[Top] [All Lists]

[TowerTalk] 20 Meter

To: <>
Subject: [TowerTalk] 20 Meter
From: (Tom Rauch)
Date: Mon, 5 Mar 2001 19:45:28 -0500
> unproper linear loading arrangements. Back to the matching circuits,
> they may introduce losses when some of their elements have losses,
> tipically a gamma capacitor with a lossy dielectric @ 28 MHz is one of
> the the worst cases for HF, or when the mechanical construction (i.e.

I'm not aware of any gamma capacitors that employ lossy 
dielectrics, mostly because lossy systems would be self-
eliminating at the power levels we run. Most dielectrics are 
polyethylene, which is one of the lowest loss materials available. 
As in a transmission line at HF, dielectric losses are not a factor.

The gamma match does not carry significant current, no matter 
what impedance element it is matching. It does not matter if the 
element's loop radiation resistance is  5 ohms or 45 ohms, the 
gamma rod sees 50 ohms. 

Current in the gamma capacitor is never more than the current in 
the 50 ohm line feeding the antenna, and voltage across the 
gamma capacitor is the reactance times the 50 ohm transmission 
line current...which has nothing to do with loop radiation resistance 
of the element. 

> with an hairpin) involves a number of different parts with several
> joints and the impedances in those points is already rather low.
> Although nothing fails and noone notices the heath, or neither a
> signal decrease is roughly appreciable, when discussing about marginal
> gains like 0.25 or 0.5 dB (a 5% or 10 % of the power) already a simple
> 0.5 Ohm series resistance in a 20 OHm radiator adds an extra .22 dB
> loss.

While it is true the current is maximum in the hairpin, if we lost 5 or 
10% of the power in a joint or a few joints in the antenna we would 
certainly have problems. Imagine 150 watts of heat distributed in 
perhaps two connections, and picture how hot the tiny surface area 
of those connections would become.

My L-network matching the common point of a phased array has a 
feed impedance of approximately 20 ohms, and it uses only a ten 
turn #8 AWG inductor bolted to a capacitor. The losses in that L 
network, which are substantially more than I expect from a 
reasonable size hairpin conductor, produce less than 30 watts of 
heat out of 1500 watts applied.

Most of the joints in a traditional hairpin are at lower current areas 
of the a hairpin, at the  50 ohm point of the system. Most hairpins 
of good design do NOT interrupt the conductor at the highest 
current area.

It makes no sense to redesign an antenna to raise the feedpoint 
impedance to 50 ohms and give up performance, when any small 
losses in matching could easily be handled through simple proper 
construction techniques.

I'm reminded of Lew McCoy's silly article a few years ago in CQ 
Magazine about terrible loss in yagi elements, and his suggestion 
that a #16 wire element has less loss than 1"diameter aluminum 
tubing because the wire "has no lossy joints".

With ten ohms, we only deal with 12.5 amperes. These are not 
small loop antennas with radiation resistances of .05 ohms and 
hundreds of amperes of current, where pressure joints do become 
a mensurable problem.

73, Tom W8JI 

FAQ on WWW:     
Administrative requests:

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