Richard Clark wrote:
"What is the source resistance of any power amplifier?"

Most power amplifiers are designed for a specified load but include
controls to adjust their outputs, both resistance and reactance, so that
a load which deviates can be made to match the 50-ohm or 600-ohm load
specified for the transmitter.

Am broadcast stations in the USA have a point at the transmitter output
just before power branches out to antenna system elements called the
"common point". Impedance here has been adjusted to a specified
resistance and a thermocouple ammeter is routinely installed which is
used to determine that the specified power output is being delivered by
the station by I squared R according to the FCC allocation. There is a
multiplier specified of 0.92 for stations of 5 KW or less. For stations
more powerful than 5 KW the multiplier is 0.947. This multiplier allows
for a slight increase in power to offset misc. errors and losses.

John E. Cunningham writes in "The Complete Broadcast Antenna Book" on
page 51:
"Suppose for example, we have a transmitter that is designed to work
into a 50-ohm load at an efficiency of 70% (Fig. 1-26). The efficiency
of this circuit expressed as a decimal, is given by

eff = RL / RS + RL

Rearranging

RS = RL [1/eff - 1]

Substituting the numbers inti this, we get

RS = 50 [1/0.7 - 1] = 50(1/0,7 - 1) = 50(0.43) = 21.4 ohms

You have an example in numbers. Hope you are happy.

Best regards, Richard Harrison, KB5WZI

On Wed, 18 Jun 2008 17:16:59 -0400, "Walter Maxwell"
wrote:

OK, Richard, is impedance 18 - j8 ohms sufficient?

Hi Walt,

It is more direct and responsive to every request I've made for
anything from anyone else so far. However, I presume it is the value
at the antenna connections of the amplifier that transforms through
the plate tuning toward the plate itself to render something more like
Rp (several KOhms?). Is it from a direct measurement, or by inference
to the load and tuning adjustments?

73's
Richard Clark, KB7QHC

Richard Clark wrote:
"I presume it is the value at the antenna connections of the amplifier
that transforms through the plate tuning toward the plate itself to
render something more like Rp (several KOhms?"

Rp has nothing to do with a Class C amplifier which acts like a switch
presenting an impedance to average current flow by being open-circuit
most of the time.

Best regards, Richard Harrison, KB5WZI

On Wed, 18 Jun 2008 16:53:43 -0500, (Richard
Harrison) wrote:

RS = 50 [1/0.7 - 1] = 50(1/0,7 - 1) = 50(0.43) = 21.4 ohms

You have an example in numbers. Hope you are happy.

Hi Richard,

As I mentioned to Walt in similar response to my oft-asked question:
direct and to the point. I see no others yet.

I know this wasn't arrived at by direct measurement and only by
inference. As I remarked to Walt (but in the inverse), I presume
this value is that transformed through the plate tuning from the
plate's Rp? Can you provide some work-up (presumably several KOhms)
from your citation for that?

73's
Richard Clark, KB7QHC

On Wed, 18 Jun 2008 17:51:34 -0500, (Richard
Harrison) wrote:

Richard Clark wrote:
"I presume it is the value at the antenna connections of the amplifier
that transforms through the plate tuning toward the plate itself to
render something more like Rp (several KOhms?"

Rp has nothing to do with a Class C amplifier which acts like a switch
presenting an impedance to average current flow by being open-circuit
most of the time.

Hi Richard,

You have no answer to what 18 - j8 ohms transforms to back through
plate tuning towards the plate. It is not responsive in kind.

As this operation is a commonplace of any treatment on the subject of
amplifier design (consult Bode, et al), then your ambiguous response
is outside of the canon of engineering.

I would intuit that you depart from Walt in this regard, and that he
will have a quantitative answer.

73's
Richard Clark, KB7QHC

"Richard Clark" wrote in message
...
On Wed, 18 Jun 2008 16:53:43 -0500, (Richard
Harrison) wrote:

RS = 50 [1/0.7 - 1] = 50(1/0,7 - 1) = 50(0.43) = 21.4 ohms

You have an example in numbers. Hope you are happy.

Hi Richard,

As I mentioned to Walt in similar response to my oft-asked question:
direct and to the point. I see no others yet.

I know this wasn't arrived at by direct measurement and only by
inference. As I remarked to Walt (but in the inverse), I presume
this value is that transformed through the plate tuning from the
plate's Rp? Can you provide some work-up (presumably several KOhms)
from your citation for that?

73's
Richard Clark, KB7QHC

Richard, am I now to understand that you haven't reviewed my Chapter 19A, which
details the entire procedure I used for the measurements that include the
resistance RL = 1400 ohms at the input of the pi-network? If you haven't I can
understand the reasons for your questions. I urge you to read 'Section 19.14
Additional Experimental Data', the last portion of the Chapter which describes
in detail the procedure is used in the measurements made on a Kenwood TS-830S.
I'm sure you'll find the answers you're looking for there.

You can find 19A by using this url from Cecil,
http://www.w5dxp.com/Chapt19a.doc. It derives from the thread directly above
'Unwin Antennas'.

Walt, W2DU

Sorry Richard, I goofed. The url I referred you to is directly above the thread
'Pornforum'.

Walt, W2DU

On Wed, 18 Jun 2008 20:39:12 -0400, "Walter Maxwell"
wrote:

Richard, am I now to understand that you haven't reviewed my Chapter 19A, which
details the entire procedure I used for the measurements that include the
resistance RL = 1400 ohms at the input of the pi-network? If you haven't I can

Hi Walt,

As I expected, you can provide quantitative data. This is a
refreshing change from the, what?, 100+ postings of "thoughts" on the
matter? It would suggest a poor rate of exchange for thinking these
days to come to so little conclusion.

You can find 19A by using this url from Cecil,
http://www.w5dxp.com/Chapt19a.doc.

Well, if you posted it through Cecil here, yes I didn't see it as I
don't read him.

Now, however, having a copy of that work from your own web site (why
didn't you reference that?) that I got after my last post, I will go
through it by the steps you lay out so carefully.

http://www.w2du.com/r3ch19a.pdf

Before doing so, and confining that to a separate posting, I would say
that I can acknowledge your fine work and care at creating a protocol
that does seem to reduce the variables, and test your hypothesis
without editorial bias. It seems like a great improvement from
earlier attempts to tackle the same issue.

A preliminary thought, however: What would you say is the Q of the
plate tank, Plate to Load, during operation under each load?

73's
Richard Clark, KB7QHC

Walter Maxwell wrote:
wrote in message
...
On Jun 14, 8:46 am, "Walter Maxwell" wrote:


I don't understand how my statement in the email above indicates that I^2*R

and

V*R could be zero. The simple ratio of E/I is not zero, yet it defines a
resistance that is non-dissipative because a ratio cannot dissipate power.


Walt


Hi Walt -

If E and I are not zero, then E*I is not zero. But you are correct
that the equations themselves do not dissipate power. :-) Resistors
do, however. If there isn't an actual resistor located where you make
your measurement, then of course there's no power being dissipated
there.

73, ac6xg

Jim, have you reviewed the new section Chapter 19A that appears on Cecil's
website that he uploaded on June 7? It appears there posted as 'Chapter 19A from
Reflections 3'. If you haven't reviewed it I urge you to do so, especially the
last portion where I report the measurements I made with a complex impedance
loading the amplifier. These measurements prove two things: 1) that the output
resistance of the amp is non-dissipative, and 2) that no reflected energy
reaches the amp tube, and in fact the measurements show that the tube doesn't
even see the reflected power. When you see the numbers and understand the
procedure I used in obtaining them you will be hard pressed to disagree with the
results.

Walt, W2DU

Hi Walt -

I have skimmed through it. It seems to me that your arguments with
this person named Bruene should be left out of it. Still, I have
enough faith in your considerable engineering talents (and enough
understanding of the underlying physics) to know that nothing in your
results would reveal dissipationless resistors having finite ohmic
resistance.

On the other hand, a virtual resistor, being the real component of a
virtual impedance, should by definition not be allowed the ability to
dissipate (or for that matter, reflect) energy. ;-)

73, ac6xg

"Richard Clark" wrote in message
...
On Wed, 18 Jun 2008 20:39:12 -0400, "Walter Maxwell"
wrote:

Richard, am I now to understand that you haven't reviewed my Chapter 19A,
which
details the entire procedure I used for the measurements that include the
resistance RL = 1400 ohms at the input of the pi-network? If you haven't I
can

Hi Walt,

As I expected, you can provide quantitative data. This is a
refreshing change from the, what?, 100+ postings of "thoughts" on the
matter? It would suggest a poor rate of exchange for thinking these
days to come to so little conclusion.

You can find 19A by using this url from Cecil,
http://www.w5dxp.com/Chapt19a.doc.

Well, if you posted it through Cecil here, yes I didn't see it as I
don't read him.

Now, however, having a copy of that work from your own web site (why
didn't you reference that?) that I got after my last post, I will go
through it by the steps you lay out so carefully.

http://www.w2du.com/r3ch19a.pdf

Before doing so, and confining that to a separate posting, I would say
that I can acknowledge your fine work and care at creating a protocol
that does seem to reduce the variables, and test your hypothesis
without editorial bias. It seems like a great improvement from
earlier attempts to tackle the same issue.

A preliminary thought, however: What would you say is the Q of the
plate tank, Plate to Load, during operation under each load?

73's
Richard Clark, KB7QHC

Hi Richard,

Sorry for the delay in answering, we were out of town all day yesterday so this
is the first opportunity I've had to read your post. Thank you for your kind
assessment of my measurements.

Now to the reason I didn't direct you to my own web page for the file 19A. That
version is about two years old, and I have made some sigificant changes since
then. Son Rick is my webmaster and at the time he was too busy to update the
file on the web page, so I gave Cecil the copy for him to put on his web page.
However, instead of adding it to this thread, for some reason he started a new
thread.

In addition to that I'm a dim-witted jerk. When I told you about accessing 19A I
was very tired and sleepy, so it didn't occur to me until the next morning that
I should have simply emailed you the most recent version. I've added some more
material even since the version I gave to Cecil, but not in the section
reporting the measurements.

Richard, I don't know of any way to determine the Q with the data I obtained,
I'd have to think about any additional measurements that would give the Q.
Unfortunately, my lab is in Florida and Jean and I are at our Summer home in
Mt. Pleasant, MI, so there is no way I can work on the Q until we return to FL
in November.

I'd appreciate any further comments you might have on the data appearing in 19A.

Walt, W2DU