Towertalk
[Top] [All Lists]

[TowerTalk] Re: Hair Pin Match-Balanced and Unbalance Blunder

To: <towertalk@contesting.com>
Subject: [TowerTalk] Re: Hair Pin Match-Balanced and Unbalance Blunder
From: K7GCO@aol.com (K7GCO@aol.com)
Date: Thu, 2 Dec 1999 16:39:35 EST
In my #21TT, there was some wrong words used I missed at 3 AM in the morning 
when I wrote it.  I added another example to #22TT also.  Delete my number 
21TT from your files.

#22TT  Recently there was a lot of talk about Balun Losses and with 
legitimate concern.  There are "Built In Baluns" by nature of the matching 
system design with no loss for coax. HyGain went out of their way to defeat 
this concept with the Hair Pin Match they used.  I had told them about it.
   Hair-Pin matches were described in WWII Technical Books.  HyGain uses hair 
pin matches but not in the same way of the original design.  It needed a 1:1 
balun on the coax to prevent RF Spill Over on to the shield.  Many were not 
used with the resulting problems RF Spill Over causes.  With WWII Hair Pins, 
the coax entered the center of the hair pin where it was grounded (?) 
connected to the boom.  The coax ran up inside the tubing to the feedpoint 
where the shield terminated at the opening on one side and the center lead 
jumped a short distance across to the other side of the hair pin.  This is 
really a "Compressed T-Match" and no series Xc is needed as the jump lead is 
short.  The balanced 50 ohm feed points are now close to each other.  No 
balun is needed with power limitations, extra cost and with all the 
discussion of Baluns and the losses, consider the following.
     There was an article about 30 years ago in QST I think it was where a 
HyGain Tri-Band Hair Pin was modified to look like the original WWII Hair 
Pin.  The larger coax cable was paralleled with the smaller diameter hair pin 
wire.  I would have thought there would have been an adjustment of spacing or 
length required. The shield connects to one side of the hair pin feed point 
and is taped to the hair pin wire on the left and right side back to the hair 
pin center connection where the shield is grounded (?) there and then on to 
the rig.  Seal coax from water entry at the open feedpoint. Carefully solder 
a grounding strap to the shield, seal the coax at this point and connect to 
the boom.  Another coax is paralleled and taped to the other wire just to 
make it the same diameter and retain symmetry.  Connect the shield at the 
center point on the boom and at the other feedpoint.  If the original 
configuration is maintained, this has just made the same or equivalent hair 
pin right from the coax cable--the WWII model. If just coax is used without 
the original supporting wire like on a hair pin match you have created and 
after adjusted for 1:1, the orientation still must be fixed so it can't move. 
 For test convenience make it from a 1/2 WL of coax with a temporary PL259 on 
it so that a MFJ Analyzer can be used near the feedpoint for test and 
adjustment purposes.  A type N female connector can then be installed 
permanently and then tape the connection as they can still leak. 
    The author claimed low SWR on all three bands of a tri-bander--and no 
balun losses.  I never liked the hair pin match HyGain used.  It was often 
1/8" aluminum wire in the Hi-I (high current) section which didn't radiate. 
When I first saw it, no one ever thought it would match 3 bands with one 
configuration--except hygiene or the author of the modification article.  In 
Eznec however, DE's less than or longer than a 1/2 WL barely affect the 
pattern.  My concerns were unwarranted.  It did require an insulator in the 
center of the DE as extra hardware.
     The 1/4 WL 1:1 Balun that Collins showed in their Transmitter Manuals 
worked great, was only for odd 1/4 WL's, increased the bandwidth and is 
lossless.  I still use it extensively on 50, 144 and 440.    
    The Clements Match of about 35 years ago in QST is a balanced feed system 
for unbalanced coax and used just one series Xc.  I've used several 
variations of it and is great for mono banders.  The author Clements, claimed 
that his wife's sewing machine noise went away on his 10M beam with his match 
which then passed the test.  This indicates a perfect balanced noise bucking 
feed system.  No RF Spill Over on to coax on transmit and no RF Spill Up into 
the feed point of vertically polarized noise or mostly vertically polarized 
signals being picked up on the coax shield.  
    There is a simple trick of a precise balanced adjustment of a T-Match, 
Clements Match and even the Gamma Match which has a lot of RF Spill Over. 
It's more of a "Convenience Match".  Using a MFJ Analyzer, make the 
adjustments for 1:1 SWR.  Then put your hand on the center of the DE.  If the 
SWR changes, the adjustments of the T or Clements Match aren't balanced 
assuming both DE halfs are the same length.  Move one tap or the other 1/2", 
readjust the variable Xc for 1:1 and do the hand test again.  When you get 
the hang of it, it goes fast and Presto--no RF Spill Over!  F/B often 
improves and RFI problems just go away and not many beams can make that test. 
 Fortunately your hand affects to unbalanced RF at the center of the DE and 
is detected by the MFJ. The voltage is supposed to be Zero at the center 
contact point on the boom.  Improper adjustment of a T or Clements can 
unbalance the RF distribution on a balanced DE.  The MFJ Analyzer sensitivity 
makes this needed adjustment possible.  I used to use the Antenna Scope 
Bridge and GD Meter to do the same thing back in the 60's.  Before that I 
shunted light bulbs from the center out each side the same distance and 
checked brightness.  
   The Gamma often has a lot of RF Spill Over.  On 10M years ago, ignition 
noise from cars was bad without resistance wire.  When I used balanced 
systems this was not a problem.  I used a T-Match and balanced 100 ohm coax 
right into a balanced link in the tank circuit and balanced inputs to 
receivers.  No noise.  I learned the value of this in the 30's.
   Now let me tell you how bad the Terrible Gamma really is.  A high power W6 
running 25 KW (that's just the buffer) had a call from a neighbor telling him 
his mast has sparks coming off it--not the antenna.  He had his wife modulate 
the rig and watched.  The arcs were coming from the coax shield to the larger 
mast--1/4 WL below the boom where the DE was attached and using a Convenience 
Gamma.  I said "attached" as there is no RF ground up there of any value 
which is often a misconception.  What was happening is this.  Although the 
coax shield and the larger diameter mast originated at the same connection 
point, the RF Spill Over voltages on each were much higher 1/4 WL down.  Due 
to the differences in the diameters, they were of different levels also. This 
is where the RF is at it's Hi-E or Hi-Z point and at 25 KW the Gamma RF Spill 
Over Voltage level could really Zap you.  I've lit neon lamps 1/4 WL down 
from the beam using the Gamma running a KW.  He added a T Match and the 
Fireworks went away as well as other RFI.  It took 25 KW to finally show how 
bad the Terrible Unbalancing Gamma really is.  It can be balanced by making 
one DE side longer than the other.  
    Would you believe the "RF Spill Over SWR on the tower molded a mechanical 
SWR in the tower that returned straight again with the T-Match." That is an 
absolutely true story--I just made it up!  There was a Jeeves Cartoon in QST 
showing and open wire line with Mechanical Standing Waves on it--my idea.  
FMing of the Oscillator was frequent years ago and I used to tell them it was 
from high SWR.  In the 40's I made up some open wire line with stiff wire and 
bent it into a SWR pattern and told those who asked "it had a high SWR on it, 
it reduced RFI, stopped FMing of the Oscillator, it's also reduced the very 
small losses even more and was virtually a true lossless line."  Some 
believed it.  I've had a lot of experience and fun with open wire line!  
K7GCO  Ken Open Wire Glanzer


--
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


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