Dave wrote:
The 2% WL value came from sci.physics.electromag.

then please take it back there. it makes no sense as it would force the
effect to get longer and longer at lower frequencies.

So how long does the coax have to be to force V/I to Z0
when the applied signal is DC?
--
73, Cecil http://www.qsl.net/w5dxp

Owen Duffy wrote:
An extension of that thinking is in the proposition that I have seen
that a Bird 43 cannot give valid readings unless there is at least a
quarter wave of 50 ohm line on each side of itself. In this case, the
magnitude of significantly affected line seems to be 25%, someone
else's is 2%, can they both be correct?

I think if you will recheck that posting you will find the assertion
was that a Bird 43 cannot give valid readings by sampling at a point.
The line must be at least 1/4WL, and preferably 1/2WL, so that voltage
maximums and minimums will exist and can be measured.

And that 2% of a wavelength is from my faulty memory. I'll try to
Google and find the exact quotation.
--
73, Cecil http://www.qsl.net/w5dxp

"Cecil Moore" wrote in message
t...
Dave wrote:
The 2% WL value came from sci.physics.electromag.

then please take it back there. it makes no sense as it would force the
effect to get longer and longer at lower frequencies.

So how long does the coax have to be to force V/I to Z0
when the applied signal is DC?
--
73, Cecil http://www.qsl.net/w5dxp

just enough for any fringe effects to become negligible... no more than a
couple diameters if the coax as a rough guess.

"Cecil Moore" wrote in message
t...
Owen Duffy wrote:
An extension of that thinking is in the proposition that I have seen
that a Bird 43 cannot give valid readings unless there is at least a
quarter wave of 50 ohm line on each side of itself. In this case, the
magnitude of significantly affected line seems to be 25%, someone
else's is 2%, can they both be correct?

I think if you will recheck that posting you will find the assertion
was that a Bird 43 cannot give valid readings by sampling at a point.
The line must be at least 1/4WL, and preferably 1/2WL, so that voltage
maximums and minimums will exist and can be measured.

And that 2% of a wavelength is from my faulty memory. I'll try to
Google and find the exact quotation.
--
73, Cecil http://www.qsl.net/w5dxp

i want to see a quote from a manufacturer's or good laboratory manual for
that 1/4 or 1/2 wave thing on the bird also.

Dave wrote:
just enough for any fringe effects to become negligible... no more than a
couple diameters of the coax as a rough guess.

Apparently, the 2% of a wavelength that I was remembering was
at 10 MHz. 1'/(98.4*0.66) rounded to 2%. All I was interested
in at the time was proving to Reg that one foot of coax forces
Vfor/Ifor = Vref/Iref = Z0, the boundary conditions assumed
by Bird Wattmeter designers. Since Kevin was not familiar with
PL-239's, I erred on the side of caution with the 2% estimate.

****Quote****
Newsgroups: sci.physics.electromag
From: "Kevin G. Rhoads"
Date: Tue, 07 Oct 2003 12:49:14 -0400
Subject: Transmission Line Question

Cecil wrote:
It addresses it adequately but doesn't answer any particulars.
Given PL-239 connectors and RG-213 coax, I wonder what the
answer would be for 10 MHz?

I'm not familiar with the connectors in question. Assuming
they are properly attached, they should not introduce much
mode diversion. For 10 MHz I would expect that all other modes
would be non-propagating (i.e., evanescent) even though RG-213
is a large coax (improved RG-8 apparently). The speed of propagation
is listed as 66%, so the nominal wavelength is 3/2 times the free
space wavelength for the TEM mode. 3/2 x 30m = 45m, which implies
the decay rate in space for non-TEM modes is going to be large
as the cable diameter is .405" (jacket) which implies the
spacing from inner to outer conductors will be less than .203".
For order of magnitude estimate, assuming the lowest non-TEM mode
can be approximated using a characteristic equation that really
is only applicable in Cartesian geometries:
(1/45m)**2 = (1/.203")**2 + kz**2
Clearly, kz must be imaginary to make this work. thus an
evanescant, non-propagating wave:
kz**2 = (1/45m)**2 - (1/.203")**2
To the accuracy used to date, the first term on the right
is negligible, so the decay rate, alpha, can be estimated:
alpha**2 = - (kz)**2 = (1/2.03")**2
Or, the lowest order undesired mode should reduce intensity
by a factor of 1/e (0.37) in about 2.03"; power will reduce
by that factor squared in the same distance (.135). In
about four inches, undesired mode power is down to about
0.018, in six inches, .00248, and after a foot, 6.14x10-6

You should double check my algebra, but I think the estimate
is reasonable. To put it into other terms, since the wavelength
in the coax dielectric is 45m and the conductor to conductor
spacing is about 2", any non-TEM mode will suffer attenuation
in E-field intensity with a space-rate constant rounghly
equal to the conductor to conductor spacing. INtensity
drops by 1/e = 1/2.71828 every 2 inches. Power availalbe
drops faster, being square of intensity.

So unless almost all the power diverts into an undesireable
mode (by a factor of more than a million to one), one foot
of cable should see pure TEM at the end.

HTH
Kevin
****End Quote****
--
73, Cecil http://www.qsl.net/w5dxp

Dave wrote:

"Cecil Moore" wrote in message
et...
I think if you will recheck that posting you will find the assertion
was that a Bird 43 cannot give valid readings by sampling at a point.
The line must be at least 1/4WL, and preferably 1/2WL, so that voltage
maximums and minimums will exist and can be measured.


i want to see a quote from a manufacturer's or good laboratory manual for
that 1/4 or 1/2 wave thing on the bird also.


Cecil was quoting someone else there, and is completely innocent :-)


Here's how the Bird 43 measures VSWR. It contains a pair of needle-fine
voltage probes, powered by small explosive charges. When coax is
connected at either side, it fires those probes out into the coax until
it finds a voltage maximum and a voltage minimum. Then it computes the
Voltage Standing Wave Ratio and a recoil mechanism reels the probes back
in. It's so slick, it all happens before you even know it.

Warning: when handling a Bird 43, keep all sensitive parts more than
1/2WL from those sockets!


An alternative possibility is that the Bird 43 does give valid readings
by sampling at the point where it physically is.


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

On Sat, 8 Oct 2005 14:20:46 -0000, "Dave" wrote:
i want to see a quote from a manufacturer's or good laboratory manual for
that 1/4 or 1/2 wave thing on the bird also.

Hi Dave,

Don't hold your breath waiting for that Baloney sandwich to be made.

73's
Richard Clark, KB7QHC

"Richard Clark" wrote in message
...
On Sat, 8 Oct 2005 14:20:46 -0000, "Dave" wrote:
i want to see a quote from a manufacturer's or good laboratory manual for
that 1/4 or 1/2 wave thing on the bird also.

Hi Dave,

Don't hold your breath waiting for that Baloney sandwich to be made.

73's
Richard Clark, KB7QHC

i'm not, just bored sitting in here watching it rain. over 2" so far today.

On Sat, 08 Oct 2005 14:13:02 GMT, Cecil Moore wrote:

Owen Duffy wrote:
An extension of that thinking is in the proposition that I have seen
that a Bird 43 cannot give valid readings unless there is at least a
quarter wave of 50 ohm line on each side of itself. In this case, the
magnitude of significantly affected line seems to be 25%, someone
else's is 2%, can they both be correct?

I think if you will recheck that posting you will find the assertion

Cecil, it is someone else who has on a number of occasions suggested
the quarter wave thing in email correspondence, and here in postings.

My suggestion is that the sampler inside a Bird 43 coupler section is
sufficiently far inside the 50 ohm coupler line to provide
measurements within the instrument's stated accuracy of what is
happening within the 50 ohm coupler, irrespective of whether, for
instance, a 75 ohm line is attached to the coupler on the load side.

The measurements of what is happening within the Bird 43 coupler could
then be used to model what is happening on the adjacent line, having
regard for any Zo changes, loss, length etc.

Owen
--

Owen Duffy wrote:
Cecil, it is someone else who has on a number of occasions suggested
the quarter wave thing in email correspondence, and here in postings?

Yep, it's not me, it's Reg. I have defended the Bird wattmeter
design. Reg sez one needs at least 1/4WL and preferably 1/2WL
in order to accurately ascertain the "real" SWR.
--
73, Cecil http://www.qsl.net/w5dxp