Keith Dysart wrote:
So it is settled then. There is no NEED for a forward OR reverse
travelling wave. Differential equations rule.

Yes, it is settled in your own mind. In my mind,
there is certainly a need for forward and reverse
traveling waves without which standing-waves would
not be possible. If you want to deny the existence
of the cause of standing-waves, there is nothing
I can to stop you.

Well, except for the inability to explain where the "reflected
power" goes in the transmitter. Of course this is not an issue
for carefully selected examples where no "reflected power" reaches
the transmitter. A more general analysis technique would not
require such careful selection of examples.

The more general analysis technique tells us that
the moon is 1000 miles away from the earth. I don't
know how far away the moon is but I know it is not
1000 miles away.
--
73, Cecil, http://www.qsl.net/w5dxp

On Apr 11, 1:06 pm, Cecil Moore wrote:
Keith Dysart wrote:
So it is settled then. There is no NEED for a forward OR reverse
travelling wave. Differential equations rule.

Yes, it is settled in your own mind. In my mind,
there is certainly a need for forward and reverse
traveling waves without which standing-waves would
not be possible. If you want to deny the existence
of the cause of standing-waves, there is nothing
I can to stop you.

My mistake. But it is difficult to know your position when
you don't indicate clearly that you disagree, so I thought
that with your reply you were agreeing. Apologies.

Well, except for the inability to explain where the "reflected
power" goes in the transmitter. Of course this is not an issue
for carefully selected examples where no "reflected power" reaches
the transmitter. A more general analysis technique would not
require such careful selection of examples.

The more general analysis technique tells us that
the moon is 1000 miles away from the earth. I don't
know how far away the moon is but I know it is not
1000 miles away.

This is a bit of a non sequitur. So what is it that you really
disagree with in the analyses performed by myself and others?

Just for clarity, an example problem that has been previously
analysed is the following:
A generator with a 450 Ohm source impedance drives a 450 Ohm
ideal transmission line terminated in 75 Ohms. What is the
magnitude of the re-reflected wave at the generator?

I think you object to computing the amount of the reverse wave
that is reflected at the generator by using the source impedance.
More specifically you do not agree that the reflection
coefficient at the generator can be derived using
RC = (Zsource - Zline)/(Zsource + Zline).
Also, you do not agree that superposition applies at the source.

Given this, you then do not agree with the computations of the
quantity of the reverse wave that is reflected at the source which
then invalidates any further analysis.

Have I managed to capture the essence of your disagreement with
my and others analyses?

Note that these analyses have been performed without the use of
powers or interference so these side issues are not part of this
question.

....Keith

Keith Dysart wrote:
My mistake. But it is difficult to know your position when
you don't indicate clearly that you disagree, so I thought
that with your reply you were agreeing. Apologies.

I don't disagree with anyone's metaphysics. What
you do inside your own mind is none of my business.
(In my mind, I can still dunk a basketball.)

This is a bit of a non sequitur. So what is it that you really
disagree with in the analyses performed by myself and others?

I have told you many times. Bench test measurements
performed over the past 20 years or so prove that it
works only in your mind, not in reality. The source
impedance of a typical ham transmitter remains somewhat
of a mystery during actual operation. The arguments
continue to rage after decades of bench test experiments
and measurements. The pages of QEX are filled with those
arguments.

A generator with a 450 Ohm source impedance ...

False assumption. That transmitters's source impedance
changes away from 450 ohms just as soon as the reflections
arrive incident upon the source, i.e. the source impedance
is a *variable* that depends upon the magnitude and phase
of the reflected wave. If your source impedance is constant,
it doesn't match real-world conditions.

I think you object to computing the amount of the reverse wave
that is reflected at the generator by using the source impedance.

I certainly don't object to your computations but the
results of those computations have been disproved on
the bench using real world ham transmitters over the
past 20 years or so. Your simple mental model doesn't
correspond to reality unless you take some extraordinary
steps which deviate from real-world ham transmitters.
Have you taken the time to review those experiments?
--
73, Cecil http://www.w5dxp.com

On Apr 12, 8:15 am, Cecil Moore wrote:
I have told you many times. Bench test measurements
performed over the past 20 years or so prove that it
works only in your mind, not in reality. The source
impedance of a typical ham transmitter remains somewhat
of a mystery during actual operation. The arguments
continue to rage after decades of bench test experiments
and measurements. The pages of QEX are filled with those
arguments.

So your only beef with my examples is that they do not
accurately model a "typical ham transmitter"?

A generator with a 450 Ohm source impedance ...

False assumption. That transmitters's source impedance
changes away from 450 ohms just as soon as the reflections
arrive incident upon the source, i.e. the source impedance
is a *variable* that depends upon the magnitude and phase
of the reflected wave.

When you say that source impedance is a "variable", do you
mean this for a "typical ham transmitter", or do you assert
that it applies to every generator, even those which can be
accurately modelled with a Thevenin equivalent circuit (as
many signal generators, even TVSGs can)?

I think you object to computing the amount of the reverse wave
that is reflected at the generator by using the source impedance.

I certainly don't object to your computations but the
results of those computations have been disproved on
the bench using real world ham transmitters over the
past 20 years or so.

Which results have been disproved on the bench?

Your simple mental model doesn't
correspond to reality unless you take some extraordinary
steps which deviate from real-world ham transmitters.
Have you taken the time to review those experiments?

Are the experiments documented in Reflections chapter 19 and
19a representative examples?

My read of these chapters is that they offer compelling
argument and evidence (at least for the tube style
transmitters examined) that ham transmitters are linear*,
at least over their normal region of operation.

Is it your assertion that these claims are incorrect?

....Keith

* Don't confuse this use of linear, which is that the
output stage is linear, with whether the input to output
transfer function of the transmitter is linear. The transfer
function may be non-linear even though the output stage is.

Keith Dysart wrote:
So your only beef with my examples is that they do not
accurately model a "typical ham transmitter"?

Yes, I have said so about a half-dozen times now.

When you say that source impedance is a "variable", do you
mean this for a "typical ham transmitter", ...

Yes, the discussion is about typical ham
transmitters - nothing else matters to typical
hams.

Which results have been disproved on the bench?

Please research the grand argument between Warren
Bruene, w5oly, and Walter Maxwell, w2du.

Are the experiments documented in Reflections chapter 19 and
19a representative examples?

As far as I know, the Bruene/Maxwell argument first saw
light in a QST article in the early '90s and has been
raging ever since.

My read of these chapters is that they offer compelling
argument and evidence (at least for the tube style
transmitters examined) that ham transmitters are linear*,
at least over their normal region of operation.

Linear is not the requirement for your source impedance.
Constant, fixed, and linear is the requirement for your
source impedance. Nothing you have presented had the
source impedance as a linear variable.
--
73, Cecil http://www.w5dxp.com

On Apr 12, 3:37 pm, Cecil Moore wrote:
Keith Dysart wrote:
So your only beef with my examples is that they do not
accurately model a "typical ham transmitter"?

Yes, I have said so about a half-dozen times now.

When you say that source impedance is a "variable", do you
mean this for a "typical ham transmitter", ...

Yes, the discussion is about typical ham
transmitters - nothing else matters to typical
hams.

So when a poster presents a problem in a context other "typical
ham transmitters", why do you dispute the answers. If you can't
discuss the problem in the context presented by the poster, why
not have the courtesy to stay out. Others may be interested in
learning about how things relate in contexts other than "typical
ham transmitters". Why sabotage the discussions by arguing and
arguing and then saying "Oh, I only meant my comments in the
context of "typical ham transmitters" WHICH WAS NOT THE CONTEXT
OF THE PROBLEM STATEMENT. And kindly stop using the Texas A&M
example of TVSG and 1000 feet of line. It is clearly out of
your context of "typical ham transmitters".

Perhaps, in your dissertations on optics, it would be valuable
to state that they apply only in the context of "typical ham
transmitters". This might make it clear to the reader that
your suggestions are not generally applicable and could reduce
the wasted bits.

Which results have been disproved on the bench?

Please research the grand argument between Warren
Bruene, w5oly, and Walter Maxwell, w2du.

Are the experiments documented in Reflections chapter 19 and
19a representative examples?

As far as I know, the Bruene/Maxwell argument first saw
light in a QST article in the early '90s and has been
raging ever since.

My read of these chapters is that they offer compelling
argument and evidence (at least for the tube style
transmitters examined) that ham transmitters are linear*,
at least over their normal region of operation.

Linear is not the requirement for your source impedance.
Constant, fixed, and linear is the requirement for your
source impedance. Nothing you have presented had the
source impedance as a linear variable.

No indeed, the source impedance was a constant and resistive
in all my examples. That meets the needs for linear analysis.
Remember f(a+b) = f(a) + f(b)?

You are claiming that for "typical ham transmitters" the
source impedance is undefinable. This is quite at odds with
the exposition in Reflections chapters 19 and 19a. Or maybe not,
after all, Reflections is quite precise and claims only for a
specific class of ham transmitters, which, I suppose, may not
be typical. Though they look so to me.

So I take it that you no longer agree with the analysis presented
in Reflections 19 and 19a. I am pretty sure that you have stated
agreement in the past.

It would be valuable if you were to expand on the reasons for your
change of thought. What convinced you that "typical ham transmitters"
could not have their source impedance measured?

....Keith

Keith Dysart wrote:
So when a poster presents a problem in a context other "typical
ham transmitters", why do you dispute the answers.

Uhhhhhh Keith, because you presented the problem to me,
not someone else. You asked me what was wrong with your
examples. I obliged you. If you don't want me to answer,
don't ask me to respond.

Perhaps, in your dissertations on optics, it would be valuable
to state that they apply only in the context of "typical ham
transmitters". This might make it clear to the reader that
your suggestions are not generally applicable and could reduce
the wasted bits.

Perhaps, you should learn to recognize the common misleading
logical diversions, including your reductio ad absurdum assertion
above, and avoid them in the future.

No indeed, the source impedance was a constant and resistive
in all my examples.

Did you bench test it or just dream it up and wave your hands?
Maybe your ten cent resistor can resolve the war in Iraq - in
your mind.

So I take it that you no longer agree with the analysis presented
in Reflections 19 and 19a. I am pretty sure that you have stated
agreement in the past.

Please stop putting words in my mouth. I have never stated
agreement or disagreement. It is just one possibility out
of many that have been presented over the years. The fact
that there are so many theories is proof that it has not
been settled. Why don't you whip out an article that settles
everything and see what QEX thinks about it?

It would be valuable if you were to expand on the reasons for your
change of thought.

Since I haven't changed my thought, that would be difficult.
--
73, Cecil http://www.w5dxp.com

On Apr 12, 3:35 am, "Keith Dysart" wrote:


Just for clarity, an example problem that has been previously
analysed is the following:
A generator with a 450 Ohm source impedance drives a 450 Ohm
ideal transmission line terminated in 75 Ohms. What is the
magnitude of the re-reflected wave at the generator?

Interesting to me that CM railed at me for complaining that his
trivial trumped-up non-real-world example wasn't worth considering,
and now he's unwilling to accept an example that IS quite realizable.
I can EASILY do it on my bench, though I'd prefer to use a 50 ohm
generator and 50 ohm line, and a 300 ohm load at the end of the line.
It's certainly no more difficult to get a known output source
impedance than it is to get a lossless transmission line, but we talk
about lossless lines all the time, to remove an unnecessary
complication from our discussions. Since we have to test our
generators to insure that they really are 50 ohm sources, I feel quite
confident that their source impedance doesn't depend on the load you
put on them.

With respect to the output impedance of a transmitter, it may indeed
depend on the tuning/loading of the transmitter's PA, likely even the
power level it's running, and _maybe_ even on the load impedance you
put on the transmitter's output terminals. But what I do doubt is
that it _changes_ for a given setup including a given load impedance,
and in steady state, for sure the load impedance isn't changing. So
in steady state, can we determine if there are or are not reflections
on a transmission line connected to the transmitter, at that
interface? If you can't, how much do you have to disturb steady state
to make that measurement?

If battles have raged on for years about the output source impedance
of a transmitter, I submit that the people making the measurements
either (a) don't understand what they are doing, or (b) have not fully
specified the conditions under which they made the measurement, or
both. I assume they would't battle if they agreed they measured
different values, but that the conditions were also different, but
that may not be a valid assumption--some are known to battle
regardless.

As I mentioned, we have to be very careful about input and output port
impedances in measurement equipment. It's not necessarily an easy
thing to get "right," but I'm confident that we've been doing it
right, and probably for longer than the "debate" over transmitter
output source impedance has been going on.

If you don't believe there's a solution to the example Keith posted,
you have no right to believe in the results of a measurement with a
vector network analyzer, and you should certainly not trust the
indicated output level of any signal generator.

Cheers,
Tom

K7ITM wrote:
If you don't believe there's a solution to the example Keith posted,
you have no right to believe in the results of a measurement with a
vector network analyzer, and you should certainly not trust the
indicated output level of any signal generator.

Methinks you have missed the context of the discussion.
If the model doesn't work for an IC-706 it is not much
use to amateur radio operators. I have already said that
a valid model can be had for a signal generator equipped
with a circulator load. Now do it for an IC-706.
--
73, Cecil http://www.w5dxp.com

On Apr 12, 3:23 pm, Cecil Moore wrote:
K7ITM wrote:
If you don't believe there's a solution to the example Keith posted,
you have no right to believe in the results of a measurement with a
vector network analyzer, and you should certainly not trust the
indicated output level of any signal generator.

Methinks you have missed the context of the discussion.
If the model doesn't work for an IC-706 it is not much
use to amateur radio operators. I have already said that
a valid model can be had for a signal generator equipped
with a circulator load.

The only problem with this statement is the assumption that
the result can only be achieved with a circulator. It only
takes a 10 cent resistor. You really should put down your
optics books for a few hours and crack open a basic circuit
theory or transmission line text. Or google, "'lattice
diagrams' reflection". For matching at the source,
only 10 cent resistors are used.

....Keith