Practical limit to tube based phono preamps [message #96236] |
Mon, 19 December 2022 20:12 |
|
gofar99
Messages: 1951 Registered: May 2010 Location: Southern Arizona
|
Illuminati (5th Degree) |
|
|
Hi, (this could go in either this forum or the tube one) In my never ending quest to get rid of hum and noise I seem to have hit an end point in tube based phono preamps that do not use global NFB. There seems to be what is essentially a practical limit to what can be done without going to extremes. The current preamps are in the -85 to -90dbv range for MM/MI use. That is really quiet BTW. Well below the noise level of a really good and clean LP. Depending on what study you check that value is in the -65 to -75dbv range. With a preamp 15 to 25 db quieter than the best you can get off a LP that is really quite sufficient. I however, am a fanatic about hum and noise. If I can hear it or measure it then it is too much. The limits of my test gear is right around -105dbv. I can get solid state phono preamps close to that but they use NFB. So getting to the limits...I removed everything that had AC on it and all power supply filtering from one preamp and made it into a two box design. I figured that it ought to be quieter...nope. When I compared it to one that is a single chassis I use in my main system the differences were really slight. Under a single db for noise and nearly identical for signal related 60HZ hum. Power supply hum and its harmonics were in both cases within a db of each other. Even with getting selected low noise tubes I have apparently reached the limit in this sort of application. (for those who are interested in this the difference between standard JJ tubes and three different selected low noise ones was only about 1.5db) So what I concluded was that the tubes and passive components were the remaining factors and cause for the wide band low level residual hiss. As I indicated in the beginning none of this is audible. My system has a digital (passive) preamp and the usual listening level is at about -40db according to the display. I have to go to -3 or -4 to hear anything right at the speakers. This level of gain would greatly exceed the ability of the amps to deliver. For now I will have to be satisfied and sit back and enjoy the music.
Good Listening
Bruce
|
|
|
Re: Practical limit to tube based phono preamps [message #96237 is a reply to message #96236] |
Tue, 20 December 2022 13:39 |
|
Wayne Parham
Messages: 18792 Registered: January 2001
|
Illuminati (33rd Degree) |
|
|
Dude, that's so cool. One of the most useful explorations in tube audio - getting noise down.
Assuming full audio bandwidth as highest priority, reducing noise is only superseded in importance to reducing distortion, and I'm not even sure they're in that order. Once harmonic distortion is under 3% or so - provided its spectral distribution naturally descends in orders - this small amount of distortion isn't really all that unpalatable.
Getting a low noise floor is definitely one of the holy grails in tube audio.
So from what you said, none of the noise - or very little - is related to your amplifier's power supply. To me, that's huge because I find power supply hum to be most annoying. Hiss is less distracting, especially if it is at a very low level. And I'm not sure it can be cured because it's part of the package, being an issue of colliding electrons. It's not like power supply ripple or a ground loop, it's inherent in any circuit with resistance.
At some point, white noise can't be reduced. It's not a qualitative thing - better components won't help here - the only way to reduce it is to reduce bandwidth or reduce circuit impedance or both. And since bandwidth is a sort of a constant (20kHz for audio), we are limited in what can be done at that most fundamental level.
|
|
|
|
Re: Practical limit to tube based phono preamps [message #96269 is a reply to message #96238] |
Wed, 11 January 2023 17:47 |
positron
Messages: 113 Registered: May 2020
|
Viscount |
|
|
A few thoughts for the newbies out there.
1. Tubes produces internal noise from a variety of reasons.
2. The resistors associated with said tube also produce noise.
A. The cathode resistor noise will be amplified by the tube's gain.
B. The grid resistor noise will also be amplified by the tube's gain.
This resistor is generally much larger than the cathode resistor and
will generate more noise.
C. The "stopper" resistor will produce noise that will be amplified by
the tube's gain.
D. The source, itself, will generate noise which is fed into the tube
and will be amplified by the tube's gain.
E. From my understanding, the Vishay naked resistors produce the lowest
noise. After that a good wire wound (non inductive for audio)is
preferred. Both are expensive, but reduce noise.
3. Concerning 120hz AC hum, I prefer at least 4 stages of filtering, 5 is even
better. My preamplifier and phono stage have 6 stages. I never use an
inductor anywhere as it causes non-linearities due to
2 factors. (If I did, it would be only in the early stages of filtering.)
A. One is inductance which is frequency sensitive. It would appear as
a straight diagonal line in the circuit below.
B. The other factor is dc resistance, which is not frequency sensitive.
Attached is a simple circuit. The choke/inductor has 200 ohms winding
dc resistance. Notice the curve where dc resistance becomes dominant.
C. Me and another music lover have experimented between choke/inductor and
resistor and the sonic change is quite substantial.
Distortion being defined as the altering of sound in any way.
With speaker crossovers, dc resistance tends to be minimized.
cheers
pos
|
|
|
|
|
|
Re: Practical limit to tube based phono preamps [message #96274 is a reply to message #96272] |
Fri, 13 January 2023 16:37 |
positron
Messages: 113 Registered: May 2020
|
Viscount |
|
|
From my understanding, ragged edges, rough wire surface, impurities in
the material/copper cause eddy currents and is one source of noise. I can't
remember which resistor company discussed this aspect though.
A couple of decades ago, the finite number of NOS vacuum tubes and prices
worried me. So I searched, tested and found the small signal JJs not only
had the lowest harmonic distortion I had ever measured, -79db below the
fundamental, but the sound quality was the most accurate/natural.
I have gobs of Bugleboys,PQs, and others, but the JJs won out with visitors
as well.
(I am not affiliated with JJ in any way, shape, or form. I am also retired
for some 10 years from designing, but still tinker.)
I am not saying harmonic distortion is the only criteria, but it sure
helps. Sovteks, EH and others are ok, but they are not NOS and not JJs.
Turned out the harmonic distortion of the JJ E88cc (prob others as well)
was some 1/10th (-19db) that of any other tube I have ever tested.
The materials used are also excellent.
Still, dealing with such small signals as the OP, one has to deal with
the noise and hum. For me, it is by far the noise over hum.
Tubes with high Rp and RL (plate resistances and plate resistors) tend
to have the highest noise levels as well.
I know, kind of a diatribe. Thoughts just run together sometimes.
pos
|
|
|
|
|