[Phil]

Patrick Turner wrote:


Phil wrote:


Patrick Turner wrote:

[snip]


Professor child called the mutual effect of anode AND grid voltages
in a triode a form of "self regulation".
See Radio Engineering, by Terman, 1937.

He does not call it NFB probably because it would have confused ppl at the time
who were used to NFB ONLY being some external networking to
get an amp of device to display lower Ra thd etc than otherwise it would
without NFB.

The observation of a triode as a mystery 3 terminal device could be
argued to contain NFB.

When you have a 300B with Ia = 70mA, Ea = +400V,
then if it acted like a resistor without NFB the R value
would be 400 / 0.07 = 5,700 ohms.
But where one maintains the grid bias voltage and fixes the cathode voltage,
abd simply raises the anode voltage by say 80V,
then one soon finds that the Ia change = 100mA, and
it is because the Ra = 800 ohms, and not the 5,700 ohms of quiescent Ea / Ia .
How can this be? Its because the anode has a massive effect on the
flow of electrons from the space charge around the cathode to anode despite
grid voltage's supposed control of current.

One simply cannot deny that a huge anode feedback effect exists,
and where you have a grid also able to control the Ia
with a gm of about 5mA/V then there MUST be a summing of the anode and grid voltage effects
according to some formula ( upon which Professor Child throws more light than I do ) .

So its very easy to see that NFB is alive and operational in a triode.

[snip]

I am not a qualified university educated triodologist.

But my powers of observation reveal NFB in triodes to me.

But by exactly what process in purely technical terms would you use to
allege NFB does NOT exist in triodes?

Patrick, you're simply being stubborn at this point. We have pointed
out, over and over, that a plate resistance, which, so far, you CANNOT
and HAVE NOT proven does not exist, can account for all of the "feedback
effects," by which you mean damping, that you talk about.


The Ea / Ia at the quiecent state = 5,700 ohms for the 300B I mentioned.
But why does it measure 800 ohms when small variations in Ea are made?

Because the *model* of a triode includes a voltage offset in series with
a diode, caused by the grid. Given, say, -10 Vg, you will get no
positive current until you reach a certain plate voltage, enough to
overcome the voltage offset caused by the grid. Specifically, the grid
can repel the electrons until the plate voltage more or less equals the
grid voltage times mu. You get no negative current because a triode only
conducts one way, like a diode.

Similarly, if you have two resistors in series, with a battery between
them, you get no positive current when placing a voltage after the
second resistor until its voltage equals the battery placed between the
two resistors. A triode acts like a resistor with a voltage offset which
is controlled by the grid.

if there was no FB and the tube was just a purely passive device like a resistor,
the changes in Ea would give Ra = 5,700 ohms.
but we get 800 ohms.
How come?

See above.

Yes, in the
absence of tests which possibly can distinguish between anode resistance
and triode feedback, we really don't know which it is. But to claim that
feedback must exist *because* effects equivalent to either feedback or
low plate resistance exist is unscientific. You have a choice, and that
is why the rest of us simply state that, in the absence of any logical,
intellectual, or engineering advantage for feedback, we prefer low plate
resistance as a *model*. That's our choice. Maybe the other choice is
better. But you have not given any analysis to my knowledge which proves
that such a choice does not exist, or even that a feedback model is a
better choice for audio or electronics.


You call me stubborn, and refuse to proove I'm wrong, and you cannot deal with the idea
of the mutual effect of the two changing voltages of grid and anode which BOTH
have large effects on the Ia and so who is right?
If say I am wrong, then so is Prfessor Child, and all the guys with
better minds than yours or mine.

I call you stubborn because you claim that feedback exists, when an
alternative model explains everything (admittedly, as far I know) just
as well. To say that it *might* be feedback would be okay, but until you
have proofs or experiments that rule out the resistor model, you cannot
say definitively that triodes have feedback. The plate may simply pull
electrons along in exactly the same manner as a battery pulls electrons
through an ordinary resistor. Then again, perhaps it would be *best*
described as feedback, and I can easily admit that. However, Henry has
some very convincing lines of reasoning that show no advantage or
greater accuracy or anything like that for feedback, I have not been
able to find lines of reasoning to contradict him, and again, as far as
I know, neither have you! I don't have to prove that triodes do not have
feedback to "prove you wrong," when all I am trying to claim is that you
are not necessarily right!

Let me see if I can make this clear. Let's assume, for a moment, that a
test will appear that truly can distinguish between low impedances
caused by feedback, and low impedances caused by a low impedance, a
simple resistance. I do not know for certain what that test would tell
us about triodes. As far as I can tell, neither do you. THAT appears to
me to be the honest answer! Since resistor models are less complicated
than feedback models, I understandably believe that everyone should use
the simpler model until it is proven to be a less accurate description,
since humans make fewer errors when using simpler, but equally powerful,
models. The MOMENT someone proves that feedback is the more accurate
model, I will state, publicly, that you were right, and that unless we
are working in some situation where there is no significant difference
between the two models, we should use feedback to describe triodes. But
not until then!

And honestly, Patrick, do you not agree that if triodes have feedback,
then so do resistors? Batteries affect the electric field surrounding
the atoms in resistors, and increasing the voltage increases the field
strength, just as it does in triodes and diodes, and if it weren't for
that, the resistance would be much higher, perhaps infinite. Mind you,
resistors might indeed have feedback, but until we find situations that
require us to view them in such a manner, why on Earth would we want to
replace the much simpler resistor model for resistors with a feedback
model? And the same question applies to triodes and diodes.

Where is you alternative modelling?

We have stated it over and over and over again -- well, we've stated the
resistor part, perhaps the battery part was not stated as clearly as it
should have been -- I don't know how many times we are supposed to
repeat ourselves.

I see there is NFB in a triode not because I am stubborn, but because
it hits me in the eye as being so very obvious, and its missing from
all other devices.

Yes, it appears to be feedback to you. It appears to be a resistance in
series with a voltage to everyone else. Who is right? Until we have
better proofs than anyone has presented, on either side, that "the other
side's model" cannot fully account for a triode's behavior, we do not
know which one is truly correct, and therefore should use the simpler
model, and that is the resistance model.

Phil

Patrick Turner.




Phil