Req.
Hi,
No need for that.
Correct.The noise produced by the tube was compared to the equivalent noise of a resistor.
In those days almost all resistors were carbon composition resistors.
These were reputedly noisy and this was used as a measure of comparison.
The Req spec is mainly used by European manufacturers and mostly only for SQ (special quality) tubes.
Some more recent tubes made in the former USSR also state this noise factor in the same way although I am not certain the measurements are taken under the same conditions.
Cheers,
Hi,
No need for that.
Correct.The noise produced by the tube was compared to the equivalent noise of a resistor.
In those days almost all resistors were carbon composition resistors.
These were reputedly noisy and this was used as a measure of comparison.
The Req spec is mainly used by European manufacturers and mostly only for SQ (special quality) tubes.
Some more recent tubes made in the former USSR also state this noise factor in the same way although I am not certain the measurements are taken under the same conditions.
Cheers,
As far as I know Johnson noise is the same as thermal nosie or also called white noise.
the formula for Johnson noise doesn't include any constants for the different resistor types, but everyone know that carbon composition resistors are more noiser than wirewound types.
The formula for Thermal noise is:
U=ROOT(4kTRB)
Where U= RMS noise voltage
k= 1,38*10e-23
T= temperature in Kelvin
R= resistance
B= bandwidth
As you can see in this formula, one way to decrease the noise is to decrease the bandwidth, temperature or the value of resistance.
Thermal noise is due to the random motion of electrons. It's unaffected by DC current, so a resistor generates thermal noise even when sitting in a drawer.
We also have other kinds of noise sources.
We have the shot noise and filcker noise sometimes also called pink noise.
Shot noise arises whenever charges cross a potential barrier.
Flicker noise is often greater at low frequencies than higher up in the spectrum.
I don't know very much about the noise sources in vacuum tubes, so some of you must help me a little bit here.
Do we have shot noise between the heater and the cathode?
Is this the reason why someone recommend to let the heatervoltage float at a voltage of about 40V above ground?
And what kind of noise do we have when the moving electrons crash with residues gas molecules, is that Flicker noise?
Do we have shot noise between the cathode and anode when a tube is in cut-off?
What else inside the tube is causing noise?
What about secondary emisson caused by the electrons when they are hitting the anode?
Or positive ions travelling to the cathode?
the formula for Johnson noise doesn't include any constants for the different resistor types, but everyone know that carbon composition resistors are more noiser than wirewound types.
The formula for Thermal noise is:
U=ROOT(4kTRB)
Where U= RMS noise voltage
k= 1,38*10e-23
T= temperature in Kelvin
R= resistance
B= bandwidth
As you can see in this formula, one way to decrease the noise is to decrease the bandwidth, temperature or the value of resistance.
Thermal noise is due to the random motion of electrons. It's unaffected by DC current, so a resistor generates thermal noise even when sitting in a drawer.
We also have other kinds of noise sources.
We have the shot noise and filcker noise sometimes also called pink noise.
Shot noise arises whenever charges cross a potential barrier.
Flicker noise is often greater at low frequencies than higher up in the spectrum.
I don't know very much about the noise sources in vacuum tubes, so some of you must help me a little bit here.
Do we have shot noise between the heater and the cathode?
Is this the reason why someone recommend to let the heatervoltage float at a voltage of about 40V above ground?
And what kind of noise do we have when the moving electrons crash with residues gas molecules, is that Flicker noise?
Do we have shot noise between the cathode and anode when a tube is in cut-off?
What else inside the tube is causing noise?
What about secondary emisson caused by the electrons when they are hitting the anode?
Or positive ions travelling to the cathode?
Shot noise
Triodes produce shot noise. At RF:
Req = 2.5/gm (roughly)
So, to return to the original post. For lowest noise, we need to run as much anode current as we dare, because gm increases with anode current.
At audio frequencies, other mechanisms are present. Electron collision with gas molecules will cause ions to be accelerated towards the grid/cathode region. When they collide, their charge is neutralised by an electron flowing up into the grid or cathode. Gas collisions are less frequent than the electron strikes at the anode, so this produces LF noise. Other sources of noise are DC surface currents in the micas due to contamination by getter material. DC currents due to fluff. Anything causes noise...
Tungsten filament emitters do not produce flicker noise. Don't get excited though - they are horribly microphonic instead, and their gm is low.
The 40V heater/cathode recommendation is due to two reasons. Firstly, the tungsten filament can emit electrons. Biasing it positively prevents electron flow to the cathode. But when I measured, 10V seemed quite sufficient to turn the diode off, and it was very difficult to be certain that there really was a diode there. Secondly, and this seems more likely, the nickel cathode structure looks like an anode to the tungsten filament and can cause electrolysis through the alumina insulation, transferring (conductive) tungsten and causing it to become leaky. Causing noise. I would welcome comments from more chemically minded people on this one...
Triodes produce shot noise. At RF:
Req = 2.5/gm (roughly)
So, to return to the original post. For lowest noise, we need to run as much anode current as we dare, because gm increases with anode current.
At audio frequencies, other mechanisms are present. Electron collision with gas molecules will cause ions to be accelerated towards the grid/cathode region. When they collide, their charge is neutralised by an electron flowing up into the grid or cathode. Gas collisions are less frequent than the electron strikes at the anode, so this produces LF noise. Other sources of noise are DC surface currents in the micas due to contamination by getter material. DC currents due to fluff. Anything causes noise...
Tungsten filament emitters do not produce flicker noise. Don't get excited though - they are horribly microphonic instead, and their gm is low.
The 40V heater/cathode recommendation is due to two reasons. Firstly, the tungsten filament can emit electrons. Biasing it positively prevents electron flow to the cathode. But when I measured, 10V seemed quite sufficient to turn the diode off, and it was very difficult to be certain that there really was a diode there. Secondly, and this seems more likely, the nickel cathode structure looks like an anode to the tungsten filament and can cause electrolysis through the alumina insulation, transferring (conductive) tungsten and causing it to become leaky. Causing noise. I would welcome comments from more chemically minded people on this one...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.