[TowerTalk] Windom, Matching and Clements match
K7GCO@aol.com
K7GCO@aol.com
Fri, 28 Apr 2000 15:11:40 EDT
The original Clements Match article was in '51 QST by w9ern--I don't have
the month. Orr has it in his latest books as well as the British
Publications. He cleverly made it from coax cable but no weather sealing was
shown and it's actually not practical. I use a non-metal support T-Match
type tube for the cross over wire. It's a cut and try tune for the tap
points, the length of the bar and the series Xc. I have several practical
versions I will write up when I get time with good pictures.
There is a tune up trick I will tell you that is very handy. The series Xc
is a short length of open coax--whatever length is needed. It was just an
extension of the coax the whole system was made of. This coax laid right on
the DE. The main coax feedline is run parallel to the DE starting at the
center of the DE to a tap point that establishes one of the 2-25 ohm tap
points. The shield connects there. The center lead of the coax runs back
parallel to the DE preferably through and insulated support tube to the other
tap point the same spacing from center. The inductive reactance canceling
series Xc is a short length of open coax with the shield connected to the DE
on the other side. It's run parallel to the DE back toward the DE center.
That's actually all you need but the DE is actually slightly physically
diameter unbalanced by the shorter Xc attachment which is what we don't want.
So Clements added a 1/4 WL of coax as an extension that is shorted mainly to
balance up physically and electrically the attachments to the DE where both
meet and are run away from the DE. If everything is properly placed the main
coax shield and the coax Xc extension doesn't have to be grounded at the
center of the DE. To assure electrical balance stability the coaxes should
be run inside the DE and emerge at the tap points where both shields connect.
This can weaken the DE at this point unless there is a beef up inside. This
takes cut and try to find the exact points initially and you will end up with
a lot of holes in the DE--I know.
Another way is to have a parallel tube the right length secured to the DE at
the 2 tap points and the DE center. The coax goes into the center of this
tubing which can be thin wall 3/4" (or 7/8" with a female coax connector I
make inside and at the end) out to the tap points for the shield. The Xc
coax length goes inside of this tubing also on the other side. The insulated
tubing to hold the coax center lead still has to be supported on top of this
a couple of inches also. Unless good mechanical techniques are used it can
be a bag of worms. But on a specific design where exact lengths and
construction techniques are shown, it's a great system and I will do this.
I've even used it on 2M and 440 quads with RG 58 and model airplane tubing.
The shorted 1/4 WL length of coax extension presents a high Z to the end of
the Xc coax stub and it doesn't see it electrically. But as the antenna is
operated above and below the resonant frequency, it creates the opposite
reactance as does the antenna and broadbands it. So one must determine what
the Xc coax stub length is by cut and try and add that to the shorted 1/4 WL
stub which can be grid dipped precisely. Here's the K7GCO Short Cut Method.
I make this length 1/2 WL long initially and connect it to a metal box in
which I have a variable Xc inside with the shaft connected to the box. I
jockey the tap points and vary the Xc until I get 1:1 SWR using a SWR bridge
like the MFJ Analyzer at the antenna. I then take a pin and start shorting
out the coax progressively to where I get 1:1 SWR. The final length is the
shorted 1/4 WL plus the length that represents the open length and correct Xc
value. I cut off the coax below the 1:1 SWR pin short 3/8" and solder the
center and shield together, seal it and the open other end. This may sound
complicated but is really isn't and its a fast tune up when you get the hang
of it. I also do the K7GCO Hand Test in the middle of the DE when
determining the final tap points. When you get 1:1 and can place your hand
on the center of the DE and no change of SWR occurs, you got a perfect and
balanced match. Few antennas can pass this test. If the SWR rises you have
to check the exact length of the DE. Then move the tap points slightly on
one side or the other while still tuning for and maintaining 1:1 SWR. When
you get it right it's a very cold DE center. Even if the DE is connected to
the boom it can still be hot with "Unbalanced RF" there which flows over the
boom and tends to fill in the pattern nulls. Would you believe it is so RF
Cold your skin will freeze to the center of the DE--RF wise that is!
I actually have a better and simpler way to match a coax to a yagi DE that
gives perfect electrical balance and match and a way to broadband the DE
also. The Clements Match is ideal for quads and will be using it there. You
have to have easy access to the DE for all of these on the tower. Unless you
do and don't want to play these games you are at the mercy of the Mfgs as to
what they use. I talk to them all the time and some are receptive and some
are not. I'll eventually have a web site and be able to show all this.
K7GCO
In a message dated 28.04.00 06:06:42 Pacific Daylight Time, sparks@cptnet.com
writes:<< hello
thanks for your valuable feedback. Interested in the clements match
any photo copy
to Josh Logan
POB 3513
Pagosa Springs CO
81147-3513
any photo copy on matching would be wonderful
there is absolutely no hurry
thank you for sharing the windom information
Josh "sparks" N7XM
-----Original Message-----
From: K7GCO@aol.com <K7GCO@aol.com>
To: Towertalk@contesting.com <Towertalk@contesting.com>; jreisert@jlc.net
<jreisert@jlc.net>; w7wjp@msn.com <w7wjp@msn.com>; w6kh@best.com
<w6kh@best.com>
Date: Thursday, April 27, 2000 5:09 AM
Subject: [TowerTalk] (no subject)
>
>In a message dated 25.04.00 17:01:51 Pacific Daylight Time,
jreisert@jlc.net
>writes:<< Sparks,
>
> The ORIGINAL Windom was a half wave dipole fed at 1/3 from one end with a
> single wire dropped to the transmitter (using the return through a
ground).
> Later, the single wire feedline was replaced with, I think, a 300 Ohm piece
> of twinlead. When I was first licensed, the Windom with twin lead and a 4:1
Balun using two special bifilar wound coils made by B and W. 73, Joe,
W1JR
>
> At 10:35 AM 4/25/00 -0600, sparks wrote:
>
> >I have been traveling, no antenna books. How is standard WINDOM antenna
>fed? at 1/3 point ??-1/3 length, feedpoint, 2/3 length?Is that correct?
>Sparks
> >
>******The first article describing the 300 ohm feed for the Windom was by
>WzeroWO Karl Dreher in a ''52 or 53 Radio Electronics Mag. I was in the
Armyin Denver and met him. I have the original article in my files which are
>packed. What he did was this. He graphed the current loops on log graph
>paper as I remember for 80, 40, 20 & 10M wire antennas using a length of
>about 136' plotted against R values of 72 in the center of the current
loops and 3600 ohms for the ends of each current loop. The current loops
for all
4 bands crossed at one point 1/3 from the end and at a R of 276 ohms which
was close enough for 300 ohm line.
>
>The currents are unbalanced in the feedline, are other than 180 degrees out
>of phase and it radiates some. It was hard to match at the end of some
>feedline lengths and making it longer would usually match the link coupling
>and series Xc of the day and sometimes Pie networks which were
somewhat new then. It's been 47 years since I read the article, I will
try and
find it and see how close I came to remembering the details. If I had some
log graph paper here I think I could duplicate his graph as I remember
the
concept he used. I called him in about '70 on my way through Denver and
he had passed away. I made a practice to meet a lot of the antenna
authors
of the day in my travels, wrote them or called them on the phone. He
said
"I was the only person to respond to his article." I though it was a very
clever way he came up with the common 276 ohm feedpoint for all 4 bands
that occurs at the same point 1/3 from the antenna end, I've never seen
one reference to him for his original article when the 300 ohm Windom
was discussed. Very sad.
>
>One of the few I never met was the Author of the Clements Match in QST
about 39 years ago. I have Clements article around here some place also.
Few ever referenced his article which is the best matching system of all
for
a single coax. The English Publications show it. I have used it
extensively
as it's a way to get absolutely balanced feed with coax. He commented that
when he installed the new match to his 10M beam, his wife's sewing machine
noise went away. It was that noise that inspired him to come up with a
new
way to feed a beam that didn't cause "RF Spill Over" on transmit and have
"RF Spill Up" of mostly vertically polarized noise RF picked up by the
coax
feedline shield on receive and fed right in the coax opening at the top.
Donut toroids are needed at the top for RF--going both directions. Few
recognized the significance of this advantage. The gamma is the worst
affender of this.
>
>The Clements Match had another outstanding feature. It gave more bandwidth
>than a gamma also due to the use of a shorted 1/4 wave coaxial stub
connected to the series Xc used which was a length of coax also. Some
smart mfg. should add this to their beams and use the proper fine tuning
procedure to assure a perfectly balanced feed with no balun. Clements ranks
in the "10 K7GCO All Time Best Designers and Honor Roll." Only 2 others in
print have recognized the sins of the "Dreaded RF Spill Over." Bailey in a
Radio Mag article in the late 30's and Remington Rand in a '52 QST article
on a 2M sleeved vertical addressed it and took steps to reduce or eliminate
it.
>
>I had a 6M AEA Loop fed with a gamma and no series Xc. The SWR varied all
>over the place as the loop shape was bent or the gamma wire moved slightly.
>Adding the series XC helped greatly but the coax shield and mast were still
>hot with RF Spill Over and Spill Up which deteriorated it's performance
>greatly. I added the Clements Match and by adjusting the 2 tap points and
>the series Xc I got 1:1 SWR and absolutely no RF Spill Over or Spill Up on
>the mast or shield. I could place my hand on the mechanical center of the
>loop and mast connection and there would be no change in SWR. When I get
>time I'll write this up for some mag. I use this hand affect balance test
onall my antennas.
>
>One way to visualize the Clements Match is to visualize the Hair Pin Match
>pictured in WWII Radio Handbooks. Then physically pull the ends of the
>dipole out. The coax center lead cross over wire extends between the 2 tap
>points now that gives a gamma type 25 ohms on each side of center. This
>longer crossover wire now has inductive reactance like a gamma and a series
>Xc is needed to cancel it and leave a resistive 50 ohms. It's a T Match for
>unbalanced coax. I have a variety of ways to mechanically and electrically
>create the Clements Match and the broadbanding feature that are what's
called a Robust Design. Clements design worked but needed refinement.
>
>Clements a W9 as I remember like Carl Dreher WZeroWO never got the
>recognition they should have. Perhaps someone can look up the Clements
Match and inform TT what issue it is in so it can be reviewed. I plan an
article
>with the variations I have on it --when I get time. Perhaps someone can
look the 300 Ohm Windom article in Radio Electronics in about '52 or '53
and
>inform TT of it. Some Libraries should have it. K7GCO
> >>
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search.htm