yagoolar's tube amp projects

[disco]

Low, compared to .... a 270 ohm resistor?

[SY]

A single series string of 105 series LEDs gave a noise of about 2uV over 22Hz-22kHz (ref Valve Amplifiers, 4th ed). Not sure how many this amp used, but let's say 10. So that would correspond to roughly 0.6uV before paralleling. The noise of the paralleled strings would reduce as the square root of the number of parallel strings, so you could drop that by a factor of 2 or 3. So the noise is comparable to an ideal (zero excess noise) 270R at room temperature (which a cathode resistor is unlikely to be).

All moot, of course: the tube is far noisier.

[yagoolar]

It is actually 4 paralleled series strings of 26 LEDs per tube.

[SY]

OK, so about 0.5uV Possibly a bit less because Morgan had a dodgy solder joint which contributed to his noise measurement. Eother way, errrr, lost in the noise.

Thanks for the clarification.

[disco]

Looking at the Johnson noise from a 270 ohm resistor at room temperature: Vnoise(rms) = root(4kTRB)
k= boltzmanns' constant
T= abs temperature (293 degr Kelvin)
R= resistance
B= bandwith

4kT= 1.62 x 10~-24, times 270 ohms, times 21,980 Hz= 9.6 X 10~-18
Vnoise = 3.1 x 10~-9 = 0,003 microvolt/Hz~-1/2

That would be less than the LEDs. Is this white noise on the cathode complete harmless or can it lead to 'listening fatigue'? I have no clue how to implement these figures but some preferred the simple cathode resistor for prolonged listening while others preferred the immediate gain in weight on the bassnotes and stronger attack (not faster) with LEDs. Perhaps it's a case tot case study, how it mixes in with other shortcomings of our audio creations

[yagoolar]

Originally Posted by disco

Looking at the Johnson noise from a 270 ohm resistor at room temperature: Vnoise(rms) = root(4kTRB)
k= boltzmanns' constant
T= abs temperature (293 degr Kelvin)
R= resistance
B= bandwith

4kT= 1.62 x 10~-24, times 270 ohms, times 21,980 Hz= 9.6 X 10~-18
Vnoise = 3.1 x 10~-9 = 0,003 microvolt/Hz~-1/2

That would be less than the LEDs. Is this white noise on the cathode complete harmless or can it lead to 'listening fatigue'? I have no clue how to implement these figures but some preferred the simple cathode resistor for prolonged listening while others preferred the immediate gain in weight on the bassnotes and stronger attack (not faster) with LEDs. Perhaps it's a case tot case study, how it mixes in with other shortcomings of our audio creations

I am talking about substituting RC with LEDs.

The best way would be to compare apples with apples, i.e. measure the noise from R or RC in the similar conditions compared to LED tests.

As to calculations you should take 820 ohm resistor @ say 320 K. This will roughly give 0,02 uV/Hz. Is this difference significant to hearing? I don't know. I am a cello player with some orchestral experience, so I always try to compare real sound to what is reproduced. I perceive LED bias (vs. RC) more real.

[yagoolar]

CCS's for C3g plate mounted.

[yagoolar]

After having read a post on connecting g3 of C3g (to either cathode or anode/plate) while g2 connected via resistor, say, 220 ohm to plate, I am in doubt. Which one is ...
Source: http://audioroundtable.com/GroupBuild/messages/1999.html

[yagoolar]

Originally Posted by tcqanh

why don't you use fixed bias for power stage ? it is simple and easy to adjust the operation point.

many people here gave up the LED bias....

I started thinking if there was a relation between DC across coupling cap and transferring signal via such polarized cap. DC voltages in fixed and cathode biases across a coupling cap may be significantly different.What is an impact on sound?
Started a thread in PARTS.

[euro21]

Originally Posted by yagoolar

I am in doubt. Which one is ...

Is this the correct phrase? There's no accounting for taste. :-)