Roy Lewallen wrote:
And if Cecil's work leads to the conclusion that the source impedance
impacts the line's SWR, then it's wrong.

It doesn't. In fact, just the opposite. My examples do NOT even
include a source impedance.
--
73, Cecil http://www.qsl.net/w5dxp



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----

On Tue, 12 Aug 2003 17:04:59 -0500, W5DXP
wrote:

Richard Clark wrote:
I see you still live in a dimensional aberration where you experience
22 days as 66 minutes.

I see that you still ignore the technical questions so I will repeat mine:
Are you saying that SWR doesn't equal (1+|rho|)/(1-|rho|)?

Hi Cecil,

Why don't you solve the first problem before presenting your own?

73's
Richard Clark, KB7QHC

On 12 Aug 2003 15:22:42 -0700, (Tom Bruhns) wrote:

Had a look at the refs. I'm curious, did you actually read the
sentence that Roy wrote?

Cheers,
Tom


Hi Tom,

Yes, did you?

73's
Richard Clark, KB7QHC

(Dr. Slick) wrote in message . com...
....
You're right about this, and it reminds us that if there is any
loss at all, we theoritically move away from a purely resistive
characteristic impedance into a complex one. This furthers the
complexity on the problem, as we must expect reactance in our coax.

It's not just theory, it's practice and measurable. But it's not a
requirement that loss introduce reactance to Zo; it's just that if the
line is lossless it must have a non-reactive Zo. Clearly if L/C =
R/G, the impedance will be non-reactive.

Also, consider what reactance you do get in practical lines, at what
frequencies. What effect does frequency have on the reactance? Why?
Under what conditions will it really be a problem? Might the reactive
part be so small that it will be totally swamped out by variations in
the real part? Think a bit about how much it will or won't mess up
the measurements you want to make.

Cheers,
Tom

"W5DXP" wrote in message
...
Dr. Slick wrote:
I didn't think you could tell us. I've never seen an SWR meter
that you could "calibrate" to 50 or 75 ohms, or less.

The calibration of the SWR meter is controlled by the internal sampling
load resistor, the 'R' in Peter's V + IR equation. I have a home-brewed
SWR meter that measures SWR on both balanced 450 ohm feedlines and on
300 ohm feedlines simply by changing the internal load resistors.
--
73, Cecil http://www.qsl.net/w5dxp
Cecil,

When you do this, does the scale still display the correct SWR for
conditions other than 1:1 ?

Tam/WB2TT

On Tue, 12 Aug 2003 15:52:05 -0700, Roy Lewallen
wrote:

It's admittedly hard for me to follow what you've written, but it sounds
like you're saying that:

Hi Roy,

1. Source impedance affects line SWR, and
2. It's impossible to tell by how much.

Did I get that right?
No. The sources offered and the data exhibited provide a very clear
answer. To state this yet again, it is the lack of knowledge in the
distances traversed between reflecting interfaces that introduces the
Mismatch Uncertainty. A smart lad might conspire to present any
particular Power measurement if he could withhold other details from
scrutiny. A lad who considered himself smart may do the same but
think he invented a free power amplifier (or dissipationless load).

This is simply an account of poor boundary controls that turn some
folk into magicians and others into the reincarnation of Galileo.
Either outcome is achieved through delusion.

Incidentally, you've brought up a new topic, that of an SWR meter.
Look at the Subject line.

As for the aside about SWR meter reading, I have performed SWR
measurements with a variety of NBS methods (many hundreds of times) -
none of them described here very often, and some never at all. I
doubt any here are so well versed in these methods as to challenge my
data by employing them (it would only confirm the results). I would
be happy to see as much effort put to it - in that it would represent
a technical rebuttal rather than echoed denials. I would be happy to
retract my points if someone revealed an error of commission/omission
- such has not happened and discussion of that data has been wholly
absent.

Look Roy, skip the rhetoric (from all sources including me) and
explain or refute the data I obtained. If you cannot accept it,
reveal the error. If the method is too tedious to replicate - say so.
This is not the first time I've broached the subject and I certainly
don't expect many to care for one, or follow blithely for another. It
only matters in issues of accurate Power determination.

73's
Richard Clark, KB7QHC

Tarmo Tammaru wrote:

"W5DXP" wrote:
The calibration of the SWR meter is controlled by the internal sampling
load resistor, the 'R' in Peter's V + IR equation. I have a home-brewed
SWR meter that measures SWR on both balanced 450 ohm feedlines and on
300 ohm feedlines simply by changing the internal load resistors.

When you do this, does the scale still display the correct SWR for
conditions other than 1:1 ?

The scale is calibrated using known loads so yes, it displays the
correct SWR up to 5.83:1. Above 5.83:1 the resolution of the scale
is poor because the full scale SWR is infinity. I have marks at
5.83:1 and 10:1. 5.83:1 is where the reflected power equals half
the forward power. Incidentally, the Z0 of my '450' ohm ladder-
line is closer to 388 ohms.
--
73, Cecil http://www.qsl.net/w5dxp



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----

W5DXP wrote in message ...
Dr. Slick wrote:
I didn't think you could tell us. I've never seen an SWR meter
that you could "calibrate" to 50 or 75 ohms, or less.

The calibration of the SWR meter is controlled by the internal sampling
load resistor, the 'R' in Peter's V + IR equation. I have a home-brewed
SWR meter that measures SWR on both balanced 450 ohm feedlines and on
300 ohm feedlines simply by changing the internal load resistors.


BTW, how do you know the accuracy of your homebrew SWR meter?


Slick

Dr. Slick wrote:
Would these be the termination resistors to ground, one for each
directional coupler?

Yes

I would think that you would have to adjust the
width of the traces as well, or the thickness of the dielectric
material, or the space between the couplers and the thru-line to use
it for an impedance other than 50 Ohms (talking about commercial
equipment, that is).

I use a toroid to sense the RF current and an R=Z0 resistor to turn it
into a voltage. I have a capacitive voltage divider on the RF voltage
which is adjustable for calibration purposes. The only real difference
between it and a 50 ohm SWR meter is the internal load resistor. It's
actually a bridge circuit with a Z0 reference resistor. Unfortunately,
it is not very accurate for the relatively high SWRs that I run on my
ladder-line, e.g. 4:1 to 16:1
--
73, Cecil http://www.qsl.net/w5dxp



-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----

(Tom Bruhns) wrote in message m...
(Dr. Slick) wrote in message . com...
...
You're right about this, and it reminds us that if there is any
loss at all, we theoritically move away from a purely resistive
characteristic impedance into a complex one. This furthers the
complexity on the problem, as we must expect reactance in our coax.

It's not just theory, it's practice and measurable. But it's not a
requirement that loss introduce reactance to Zo; it's just that if the
line is lossless it must have a non-reactive Zo. Clearly if L/C =
R/G, the impedance will be non-reactive.


True. Never considered this before, thank you. Makes sense that
if the ratio of the series resistance and shunt conductance are the
same as the ratio of the series inductance and shunt capacitance, that
the transmission line will still be non-reactive.


Also, consider what reactance you do get in practical lines, at what
frequencies. What effect does frequency have on the reactance? Why?
Under what conditions will it really be a problem? Might the reactive
part be so small that it will be totally swamped out by variations in
the real part? Think a bit about how much it will or won't mess up
the measurements you want to make.

Cheers,
Tom

In general, using an MFJ antenna analyzer to get a rough idea of
what the series equivalent complex impedance is (these are not $80,000
vector network analyzers!), it seems that cheap 3' RG-8x jumper coax
cables mainly add series inductance to the system, as the reactance
gets higher with increasing frequency.
We try to design 9 element chebychev low-pass filters, which is
not difficult at very low power levels, as two 1/4 watt 100 Ohm
resistors make a decent dummy load all the way out to 200 megs or so.

The problem is characterizing insertion loss using higher power
transmitters, when we know that the 1000 watt cantenna swings from 40
to 70 Ohms (with reactance too) as you get above 80 megs or so. It
become difficult to know if you are moving in the right direction or
not.

Slick