[Patrick Turner]

I had a nice cycle of about 50km this morning before breakfast.

Blue skies, not many idiots on the roads, and a body that worked without

too many aches and pains.

Then I turn on my computer to see what is going on in the wider world,
and all I seem to want to do is turn it off again, and stay busy
elsewhere, because
after I laid down the challenge for someone to
PROOVE THERE AIN'T NFB IN A TRIODE,
I see not one witty and convincing tome on the subject.

I live in hope.


A fabulous live concert awaits me this evening warranting the wearing of
a suit,
and a good night out beckons.

Have a good weekend all.

Patrick Turner.

[Chris Hornbeck]

On Sat, 19 Aug 2006 04:03:37 GMT, Patrick Turner
wrote:

PROOVE THERE AIN'T NFB IN A TRIODE,

*Feedback* has a specific meaning, and includes several
easily testable results.

Voltage or current from the output
terminal appears at the input terminal.

As a result of this, transfer functions change in
defined proportions, input terminal impedance changes,
noise and distortion referred to the input change,
closed loop bandwidth changes, etc.

None of these things occur in your analogy, because
it's just that, an analogy. It's not a real model
or a real (testable) theory. A real theory is
disprovable by testing. Easy, in this case.

Primary effect: Does your model show any
observable voltage or current from the output
appearing and measurable at the input? (Not counting
Black Holes, Virtual Grids or poltergeists.)

Secondary effect: Does it show any reduction in noise
or increase in bandwidth resulting from *feedback* and
not from local circuit impedances and device characteristics?

The reason why this idea exists only on this particular
newsgroup is twofold. First, it would be embarrassing
in a technical setting. Second, there's apparently nobody
left here who gives a ****. I'm beginning to understand
why.

Good Night and Good Luck,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau

[Patrick Turner]

Chris Hornbeck wrote:

On Sat, 19 Aug 2006 04:03:37 GMT, Patrick Turner
wrote:

PROOVE THERE AIN'T NFB IN A TRIODE,

*Feedback* has a specific meaning, and includes several
easily testable results.

Voltage or current from the output
terminal appears at the input terminal.

These criteria are not necessary for a NFB network where the combined
field effect
of input grid voltage and output anode voltage have a mutual efect on
the output.


As a result of this, transfer functions change in
defined proportions, input terminal impedance changes,
noise and distortion referred to the input change,
closed loop bandwidth changes, etc.

As a result of the electrostatic NFB network within the triode,
the input terminal impedance remains high with a triode because
there is such a miniscule change in input current.




None of these things occur in your analogy, because
it's just that, an analogy. It's not a real model
or a real (testable) theory. A real theory is
disprovable by testing. Easy, in this case.

But unless stated other wise, you will pick tests to give the outcome
you want.



Primary effect: Does your model show any
observable voltage or current from the output
appearing and measurable at the input? (Not counting
Black Holes, Virtual Grids or poltergeists.)

As stated above, the triode input impedance remains high at all times
regardless of the NFB because the FB is a charge effect, and not a
current effect,
or like a typical shunt FB path around an opamp for example
where one input port is grounded, the other live input is at the
junction of
two resistors where the net effect of output and input voltages
trying to shunt each other is applied to the opamp.




Secondary effect: Does it show any reduction in noise
or increase in bandwidth resulting from *feedback* and
not from local circuit impedances and device characteristics?

If the NFB acting as a charge effect from the anode is interupted by
placing a screen in the
triode between the grid and anode then the Ra, gain, and noise are all
greater than a triode
and bandwidth reduced due to the action of stray output C shunting the
signal from the high Ra
and perhaps also any L component in the load shunting the LF.




The reason why this idea exists only on this particular
newsgroup is twofold. First, it would be embarrassing
in a technical setting. Second, there's apparently nobody
left here who gives a ****. I'm beginning to understand
why.

I think you don't understand as much as you need to.

There are some people here who do give a ****.

You have not proved that there is no NFB in a triode.

Patrick Turner.





Good Night and Good Luck,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau