Walter Maxwell wrote:
Cecil, you keep saying Steve and I get the same outcome. Where do you see that
result ?

You both get the same measurable forward and reflected powers,
forward and reflected voltages, and forward and reflected currents.

Why do you keep insisting on an S-parameter analysis?

Because that is essentially what Steve used and it works. It has
worked for decades. Please download the HP AN 95-1 and see for yourself.

When
properly adjusted to match the output to the input in his T-network the
reflected power reaching the input is totally re-reflected. This results only
from a reflection coefficient of 1.0.

Only in your model, Walt, not in Steve's. Reflection coefficients of 1.0
are INVALID in Steve's quasi-S-parameter model.

Saying that it is impossible for a 1.0 to
exist in Steve's model is simply not true.

Yes, it is true, for the configurations discussed so far. The ONLY time a
reflection coefficient of 1.0 exists in Steve's model is for a *physical*
short, a *physical* open, or a *physical* pure reactance. In Steve's model,
virtual stuff is invalid.

It's true that the physical reflection coefficient is 0.5. How then do you
account for ALL the reflected energy being re-reflected to the load? The fact
is that a reflection coefficient of 1.0 is also established there by wave
interference. You really must come to the grips with the fact that a reflection
coefficient of 1.0 can be established by wave interference. You are now the one
who won't budge.

You are asking me to budge away from the rules of an S-parameter analysis?

Sorry, Walt, virtual reflection coefficients are INVALID in an S-parameter
analysis like Steve is using. The s11 reflection coefficient that Steve uses
is defined as the "input reflection coefficient looking into port 1 with
the output port terminated by a matched load." Since there is no such thing
as a Z0 = 0 or a Z0 = infinity, there is no such thing as a reflection
coefficient equal to 1.0. The reflection coefficient, s11, that Steve is
using can *NEVER* be zero or one - *NOT EVER*.

I haven't seen anything in Steve's paper that shows he's using an S-Parameter,
can you show me where? Not that it would make any difference in the outcome.

His reflection coefficient is (Z2-Z1)/(Z2+Z1). That is identical to the S-parameter
reflection coefficient, s11. It is a constant and never changes to 1.0, no matter
what happens to the reflected power. Dr. Best simply doesn't use virtual
reflection coefficients. Using virtual reflection coefficients in an S-parameter
analysis is *INVALID*, i.e. they simply do not exist.

In addition, referring to your paragraph immediately above, where did you get
the idea that I said Steve's rho = SQRT(Pr/Pf)? Cecil, that's simply SWR, not
rho.

Let's take the 133.33W forward and 33.33W reflected example. rho equals the square
root of (33.33/133.33) = 0.5. That's not SWR, Walt. SWR cannot be less than unity.
That is indeed the physical voltage reflection coefficient.

At this point, Cecil, if you are still unable to accept the concept of
establishing a reflection coefficient of 1.0 through wave interference then
there is no use of continuing this discussion.

I accept the concept of a reflection coefficient of 1.0 for your model, Walt.
But a reflection coefficient of 1.0 is simply NOT allowed in the quasi-S-parameter
analysis that Steve uses. The S-parameter analysis works and has been used for
decades but an s11 of 1.0 simply never happens (except at a short or an open).
The reflection coefficients in an S-parameter analysis and in Steve's analysis
are ***PHYSICAL***. They are ***NEVER*** virtual. I'm sorry if that upsets you.
I don't know what else to say.
--
73, Cecil http://www.qsl.net/w5dxp



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----