On Thu, 28 Aug 2003 12:18:44 -0500, W5DXP
wrote:

My basic approach is to achieve a Z0-match and therefore
forget about source impedance.

Hi Cecil,

This is a cavalier attitude if you can afford it. Otherwise, those
who so desperately hammer out the last 0.1 dB antenna gain are going
to fall to their knees in wrack when they discover that their rig's
characteristic Z of, say, 70 Ohms meeting the discontinuity of their
low pass filter's 50 Ohms turns that effort into heat behind the
antenna jack.

I have long since stopped being surprised by those who spin on like
whirling dervishes over trivial matters in the face of 10 fold losses
in front of them. This, of course, is even more trivial when they
gush on about their premium equipment that behind the knobs
"efficiently" transforms 20 - 25 Amperes of DC current into 100 Watts
RF. Now, that puts perspective to the topic: smoke and reflection
coefficient.

73's
Richard Clark, KB7QHC

(Dr. Slick) wrote in message . com...
"Tarmo Tammaru" wrote in message ...

According to Adler, Chu, and Fano, "Electromagnetic Energy Transmission and
Radiatin", John Wiley, 1960, (60-10305),
when they talk about lossy lines, and say that Zo is complex in the general
case, they come up with a maximum value for the reflection coefficient of (1
+ SQRT(2)). Eq 5.14b. Remember, it is a lossy line; so, the reflected
voltage gets smaller as you move away from the load. Somebody might want to
check this out, in case I misunderstood something. BTW, the three authors
were all MIT profs.

Tam/WB2TT


Les Besser, Pozar, and Kurokawa are all PhDs too.

But a PhD is only human too.

Could you send me that paper?

Garvin, it's a book, not a paper. Your bookstore can probably help
you with it. Besser, at least, apparently came to understand the
error of his ways on this point, and Besser Associates is now
presenting the correct formula in their classes, per the email I
posted from them. I haven't checked with Pozar or Kurokawa, but
strongly suspect you are taking their works out of context and
misinterpreting them.

Cheers,
Tom

Richard Clark wrote:
But, again, this discussion is generally reserved only for those
interested in accuracy. :-)

Like I say, my solution is to block any reflections from being
incident upon the source. But I have a question. Since we are
discussing coherent sine waves, it seems to me that any reflection
from the source impedance will become indistinguishable from the
generated wave. In fact, the present convention of generated power
equals forward power minus reflected power is designed to overcome
that very problem.
--
73, Cecil, W5DXP

Richard Clark wrote:

W5DXP wrote:
My basic approach is to achieve a Z0-match and therefore
forget about source impedance.

This is a cavalier attitude if you can afford it.

It's all part of my "Work Smarter, Not Harder" nature. The elimination
of reflected energy incident upon the source is extremely rewarding
in multiple ways.
--
73, Cecil, W5DXP

Tom, to save everybody a lot of trouble -

The greatest theoretical value of the magnitude of the
reflection coefficient occurs when the angle of Zo is
-45 degrees, and the terminating impedance is a pure
inductive reactance of |Zo| ohms.

Do you think I should have mentioned this when I
began this and other threads by saying a reflection
coefficient greater than unity can occur?

The riot police can now return to barracks.
----
Reg, G4FGQ.

====================================
---
"Tom Bruhns" wrote
"Reg Edwards" wrote
By the way, you've told us only half the story.
What's the value of the
load impedance which maximises the reflection
coefficient?
====================================
Hey, Reg, it's just a simple high-school (well,
maybe first-year
college) differential calculus problem. Just let
Garvin work through
it for us. Hey, good Dr., could you do that for
us? Just write an
expression for |Vr/Vf| = |(Zl-Zo)/(Zl+Zo)| in terms
of Rl and Xl and
find the partial derivatives with respect to those
two variables, and
set both equal to zero, while letting Ro=Xo. It's
mostly just a bunch
of bookkeeping. You should come up with values of
Rl and Zl in terms
of Ro, and you can check to be sure that's actually
a maximum and not
a minimum or saddle point. You should see a
symmetry for Ro=-Xo, the
more usual limiting case.

(Of course, that's not quite right, as I'm sure the
good Dr. and Reg
both know. Since we're talking passive here, you
need to insure that
Rl stays positive, so you just may need to check
along the boundary
where Rl=0. And you should convince yourself that
the most reactive
possible line really does yield the largest possible
|Vr/Vf|. So it
becomes a task of finding the maximum value of a
function f(Rl, Xl,
Xo) with Ro fixed positive non-zero, under the
constraints that Rl=0
and |Xo|=Ro.)

On Thu, 28 Aug 2003 12:33:53 -0700, W5DXP
wrote:

Richard Clark wrote:
But, again, this discussion is generally reserved only for those
interested in accuracy. :-)

Like I say, my solution is to block any reflections from being
incident upon the source. But I have a question. Since we are
discussing coherent sine waves, it seems to me that any reflection
from the source impedance will become indistinguishable from the
generated wave. In fact, the present convention of generated power
equals forward power minus reflected power is designed to overcome
that very problem.

Hi Cecil,

So you DO want to perform this test?

Your presumption of coherency is false unless you engineer the
solution.

I got there first and made sure that wasn't gonna happen. :-)

Any random attempt has only a one in 360 chance of being correct
within one degree of coherent. This is simple interference math after
all. Most individuals would just notice a 10 degree error which would
boost your chances to slightly less than 3% - not very good coherency
wise.

73's
Richard Clark, KB7QHC

On Thu, 28 Aug 2003 12:39:25 -0700, W5DXP
wrote:

Richard Clark wrote:

W5DXP wrote:
My basic approach is to achieve a Z0-match and therefore
forget about source impedance.

This is a cavalier attitude if you can afford it.

It's all part of my "Work Smarter, Not Harder" nature. The elimination
of reflected energy incident upon the source is extremely rewarding
in multiple ways.

Hi Cecil,

If smarter were hotter, then you could toast bread at 10 feet.
Casting back ten watts by burning 20 hardly qualifies for more.

73's
Richard Clark, KB7QHC

(Tom Bruhns) wrote in message om...

Hey, Reg, it's just a simple high-school (well, maybe first-year
college) differential calculus problem. Just let Garvin work through
it for us. Hey, good Dr., could you do that for us? Just write an
expression for |Vr/Vf| = |(Zl-Zo)/(Zl+Zo)| in terms of Rl and Xl and
find the partial derivatives with respect to those two variables, and
set both equal to zero, while letting Ro=Xo. It's mostly just a bunch
of bookkeeping. You should come up with values of Rl and Zl in terms
of Ro, and you can check to be sure that's actually a maximum and not
a minimum or saddle point. You should see a symmetry for Ro=-Xo, the
more usual limiting case.

Cheers,
Tom


Well, Tom, as your education is "not worth mentioning", i don't
feel a need to impress or bore you with a lot of math. And a lot of
math is not needed in this case, although it would be interesting to
see the derivation of the conjugate equation.

Maybe you should play some more chess...


Slick

(Tom Bruhns) wrote in message om...

Garvin, it's a book, not a paper. Your bookstore can probably help
you with it. Besser, at least, apparently came to understand the
error of his ways on this point, and Besser Associates is now
presenting the correct formula in their classes, per the email I
posted from them. I haven't checked with Pozar or Kurokawa, but
strongly suspect you are taking their works out of context and
misinterpreting them.

Cheers,
Tom


You are full of sh**. You never emailed Besser, and they are
teaching it correctly.

What about the ARRL? Check Pozar and Kurokawa yourself, don't
take my word for it.

Lemme guess, you are a lousy chess player, aren't you?

Sounds like you lost one too many games...


Garvin (Slick)

"George, W5YR" wrote in message ...


Finally, he clearly shows how terminating an actual physical line
appropriately can result in a reflection coefficient as large as 2.41.

This revelation DOES NOT imply that the reflected wave would bear more power
than the incident wave. For a line to display this behavior, it must first
of all have a high attenuation per wavelength. Due to this high attenuation,
the power in the reflected wave is high for only a short distance from the
termination.


George, with all due respect, even if the SWR measurement was
done right at a short or open, the highest rho you could get would be
1.

If the power reflection coefficient is the square of the
MAGNITUDE of the voltage reflection coefficient, how can you have a
voltage RC greater than one without the power RC being also greater
than one??


Slick