Keith Dysart wrote:


It takes many more reflections of reflections before the
impedance at the input starts to look like a short.

An excellent point, Keith. And applications for it abound.

Problem is that it conflicts with the predictions made by the power
propagation model. (The politically correct name is the Joules/sec
propagation model.) That model shows that only a single reflection of
power is needed to explain the whole thing. Of course in some cases
the wave of power has to figure out how to turn around and go back the
other direction after it's been cancelled in order to conserve energy.
(A problem it wouldn't have to solve had it not violated it in the
first place.) Admittedly, some of the details have yet to be worked
out. :-)

73, ac6xg

Jim Kelley wrote:
The question was excellent. 'Impedance discontinuity points' is nonsense.

It is common practice during an analysis to consider
the impedance discontinuity where Z0 changes to be point
connections at a plane. All of my technical books draw
the connections that way. None of them attempt to
treat the connection as anything other than points.
If you choose to analyze what happens at a 1 mm drop
of solder, be my guest.
--
73, Cecil http://www.w5dxp.com

On Mon, 10 Dec 2007 21:04:06 GMT, Owen Duffy wrote:

I can't quickly think of a sampling technique that truly takes a
point sample.

Slotted line.

73's
Richard Clark, KB7QHC

Richard Clark wrote:
On Mon, 10 Dec 2007 21:04:06 GMT, Owen Duffy wrote:


I can't quickly think of a sampling technique that truly takes a
point sample.


Slotted line.

Doesn't take current sample

Probe is of finite size (albeit small fraction of lambda in most cases)

The difficulty of making accurate RF current measurements at a point is
why things like slotted lines were invented. You can measure voltage at
a series of points and calculate what the current *must* be.

Richard Clark wrote in
:

On Mon, 10 Dec 2007 21:04:06 GMT, Owen Duffy wrote:

I can't quickly think of a sampling technique that truly takes a
point sample.

Slotted line.

73's
Richard Clark, KB7QHC


Thanks Richard and Jim,

I mean't literally a point sample, and I doubt that we can manufacture a
probe for either voltage or current that has zero dimensions (ie is
influenced ONLY by conditions at a point of no physical size.).

I go on to say (if you read on) in different words that notwithstanding
that, practical probes can be made that give acceptable accuracy.

They unsaid key thing is that these probes do not depend on averaging
over a length of line (as Roger suggested), that is actually a defect
that limits their usable upper frequency.

Owen

Jim Kelley wrote:
Problem is that it conflicts with the predictions made by the power
propagation model. (The politically correct name is the Joules/sec
propagation model.) That model shows that only a single reflection of
power is needed to explain the whole thing. Of course in some cases the
wave of power has to figure out how to turn around and go back the other
direction after it's been cancelled in order to conserve energy. (A
problem it wouldn't have to solve had it not violated it in the first
place.) Admittedly, some of the details have yet to be worked out. :-)

Wow Jim, you need to repeat Fields and Waves 310. :-)
You have misunderstood virtually every principle
of the wave reflection model.
--
73, Cecil http://www.w5dxp.com

Cecil Moore wrote:
Jim Kelley wrote:

Problem is that it conflicts with the predictions made by the power
propagation model. (The politically correct name is the Joules/sec
propagation model.) That model shows that only a single reflection of
power is needed to explain the whole thing. Of course in some cases
the wave of power has to figure out how to turn around and go back the
other direction after it's been cancelled in order to conserve energy.
(A problem it wouldn't have to solve had it not violated it in the
first place.) Admittedly, some of the details have yet to be worked
out. :-)


Wow Jim, you need to repeat Fields and Waves 310. :-)
You have misunderstood virtually every principle
of the wave reflection model.

I wouldn't presume to take credit for any of the above. I learned it
on r.r.a.a. from someone who I think needs to take Fields and Waves 1.
:-)

73, ac6xg

On Mon, 10 Dec 2007 16:16:15 -0800, Jim Lux
wrote:

Richard Clark wrote:
On Mon, 10 Dec 2007 21:04:06 GMT, Owen Duffy wrote:


I can't quickly think of a sampling technique that truly takes a
point sample.


Slotted line.

Doesn't take current sample

Hi Jim,

Never needed to with a slotted line.

Probe is of finite size (albeit small fraction of lambda in most cases)

And the time to measure anything is finite too.

The difficulty of making accurate RF current measurements at a point is
why things like slotted lines were invented. You can measure voltage at
a series of points and calculate what the current *must* be.

At worst, only three points, and generally for SWRs folks here would
agree were astronomical (and incalculable, but I could measure them
anyway).

With four fixed points (nominally eighth wave, but a range of
frequencies can be accommodated), I could determine the complex
impedance of any load. [re. "Microwave Measurements," Ginzton, Sec.
5.12] There are a myriad of other slotted line techniques, but I will
leave them for a suitable occasion.

I have never had to consider what the current *must* be - an
unnecessary elaboration.

73's
Richard Clark, KB7QHC

Jim Kelley wrote:

Cecil Moore wrote:
Jim Kelley wrote:
Problem is that it conflicts with the predictions made by the power
propagation model. (The politically correct name is the Joules/sec
propagation model.) That model shows that only a single reflection
of power is needed to explain the whole thing. Of course in some
cases the wave of power has to figure out how to turn around and go
back the other direction after it's been cancelled in order to
conserve energy. (A problem it wouldn't have to solve had it not
violated it in the first place.) Admittedly, some of the details
have yet to be worked out. :-)

Wow Jim, you need to repeat Fields and Waves 310. :-)
You have misunderstood virtually every principle
of the wave reflection model.

I wouldn't presume to take credit for any of the above. I learned it on
r.r.a.a. from someone who I think needs to take Fields and Waves 1. :-)

If you are talking about me, you have either misunderstood
what I said or you enjoy bearing false witness. Here's a quote
from my 2005 magazine article at:

http://www.w5dxp.com/energy.htm

"The author has endeavored to satisfy the purists in this
series of articles. The term "power flow" has been avoided
in favor of "energy flow". Power is a measure of that energy
flow per unit time through a plane. Likewise, the EM fields
in the waves do the interfering. Powers, treated as scalars,
are incapable of interference. Any sign associated with a power
in this paper is the sign of the cosine of the phase angle
between two voltage phasors."

Also, here is an EXCEL spreadsheet version of what happens
during the transient buildup to steady-state.

http://www.w5dxp.com/1secsgat.gif

Do you really consider 30 iterations to be only a single
reflection?
--
73, Cecil http://www.w5dxp.com

On Tue, 11 Dec 2007 00:36:02 GMT, Owen Duffy wrote:

I mean't literally a point sample, and I doubt that we can manufacture a
probe for either voltage or current that has zero dimensions (ie is
influenced ONLY by conditions at a point of no physical size.).

Hi Owen,

I would say that is arguable given my exposure to research in
Nanotechnology. I've seen one molecule probes (illustrated actually,
by microscopes that do enjoy vastly higher resolutions than were state
of the art 10 years ago; and they can image complex atoms and small
molecules).

These scales are commonly referred to as 1D, or quantum dots. Now,
given a measurement must be made, it demands a probe. Further, given
a quantum dot sized probe, it necessarily describes the smallest size
for which any smaller size yields no more information.

The practicality of it is that this quantum dot would have to reside
at the end of a very thin lead. A carbon nanotube so qualifies, and
they can be grown to considerable length (for this field), and they
themselves occupy no more than 1nM diameter (for a single walled
construction which is not terribly difficult to obtain).

In the practical field of RF, I seriously doubt such a probe/lead
combination would perturb a slotted line. In fact, their influence
would be lost in the decimals far to the right of the accuracy of the
line's construction. Or, I could easily imagine that most computers
would suffer rounding errors long before the probe/lead's influence
could be calculated if the line were perfect.

Yes, for the absolutist (no speaking of you Owen), a probe could never
be small enough, but if it were smaller it would be impossible to
obtain a reading. Sort of self-extermination of the absolutist
argument.

73's
Richard Clark, KB7QHC