or its relatives beats the crap out of any higher-mu tube.

Peace,
Paul


  #38   Report Post  
Posted to rec.audio.pro
Chris Hornbeck Chris Hornbeck is offline
external usenet poster
 
Posts: 1,744
Default Hi Z Buffer Distortion

On Tue, 14 Apr 2009 05:26:21 GMT, "Paul Stamler"
wrote:

type 6AV6. Its triode is a half of
a 12AX7, probably the most linear device (into a suitably high
impedance, but also into a somewhat reactive load) that *can* be
made, and "NOS" devices shouldn't cost much of anything.


Naah; the 12AX7 isn't anything like the most linear tube out there. A 6SL7
can beat it, and so can a 5751, but if you want real clean behavior, a good
6SN7 or its relatives beats the crap out of any higher-mu tube.


I'll not argue too much - the last vacuum valve amplifier that
I built (in 1994) used 1/2 of a type 6SL7 driving paralleled sections
of a type 6SN7 driving a type 845, famous for its drive neediness
(over 300 volts peak-to-peak; yikes).

Good bottles of both these types are really, really good. My
favorites are the old Sylvania's with cylindrical cathodes, but
these days the hot tip is to use the 12 volt heater versions
(for availability!) or other un-hip voltages. And plain vanilla
Russians are surprisingly good - not much to go wrong with
these types really.

A good 12AX7/ECC83 is in this same ballpark, and IME more reliably
linear (into very high Z optimised loading, yada yada). "Good" here
means Telefunken's and large geometry Amperex's. Also the old
large geometry TungSol's and RCA's if sorted for noise are great.

There are lots of crappy, inexcusable 12AX7 type devices these
days - they are *not* 12AX7's - anything that looks Chinese is
probably a 12AT7 geometry (and worthless for linear audio).

Never seen a really linear 5751, but then I've sure not seen
everything! And they have an excellent rep.


Much thanks, as always,
Chris Hornbeck
  #39   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default Hi Z Buffer Distortion

Chris Hornbeck wrote:
There are lots of crappy, inexcusable 12AX7 type devices these
days - they are *not* 12AX7's - anything that looks Chinese is
probably a 12AT7 geometry (and worthless for linear audio).


As far as I know, all of these tubes are from the same Chinese factory.
It's the same one that makes the "KT88" tubes that have a maximum 250V
plate voltage on the data sheet.

Never seen a really linear 5751, but then I've sure not seen
everything! And they have an excellent rep.


To be honest, there are a LOT of compactrons out there that are intended
for use in TV tuners, with very low noise and very low input capacitance,
which sell for nothing and are excellent performers in high-Z circuits.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
96 khz/DMA buffer [email protected] Pro Audio 145 June 30th 08 11:47 AM
Tube Buffer for CDP Goh Ah Soon Vacuum Tubes 1 December 29th 07 01:06 AM
Buffer size...? Phil Gardner Pro Audio 1 February 7th 05 10:39 PM
Buffer Underrun Problem Brian Standefer Pro Audio 0 October 15th 04 02:16 PM
Buffer underrun and DAO audio discs Mike Pro Audio 3 August 12th 04 11:24 AM


All times are GMT +1. The time now is 10:44 AM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"