On Jun 20, 1:33*pm, Owen Duffy wrote:
" wrote in news:6d1a078a-e5f4-
:

...

I think the problem with using the double coax is the very large
capacitance it adds to the feed line, effectively becoming a low pass
filter. Could be mistaken about the cause, but not the symptoms. I
have actually used the system to make some local 6 meter contacts.

The coax is, and always is a transmission line, and at the length you
described cannot be approximated well as a shunt capacitance.

What you have is a cascade of two line sections, one of say 600 ohms,
then one of 100 ohms, and each is probably operating with standing waves,
so there is impedance transformation.

To illustrate, lets say your feedpoint at 3.6MHz with a certain loop
antenna was 100+j0, and you had say 30m (100') of 600 ohm open wire, the
impedance looking into that would be around 240-j675. If you feed that
with say 6m (20') of LMR400 twin, then input Z would be around 60+j260
and loss would be about 40%.

The synthesised shielded pair is relatively lossy, and low Zo. Most
people use this configuration thinking that the shielding prevents
external fields from common mode current, but they are quite wrong. Seehttp://www.vk1od.net/transmissionline/stcm/index.htm.

Though the traditional approach has been to use a 4:1 voltage balun at
the rig to feed these things, there is good argument to use a 1:1
Guanella balun (current balun), and it can be located outside the shack
and inboard shield effectively grounded to deal with common mode current.
You still need to minimise the length of coax operated at high VSWR, and
it would not be necessarily absurd to think about low loss coax.

Approximating coax as a shunt capacitance might be reasonably accurate
for some applications at audio frequencies, but it is probably not for
most RF applications.

Owen

Thanks, Owen. I made up my test after reading the june 2008 QST
story.

Paul, KD7HB

Looks like most of the information I needed was already
written up on the DX Engineering webpages. I'll have to
go do a bit more reading before committing to anything.

Thanks for all the comments so far....

Pete

"Pete Bertini" wrote in
news
....
So, I would have balanced open-wire feedline from the dipole to
the 4:1 balun followed by a 1:1 common-mode RF choke with

I should have mentioned that 4:1 baluns integrated into ATUs are most often
voltage baluns.

When operated near a voltage maximum with high standing waves, they can be
very lossy. Sometimes it is claimed that they 'tame' difficult antennas
better than current baluns, which is often due to their internal loss. The
problem as such is not the balun, it is the extreme load presented by a
very poor antenna, and the lossy balun is a poor (grossly inefficient)
circumvention.

So, when a 4:1 voltage balun allows a match where a 1:1 current balun
doesn't, the problem is probably the antenna, not the balun.

Owen

That's pretty much what I've been reading on various websites.
At least some of the mystery is being cleared up A fan dipole
is starting to look better; at least I can control the feedpoint
impedance to within reason and keep a decent pattern on
the bands.

Pete

"Owen Duffy" wrote in message
...
"Pete Bertini" wrote in
news
...
So, I would have balanced open-wire feedline from the dipole to
the 4:1 balun followed by a 1:1 common-mode RF choke with

I should have mentioned that 4:1 baluns integrated into ATUs are most
often
voltage baluns.

When operated near a voltage maximum with high standing waves, they can be
very lossy. Sometimes it is claimed that they 'tame' difficult antennas
better than current baluns, which is often due to their internal loss. The
problem as such is not the balun, it is the extreme load presented by a
very poor antenna, and the lossy balun is a poor (grossly inefficient)
circumvention.

So, when a 4:1 voltage balun allows a match where a 1:1 current balun
doesn't, the problem is probably the antenna, not the balun.

Owen

Pete Bertini wrote:
Just curious if this has ever been done? I'm thinking about
putting up a 90 foot dipole feed with homemade open wire
line. I'd like to bring it into the shack using paralled runs
of LMR-400 cable, since the final 25 feet is via 3" electrical
conduit that also has rotor and other cables.

I believe the parallel cables with give me a 100 ohm impedance.
The open wire will be using #10 with homemade spreaders,
I'm going to try for 600 ohms at the feedpoint. I was wondering
if tapering the spacing on the feedline would give me a smoother
impedance where the open wire, arrestor, and twin coax arrange-
meet?


Yes..

Tapered transmission lines have been around for decades, and work just
like you'd expect.

For the ultimate in performance, you actually want to use an exponential
taper (spacing increases by the same percentage per unit distance), but
a linear taper (spacing increases by the same distance per unit
distance) works pretty much the same, especially if the taper is "long"
(10x) relative to the spacing. that is, going from 1" to 3" spacing
over a foot or two..


Pete k1zjh

As an academic exercise, for example using a tapered
line to match an OCD with a known FP impedance of 200 ohms
down to a 1:1 balun at the 50 ohm coax line transition point.
What is the recommended minimum length of cable needed to
perform a four to one characteristic impedance change in the
open wire for relating to the wavelength?

Pete


"Jim Lux" wrote in message
...
Pete Bertini wrote:
Just curious if this has ever been done? I'm thinking about
putting up a 90 foot dipole feed with homemade open wire
line. I'd like to bring it into the shack using paralled runs
of LMR-400 cable, since the final 25 feet is via 3" electrical
conduit that also has rotor and other cables.

I believe the parallel cables with give me a 100 ohm impedance.
The open wire will be using #10 with homemade spreaders,
I'm going to try for 600 ohms at the feedpoint. I was wondering
if tapering the spacing on the feedline would give me a smoother
impedance where the open wire, arrestor, and twin coax arrange-
meet?


Yes..

Tapered transmission lines have been around for decades, and work just
like you'd expect.

For the ultimate in performance, you actually want to use an exponential
taper (spacing increases by the same percentage per unit distance), but a
linear taper (spacing increases by the same distance per unit distance)
works pretty much the same, especially if the taper is "long" (10x)
relative to the spacing. that is, going from 1" to 3" spacing over a foot
or two..


Pete k1zjh

Pete Bertini wrote:
As an academic exercise, for example using a tapered
line to match an OCD with a known FP impedance of 200 ohms
down to a 1:1 balun at the 50 ohm coax line transition point.
What is the recommended minimum length of cable needed to
perform a four to one characteristic impedance change in the
open wire for relating to the wavelength?


What's the VSWR and frequency spec?

50 ohm balanced pair might be hard to come by.. the usual 276*log10(D/d)
equation is only valid when Dd. (D = spacing, d = diameter) and just
ballparking 50 ohms means that the spacing between the wires is about
10% more than the diameter (D/d 1.1). 200 ohms is more like D/d = 3

"Pete Bertini" wrote in
:

As an academic exercise, for example using a tapered
line to match an OCD with a known FP impedance of 200 ohms

Pete,

You really don't disclose very much, certainly not in a clear way. But,
some of your hints (eg OCD with open wire feed) leave me asking why? I am
not going to waste time writing a lot of words based on a hint though.

Owen

Pete Bertini wrote:
As an academic exercise, for example using a tapered
line to match an OCD with a known FP impedance of 200 ohms
down to a 1:1 balun at the 50 ohm coax line transition point.
What is the recommended minimum length of cable needed to
perform a four to one characteristic impedance change in the
open wire for relating to the wavelength?

Pete

It's not practical to make a 50 ohm open wire line. The wires would have
to be extremely close together.

But as to "recommended" minimum length, it depends on who does the
recommending and what your criteria are. The length of line necessary to
transition from one impedance to another depends on at least three
things: the amount of transformation required, the type of taper (e.g.,
linear, exponential, Klopfenstein), and how close the impedance match
has to be (usually specified as input reflection coefficient). There's a
brief treatment in Johnson, _Antenna Engineering Handbook_.

In Brown, Sharpe, Hughes, and Post, _Lines, Waves, and Antennas_, it's
stated that the tapered section must be at least a quarter wavelength
long, and that if less than that, "it will fail in its matching function".

In Skilling, _Electric Transmission Lines_, ". . . a 2 to 1 change of
characteristic impedance many be accomplished by using a tapered section
that is 1 wavelength long. A good impedance match results at this
frequency and at all higher frequencies. Results are fairly good if the
tapered section is as little as 1/2 wavelength long, particularly if the
impedance ratio required is less than 2 to 1. It is not worth while to
use a tapered section much shorter than 1/2 wavelength, for a taper that
is 1/4 wavelength or shorter will give nearly as much reflection as will
the abrupt change of impedance at a mismatch. Those frequencies for
which the length of the taper is a multiple of a half-wavelength are
transmitted with the least reflection. This may be a guide in choosing
the length of the taper." The author goes on to state that linear and
exponential tapers give practically the same results, and summarizes
with a recommendation that a tapered section be between 1/2 and 1
wavelength long at the lowest frequency of use.

Roy Lewallen, W7EL

"Roy Lewallen" wrote in message
...
Pete Bertini wrote:
As an academic exercise, for example using a tapered
line to match an OCD with a known FP impedance of 200 ohms
down to a 1:1 balun at the 50 ohm coax line transition point.
What is the recommended minimum length of cable needed to
perform a four to one characteristic impedance change in the
open wire for relating to the wavelength?

Pete

It's not practical to make a 50 ohm open wire line. The wires would have
to be extremely close together.

But as to "recommended" minimum length, it depends on who does the
recommending and what your criteria are. The length of line necessary to
transition from one impedance to another depends on at least three things:
the amount of transformation required, the type of taper (e.g., linear,
exponential, Klopfenstein), and how close the impedance match has to be
(usually specified as input reflection coefficient). There's a brief
treatment in Johnson, _Antenna Engineering Handbook_.

In Brown, Sharpe, Hughes, and Post, _Lines, Waves, and Antennas_, it's
stated that the tapered section must be at least a quarter wavelength
long, and that if less than that, "it will fail in its matching function".

In Skilling, _Electric Transmission Lines_, ". . . a 2 to 1 change of
characteristic impedance many be accomplished by using a tapered section
that is 1 wavelength long. A good impedance match results at this
frequency and at all higher frequencies. Results are fairly good if the
tapered section is as little as 1/2 wavelength long, particularly if the
impedance ratio required is less than 2 to 1. It is not worth while to use
a tapered section much shorter than 1/2 wavelength, for a taper that is
1/4 wavelength or shorter will give nearly as much reflection as will the
abrupt change of impedance at a mismatch. Those frequencies for which the
length of the taper is a multiple of a half-wavelength are transmitted
with the least reflection. This may be a guide in choosing the length of
the taper." The author goes on to state that linear and exponential tapers
give practically the same results, and summarizes with a recommendation
that a tapered section be between 1/2 and 1 wavelength long at the lowest
frequency of use.

Roy Lewallen, W7EL

Roy

Thank you Roy for taking the time to post. That's exactly what I
was looking for, and answers my questions completely. My
curiousity is satisfied.

Pete