On 25/05/2010 16:10, lu6etj wrote:
....
I think the mixture or combination of models -maybe- it would not be
"elegant" or consistent although it can arrive to the same numerical
results, but I don not dare to advance more than that in my
speculations :)

Miguel,

My explanation uses standard linear circuit theory, and the RLGC model
of a transmission line (captured in the Telegrapher's equation). It is a
frequency domain model, and it is complete and consistent.

One of the things that creates confusion in some peoples minds is that
they want one foot in the frequency domain (where you can talk about
concepts like reactance, complex impedance, VSWR) and simultaneously,
one in the time domain taking about re-re-reflected waves.

You can work in either domain, and you can transform between domains,
but trying to be in both at the same time creates problems.

BTW, if you think the problem is challenging to solve in the frequency
domain, don't even think about trying to solve it in the time domain.

So, do not worry about re-reflection, it is dealt with as you have
discovered by the steady state solution when you load the source with
the (steady state) impedance seen looking into the line. The resolution
of the wave component voltages and currents with KVL and KCL at the
circuit nodes gives the steady state solution. The Telegrapher's
equation gives you the amplitude and phase relationship of the wave
components for the transmission line, not just for fictitious lossless
lines, but for practical lines as the example demonstrates.

A steady state frequency domain analysis is quite adequate for most ham
problems. You don't see it spelled out as such, but that is how the ham
handbooks describe and solve problems.

As far as the myths about PAs destroyed by absorbing reflected power,
see "Does SWR damage HF ham transmitters?" at
http://vk1od.net/blog/?p=1081 .

Owen

On 25 mayo, 03:56, Owen wrote:
On 25/05/2010 16:10, lu6etj wrote:
...

I think the mixture or combination of models -maybe- it would not be
"elegant" or consistent although it can arrive to the same numerical
results, but I don not dare to advance more than that in my
speculations :)

Miguel,

My explanation uses standard linear circuit theory, and the RLGC model
of a transmission line (captured in the Telegrapher's equation). It is a
frequency domain model, and it is complete and consistent.

One of the things that creates confusion in some peoples minds is that
they want one foot in the frequency domain (where you can talk about
concepts like reactance, complex impedance, VSWR) and simultaneously,
one in the time domain taking about re-re-reflected waves.

You can work in either domain, and you can transform between domains,
but trying to be in both at the same time creates problems.

BTW, if you think the problem is challenging to solve in the frequency
domain, don't even think about trying to solve it in the time domain.

So, do not worry about re-reflection, it is dealt with as you have
discovered by the steady state solution when you load the source with
the (steady state) impedance seen looking into the line. The resolution
of the wave component voltages and currents with KVL and KCL at the
circuit nodes gives the steady state solution. The Telegrapher's
equation gives you the amplitude and phase relationship of the wave
components for the transmission line, not just for fictitious lossless
lines, but for practical lines as the example demonstrates.

A steady state frequency domain analysis is quite adequate for most ham
problems. You don't see it spelled out as such, but that is how the ham
handbooks describe and solve problems.

As far as the myths about PAs destroyed by absorbing reflected power,
see "Does SWR damage HF ham transmitters?" athttp://vk1od.net/blog/?p=1081.

Owen

Hi Richard and Owen

To Richard: What I mean is irrelevant :) relevant is what Walt
wanted to say in this sentence: "Because of the absorption of the
pad, the generator sees a nearly perfect match for all load conditions
and all reflected power is lost "
Pllease, tell me what in english means "all reflected power is lost"?
I understood (or translate or interpret) that reflected power is
dissipated in the pad: Is it a bad translation/interpretation?

To Owen: Sincerely thanks for your reasons. You can be sure I will
take note about your explanation an take some time tu analize it, but
I am not sure about to arrive at at the right conclusion because what
I read in this newsgroup is a long-standing discussion here.

Miguel

Another bit of reading that might help shed light on the matter is
http://eznec.com/misc/Food_for_thought.pdf, written eight years ago
during one of the many times the subject has come up before on this
newsgroup.

The chart and discussion in the "Forward and reverse power" section show
that the concept of "reflected power" being absorbed in or dissipated by
the source is incorrect. I find the concept of traveling waves of
average power to be misleading at best, and analyses using this concept
lead to impossible conclusions like the supposed absorption of power in
the source resistance.

Roy Lewallen, W7EL

On 25/05/2010 18:45, lu6etj wrote:
....
To Owen: Sincerely thanks for your reasons. You can be sure I will
take note about your explanation an take some time tu analize it, but
I am not sure about to arrive at at the right conclusion because what
I read in this newsgroup is a long-standing discussion here.


Yes, it is a recurrent discussion item. No doubt someone will be along
shortly to add some confusion to the pot.

A parting thought, do not confuse the process of establishment of steady
state with steady state. The only reason that a steady state solution is
valid, is that the system spends most of its time in steady state, or
substantially so. If the solution needs to focus mainly on establishment
of steady state (in other words, the system never substantially
settles), then you should be doing a time domain solution, and VSWR,
reactance, complex impedance are not a time domain concepts.

If you need to convince yourself, do a hand workup of five of ten
transits with a sinusoidal excitation. See that is does converge, and
quickly, and that at the load end, the reflected wave relative to the
forward wave is a fixed ratio (and hence VSWR) that depends ONLY on Zo
and Zl, there is NO influence by the source or any perceived source
reflection on the steady state. There are some animations of this on the
net, but they seem to confuse people more than enlighten them.

Owen

On May 25, 2:56*am, Roy Lewallen wrote:
The chart and discussion in the "Forward and reverse power" section show
that the concept of "reflected power" being absorbed in or dissipated by
the source is incorrect.

I'm sorry, Roy, but that is a very misleading statement. There are
THREE things that can possibly happen to the reflected energy.

1. It can indeed, be absorbed/dissipated by the source but it
certainly doesn't have to be. The conditions governing absorption,
reflection, and redistribution of EM wave energy are well understood
in the field of optics. RF gurus seem to be myopic about interference
effects.

2. It can obey the laws of the reflection model for EM waves, e.g. if
the source impedance is different from the Z0 of the transmission
line, some reflected energy will be re-reflected. (I know that doesn't
apply to your "food for thought" examples because the source resistor
is equal to the Z0 of the feedline.)

3. The interference phenomenon that you completely ignored in your
discussion, presumably through ignorance. This third possibility for
the reflected energy redistribution is associated with constructive
and destructive interference. It is the same phenomenon that is in
operation with the 1/4WL coating on non-reflective glass. Here's a
quote from a Florida State University web page with capitals added by
me for emphasis:

micro.magnet.fsu.edu/primer/java/scienceopticsu/interference/
waveinteractions/index.html

"... when two waves of equal amplitude and wavelength that are 180-
degrees ... out of phase with each other meet, they are not actually
annihilated, ... ALL OF THE PHOTON ENERGY present in these waves must
somehow be recovered or REDISTRIBUTED IN A NEW DIRECTION, according to
the law of energy conservation ... Instead, upon meeting, the photons
are REDISTRIBUTED TO REGIONS THAT PERMIT CONSTRUCTIVE INTERFERENCE, so
the effect should be considered as a REDISTRIBUTION OF light (and RF)
waves and PHOTON ENERGY rather than the spontaneous construction or
destruction of light."

This is the concept that you have been missing for eight years.
"Redistribution of energy" in a transmission line is simply a reversal
of energy flow. Each of your examples have different magnitudes of
interference (depending on the phasing between the forward energy and
the reflected energy). The phasing is a variable for all of your
examples. Why did you completely ignore the interference? Why would
you expect a CONSTANT magnitude of reflected energy to be dissipated
in examples with VARIABLE phasing and VARIABLE levels of
interference??? Here are the possibilities:

1. No Interference at the source resistor - this happens when the
forward wave is 90 degrees out of phase with the reflected wave at the
source resistor. In this case, the reflected energy is indeed
dissipated in the source resistor. I have written a short article on
the no interference case at:

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

2. Constructive Interference at the source resistor - This happens
when the phase angle between the forward wave and reflected wave is
less than 90 degrees at the source resistor. The voltage across the
source resistor increases and therefore more power is dissipated in
the source resistor. In fact, for total constructive interference, all
of the forward power and all of the reflected power is dissipated in
the source resistor. This happens when the SWR is infinite and the
reflected wave arrives at the source resistor in phase with the
forward wave from the source.

3. Destructive interference at the source resistor - This happens when
the angle between the forward wave and reflected wave is between 90+
degrees and 180 degrees. In this case, reflected energy is
redistributed back toward the load and less is dissipated in the
source resistor. In fact, for total destructive interference, zero
reflected power is dissipated in the source resistor and all of the
reflected energy is redistributed back toward the load. This happens
when the SWR is infinite and the reflected wave arrives at the source
resistor 180 degrees out of phase with the forward wave from the
source.

I explained all of this in my "WorldRadio" article way back in 2005.
Presumably, you have never read it. In order to understand the role
that interference plays during the superposition of EM waves, you
might want to read it and take a look at the references. The question
of where reflected energy goes was answered long ago by optical
physicists. Too bad that RF gurus remain so ignorant of that
knowledge. Here's that five year old "WorldRadio" article.

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

You obviously know how to add voltage phasors but you are obviously
ignorant as to what happens to the ExH power in each of those waves.
In any optics reference book, you will find the following irradiance
equation. Since irradiance uses the same units as the Poynting vector,
I have changed I (irradiance) to P (power density) in the irradiance
equation. I first saw this equation in Dr. Best's QEX article[1].

Ptot = P1 + P2 + 2*SQRT(P1*P2)cos(theta)

When any two EM waves are superposed, whether light waves or RF waves,
this is what happens to the component powers. The phase angle, theta,
is the relative phase angle between the two electric fields before
superposition. The last term in the equation is called the
"interference term" and it is easy to see how it modifies the total
power in the wave after superposition. Energy must be conserved. If
destructive interference occurs between two waves in a transmission
line, constructive interference must occur in the opposite direction -
and vice versa.

By completely ignoring the destructive and constructive interference
occurring in your "food for thought" examples, you have ignored the
accepted laws of physics (available in any physics optics reference)
and arrived at completely false conclusions. Roy, every one of your
examples can be explained simply by using the above equation from the
field of EM wave optics. Differing types/levels of interference
explains every one of your examples perfectly.

[1] Best, Steven R., "Wave Mechanics of Transmission Lines, Part 3",
QEX, Nov/Dec 2001
--
73, Cecil, w5dxp.com

On May 24, 10:31*pm, lu6etj wrote:
Anyway, my question is about validity of the assertion that reflected
wave -in that example- IS ABSORBED by the pad. According to my simple
calculations this hipothesis, as I see it, it does not coincide with
my early learnings.

Miguel, let's switch your example over to an easier to understand
example. Assume an ideal signal generator equipped with a resistive
circulator load. Let's call such a device an SGCR, a Signal Generator
equipped with a Circulator and a Resistor. Assume that 100% of the
reflected energy is dissipated in the circulator load resistor (none
re-reflected) and none of the reflected energy reaches the source. So
here is the block diagram.

SGCR--------feedline--------load

That model should be easier to discuss than the pad attenuator model.
What do you think?
--
73, Cecil, w5dxp.com

On Tue, 25 May 2010 00:45:36 -0700 (PDT), lu6etj
wrote:

To Richard: What I mean is irrelevant :) relevant is what Walt
wanted to say in this sentence: "Because of the absorption of the
pad, the generator sees a nearly perfect match for all load conditions
and all reflected power is lost "
Pllease, tell me what in english means "all reflected power is lost"?
I understood (or translate or interpret) that reflected power is
dissipated in the pad: Is it a bad translation/interpretation?

Hi Miguel,

Your translation is fine.

However, I have no idea what the pad design looks like, nor do I know
the component values. I have calibrated thousands of standard pads at
frequencies up to the 12 GHz. They came in either a Pi design, or a T
design. Their intended use is in system isolation. That is, they
isolate the source from the load OR isolate the load from the source
OR isolate both. For certain component values, you can replace the
"OR" with "AND."

You would isolate the source to keep its frequency and power constant.
You would isolate the load to keep line SWR flat. For this line
application, it is assumed you are calibrating either a load equal or
nearly equal to Zc, or you are measuring RF power. These are the
purpose of pads (they also serve the same function in audio circuits).
Measuring power in the presence of SWR other than 1:1 requires
sophisticated math that is rarely discussed here. Most discussion
usually accepts the presumptions of special cases (which are often
sufficient) and employ less rigorous formulas (which serve well within
the unstated presumptions).

In conventional Kirchoff analysis, the resistor that bridges the
transmission line opening becomes the source (that is Vs and Rs). Pad
design usually makes that one resistor for the Pi pad, or two
resistors for the T pad. If you are working in accurate and precise
measurement, you then account for the input (source) resistance in
parallel/series combinations. This second computation is the numeric
analysis of isolation. The higher the ratios of these pad resistors,
the higher the isolation.

It doesn't normally serve any use to have the pad apparent resistance
(what I called Rs above) different from Zc or from Zload, but as this
component is a sacrificial one, the designer may choose to put it to
use to achieve a desired goal. Pad performance suffers with heat due
to energy combinations that come from multiple/single sources.

73's
Richard Clark, KB7QHC

On 25/05/2010 18:56, Roy Lewallen wrote:
Another bit of reading that might help shed light on the matter is
http://eznec.com/misc/Food_for_thought.pdf, written eight years ago
during one of the many times the subject has come up before on this
newsgroup.


Roy's notes lead you into thinking about instantaneous power, and the
behaviour over a complete AC power cycle. Thinking that through, and the
implication about the power/time graphs at various points on the
transmission line gives insight into average energy flow, and energy
exchange between line sections, line and load and line and source, and
the physical bounds of exchange of stored energy during a cycle.

The Telegraphers Equation that I mentioned earlier captures the
behaviour of the line, and basic AC circuit theory, Joule's law etc
takes you the rest of the way.

Sure, working these things through takes time... and there isn't much
afforded in today's world of instant gratification where appealing
explanations are more accepted, irrespective of whether they are correct.

Miguel, invest in yourself and your knowledge, don't accept explanations
because of a vote count, accept them because they reconcile with things
that you truly know as facts (which sometimes means challenging what you
think you know).

Owen

On 25 mayo, 12:51, Richard Clark wrote:
On Tue, 25 May 2010 00:45:36 -0700 (PDT), lu6etj
wrote:

To Richard: What I mean is irrelevant :) *relevant is what Walt
wanted to say in this sentence: *"Because of the absorption of the
pad, the generator sees a nearly perfect match for all load conditions
and all reflected power is lost "
Pllease, tell me what in english means "all reflected power is lost"?
I understood (or translate or interpret) that reflected power is
dissipated in the pad: Is it a bad translation/interpretation?

Hi Miguel,

Your translation is fine.

However, I have no idea what the pad design looks like, nor do I know
the component values. *I have calibrated thousands of standard pads at
frequencies up to the 12 GHz. *They came in either a Pi design, or a T
design. *Their intended use is in system isolation. *That is, they
isolate the source from the load OR isolate the load from the source
OR isolate both. *For certain component values, you can replace the
"OR" with "AND."

You would isolate the source to keep its frequency and power constant.
You would isolate the load to keep line SWR flat. *For this line
application, it is assumed you are calibrating either a load equal or
nearly equal to Zc, or you are measuring RF power. *These are the
purpose of pads (they also serve the same function in audio circuits).
Measuring power in the presence of SWR other than 1:1 requires
sophisticated math that is rarely discussed here. *Most discussion
usually accepts the presumptions of special cases (which are often
sufficient) and employ less rigorous formulas (which serve well within
the unstated presumptions).

In conventional Kirchoff analysis, the resistor that bridges the
transmission line opening becomes the source (that is Vs and Rs). *Pad
design usually makes that one resistor for the Pi pad, or two
resistors for the T pad. *If you are working in accurate and precise
measurement, you then account for the input (source) resistance in
parallel/series combinations. *This second computation is the numeric
analysis of isolation. *The higher the ratios of these pad resistors,
the higher the isolation.

It doesn't normally serve any use to have the pad apparent resistance
(what I called Rs above) different from Zc or from Zload, but as this
component is a sacrificial one, the designer may choose to put it to
use to achieve a desired goal. *Pad performance suffers with heat due
to energy combinations that come from multiple/single sources.

73's
Richard Clark, KB7QHC

Dear friends

Sincerely it was not my desire to vivify old polemics but to tell the
truth, eight years it is a lot of time for not having arrived to a
consent!; hey boys this is science non religion! We must have a way
of leaving the well! :)

Is not possible you are using different models to describe an only one
phenomenon?, as looking at the same cat from their muzzle or from his
tail believing each one his cat is the true or real "cat" :)

I finished reading Cecil's article (http://www.w5dxp.com/nointfr.htm)
and I took of his example that of the 12,5 ohm load. I took a Smith's
chart and obtained the line input impedance, then I applied basic
circuits theory and I obtained the same value of power dissipated in
Rs -exactly-
As I see, if we use a simple electric model of generators and
impedances to solve the problem (maybe like Owen suggests), we can
explain the dissipation in Rs without appealing to any reflected power
returning into the generator because the interference phenomenon that
Cecil describes takes place to form the impedance that generator
see.
Or alternatingly the dissipation can be described by means of the
equations that Cecil shows in its page. In such case we should
consider both powers (direct and reflected) operating simultaneously
on generator resistance. Same cat, different points of view... :)

Perhaps my vision is naive but this situation reminds me an example of
Sears and Semansky book "University physics" (third edition) where he
explains that energy can be thought as not transported by charges in
movement, but for the electromagnetic field associated to them. Last
is a little hard to see -Poynting vector et al- :) but it is
applicable.

Always has been a pleasure for me to read you. I have learneing very
much from your enthusiastic discussions. You made me think of things
that I never thought without your help. Thank you.

Miguel Ghezzi . LU6ETJ

PS: Meanwhile I take the Owen advice and I am still studying!

On Tue, 25 May 2010 20:41:45 -0700 (PDT), lu6etj
wrote:

eight years it is a lot of time for not having arrived to a
consent!

Hi Miguel,

You got on this train rather late if all you see is eight years of it.
The circular references have entertainment value - so did vaudeville.

73's
Richard Clark, KB7QHC