It has always been my understanding that vacuum tubes actually
generate noise when used in circuits and this noise prevents tubes
from achieving the same signal to noise as a solid state or transistor
circuit can achieve. I recall that during my early days in amateur
radio someone even developed a procedure whereby some of the tube in
the receiver circuit were converted to a solid state tube to help
reduce the noise inherent in tubes so that extremely weak radio
signals were not covered by not only normal atmospheric noise but that
PLUS the noise added by the tube circuits. I asked a former NASA
engineer about tube noise and he dismissed this by saying that it was
not so much the tubes themselves but other noisy components in the
circuit like capacitors and resistors and that by carefully choosing
the appropriate devices for the circuit that today you can now design
a very quiet tube circuit that equals solid state. This is contrary
to everything I have ever heard before. Is his statement accurate or
not and why?


--

"Richard Staples" wrote in message
...
It has always been my understanding that vacuum tubes actually
generate noise when used in circuits and this noise prevents tubes
from achieving the same signal to noise as a solid state or transistor
circuit can achieve. I recall that during my early days in amateur
radio someone even developed a procedure whereby some of the tube in
the receiver circuit were converted to a solid state tube to help
reduce the noise inherent in tubes so that extremely weak radio
signals were not covered by not only normal atmospheric noise but that
PLUS the noise added by the tube circuits. I asked a former NASA
engineer about tube noise and he dismissed this by saying that it was
not so much the tubes themselves but other noisy components in the
circuit like capacitors and resistors and that by carefully choosing
the appropriate devices for the circuit that today you can now design
a very quiet tube circuit that equals solid state. This is contrary
to everything I have ever heard before. Is his statement accurate or
not and why?


--


From my experience the converse is true. Selection of low noise vacuum
tubes is important especially for tubes in the earlier amplification stages
e.g. phono and 1st or even 2nd preamplifier stages where tube generated
noise can be more apparent, but this noise is also effected by circuit
design less than just discrete components per see. IMO low noise tube
selection is just one criteria amongst many in achieving sonically good
performance from tubed audio equipment.

Mike



--

Richard Staples wrote:
It has always been my understanding that vacuum tubes actually
generate noise when used in circuits and this noise prevents tubes
from achieving the same signal to noise as a solid state or transistor
circuit can achieve. I recall that during my early days in amateur
radio someone even developed a procedure whereby some of the tube in
the receiver circuit were converted to a solid state tube to help
reduce the noise inherent in tubes so that extremely weak radio
signals were not covered by not only normal atmospheric noise but that
PLUS the noise added by the tube circuits. I asked a former NASA
engineer about tube noise and he dismissed this by saying that it was
not so much the tubes themselves but other noisy components in the
circuit like capacitors and resistors and that by carefully choosing
the appropriate devices for the circuit that today you can now design
a very quiet tube circuit that equals solid state. This is contrary
to everything I have ever heard before. Is his statement accurate or
not and why?



I'd ask said "NASA engineer" precisely which tubes he belives have
inherent noise generation levels equal or less than resistors or
capacitors, and which he believes are equal or lower noise than a
typical JFET. That's a start. Let us know?

There are low noise thermionic devices like some Nuvistors, but still
JFETs and certain bipolar transistors (many) are still quieter by a
fairly wide margin.

For audio applications though, achieving extrememly low noise levels
isn't really a tremendously important issue, the exception being when
amplifying really low level signals, like condensor mics and phonograph
cartridges... assuming relatively good low noise circuits using tubes
are what we're talking about - there are some very noisy tube units out
there, and that's not a good thing. Usually the inherent noise in a
noisy line stage can be reduced through better design/circuit changes,
or changing out the tube type or selecting for a low noise example of
that given type.

_-_-bear


--

"Richard Staples" wrote in message
...
It has always been my understanding that vacuum tubes actually
generate noise when used in circuits and this noise prevents tubes
from achieving the same signal to noise as a solid state or transistor
circuit can achieve. I recall that during my early days in amateur
radio someone even developed a procedure whereby some of the tube in
the receiver circuit were converted to a solid state tube to help
reduce the noise inherent in tubes so that extremely weak radio
signals were not covered by not only normal atmospheric noise but that
PLUS the noise added by the tube circuits. I asked a former NASA
engineer about tube noise and he dismissed this by saying that it was
not so much the tubes themselves but other noisy components in the
circuit like capacitors and resistors and that by carefully choosing
the appropriate devices for the circuit that today you can now design
a very quiet tube circuit that equals solid state. This is contrary
to everything I have ever heard before. Is his statement accurate or
not and why?

It appears that hardly 2 tubes were created the same. Some of the same
manufacture and type are more "noisy" (and "microphonic") than others.
Nevertheless all tubes display some "tube rush". I can see it on the meters
in my recording devices and can hear it with headphones or when I walk up
close to my Maggie ribbons. Nevertheless normal tube rush doesn't present
any problem when listening to music.


--

In article , BEAR
wrote:

--snippage--

For audio applications though, achieving extrememly low noise levels
isn't really a tremendously important issue, the exception being when
amplifying really low level signals, like condensor mics and phonograph
cartridges... assuming relatively good low noise circuits using tubes
are what we're talking about - there are some very noisy tube units out
there, and that's not a good thing. Usually the inherent noise in a
noisy line stage can be reduced through better design/circuit changes,
or changing out the tube type or selecting for a low noise example of
that given type.

I would argue that even a phono stage isn't all that critical. It's easy
to design a preamp that's better than 10 dB quieter than the noise level
of a stylus in a groove; shoot, that doesn't even take "low noise"
technique. I don't think it matters if the preamp noise is audible when
the stylus is *not* in the groove, BTW; not too many folks spend much
time listening to that.

More important would be low distortion, and that can be traded off
against lowest possible noise by using, say, a differential input stage.

Isaac


--

"Mike Gilmour" wrote in message
...
"Richard Staples" wrote in message
...
It has always been my understanding that vacuum tubes actually
generate noise when used in circuits and this noise prevents tubes
from achieving the same signal to noise as a solid state or transistor
circuit can achieve. I recall that during my early days in amateur
radio someone even developed a procedure whereby some of the tube in
the receiver circuit were converted to a solid state tube to help
reduce the noise inherent in tubes so that extremely weak radio
signals were not covered by not only normal atmospheric noise but that
PLUS the noise added by the tube circuits. I asked a former NASA
engineer about tube noise and he dismissed this by saying that it was
not so much the tubes themselves but other noisy components in the
circuit like capacitors and resistors and that by carefully choosing
the appropriate devices for the circuit that today you can now design
a very quiet tube circuit that equals solid state. This is contrary
to everything I have ever heard before. Is his statement accurate or
not and why?


--


From my experience the converse is true. Selection of low noise vacuum
tubes is important especially for tubes in the earlier amplification
stages
e.g. phono and 1st or even 2nd preamplifier stages where tube generated
noise can be more apparent, but this noise is also effected by circuit
design less than just discrete components per see. IMO low noise tube
selection is just one criteria amongst many in achieving sonically good
performance from tubed audio equipment.

If all sources of excess noise were to be eliminated--shot noise,
microphonics, etc.--leaving nothing but thermal (Johnson) noise, tubes would
still be noisier simply because they're hotter.

Norm Strong

--
NewsGuy.Com 30Gb $9.95 Carry Forward and On Demand Bandwidth

isw wrote:
In article , BEAR
wrote:

--snippage--


For audio applications though, achieving extrememly low noise levels
isn't really a tremendously important issue, the exception being when
amplifying really low level signals, like condensor mics and phonograph
cartridges... assuming relatively good low noise circuits using tubes
are what we're talking about - there are some very noisy tube units out
there, and that's not a good thing. Usually the inherent noise in a
noisy line stage can be reduced through better design/circuit changes,
or changing out the tube type or selecting for a low noise example of
that given type.


I would argue that even a phono stage isn't all that critical. It's easy
to design a preamp that's better than 10 dB quieter than the noise level
of a stylus in a groove; shoot, that doesn't even take "low noise"
technique. I don't think it matters if the preamp noise is audible when
the stylus is *not* in the groove, BTW; not too many folks spend much
time listening to that.

I don't think you'd want to listen to a phono stage that was only 10dB
quieter than typical stylus noise.

Hum is part of that equation, and getting the hum out is important too.

Solid state is easier than tubes for noise, but still an issue.


More important would be low distortion, and that can be traded off
against lowest possible noise by using, say, a differential input stage.

Actually, "low distortion" is relative.

Differential circuits tend to end up sounding rather more "hard" and
"sterile" than non-differential circuits due to the spectra of
distortion products being mostly odd order, and cancelling strongly the
second harmonic.

It has been shown in peer reviewed Journal publications that merely low
THD does not translate directly to audibility of distortion.

Consider that your signal is *modulated* by all low level artifacts,
noise, hum, distortions etc... so the actual level of the nominal signal
you are listening to being *above* the nominal noise/hum/distortion
floor does not mean that it is unaffected by same.

But as long as one is happy with what one is hearing, then all is well...

As Count Basie said (iirc) "If it sounds good, it is good..." or was
that Duke Ellington??

_-_-bear



Isaac




--

"BEAR" wrote in message
...
isw wrote:
In article , BEAR
wrote:

--snippage--


For audio applications though, achieving extrememly low noise levels
isn't really a tremendously important issue, the exception being when
amplifying really low level signals, like condensor mics and phonograph
cartridges... assuming relatively good low noise circuits using tubes are
what we're talking about - there are some very noisy tube units out
there, and that's not a good thing. Usually the inherent noise in a noisy
line stage can be reduced through better design/circuit changes, or
changing out the tube type or selecting for a low noise example of that
given type.


I would argue that even a phono stage isn't all that critical. It's easy
to design a preamp that's better than 10 dB quieter than the noise level
of a stylus in a groove; shoot, that doesn't even take "low noise"
technique. I don't think it matters if the preamp noise is audible when
the stylus is *not* in the groove, BTW; not too many folks spend much
time listening to that.

I don't think you'd want to listen to a phono stage that was only 10dB
quieter than typical stylus noise.

Hum is part of that equation, and getting the hum out is important too.

Solid state is easier than tubes for noise, but still an issue.


More important would be low distortion, and that can be traded off
against lowest possible noise by using, say, a differential input stage.

Actually, "low distortion" is relative.

Differential circuits tend to end up sounding rather more "hard" and
"sterile" than non-differential circuits due to the spectra of distortion
products being mostly odd order, and cancelling strongly the second
harmonic.

It has been shown in peer reviewed Journal publications that merely low
THD does not translate directly to audibility of distortion.

Consider that your signal is *modulated* by all low level artifacts,
noise, hum, distortions etc... so the actual level of the nominal signal
you are listening to being *above* the nominal noise/hum/distortion floor
does not mean that it is unaffected by same.

But as long as one is happy with what one is hearing, then all is well...

As Count Basie said (iirc) "If it sounds good, it is good..." or was that
Duke Ellington??

_-_-bear



Isaac


Duke Ellington

Mike



--

"Richard Staples" wrote in message
...
It has always been my understanding that vacuum tubes actually
generate noise when used in circuits and this noise prevents tubes
from achieving the same signal to noise as a solid state or transistor
circuit can achieve. I recall that during my early days in amateur
radio someone even developed a procedure whereby some of the tube in
the receiver circuit were converted to a solid state tube to help
reduce the noise inherent in tubes so that extremely weak radio
signals were not covered by not only normal atmospheric noise but that
PLUS the noise added by the tube circuits. I asked a former NASA
engineer about tube noise and he dismissed this by saying that it was
not so much the tubes themselves but other noisy components in the
circuit like capacitors and resistors and that by carefully choosing
the appropriate devices for the circuit that today you can now design
a very quiet tube circuit that equals solid state. This is contrary
to everything I have ever heard before. Is his statement accurate or
not and why?


Well, it shouldn't be that difficult to measure -- both Linear Technology
and Texas Instruments have application notes on noise measurements for their
low-noise regulators -- and the same procedure can be applied to measuring
the noise in a tube preamplifier, an amplifier, or the noise in an LED used
to bias a transistor amplifier -- seems that the favorite instrument is an
HP3403C thermopile true RMS meter with about 60dB of gain ahead of it -- you
bandpass filter the signal from 100 mHz to 10 Hz and integrate the
measurement over a 10 second interval.

With respect to low noise receivers -- jitter = noise



--