[TowerTalk] TowerTalk Digest, Vol 33, Issue 32

[Dudley Chapman]

Ian wrote.......

Message: 1
Date: Fri, 9 Sep 2005 15:35:22 +0100
From: Ian White G/GM3SEK <gm3sek at ifwtech.co.uk>
Subject: [TowerTalk] Reduced size four-squares?
To: towertalk at contesting.com
Message-ID: <1TDF5HVq2ZIDFAR4 at ifwtech.co.uk>
Content-Type: text/plain;charset=iso-8859-1;format=flowed



At 01:35 PM 9/1/2005, Vojtech Ritzko wrote:

 >>I am planning 4SQ for 80M. I have problem with space, due to property
limitation.
 >>Does anybody compared:
 >>
 >>4SQ with ?/4 spacing, but only ?/8 long radials  >>  >>Versus  >>  >>4SQ
with ?/8 spacing with ?/4 long radials  >>  >>TNX  >>  >>OM8AW

Jim Lux replied:
 >I'll give you a theoretical take.  Others will no doubt give you  >some
practical experience.
 >
 >Closer spacing makes for higher interelement coupling, the  >feedpoint
impedances will probably be lower, so the IR losses  >(assuming everything
else is identical) will be higher. The resonance  >will be higher Q, and the
coupling changes faster with frequency, so  >it will be harder to get good
phasing. Some depends on: would you  >use lambda/8 phasing or lambda/4
phasing (or some variant with  >a bit more phase shift than the physical
distance (like W8JIs  >scheme), which increases the forward gain a bit).
[...]
 >On the other hand, the length of the radials (assuming you're  >planning
on putting them on the ground) isn't really frequency  >dependent [...]
>Lots of wires close to the base is probably more important than  >few wires
farther out.
 >
 >If I had the choice, I'd go for 1/4 wave spacing and short radials.
 >Easier tuning and operation, mostly.

...and there the topic ended.

Could I revive it with a slightly different question, please?

I'm planning a 40m 4-square on a lot that isn't quite big enough. 
Specifically, there isn't enough space for both 0.25wl element spacing
*and* 0.25wl radials all around. The practical options are either:

0.25wl spacing with radial lengths for two of the elements limited  to
0.18wl in some directions; or

0.20wl spacing with 0.25wl radials all around.

Which to choose? Very soon it's going to be decision time.

Here are my thoughts so far, but this is my first 4sq so it's all
theoretical. What I need is some practical input!

The conventional 0.25wl spaced option is easy to feed with 90deg phasing, so
it would get me on the air quickly. But the radial layout would be shortened
and non-symmetrical, and that might make it harder to obtain good nulls
(good rear and side nulls are very important here in G-land). Alternatively,
I could shorten all the radials to the same length of 0.18wl to keep the
layout symmetrical... but then the ground losses would increase.

The closer-spaced 0.20wl option can be done, but it requires a special feed
network (eg a Lewallen-Lahlum setup as described recently by ON4UN).  This
non-standard element spacing will require "custom" element currents and
phasing, which can also be optimized to give almost the same performance as
a full-sized 4sq with the basic 90deg feed. "We have the technology" to do
all that (I hope) but it would be a challenge, and would obviously take much
longer to get on the air.


All suggestions welcomed...


--
73 from Ian G/GM3SEK

..................................................

Ian,
   Same answer as before.  The 0.25 wl spacing with 0.18 radials all around
is clearly superior to your alternative.  Using 0.18 wl radials is a very
small compromise such that you would probably not be able to measure the
difference between that and 0.20 wl radials.   It is not a linear
relationship.  Since the current density in the ground, going radially
outwards from the base of the vertical element is very similar to the
current density on the vertical element itself.  As you go outwards, the
current drops sinusoidally to zero at 0.25 wl.  Since the current is already
low at 0.18 wl, extending the radials to 0.20 wl buys you very little.  

There is a rule of thumb that says for a given length of radials, you reach
a point of diminishing returns when the tip to tip chordal distance between
adjacent radials is 0.025 wavelengths.  Adding more radials beyond that buys
you very little.

Dudley - WA1X