On Tue, 11 Oct 2005 17:45:51 GMT, Cecil Moore wrote:
Exactly what did you find technically wrong with that third hand quote?
Why would I bother? It reveals any one of several many problems:
1. Incorrect reporting through poor transcription;
2. Presumed ascription of source;
3. No access to the original (as third hand offers no citations, or
if it did, it renders third hand reporting as immaterial);
4. No access to the presumed source of the third hand-off;
5. No access to the original (op. cit.) author;
6. Discussion does not attend the topic (and is of no interest to me,
nor others barring their continuing that line of inquiry - I won't
hold my breath for that);
7. Third hand-off reporting is the lame equivalent of celebrity news.

Cecil Moore wrote:
Ian White G/GM3SEK wrote:
The Bird doesn't require any upstream and downstream boundary
conditions.

When Bird requires a 50 ohm environment, they are requiring
50 ohm boundary conditions for the reading to be valid.

No, they're not. They are requiring a 50-ohm system reference impedance.

What you call the "impedance environment" consists of physical things
like the source impedance, line impedance and load impedance. You're
confusing those with the system reference impedance, which something
completely different.

System reference impedance is purely a matter of definition. The most
common choice is 50 ohms... and by definition, that means 50 ohms
exactly.

Having made that choice, then you obviously design and calibrate your
instruments to give correct readings in an impedance environment that is
as close to your chosen reference impedance as you can practically make
it.

Your example shows the difference between impedance environment and
reference impedance most clearly.

If
you install the Bird in a 450 ohm environment on both sides
of the wattmeter, for instance, it will NOT read a valid forward
power and reflected power. In a matched-line 450 ohm environment
with absolutely zero reflected power, the Bird will indicate an
SWR of 9:1, a |rho| of 0.8 and a ratio of reflected power to
forward power of 0.64 even when the reflected power is zero.

You have changed the impedance environment to 450 ohms, and that's
fine... but all of the Bird's readings are perfectly correct if the
system reference impedance remains defined at 50 ohms. The reason why
say they are incorrect is that you also changed your definition of
system reference impedance to 450 ohms, without acknowledging that you
did it.

It's like doing a financial calculation without mentioning that you
switched into another base currency... darn right the results are not
valid.


--
73 from Ian G/GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Tue, 11 Oct 2005 21:11:55 +0100, Ian White G/GM3SEK
wrote:

Thank you, you're right. The key difference between direct and indirect
measurements is not about the need for mathematics; it's about the need
for additional input from theory.

What I should have said is:

Hi Ian,

Remarkable touch of admission - especially to my over-arching method
of criticism.

Another point needs to be attended; the discussion of the measurement
of SWR seems quite, and absurdly, drawn in kindergarten terms of
mathematics as if the determination were king and the numbers simply
fell out be virtue of cranking the equation.

Nothing could be further from the truth.

As much as Reg pines away about what a SWR does not measure,
absolutely the same could be said for his unspoken inference that
probe measurement along a line does measure it.

True, probing the line reveals a pattern that in the mind conforms to
the expectation of standing waves, but this is simply trying to
measure your own shadow when each time you stretch out the rule your
shadow moves further out. What size is your shadow - when?

Using the formula everyone here leans upon as the archetypal equation
for SWR, and claiming they've measured at the appropriate points along
the line (undoubtedly only in their imagination) easily leads to
errors above 20%. Worse yet is that this same formula fails utterly
at the bench (am I embarrassing anyone?) for any but the most
pedestrian of SWRs which are easily resolved by a SWR meter in the
first place.

Beyond this issue of accuracy (certainly no one is interested in that
are they?) stands the fictions of requiring slightly more than a
quarterwave length, or access along the line to both the trough and
the peak. Anyone so hamstrung to NEED these criteria, hasn't ever
really faced the problem of measuring SWR on an open line in the first
place. The double-minima method offers an exceptional accuracy for
high SWRs and occupies a smaller region of line than otherwise
demanded. Measuring at the minima also reduces the error introduced
in the very act of measuring SWR.

To this last, how many here can guarantee their probes will approach
the line with the same offset? There's a very good reason why SWR
probes are mounted on vernier carriages. How many would recognize
when the probes were too deep, or not deep enough to justify the
measurement? To imagine any approaching the line with hand held leads
raises the prospect of shooting marbles without thumbs.

As to these meters that everyone is rushing to use - Square Law or
Linear? Don't know the difference? You don't know accuracy or how to
obtain it when you have no choice. How do you render a Square Law
detector linear? How do you linearize a Square Law detector
measurement? No concern? You aren't measuring SWR then either. The
method of measurement for low, medium, and high SWRs is not the same.
One size does not fit all as the discussion in this group might imply
(from that same lack of actually having done it). Even the math is
different - and if any argue that this observation flies in the face
of simple transmission line equations, then these casual tourists are
comfortably remote from actually measuring the rough terrain of SWR.

However, none of this practicality is going to disturb the armchair
SWR analyzer. It comes to their great fortune that one simple
instrument will probably offer far more accuracy than they could ever
obtain by trying to be literal about SWR "on the line."

73's
Richard Clark, KB7QHC

Owen Duffy wrote:
Cecil Moore wrote:
Not true, Reg. My question was specified using RG-213 at 10 MHz.

True enough, but in the context of the question as to whether the Bird
43 reads sufficiently accurately, the transmission line on which one
is interested in the decay of the evanescent modes is the Bird
Thruline coupler section, not Rg-213 or any other cable that might be
attached to the Bird.

I didn't read it that way, Owen. IMO, the real question is:
What length of 50 ohm coax needs to be attached to the Bird
input and output to ensure that a 50 ohm environment is
present?
--
73, Cecil http://www.qsl.net/w5dxp

Owen Duffy wrote:
I don't think anyone is suggesting that the Bird could be used in a
general sense to estimate the VSWR on your 450 ohm line.

I thought that was the subject of the discussion.
--
73, Cecil http://www.qsl.net/w5dxp

Richard Clark wrote:
Certainly anything that is third hand and name dropping - but you
already knew that from my previous posting.

It's obvious that you cannot bring yourself to believe that e=mc^2.
--
73, Cecil http://www.qsl.net/w5dxp

On Wed, 12 Oct 2005 02:04:52 GMT, Cecil Moore wrote:

Owen Duffy wrote:
I don't think anyone is suggesting that the Bird could be used in a
general sense to estimate the VSWR on your 450 ohm line.

I thought that was the subject of the discussion.

From an earlier post:

In the case of the Bird 43, I suggest that if had, say, at 1MHz, 75
ohm line and a 75 ohm load on the load side, that the V/I raio for the
travelling waves in the region of the sampling element would be so
close to 50 ohms as to not materially affect the accuracy of
measurements on the 50 ohms coupler section, irrespective of the fact
that the sampling element has only 0.02% of a wavelength of 50 ohm
line on its load side.

(For avoidance of doubt, nothing in the foregoing is to imply the Bird
43 would be directly measuring or indicating the conditions on the 75
ohm line.)

Owen
--

Ian White G/GM3SEK wrote:

You have changed the impedance environment to 450 ohms, and that's
fine... but all of the Bird's readings are perfectly correct if the
system reference impedance remains defined at 50 ohms.

I have changed the system reference impedance to 450 ohms. Assuming
a tube PA with a pi-net output, 50 ohms doesn't exist anywhere anymore.
The system reference impedance is no longer 50 ohms so the Bird wattmeter
is being abused and misused. You can do the same thing by using a DC
voltmeter on an RF voltage or by using a hammer on a screw. If you
want to know the SWR on 450 ohm line, use a 450 ohm SWR meter.
--
73, Cecil http://www.qsl.net/w5dxp

Owen Duffy wrote:
In the case of the Bird 43, I suggest that if had, say, at 1MHz, 75
ohm line and a 75 ohm load on the load side, that the V/I raio for the
travelling waves in the region of the sampling element would be so
close to 50 ohms as to not materially affect the accuracy of
measurements on the 50 ohms coupler section, irrespective of the fact
that the sampling element has only 0.02% of a wavelength of 50 ohm
line on its load side.

If there is 75 ohm coax on the input of the Bird, the reflected
power reported by the Bird on the coax will be off by an infinite
percent. That's pretty inaccurate.
--
73, Cecil http://www.qsl.net/w5dxp

On Wed, 12 Oct 2005 02:06:22 GMT, Cecil Moore wrote:
It's obvious that you cannot bring yourself to believe that e=mc^2.
If "that is obvious" to you, then such are your problems of a third
hand education resulting in uncontrolled topic inflation.