[TowerTalk] (no subject)

[K7GCO@aol.com]

In a message dated 8/13/00 5:03:23 PM Pacific Daylight Time, w7ti@jps.net 
writes:
<< 
 I'm installing a triband quad on my tower, and I'm wondering how far
 above the tower the quad should be to have little or no interaction
 with the tower.  I modeled it with EZNEC, using a 18 inch diameter,
 80-foot long "wire" projecting from ground halfway up inside the quad,
 as if the center of the quad was even with the top of the tower.
 EZNEC showed essentially no interaction at all.  
 
 Being naturally a suspicious type, this seems too good to be true.  It
 means I don't have to put the quad up on a long mast.
 
 Any comments?
 73, Bill W7TI
  >>
I've found about the same thing except for vertical polarization. I did find 
a way to reduce (only) the severe coupling with the equivalent of a 
fiberglass mast and decoupling stubs (trick sticks).  It also depends how 
resonant the tower is, if it's grounded or not.  I had a metal boom that I 
replaced and for certain tower lengths that was beneficial.

The horizontal quad has the advantage of the feedline attaching to the bottom 
of the quad most out of it's direct field but I will still be using a 
fiberglass masts.  If you maximize the current loops in Eznec as seen on the 
screen, you can still see RF absorption the to tower, metal mast and boom 
that goes away with a non metal mast and boom.  Even the boom down the middle 
picks up some RF depending also on the length of the mast and tower connected 
to it.  That's another reason I like tapered boomless quads--another variable 
eliminated.  Lay some Florissant lamps on the tower, mast and give it a KW at 
night and go take a look.  Be prepared for some surprises.

On a multiband yagi with or without traps where they have elements grounded 
to the boom, there is evidence that although each off band element is mostly 
off resonance, it can selectively connect to other elements with certain boom 
spacings and connections near resonance.  The solution I found was to 
insulate the elements from the boom.  When yagi joints go bad in an 
unbalanced manner and are connected to the boom, the boom gets a strong dose 
of RF also and that will up set F/B more than gain usually.  Both of these 
concepts can be proven in Eznec also for the doubters.  Having a DE slightly 
unbalanced (like many are) will excite the coax shield, boom and mast also.  
With a 1/2 WL of coax connected to the feedpoint of a yagi DE grounded to the 
boom like with a T or gamma, do the "Magic K7GCO Handy Hand Test Instrument 
Test" with a MFJ Analyzer.  Just touch your hand to the center of the DE and 
see if the SWR changes.  If it does, change the tip lengths of the DE until 
you have 1:1 SWR along with the matching section and no change of SWR when 
you place your hand on the center of the DE--connected to the boom or not.  
Would you believe when you "get it right or perfectly balanced and none of 
the dreaded RF spill over present," give it a 5 minute 1 KW carrier test late 
at night and frost will appear.  It's a reverse heat process phenomena seldom 
seen almost on a par with the anti gravity pulpulsion systems of flying 
saucers. The maximum potential F/B ratios will be seen for the first time 
without the RF running down the mast and tower and certain nosies won't be 
heard for the first time.  The somewhat misleading ideal concept that occurs 
in software is the RF source used in the DE has no feedline for potential RF 
Spill over.  It's an "Ideal World RF Supply" that doesn't really exist unless 
the best of feed systems are used.  The 2nd and 3rd best feed system I've 
used is open wire line and 100 ohm balanced coax.  The 1st best is installing 
the transmitter right at the DE with the proper techniques.  An all band 
xmitter there that is automatically controlled by the frequency of the RF 
sent to it can really do a great job if far away and the driver has enough 
power to over come the long feedline losses to the final.  This is a great 
idea for contesters who would like to recover the feedline losses on xmit.  
Those who do this don't like to talk about it unless you can trade an equal 
idea back to them.  k7gco


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