Topband
[Top] [All Lists]

TopBand: 160m shunt feeds

To: <topband@contesting.com>
Subject: TopBand: 160m shunt feeds
From: w8jitom@postoffice.worldnet.att.net (w8jitom@postoffice.worldnet.att.net)
Date: Thu, 31 Jul 1997 09:44:11 +0000
> From:          km1h@juno.com (km1h @ juno.com)
To: <topband@contesting.com>
> Date:          Wed, 30 Jul 97 17:24:33 +0000

Hi Carl,

> Perhaps I used the wrong word but we mostly agree here Tom. My concern
> was indeed with bandwidth and the High Q, narrow bandwidth, high voltage
> that a single #12 would bring. The #12 was my first attempt back in 1984.

The inductive reactance of an additional  #12 wire fifty or a hundred 
feet high can be pretty high, requiring a large amount of capacitive 
reactance ( a low value C) to tune it out. Since it is a series 
resonant circuit, the lower the C the higher the operating Q and the 
higher the loss.

> 12-16" gave the best FS with both the #12 and the .75" .  The FS dropped
> rapidly over 24"; this was possibly due to the FS site being directly
> behind the tower/feed sight line and the feed beginning to actually
 radiate.  

KE3EG reported something similar, a pattern skewing rather than an 
actual loss. I've never measured FS, since I always use four drop 
wires spaced every 90 degrees.

You could also have  run into both dielectic losses AND  "current 
bunching"  when the wire moves close to the tower. Modeling programs 
are pretty good in most normal antenna, but  Eznec (and to the best 
of my knowledge NEC itself) does NOT consider current bunching.

Current bunching occurs because strong induction fields force 
current to one side of the tower. Current bunching reduces the 
effective conductor area and increases the actual loss resistance. 
This effect can be very pronounced if current is high and spacing 
close.

This effect is VERY evident in multi-turn loop antennas, and even 
the common inductors we used in RF tank systems. In every case it 
lowers Q by increasing loss resistance. This is one reason why 
inductors and multi-turn loops always have lower Q than predicted 
by considering a conductor with uniform current across the entire 
surface. I see no reason why it wouldn't be an effect in a shunt fed 
tower, but have no idea how large the effect would be.

> This could be the beginning of another topic, but what is the ideal # of
> elevated radials? Use 15-20' elevation as an example. The original
> article in QST said one to four.  I started with 4 and now am up to 12
> based on comments from K1ZM and others. 

WVNJ radio used six elevated radials at 35 feet on 1160 kHz, and 
when we pulled in 50 or 60 temporary radials on the ground....with 
no other changes.... their average non-directional field strength 
went up about five dB. 

When I did a test on 3.7 MHz with a 1/4 vertical, the difference 
between 4 elevated radials (6-8 ft) and 60 radials (either on the 
ground or elevated) was about four or five dB.  N7CL has posted that 
the military measured systems and got the same results. I think he 
reported about 5 to six dB.

On a positive note, there was an IEEE report that indicated six 1/4 
drooping radials 1/4 above earth was only slightly weaker than a 
conventional system using the same overall height structure, so if 
you can get six radials up 130 ft you'd certainly be in business.
 
Five dB isn't "a lot", that amount is easily lost or gained in day to 
day propagation changes or made up for by having a superior 
location. So maybe it isn't worth worrying about.

> >That does not give an accurate measurement of self-resonant 
> >frequency, because the gamma wire forms a stub with the 
> >tower, and acts like a transmission line.
> 
> Only when there are no radials was my experience. In a folded monopole
> the only result is a step up in impedance which made loose GDO coupling
> simpler.  The radials removed the stub effect .

My concern is the  gamma wire has inductive reactance, and certainly 
perturbs the system. I can't remove my radials here, but I do know 
when I dip the gamma wire it indicates the tower is much lower in 
resonance than it really is.  I also can't understand why you'd 
remove the radials, since they help determine the resonant frequency 
of the structure.

We used gamma wires to de-tune power lines near broadcast 
stations, if the pattern was distorted by a vertical ground wire 
or metal pole. The solution was to run a "gamma wire" up the side 
and fan it out and back in. By finding the right tap point a 
pole could be de-tuned with just a simple wire, although some cases 
required a series gamma capacitor.

I am very suspicious of a gamma match dip telling you anything 
useful about the structure's self-resonant frequency. Consider this 
point.....how many 20 meter beams do you see built WITHOUT a series 
C in the gamma? The only ones you'll find are those who's driven 
element is, without the gamma, resonant far above 20 meters! In 
those cases the gamma will "dip" on 20 meters, while the element is 
too short to be on 20.

> What did "ground" consist of ?

About 110 or so radials that vary between 125 and 250 feet long 
on that tower, but it's also bonded into other radial systems, and I 
just got another 15000 feet of wire for a new tower to add a new 
160 element toward Eu.

I need all the help I can get to multi-hop signals up that lousy 
Appalacian range towards Eu.

73, Tom W8JI 

--
FAQ on WWW:               http://www.contesting.com/topband.html
Submissions:              topband@contesting.com
Administrative requests:  topband-REQUEST@contesting.com
Problems:                 owner-topband@contesting.com

<Prev in Thread] Current Thread [Next in Thread>