On Feb 11, 5:38*pm, Jim Lux wrote:
K7ITM wrote:
On Feb 11, 8:02 am, phaedrus wrote:
I seem to recall that ladder/open-wire feeders are relatively high
impedance ~500 ohms. Is it possible to construct 50 ohm open wire
feeder and if so, what would the spacing be?

Thanks!

As hinted by others, especially Owen, it's quite possible to make
parallel-conductor line of very low impedance. *There are two ways
that are easy to describe, and these may well bring others to your
mind:

First, consider a line made of broad, flat conductors: *copper
ribbon. *If the spacing is close compared with the conductor width,
the impedance will be low. *This is not particularly practical for
"open-wire feeders," but is commonly used on printed circuit boards.
In the extreme, you can make transmission lines with impedances under
10 ohms this way, if you're of a mind to. *You can use thin, flexible
Kapton tape, perhaps 0.1mm thick or less, and have conductors 10 or
more mm wide. *According to a formula in Sams' "Reference Data for
Engineers," the impedance will be approximately 377*s/
(sqrt(epsilon)*w), where w is the conductor width and s is the
spacing. *For air dielectric, where epsilon is 1, you can make a 10
ohm line if the width is about 38 times the spacing.

You see this in microstrip construction on various substrates,
particularly where it's part of a matching network for transistor
amplifiers (which can have Zs of a few ohms)



On the other hand, there's generally not much advantage to making a
low-impedance balanced line for RF work, so they are pretty uncommon
in practice.

in RF communications, that's true.

In other high power RF applications, low Z can be desirable (because
you'd rather take the IR hit than deal with the high voltage).

In the pulsed nuclear research business you see lines with Z of a few
ohms in pulse forming networks ften using distilled water as the
dielectric.. that epsilon of 80 gets the C right up there so Sqrt(L/C)
is small.



Cheers,
Tom



Yes, indeed a low impedance can be helpful sometimes...but the times
I've been involved with low-Z lines, they've almost always been
unbalanced (e.g. coaxial). For a PFN, the high epsilon is also an
advantage in storing a lot of energy. Since even very pure water is
pretty conductive (relative to seriously good dielectrics), I'm a bit
surprised it's used. Are the cases you're thinking of balanced lines,
or coaxial? Is the impedance also chosen to match the load (in some
sense), or is there also a transformer at the output to match to a
higher-impedance load?

It occurs to me that for a PFN, energy storage is a good thing...in
fact, it's half of what a PFN is all about, the other half being
delivering that energy in a controlled manner. But in power
transmission, (excess) energy storage is generally a disadvantage.
Lines with dielectric with high epsilon are lossier, and delay the
signal more, than lines with low epsilon dielectric.

Cheers,
Tom

On Feb 11, 7:21*pm, K7ITM wrote:
On Feb 11, 5:38*pm, Jim Lux wrote:


In other high power RF applications, low Z can be desirable (because
you'd rather take the IR hit than deal with the high voltage).

In the pulsed nuclear research business you see lines with Z of a few
ohms in pulse forming networks ften using distilled water as the
dielectric.. that epsilon of 80 gets the C right up there so Sqrt(L/C)
is small.

Yes, indeed a low impedance can be helpful sometimes...but the times
I've been involved with low-Z lines, they've almost always been
unbalanced (e.g. coaxial). *For a PFN, the high epsilon is also an
advantage in storing a lot of energy. *Since even very pure water is
pretty conductive (relative to seriously good dielectrics), I'm a bit
surprised it's used.

Both coaxial and parallel plate balanced.. For short pulses, the water
isn't all that conductive (because it takes time for the water
molecules to dissociate); you worry more about particulate or gas
bubble contaminants cause a dielectric breakdown.

For a spectacular example, look at the Z-machine.

On Feb 11, 4:02*pm, phaedrus wrote:
I seem to recall that ladder/open-wire feeders are relatively high
impedance ~500 ohms. Is it possible to construct 50 ohm open wire
feeder and if so, what would the spacing be?

Thanks!

Some simple experimental work he

http://www.karinya.net/g3txq/twin_feed/

73,
Steve G3TXQ

steveeh131047 wrote in news:cd964a18-6932-407b-86fd-
:

http://www.karinya.net/g3txq/twin_feed/

An interesting read.

I think that if you had measured the loss, none of the options tried would
be very attractive. Measuring a short stub with your new AIM4170 will allow
calculation of the effective loss of the line.

I don't understand the detail of the connection neatness thing, but one way
to construct a 50 ohm 'balanced line' line is with four 50 ohm coax lines.
Cut equal lengths, tie all of the shields at each end together, tie pairs
of inner conductors together, and use the now paired inner conductors as a
50 ohm 'balanced line'. Of course, the line is only balanced (in current
terms) by virtue of its usage rather than its construction.

Owen