For/From OZ - the magazin for EDR, Danish Radioamateurs association!
Four Element antenna for 40-meters
- without rotor!
Or the story of how a logbook analysis turns into an antenna project.
OZ1IKY, Kenneth Hemstedt, member of Copenhagen Contest
Club (OZ7A)
If you think you will find a full finished description of
how to build this antenna in this article - so do not read on, and hurry on to
something else.
Background
for the project, what made it all happen.
At one point just after both Scandinavian Activity Contest and the CQ WW SSB
Contest, I came to look at our results. Not only the
overall result, but the individual bands and their contribution to the Total
Score. Overall, had we run lots of QSO'er in
virtually every band we could activate as a Multi-2 station. Even 10 - and
15-meter had gifted us with openings - and thus some score. So which band could
we be pulling more out of?
For a detailed review of the 40 meter band came up
something that gave cause for some reflection. There was plenty of QSO's.
Indeed, as many as 20 and 80
meter bands. But point-numbers where more similar to the 80 meter band.
What was this all about? I made a comparison
with the log from when we worked as OZ1HQ on 40 meters SSB. AHA!
The difference lay in the overwhelmingly in the
number of a point QSO's!
There were some thoughts and a lot of discussions in the
EDR Amager section. We
landed almost every time on the antenna, because the output power we had.
But how do you get a better antenna, with
directionality, when you have rooms at a school in Tårnby and not at all can
make a big 2 - or 3-element beam for 40 meters? We have not as much space as our
colleagues in Finland
... (OH8X). Can we learn something from elsewhere? In
a normal QSO our dipole, which is a part of our Mosley beam, is very good.
But with the strong east and southeast
European contest stations, it is difficult to hear a 'small' North or South
American, or just think of an Asian or South African signal, right?
Then
it was, I came to think back on both my start on HF back in the 1980s when I
made a couple of slooping dipoles for 20 - and 15-meters to run North America,
and my time in Holstebro OZ9HBO where there was the usual friendly competition
between local branches around the HF field day. What was that Skive OZ7SKV (1)
had made for their 40 meter set-up?
[OZ7SKV 40 mtr Ant.]
A
mail to Martin, OZ3MC. Fast feedback, I had remembered correctly - thanks
Martin! It was something with four dipoles and
the coax cables should be in a specific length, he just remembered - and of
course a switchbox in the middle. But I was missing just a little on the
theoretical and practical level here before I would throw myself into
constructing anything. How about the specific cable length? How much
front-back has it? Why does it work at all? Have others made something like
this; I remember some articles somewhere.
ON4UN's book was turned apart (2).
Yes, who however only had space and opportunity
to permanently put a four-square array for 80-meters! But definitely a book
worth reading must say, even if you do not have just the space. There is
inspiration for a couple of other exciting antenna projects.
I nevertheless found a clue in the right
direction for an old ARRL antenna book. Truly, in ARRL's Antenna Handbook
edition 19 from 2000 (3) I found just what I was looking for. Later, several
more articles by searching online came along.
You
take a bunch of relays, some antenna wire and some coaxial cable, all in the
right lengths - and just put it up?
The description in ARRL's Antenna Handbook is as such reasonable simple. There
is not the major theoretical rhetoric, but a nice practical approach. It is
originally K1WA, David, who has tried it in a solution with "slooping dipoles".
His version is even made to 40-meters as a starting
point. You need a tower about 20-22 meter high,
a switchbox with a couple of relays in - a bit of wire and coax.
The displayed pattern is
shown with "open" terminals for coaxial cables - both inner conductor and
shield. The box I bought is not of the sort, so we will see the results are the
same.
K1WA, David has tried different setups with both three,
four and five dipoles. He believes that
five dipoles are the best, but there is no big difference between four and five
dipoles, from what he originally tested.
[K1WA’s system from the ARRL handbook]
Each
dipole has a piece of coaxial cable at 36 feet, or 10.5 to 11 meters - or
equivalent to 3 / 8 wavelength; please remember the shortening factor! The cable
provides an extension of the dipoles of about 5 percent and so the dipoles
behave as a reflector. And three or four reflectors
and the tower itself impacts on the system, it should be able to give a 4 dB
gain to the front and a front-back of about 15 to 20 dB. It was K1WA David's
message at least. Our focus is not so much the front gain, but to create a
better attenuation to the rear, as we have on our dipole for 22 meters.
I'm still not a master in
all the fine features of EZNEC; instead I simulated it with a dipole and three
reflectors, which are 5% longer than dipoles, and a tower. EZNEC agreed on
front-gain, but not entirely agreed with the front-back attenuation. But if we
do place them in the northeast, southeast, southwest and northwest, you would
get the desired effect in the desired direction. And since it can not be done
with a pure south-east-north-west anyway at our QTH it's ok. West dipole would
stand in the middle of the football field of the school ...
Indeed, the simulation showed that a version with a
dipole and a reflector, and mast, gave an almost heart-shaped charm, as in a few
phase ground planes.
[EZNEC radiation view of the antenna system]
While EZNEC reveals that there may be something funny about impedance, it looks as if it lies between 75 and 100 ohms. There was also a once in a while a few funny conclusions about the dipoles lengths. The bandwidth of the antenna shows as the whole 200 KHz of the 40 meter band. But let the practical test show how we end up.
[EZNEC SWR-pattern of the system]
[EZNEC view of the antenna system, including tower]
The
theoretical parameters that I get out of the program, in fact, goes well with
what K8UR, Dennis (4) has arrived at in his calculations.
The article I found online, through a search in fact exactly this topic!
K8UR goes as further and sets a Phasing-box on.
It was initially something I didn’t quite believed in - right until
someone told me that this indeed is what OY6FRA is running with!
A further article, which
confirms many of these things - and actually have something more precise about
K1WA's brief statement about 3 / 8 wavelength, is OE5CWL, Jürgen’s (5) article
which actually affects all this straight in the eye. One of OE5CWL's main points
is exactly that with 3 / 8 wave length, as the cornerstone of the whole
structure.
The length of coaxial cable: l = 0,375 x lambda x k
0375 = 3 / 8 wavelength
l = physical length of
coaxial cable
lambda = wavelength
k = velocity of the coax / shortening factor
Their observations back
in 1988, was the piece of a 3 / 8 wavelength, acts inductively on the unused
dipole(s), and thus moves the resonance frequency down by 5%. That is, in this
case we move it from around 7.1 MHz to about 6.7 MHz. He also uses only four
slooping dipoles, as we have planned here. The article may be of interest to
those who might engage in something with an 80 meter version for their next
field-day. OE5CWL gave a few tips for this because his version is an
abridged version, so you do not need a huge tower standing in the open Danish
landscape. He writes something on his wish by
email send further details. I have not tested this statement yet.
A somewhat more in-depth articles in the more theoretical type for this type of antenna has N6LF, Rudy (6) (7), committed. In these two articles he looks at the difference between K1WA's and K8UR's setups. He has actually tried both design types, and also looks closely at the "Spitfire" design. One thing I did notice in the articles was the adjustment/impedance problems which I subsequently stumbled on (see "On-The-Air"). W2VJN is run into the same experience with impedance, which was significantly below 50 ohms, and that the best SWR therefore was 1,8:1. Simple adjustments can solve the problem very simply. He also comes around at the front-back problems discussed, and optimization as I should have looked at retrospectively. These two articles provide, as I said a lot more inspiration - primarily for the low bands (80 and 160 meters).
The good thing about these descriptions is that if you like them, they can easily scale to another band without too much mathematical calculation exercises! Why not try it at either 20 - or 15-meters?
The practical part of the exercise
Since Flemming, 5P5N / OZ4P
had to go to the USA, he became strongly convinced that it was in both his and
the club's best interest that there was room in his luggage for a switchbox.
The advantage of this box is that - besides
being cheap – the power is fed through the antenna cable. No extra wires that
flutters around the tower.
[The Ameritron switchboks]
It is set to cope with a 2.5 KW
PEP. We have obviously not had it opened, and glanced at the relays which are 12
VDC, and an 8 amps 24/240VAC and looks like a standard relay of some sort. So if
we had finally 'fired' one of them off, they are relatively easy to replace with
a 'standard' relay with a few more amps from one of the major 'catalogue
distributors' of electronics stuff in Denmark.
One can usefully look at a few sites where the amateur
has dealt with a little something. It's actually easier than you think. K5LXP
(8) have made a copy of the Array Solutions large KW switch two times six for
the use of a 100 watt station – but actually it looks like it could take a bit
more output power. EI7BA (9) has a bit of something bigger. His version, he
emailed back to me, is tested in a contest station with well over about a 1.5 KW
... - whatever that meant, but I’m confident that it can take the heat.
He uses 24-volt relays, as they
were in his ‘junk-box’, but I've seen the same type of RS in a 12-volt version.
The actual dipoles, is
four equal dipoles, with resonance at 40-meter band. Since a dipole without
traps and other shortening factors have a reasonable bandwidth, I decided that
they should cut to the middle of the band so we could enjoy both to run CW and
SSB contest at the same antenna without having to adjust anything.
A few small parts from the local DIY store to pull it all
up and down with pulleys, wire lock, etc., is probably a good idea. Think of it
as a double flagpole, where you have to pull up 2 times 2 flags! This did indeed
make life easier in the setup phase, where you can not avoid having to pull it
up and down a few times until everything is in perfect shape. Especially when
one end is in 18 to 20 meters and it is not always so exciting weather in
Denmark that you always want to
climb the tower. Because I have a feeling that Mr. Murphy would love to look
just in when it is most inappropriate.
The originally purchased copper wire could not handle the
weight and pull. So I
quickly procured a 100 meters of steel antenna wire instead.
It keeps quite well so far and is not so
terribly expensive. And if you buy the right one, you can actually solder on it.
[4 dipoles for 40 meter]
The
length of each of the wires; l = 71.25 / f
l = length in meters
71.25 = 1 / 4 wavelength
f = frequency in MHz - in this case around 7.1 MHz
A total of eight pieces of wire at 10.05 meters - and
just a bit to adjust to, and low clip points buckle for wire locks etc. You
never know what happens to the resonance and lengths when it comes close to the
tower. EZNEC showed indeed some funny little results.
Feeder dipole was indeed below 20 meters in total length.
We had a man in the tower
to fix a pair of pulleys and we got pulled a good string through, to pull it all
up and down. One must, of course, despite good theoretical preparation and
calculation could predict that it should be up and down a few times before it is
fully adjusted to where it is supposed to be. Here, after it went really quickly
in getting things up in the air.
[A man in the tower …]
I got lucky enough that I only should have it up and down just once to adjust the length of the dipoles. True enough - resonance was at 6.25 MHz. A little backwards expense; 1.2 meters from each end of all four dipoles. Up and down went really easy! "Flagpole idea" paid off. One man could handle this in virtually no time.
[The finished antenna system]
ON-THE-AIR!
And how does it work in reality? Apart from the
impedance are below 50 ohms - not as EZNEC predicted - and thus provides a
reasonably high SWR – but seems that resonance and bandwidth very well fits with
the previously calculated. Maybe we should try to put the HF and 2 meter beam's
into the EZNEC? So there must be produced something to transform from about 30
ohms to 50 ohms. That in itself is no surprise
since it was one of the things mentioned in N6LF Rudy articles. But it was
someone in the club, said to be considered as a minor unimportant detail.
So we take it from here so far. SWR at the band edges are 1:2, and in the band-center
1:1,79.
[Screen dump from my miniVNA of the North West Dipole]
I
resorted to the good old listening method and return reports from other
stations. The result was in fact - even though the used switch box has a common
connection of all four dipoles to the shield - distinct differences in the
stations where there when changing between the four antennas. The Russian
beacons at 7.039 MHz was genius, because they're just running all the time;
didahdahdit, dididit, and so on. From largest to
smallest signal I observed about a two S-grades.
One of them I could hardly hear before I switched to the best direction in
relation to its location. The dipole in 20-22 meters up, were at the same
time served as a sort of check and / or reference signal forces. Two stations
standing side by side, so there was visibility of both their S-meters
simultaneously.
Reports from a couple of
friendly and patient stations a sunny Sunday afternoon in both Russia, Belgium,
Italy and France, confirmed this observation on the transmitter side. It
is not the same great differences as with a yagi antenna. But it's a nice piece
along the way, right?
A few experiments with
one of the dipoles set and the other three floating, as in the original
articles, merely confirmed that this was useful - just that there had to be made
a switchbox which has both inner conductor and screen hovering in the three
dipoles are not used. There was a
difference to hear, although I must admit that in my ear sounded minimal.
So there are now only two open issues that
must be addressed. It is impedance
matching and constructing a switchbox that lets coaxial cables suspended when
not in use for the transmitter.
But initially, it has been worth the hassle. There
is relatively little money made a directional antenna for 40 meters, which fits
into an area where a 2-or 3 element yagi for 40 meters not immediately is
possible. So I am already looking forward to the
first contest where we can set the new antenna loose on the hunt after all
multiplier'ne, and this time hopefully some more three-point QSO’s than
previously.
The aim of
this article?
The aim has not been to
provide a complete finished description of how to build this antenna system.
The aim of this article
was to talk about what a logbook analysis may cause, and give a little
inspiration for both this and another antenna project for both the individual
radio amateur, but perhaps also to the clubs out and about. Was this
perhaps something to your next field-day? Was
this something that could be used for your current club station, as a cheaper
alternative to the big beam antenna for 40 meters - which probably not fits in
anyway, if one can be allowed to set it up.
... And where do we go next?
Well, one does not know.
But there lurks a transceiver båndpas filter
project for the big contests as an add-on to our Multi-2 station, which also can
be used for field-days. The previously mentioned switchbox for a
multi-two contest station could also be considered.
But perhaps theVHF-UHF fieldday comes first in
line? Here I think of a little something special aerial work for 2 meters.
Let’s see what comes first. See you further down the log dear
OM.
Reference:
1 - www.oz7skv.dk, Look under ”fotogallerier”.
2 - ARRL ON4UN’s Lowband DXing 4’th edition, 2005, kapitel 8, 9, 11 og 12, ISBN 0-87259-914-0
3 - ARRL Antenna Handbook 19’th edition, 2000, (6-32 til 6-35) ISBN 0-87259-817-9
4 - K8UR, Dennis C. Mitchell, CQ December 1989, The K8UR Low-Band Vertical Array.
5 - OE5CWL, Dipl. Ing. Jürgen A. Weigl. A Shortened 40-meter Four-Element Sloping Dipole Array link: http://www.oe5.oevsv.at/export/sites/oe5/Technik/Antennen/sloperen.pdf Oprindeligt publiceret i Ham Radio Magazine, maj 1988.
6 - ARRL NCJ amrch/april 2003, N6LF Rudy Severns, ”Single Support Gain Antenna for 80 and 160 meters”
7 - ARRL QEX january/february 2004, N6LF, Rudy Severns, ”Getting the most from half-wave sloper arrays”
8 - http://www.qsl.net/k/k5lxp/projects/CoaxSwitch/CoaxSwitch.html K5LXP has made his own 100 watts version of Array Solutions big kilowatt switchbox.
9 - http://www.qsl.net/ei7ba/remote.htm EI7BA has also done something in switchboxes- he has in a mail confirmed that is has been tested at at least 1,5 KW