W5DXP wrote:
wrote:
We have a choice of two rho for this situation:
Correction: We have a choice of two reflection coefficients
each with its own unique definition.
black box - 0, computed from the surge impedance of the line and
the steady state impedance of the load
Actually, Sqrt(Pref/Pfwd), the definition of rho.
This definition seems incomplete. There is a choice of sign. How
do you pick?
open box - 0.5, computed from the surge impedance of the line and
the surge impedance of the load
Actually, (150-50)/(150+50), the definition of s11.
In any case, what we have in this experiment is a case where there
IS an impedance discontinuity and yet there is no reflection (if
you use the "black box" rho, as is often done).
This is technically not true.
None-the-less, no one seems to have difficulty treating it as if it is.
The NET reflections are zero. There
are two non-zero component reflections as seen from the s-parameter
equation:
b1 = s11*a1 + s12*a2
These three terms are all reflections. b1 is the NET reflections
toward the source. Since b1 = zero, s11*a1 = -s12*a2, i.e. the two
component reflections are of equal magnitude and opposite phase and
therefore cancel. This is explained in the last three paragraphs on
the Melles-Griot web page.
Yes, but only if the box is open. How do you analyze the black box
when you are only permitted to know the impedance at the inteface
looking
towards the load? rho is just computed from the only information
available. No one complains that the problem can't be solved.
So, if we are allowed to say in the first experiment that rho is 0
despite an impedance discontinuity, we are equally allowed to say
for the second that rho is -1 despite the absence of a discontinuity.
There is NOT an absence of a discontinuity.
There is no discontinuity at the interface in question.
There is NO physical discontinuity at the black box.
Exactly, and people argue that it is inappropriate to claim that a
reflection occurs at this interface.
And yet, in the symmetrical case where there IS a discontinuity,
many are quite comfortable talking about the lack of reflections
at the inteface.
So, if you don't wish to permit reflections at interfaces without
an impedance discontinuity, please never speak of an absence
of reflections at an interface with a discontinuity.
If it's good for the goose, it should be good for the gander.
....Keith