[John Byrns]

OK, unlike yesterday when I was not in the mood for ASCII art, today I
am happy to do an ASCII sketch of the triode equation used in my model.

Please remember that you must use a fixed pitch font in your news reader
to properly view this sketch.

Here is the equation for the triode model as I first posted it several
days ago.

Ipk = c * (u * Vgk + Vpk) ^ 1.5

For the purposes of this sketch I am going to replace the constant "c"
with a new constant "k" which is related to "c" by the following
function.

c = k / (u ^ 1.5)

Making this substitution we get the following equation which I will
sketch.

Ipk = k * (Vgk + Vpk / u) ^ 1.5

OK, finally here is the sketch showing how the negative feedback works.


__________
| |
| divide |
+----------------| by |---------------------+
| | u | |
| |__________| |
| |
| |\ |
| |\ | \ |
| | \ __________ | \ |
+--| \ | | | \ |
| \ | 3/2 | | v-c \ | Output
| sum |---| power |---| transcon- |---+-------
Input | / | function | | ductance / (plate)
--------| / |__________| | (k) /
(grid) | / | /
|/ | /
|/


I hope this sketch makes the negative feedback in the equation, and the
spice model derived from it, clear to all. This does not imply one way
or the other whether or not there is actually negative feedback involved
in the operation of a triode vacuum tube.

To properly understand this sketch it is necessary to understand that
the transconductance block is connected such that a positive input to
the transconductance block causes positive current to flow from the
output of the model into the transconductance block. This may be easier
to understand if you consider the value of the transconductance to be -k
rather than k.

I apologize in advance if this sketch doesn't reproduce properly on your
computer even when using a fixed pitch font.


Regards,

John Byrns

[John Byrns]

In article . com,
"Bret Ludwig" wrote:

John Byrns wrote:
OK, unlike yesterday when I was not in the mood for ASCII art, today I
am happy to do an ASCII sketch of the triode equation used in my model.


I'm not trying to be prickish, but ASCII art is a pain in the ass to
read. Consider starting a web site and making drawings with CAD if you
wish to communicate your ideas with the best clarity.

You are obviously a member of the Powerpoint generation.


Regards,

John Byrns

[Patrick Turner]

Bret Ludwig wrote:

John Byrns wrote:
OK, unlike yesterday when I was not in the mood for ASCII art, today I
am happy to do an ASCII sketch of the triode equation used in my model.


I'm not trying to be prickish, but ASCII art is a pain in the ass to
read. Consider starting a web site and making drawings with CAD if you
wish to communicate your ideas with the best clarity.

It looked OK on my screen.

It allows me to better guess what is is in the picture such as
the signage of the triangular amp representations.

One should be able to see at a glance what is completely unambiguous with
all wave forms so that the level of bickering about WTF was meant is reduced
to zero.

If someone presents an irrefutably good drawing, I'll post it at my website
where i have a spare 180MB of space.

Patrick Turner.

[Ian Iveson]

Bret wrote:

I'm not trying to be prickish, but ASCII art is a pain in the ass to
read. Consider starting a web site and making drawings with CAD if
you
wish to communicate your ideas with the best clarity.

It wouldn't be worth it in this case. I am sure you can imagine your
own diagram by glancing at the equation.

It simply compounds John's trivialisation of the notion of feedback.

Since when has it made sense to draw a diagram of an equation? What is
it supposed to mean? What does it add to understanding? Nothing at all
in this case.

A similar diagram could be drawn of an equation describing Henry's
example of a falling parachute. Or pretty much any real process.

OTOH, the SPICE code from which this fragment comes is a circuit
listing: what SPICE calls a subcircuit model. Perhaps it would be
interesting if John were to post the circuit, so it can be compared to
the diagram? If it is the same, then the diagram has a little bit of
meaning at least, however divorced from reality.

But....

cheers, Ian

[Patrick Turner]

Ian Iveson wrote:

Bret wrote:

I'm not trying to be prickish, but ASCII art is a pain in the ass to
read. Consider starting a web site and making drawings with CAD if
you
wish to communicate your ideas with the best clarity.

It wouldn't be worth it in this case. I am sure you can imagine your
own diagram by glancing at the equation.

It simply compounds John's trivialisation of the notion of feedback.

Since when has it made sense to draw a diagram of an equation? What is
it supposed to mean? What does it add to understanding? Nothing at all
in this case.

A good picture speaks one thousand words of wisdom.

But a thousand of your words could only be described with scribbles.

Patrick Turner.


A similar diagram could be drawn of an equation describing Henry's
example of a falling parachute. Or pretty much any real process.

OTOH, the SPICE code from which this fragment comes is a circuit
listing: what SPICE calls a subcircuit model. Perhaps it would be
interesting if John were to post the circuit, so it can be compared to
the diagram? If it is the same, then the diagram has a little bit of
meaning at least, however divorced from reality.

But....

cheers, Ian

[Ian Iveson]

Patrick Turner wrote

Bret wrote:

I'm not trying to be prickish, but ASCII art is a pain in the ass
to
read. Consider starting a web site and making drawings with CAD
if
you
wish to communicate your ideas with the best clarity.

It wouldn't be worth it in this case. I am sure you can imagine
your
own diagram by glancing at the equation.

It simply compounds John's trivialisation of the notion of
feedback.

Since when has it made sense to draw a diagram of an equation? What
is
it supposed to mean? What does it add to understanding? Nothing at
all
in this case.

A good picture speaks one thousand words of wisdom.


Whereas this diagram, which is not good, complicates a single simple
equation.

In what way do you find it helpful?

Does it help you to calculate the result of the equation? Start with
an input, compute an output, compute the resulting input, and repeat
until you arrive at a stable solution?

Unless you naturally, like computers and hence SPICE, use numerical
computation rather than analytical methods, that would be the long way
round. It's easier to work from the equation after doing a spot of
algebra to include the load.

Systems theory, including feedback analysis, is not necessary in this
case. Unlike with real feedback.

Clearly it has nothing to do with how a real triode works. A real
triode only has one path between anode and grid, or from any electrode
to any other, for that matter. Hence there cannot be a separate
feedback path, and definitely not with a different function, like John
has put in his diagram of Duncan Munro's model.

It's a pity I can no longer give a link, BTW, to the documentation for
the much-improved, later version of the model John took his equation
from. Unfortunately it is no longer available from the author, and I
don't want to breach copyright in this case.

I can tell you that the diagram is rather different.

If you want a good diagram to put on your site, perhaps you could ask
Duncan Munro if he doesn't mind you making commercial use of his work?
But then you don't bother about copyright do you?

But a thousand of your words could only be described with scribbles.

Yep. Like those Japanese drawings. Every word a work of art.

Er...or are they Chinese?

Ian


A similar diagram could be drawn of an equation describing Henry's
example of a falling parachute. Or pretty much any real process.

OTOH, the SPICE code from which this fragment comes is a circuit
listing: what SPICE calls a subcircuit model. Perhaps it would be
interesting if John were to post the circuit, so it can be compared
to
the diagram? If it is the same, then the diagram has a little bit
of
meaning at least, however divorced from reality.

But....

cheers, Ian

[Lost'n Found]

If the ASCII is not clear, do this:

Copy the diagram and paste in notepad.

Should be very very clear to understand
"John Byrns" wrote in message
...

OK, unlike yesterday when I was not in the mood for ASCII art, today I
am happy to do an ASCII sketch of the triode equation used in my model.

Please remember that you must use a fixed pitch font in your news reader
to properly view this sketch.

Here is the equation for the triode model as I first posted it several
days ago.

Ipk = c * (u * Vgk + Vpk) ^ 1.5

For the purposes of this sketch I am going to replace the constant "c"
with a new constant "k" which is related to "c" by the following
function.

c = k / (u ^ 1.5)

Making this substitution we get the following equation which I will
sketch.

Ipk = k * (Vgk + Vpk / u) ^ 1.5

OK, finally here is the sketch showing how the negative feedback works.


__________
| |
| divide |
+----------------| by |---------------------+
| | u | |
| |__________| |
| |
| |\ |
| |\ | \ |
| | \ __________ | \ |
+--| \ | | | \ |
| \ | 3/2 | | v-c \ | Output
| sum |---| power |---| transcon- |---+-------
Input | / | function | | ductance / (plate)
--------| / |__________| | (k) /
(grid) | / | /
|/ | /
|/


I hope this sketch makes the negative feedback in the equation, and the
spice model derived from it, clear to all. This does not imply one way
or the other whether or not there is actually negative feedback involved
in the operation of a triode vacuum tube.

To properly understand this sketch it is necessary to understand that
the transconductance block is connected such that a positive input to
the transconductance block causes positive current to flow from the
output of the model into the transconductance block. This may be easier
to understand if you consider the value of the transconductance to be -k
rather than k.

I apologize in advance if this sketch doesn't reproduce properly on your
computer even when using a fixed pitch font.


Regards,

John Byrns