In article . com,
wrote:

I've experimented with numerous low noise antenna receiving
configurations at this QTH and finally settled on a 90 foot "Windom"
configuration where the antenna is fed with 50 ohm coax at the 35%
length point with a 6:1 impedance transformer. The high-Z winding is 10
turns on a ferrite toroid and the coax is connected to 4 turns on the
toroid. There are no earth grounds and because of the separate windings
the antenna is galvanically isolated from the coax. The wire is
suspended about 6 inches above my west concrete wall which is about 5
feet high. This antenna has be the most effective antenna with minimal
noise for the 2 to 30 MHz bands.


As an example if your wire antenna is #18 wire about 10 feet off the
ground then the impedance of the wire is around 500 to 600 ohms. You
could probably use a higher ratio transformer like 11:1 instead of ~ 6:1.

I would use 13 turns to 4 turns instead of 10 to 4 turns.

--
Telamon
Ventura, California

As an example if your wire antenna is #18 wire about 10 feet off the
ground then the impedance of the wire is around 500 to 600 ohms. You
could probably use a higher ratio transformer like 11:1 instead of ~
6:1.

I would use 13 turns to 4 turns instead of 10 to 4 turns.

This will vary to freq though. I think quibbling over the transformer
ratio, or actually, even using a matching transformer at all, is
usually overkill. Often a 1:1 will be better on some bands.

I modeled his antenna just now to see what happens per band
changes. I fixed the feed at 33% offset, as that was about
the 1/3 point across the wire, and should be close to his 35%
as far as the example.

Band-------Impedance-------SWR if using 50 ohm coax to feed.

2 mhz , 0.0894 -J 1762.870, 100-1
5 mhz , 6.2812 -J 140.548, 70-1
10 mhz, 23.7054 -J 248.378, 54-1
15 mhz, 4523.4000 +J 3621.610, 100-1
20 mhz, 111.780 -J 461.406, 40-1
25 mhz, 87.0032 -J 302.004, 23-1
30 mhz, 1584.920 -J 1004.170 , 44-1

The feed impedances are all over the map just
depending what band you are on. Trying to get
the "perfect" transformer ratio is generally a waste
of time. Not to mention that adding the transformer
will rarely increase the actual s/n ratio. On some of
those bands, a 1:1 would actually be better... His
case is a good example of good decoupling. No ground
was used either. I love it...:/ It's not the use of a transformer
that makes the antenna good, or even gives any improvement
in actual s/n ratio. It's the decoupling. He could use a 1:1,
and it would probably work just as well in general.
I could give the numbers concerning coax feedline loss vs freq
vs length, to prove my point , but I'd have to recalculate all the
numbers.
I did that once, and posted it here, but I'd have to google it
up to find it. That was 2-3 years ago... A transformer is
rarely needed for any kind of decent radio. The feedline
loss is just not enough to actually degrade the s/n ratio.
Only in rare cases of totally lousy, thin, waterlogged coax
that was 1000 ft long might this be the case. Transformers
are nifty devices, but for general HF use, all they do is
pump up the S meter. On the lower HF bands, coax loss
is so low as to be practically ignored as part of the equation
when receiving with a 90 ft wire. Ditto for 30 mc, with any
average setup. The loss is greater at the upper end of the band,
but it's still rarely large enough to reduce the s/n ratio.
MK

In article .com,
wrote:

As an example if your wire antenna is #18 wire about 10 feet off the
ground then the impedance of the wire is around 500 to 600 ohms. You
could probably use a higher ratio transformer like 11:1 instead of ~
6:1.

I would use 13 turns to 4 turns instead of 10 to 4 turns.

This will vary to freq though. I think quibbling over the transformer
ratio, or actually, even using a matching transformer at all, is
usually overkill. Often a 1:1 will be better on some bands.

I modeled his antenna just now to see what happens per band
changes. I fixed the feed at 33% offset, as that was about
the 1/3 point across the wire, and should be close to his 35%
as far as the example.

Band-------Impedance-------SWR if using 50 ohm coax to feed.

2 mhz , 0.0894 -J 1762.870, 100-1
5 mhz , 6.2812 -J 140.548, 70-1
10 mhz, 23.7054 -J 248.378, 54-1
15 mhz, 4523.4000 +J 3621.610, 100-1
20 mhz, 111.780 -J 461.406, 40-1
25 mhz, 87.0032 -J 302.004, 23-1
30 mhz, 1584.920 -J 1004.170 , 44-1

The feed impedances are all over the map just depending what band you
are on. Trying to get the "perfect" transformer ratio is generally a
waste of time. Not to mention that adding the transformer will
rarely increase the actual s/n ratio. On some of those bands, a 1:1
would actually be better... His case is a good example of good
decoupling. No ground was used either. I love it...:/ It's not the
use of a transformer that makes the antenna good, or even gives any
improvement in actual s/n ratio. It's the decoupling. He could use a
1:1, and it would probably work just as well in general. I could give
the numbers concerning coax feedline loss vs freq vs length, to prove
my point , but I'd have to recalculate all the numbers. I did that
once, and posted it here, but I'd have to google it up to find it.
That was 2-3 years ago... A transformer is rarely needed for any kind
of decent radio. The feedline loss is just not enough to actually
degrade the s/n ratio. Only in rare cases of totally lousy, thin,
waterlogged coax that was 1000 ft long might this be the case.
Transformers are nifty devices, but for general HF use, all they do
is pump up the S meter. On the lower HF bands, coax loss is so low
as to be practically ignored as part of the equation when receiving
with a 90 ft wire. Ditto for 30 mc, with any average setup. The loss
is greater at the upper end of the band, but it's still rarely large
enough to reduce the s/n ratio.


There maybe no perfect transformer but you might as well shoot for the
right impedance target if you are building it.

--
Telamon
Ventura, California