[jqp]

Could someone explain how magnets work?   

[G Georgopoulos]

distortion in tubes has to do with the transfer charactiristic and negative feedback,the transfer charactiristic is somewhat curved especially in high gain this makes the most of tube distortion,negative feedback linearises this to some extend,negative feedback can be local or global,local feedback is say the voltage developed on the cathode self biasing resistor,many people who like tube distortion drive the tube hard by bypassing this resistor with a capacitor,then the tube operates with the most curved tranfer charactiristic and high  distortion,if you want less distortion and less transfer charactiristic you dont bypass the rsistor thus local feedback,global feedback goes one step further in reducing distortion...

[Steve]

JJ has developed some small signal tubes that have exceptionally low distortion, some -76 to -80db down with only small amounts ofGNF.
Nelson Pass has an interesting article, with measurements, explaining higher orders of HD with moderate to fairly high levels of global negative feedback (gnf). Here is a link.

https://www.passdiy.com/pdf/distortion_feedback.pdf

Cheers
Steve

[FullRangeMan]

I disagree in general terms.
The impedance's of a tube, do not have a direct relationship to colour! of its sound.
If designed properly, it should not add any colour (= distortion).
Colours and textures in tubeamps are in guitar amps, where these distortions are deliberately allowed to make that distinct guitar sound.
in HiFi we do't want any colours.

I did not say it was, perhaps not have to do.
Because at both ends there are valves with good sound.

[FullRangeMan]

Could someone explain how magnets work?   

With magnetism, this force came from some place (from the Free Energy).

[G Georgopoulos]

back to learning tubes...

a triode is usually biased at half the power supply rail voltage
for example 300v rail 150v across plate to cathode
250v rail 125v across plate to cathode
this allows for maximum positive going and negative going
swing
it also makes easy power dissipation calculations of the tube
the power dissipation then of the tube is
Ibias x Vplate to cathode =Pdissipation
for example
2mA x 150v = 0.3watts
2mA x 125v = 0.250watts
to understand power dissipation of triodes
i will describe what happens with operation
when the plate voltage increaces the current through the tube decreaces
you can check this with ohms law
what it means is there is no increase in Pdiss because even though
the plate voltage has increased the current has decreased
now next
when the plate voltage has decreased and the current increased
again using ohms law
means the Pdiss hasnt decreased because the current has  increased
so the Pdiss of
the triode is always Ibias x 150v or Ibias x 125v depending on
power supply rail voltage you choose as Vbias plate to cathode

you're close to learning everything about triode amplifiers...

[Steve]

back to learning tubes...

a triode is usually biased at half the power supply rail voltage
for example 300v rail 150v across plate to cathode
250v rail 125v across plate to cathode
this allows for maximum positive going and negative going
swing
it also makes easy power dissipation calculations of the tube
the power dissipation then of the tube is
Ibias x Vplate to cathode =Pdissipation
for example
2mA x 150v = 0.3watts
2mA x 125v = 0.250watts
to understand power dissipation of triodes
i will describe what happens with operation
when the plate voltage increaces the current through the tube decreaces
you can check this with ohms law
what it means is there is no increase in Pdiss because even though
the plate voltage has increased the current has decreased
now next
when the plate voltage has decreased and the current increased
again using ohms law
means the Pdiss hasnt decreased because the current has  increased
so the Pdiss of
the triode is always Ibias x 150v or Ibias x 125v depending on
power supply rail voltage you choose as Vbias plate to cathode

you're close to learning everything about triode amplifiers...

I hate to disagree, but when the plate voltage increases, with either cathode bias or
fixed bias, the plate current and plate dissipation increases as well. 
When the plate voltage decreases, the above is opposite.

I think what you are referring to, George, is when fixed bias increases, the plate voltage increases while
the plate current decreases, thus plate dissipation remains basically static. If the fixed bias decreases,
then the plate voltage decreases while the plate current increases. This assumes the plate resistor
is of sufficient value so as the tube is basically in the linear region of the plate curves.

It can get somewhat confusing at times.

Cheers
Steve

[G Georgopoulos]

 a negative going voltage at the grid reduces tube current,and increases tube impedance, this increases plate voltage and reduces the voltage drop on the plate resistor...

[Steve]

a negative going voltage at the grid reduces tube current,and increases tube impedance, this increases plate voltage and reduces the voltage drop on the plate resistor...

Good point George, I need to clarify as most are not aquainted with tubes. When I state the grid voltage increases in reply #26, I mean it increases in negative value compared to the cathode.

A little more information concerning the common term plate resistance.

The plate resistance of a single tube creates an interesting phenomenom, it is non-linear vs signal level.

As you indicate, as the grid becomes more negative with respect to the cathode, the plate resistance increases, until at tube/plate cut off the plate resistance becomes infinite.

As the grid becomes more positive with respect to the cathode, the plate plate voltage drops, plate current increases, but the plate resistance decreases very little in comparison. This non-linearity in plate impedance over the positive and negative signal cycle causes the damping factor to vary (if driving a speaker). This non-linearity is in relation to the signal level when used as a single tube. As PP, the damping factor remains more constant over the signal cycle.

This subject can be covered in more depth, but it is not necessary for the average reader.

Cheers
Steve

[G Georgopoulos]

I hate to disagree, but when the plate voltage increases, with either cathode bias or
fixed bias, the plate current and plate dissipation increases as well. 
When the plate voltage decreases, the above is opposite.

I think what you are referring to, George, is when fixed bias increases, the plate voltage increases while
the plate current decreases, thus plate dissipation remains basically static. If the fixed bias decreases,
then the plate voltage decreases while the plate current increases. This assumes the plate resistor
is of sufficient value so as the tube is basically in the linear region of the plate curves.

It can get somewhat confusing at times.

Cheers
Steve

i think you confuse this with a pentode output stage...

[Steve]

i think you confuse this with a pentode output stage...

I suggest to viewers to check out and study some books for more information. Here are just a few to check out. My below explanation
is not exhaustive.

1) Semiconductor and Tube Electronics by James G Brazee (My engineering textbook.)
2) RCA Radiotron Designers Handbook
3) Radio Amateurs Handbook
4) RCA receiving tube manual

A pentode operates quite differently than a triode. In the pentode's case, raising or lowering the plate voltage
has little effect on the the plate current, assuming the negative DC bias remains constant, because of the screens electrostatic shielding effect.
Since the plate voltage raises, the plate current rises only slightly, the plate dissipation increases.

The screen grid has the same/similar effect on the plate current as the plate does in a pentode. In fact, back in the "olden"
days, the screen grid was sometimes used as the plate, while the plate was grounded to form a shield, used in the low signal arena such
as phono stages etc to lessen hum.

Below is what is called the plate family of curves, a graph, for the 417A/5842 triode. Let's examine it.


Notice in the graph the negative DC bias "lines". If one follows a line, let's say -1 volt DC bias "line", as the plate voltage
rises, say from 100 to 150 volts, the plate current rises, thus the plate dissipation rises. The same can be said for any of the DC bias "lines".
Notice as each line approaches zero plate current, the curvature increases even more. The plate resistance is increasing until cutoff, zero
current, in which the plate resistance is infinite.

Now if we had a perfect triode, in the case that the negative DC grid voltage rises, the plate voltage rises while the plate current decreases, assuming
we have a plate resistor/plate resistance or plate reactance such as an output transformer. This is a simple explanation.

This is how amplification works. If we have a negative DC "bias" voltage on the grid, and feed a signal onto that grid, modulating the grid, then we will vary the bias voltage from its DC value, and the plate current and plate voltage will vary. The gain we experience is determined by the gain of the tube under design conditions. So we could have a gain of any number, generally until we reach about 50 or so. Less is much more common though. By the way, the "amplification factor" listed in the tube specs is not the actual gain of the circuit, but the Mu. Mu is defined as "minus" plate voltage change divided by grid voltage applied/change, with constant plate current. (The "minus" sign simply means the signal is inverted.)

Cheers
Steve