On Apr 20, 9:13 am, Cecil Moore wrote:
Keith Dysart wrote:
But a key realization is that the behaviour on the line
can not depend on arcane details about the construction of the
generator.

Agreed, but the behavior on the line also doesn't depend
upon where the forward and reverse traveling wave came from.
All that is important is their existence. The reverse
traveling wave might come from a separate source and the
line doesn't know the difference.

All true, and as long as you think of voltage and current waves
you won't get into trouble.

So when such statements are made, it is instructive
to explore the situation with a different generator. If the outcome
is different, then there is something wrong with the theory.

No, if the outcome is different, something different is
happening while abiding by the wave reflection model rules.
Why are you surprised that changes occur when something is
changed?

The key point is that the line conditions did not change, so the
same reflections must be occuring and yet your explanation
claims that sometimes nothing is reflected and sometimes
all is reflected.

Note
that the standard explanations only require that the impedance
of the source be known.

If the "standard explanations" were correct, this would
have been put to bed a long time ago.

And it was. But there are always a few stragglers.

But how does it know, since the output (source) impedances are
the same for both cases?

Conclusion: The output source impedance is NOT the impedance
encountered by the reflected wave. That's essentially what
Walter Maxwell's paper says.

Did you read the same paper I read? I recall that the claim was a
conjugate match of the effective impedances AFTER tuning. The
systems under discussion here have not been tuned for maximum
power transfer.

I think you have made this mistakte before. In my expression P=VI,
V and I are simultaneous instantaneous values and P is the
instantaneous power. If you want average you need to integrate
and divide.

No, if you want to average sinusoidals , use RMS values, which
is the result of said integrating and dividing. Somebody else
already did all the work.

This is true, but reflection is an instantaneous behaviour.
Simplifying
the analysis to RMS or average values, while often effective, bears
many risks in misleading the understanding, as demonstrated here.

Your expression applies only to sine waves and V and I are peak
voltages.

I is a peak voltage??????? You are more confused than
I ever realized. :-)

Oh darn. A typo. I know it will be quoted over and over in subsequent
posts. So be it.

Well, actually two typos. It should read "V and I are RMS voltages
and currents".

Interestingly, the expression P=V*I*cos(theta) for sine
waves is always derived by starting with Pinst=Vinst*Iinst,

Interestingly, you forgot to say you were talking about
instantaneous power until now.

My apologies. I did not mean to confuse you. Since reflection is
an instantaneous phenomenon, I just tend to think of it that way.

True. But there IS energy flowing at every point that is not a voltage
minimum or maximum. At these points there is NEVER any energy
flowing.

Every half cycle in every EM traveling wave, the instantaneous
power is zero. Would you therefore argue that traveling waves
cannot transfer any energy? Please get real.

Of course not. But when the instaneous power is 0 for all instants
then no energy can be flowing.

Given two parallel water pipes carrying 100 gallons/minute
in opposite directions. I can see you arguing that there
is no NET flow of water. But I cannot see you arguing that
there is no flow of water at all when the pressure of just
one stream can knock you off your feet.

Given one pipe of water, I also would not argue there are 10**9
gallons flowing per minute in one direction and 10**9 gallons
flowing per minute in the other, but the net is 0 so all is well.

That is analagous to the claim being made for forward and
reverse energy carrying travelling waves. Are you arguing
that for the water?

Well that does leave you with a bit of a conundrum since on
an open ended line, zero energy is flowing at every quarter
wavelength point back from the load.

The conundrum is all yours. Every half cycle, every EM
traveling wave is at a zero instantaneous power point.
If EM traveling waves can transfer energy while the
instantaneous power is zero every half cycle, superposing
two of them doesn't present a conundrum at all.

Except at those points where for every instant the instantaneous
energy transfer is zero.

At such a point a real instantaneous wattmeter would always
indicate zero.

A real instantaneous wattmeter would always indicate zero
at the zero crossing of every EM traveling wave in the
universe so exactly how does the light from Alpha Centauri
ever make it to Earth when the instantaneous power is zero
every half cycle? Please get real.

There is quite a difference between the instanteous power being
occasionally zero and being zero for all instances. Real enough?

....Keith

On Apr 20, 9:23 am, Cecil Moore wrote:
Keith Dysart wrote:
For example, in all your examples you now provide a circulator
because you think this is the only way to control reflections,
when in fact, all that is necessary is a source impedance that is
equivalent to the line impedance and there will be no reflection.

Your above statement has been proved wrong in any number
of bench experiments.

Actually, most bench experiments demonstrate the opposite.
Were you to take a moment and look at the schematic for any
typical signal generator, you would find the resistor that matches
the generator to the line and prevents reflection of signals
incident upon the generator.

In other cases you claim that problems are insolvable because
insufficient information is provided about the generator. In fact,
all you need to do is use the source impedance to compute
the reflection coefficient and the problem is solved.

Again, proved wrong by any number of bench experiments.

It would be educational if you could describe one of these
experiments.

Remember, you are looking for a re-reflection of the reverse signal
at a generator whose source (output) impedance matches the line
characteristic impedance.

You claim that superposition does not work in a generator.

Not that it doesn't work - just that it is impossible
for a reflected wave to compete with an active dynamic
source. The source essentially ignores the reflected
wave like a fire hose ignores you trying to spit up
the hose against the flow.

Either it works, or it doesn't. Superposition in circuit theory
is not dependant on the relative magnitudes of any of the
sources.

You will find that the alternatives work better, ...

Sorry Keith, I stopped listening to you when you asserted
that I is voltage.

Darn. I knew that typo would show up again. On the other
hand, I can take some pleasure in accurate predictions.

....Keith

Cecil Moore wrote:
Gene Fuller wrote:
Where are the equations that describe this "delta-t" stuff that you
keep bringing up?

I made an error in my last reply and have canceled that reply.

delta-t is a mathematical term, Gene, related in the limit
to the differential dt. If you have to ask, I'm sure you
wouldn't understand the explanation.

Given the following experiment with two signal generators
equipped with circulators and load resistors - the generators
are phased-locked to ensure coherency:

[random s-parameter babble snipped]

How
long does it take to achieve b1 = s11(a1) + s12(a2) = 0?
I estimate that time to be delta-t which becomes dt
in the differential equation limit.

The principle of superposition allows us to observe that
s11(a1) and a12(a2) actually existed before they were
canceled.

Cecil,

That's typical; you don't have a clue about what you have dragged
yourself into.

Let's see. Delta-t is, uhhhh, delta-t. And of course we can replace
delta-t by the differential form, dt, if we choose. You haven't offered
the slightest justification for this bizarre formulation. Where are the
equations that include delta-t or dt? Sure, Maxwell's equations include
time derivatives, but somehow that does not seem to offer much help for
your delta-t dilemma. I always though that part of the utility of
s-parameters is that they do not have any time dependence in steady
state conditions. Is there some equation for ds11/dt that I missed?

I would sure like to see your definition of "the principle of
superposition" if it somehow says that something existed before it was
canceled. I cannot find that definition in any of my sources.

8-)

73,
Gene
W4SZ

On Apr 19, 3:24 pm, Cecil Moore wrote:

Interaction is certainly required for them to cancel forever.

So let it be written, so let it be done. :-)

Otherwise, we have all these energyless phantom ghost waves
existing forever.

Apparently you intend to concern yourself with waves which never
existed, forever.

73, AC6XG

On Apr 19, 7:47 pm, Richard Clark wrote:
On Thu, 19 Apr 2007 17:51:10 -0700, Jim Kelley
wrote:

So I gotta ask: what do waves do instead of superposing when there
isn't a load somewhere? Whistle and look the other way? :-)

Hi Jim,

What do they DO?

I was hoping to persuade you to tell us, Richard. In your post, you
only told us what they don't do.

73, Jim AC6XG

Keith Dysart wrote:
All true, and as long as you think of voltage and current waves
you won't get into trouble.

I don't recall Maxwell's equations relying on "voltage and
current waves".

The key point is that the line conditions did not change, so the
same reflections must be occuring and yet your explanation
claims that sometimes nothing is reflected and sometimes
all is reflected.

If you put two signal generators equipped with circulator
loads at each end of a transmission line, there are absolutely
no reflections anywhere. Yet there is a forward wave and a
reverse wave. We can cause a reverse wave when "sometimes
nothing is reflected and sometimes all is reflected". The
transmission line will not know the difference.

Did you read the same paper I read? I recall that the claim was a
conjugate match of the effective impedances AFTER tuning. The
systems under discussion here have not been tuned for maximum
power transfer.

The existence of a conjugate match is irrelevant to our
discussion. All that is relevant to our discussion is
that the reflected waves does not see the generator
source impedance.

Oh darn. A typo. I know it will be quoted over and over in subsequent
posts. So be it.

Keith, arrogant omniscient beings, as you present yourself
to be :-), do not make typos.

Of course not. But when the instaneous power is 0 for all instants
then no energy can be flowing.

The instantaneous power is zero every 1/2WL in an EM wave.
Therefore, according to you, EM waves cannot transfer energy
or power. Good luck on that one.

There is quite a difference between the instanteous power being
occasionally zero and being zero for all instances. Real enough?

If the instantaneous power is zero for all space and time,
an EM wave cannot exist.
--
73, Cecil, w5dxp.com

Keith Dysart wrote:
It would be educational if you could describe one of these
experiments.

I already did - Bruene's early 1990's QST article.

Remember, you are looking for a re-reflection of the reverse signal
at a generator whose source (output) impedance matches the line
characteristic impedance.

It happened with your pet generator from which you quickly
tried to divert attention. Zero power dissipation inside
a "matched" source is hard to sweep under the rug, huh?

Either it works, or it doesn't.

Superposition works in some situations and doesn't work
in others. For instance, it doesn't work with power.
--
73, Cecil, w5dxp.com

On 20 Apr 2007 09:34:11 -0700, Jim Kelley wrote:

On Apr 19, 7:47 pm, Richard Clark wrote:
On Thu, 19 Apr 2007 17:51:10 -0700, Jim Kelley
wrote:

So I gotta ask: what do waves do instead of superposing when there
isn't a load somewhere? Whistle and look the other way? :-)

Hi Jim,

What do they DO?

I was hoping to persuade you to tell us, Richard. In your post, you
only told us what they don't do.

Hi Jim,

They do nothing. It is still an odd question - unless it was posed
for the benefit of lurkers who are in doubt.

73's
Richard Clark, KB7QHC

Richard Clark wrote:

On 20 Apr 2007 09:34:11 -0700, Jim Kelley wrote:


On Apr 19, 7:47 pm, Richard Clark wrote:

On Thu, 19 Apr 2007 17:51:10 -0700, Jim Kelley
wrote:


So I gotta ask: what do waves do instead of superposing when there
isn't a load somewhere? Whistle and look the other way? :-)

Hi Jim,

What do they DO?

I was hoping to persuade you to tell us, Richard. In your post, you
only told us what they don't do.


Hi Jim,

They do nothing. It is still an odd question - unless it was posed
for the benefit of lurkers who are in doubt.

Hi Richard,

Actually, it was posted because of doubt about your claim "Remove the
load, and you remove interference." Please describe this phenomenon
in more detail. The implications are huge.

Thanks,

Jim AC6XG

On Fri, 20 Apr 2007 11:23:24 -0700, Jim Kelley
wrote:

Actually, it was posted because of doubt about your claim "Remove the
load, and you remove interference." Please describe this phenomenon
in more detail. The implications are huge.

Hi Jim,

Implications aside, it would help us all if you simply describe your
doubt instead of having me fish for your problem.

73's
Richard Clark, KB7QHC