"Dr. Slick" wrote in message
om...
"David Robbins" wrote in message
...
"Dr. Slick" wrote in message
m...
Cecil Moore wrote in message
...
Dr. Slick wrote:
If you agree that the Pref/Pfwd ratio cannot be greater than
1
for a passive network, then neither can the [Vref/Vfwd]= rho be
greater
than 1 either.

Sqrt(Pref/Pfwd) cannot be greater than one. (Z2-Z1)/(Z2+Z1) can be
greater than one. Both are defined as 'rho' but they are not
always equal. (Z2-Z1)/(Z2+Z1) is a physical reflection coefficient.
Sqrt(Pref/Pfwd) is an image reflection coefficient.


I agree that Sqrt(Pref/Pfwd) cannot be greater than one for a
passive network.

(Z2-Z1)/(Z2+Z1) can be greater than one, for passive networks
and certain combinations of complex Z1 and Z2. I feel this is
incorrect
usage of this formula, which should be limited to purely real Zo.
A [rho] that is greater than one gives meaningless negative SWR
data, and is limited to active devices.

it only gives negative swr values if you incorrectly use the lossless
line
approximation to calculate vswr from rho. that is the incorrectly
applied
formula in this case. that formula is not valid for a lossy line, you
must
go back to the original definition of VSWR=|Vmax|/|Vmin|. which as we
have
also noted is not meaningful on a lossy line as Vmax and Vmin are
different
at each max and min point because of the losses in the line affecting
both
the forward and reverse waves.


What if Z1 represents the impedance at the end of the line? Then
it doesn't matter what the losses are. Then you are trying to find
the
RC right at the meeting of the line and the load, so you don;t care
what the
losses of the line are.

Slick

but Vmax and Vmin don't occur at the same location on the line. when
measuring Vmax and Vmin you have to look 1/4 wave apart on the line... so by
definition it is a distributed measurement so losses do matter.