W6SDO.COM                                                    SAN DIEGO, CALIFORNIA USA   
LOOP ANTENNA FOR 17 METERS

I have been considering for some time the
need for a useful antenna that would utilize
the wasted deck space above my three-car
garage.

Since the deck faces South it has turned
out to be too hot for both patio lounging
and for most plants. We’ve tried both and
have failed. Both of these functions have
been relocated to the patio space at the
rear of the house.  So far the only
successful occupant of the deck is a
satellite TV dish. I even considered
sheeting the deck with aluminum and
putting up a vertical. However, I already
had a very nice multi band vertical dipole
mounted on the deck at the rear of the
house and hardly needed another.

Then I read the paper by Steve Cerwin,
WA5FRF, titled “The Long Wire Loop: an
Omni-directional, Multiband, Low Angle
Radiator”. I must give Steve full credit for
my discovery of the perfect antenna for this
underutilized location. Until I read his paper,
I only knew that a full wave loop made a
great basis for a quad beam and could be
used for NVIS purposes (as is had
considered a full wave loop when I was
designing my own NVIS antenna installation).
However, it was only after reading his paper
that I learned that a two wavelength loop
would produce almost no wasted radiation
at 90 degrees (vertical) since what goes
straight up at 18 megahertz almost never
comes down. Further, this two wavelength
loop antenna design produces a relatively
uniform omni-directional radiation pattern
with a moderately low angle of radiation.

The figures below show typical modeling
results for a 2 wavelength loop compared
to a 1 wavelength loop. The loops are
similarly fed at the center of one of the four
legs. The first figure shows (in red) the large
amount of skyward radiation from a full
wave loop whereas the radiation from a
two wavelength loop (in black) peaks at
about 45 degrees and has very little
skyward radiation. The larger loop is
reported to have a gain of around 5 dBi
which can potentially provide some
advantage over a horizontal dipole. The
second figure shows how the azimuthal
pattern for the full wave loop (in red)
compares to that of the two wavelength
loop (in black). Results for the slightly
rectangular loop configuration that I am
installing may not be exactly the same as
these illustrations, but should be close.


































A quick check of the garage deck
dimensions revealed the deck was a
perfect match to a 17 meter two wavelength
horizontal loop antenna. This antenna
should provide a nice compliment to my
hex beam and off center fed Windom
antenna on 17 meters. This antenna was
not a part of my original antenna plan but it
is such an easy build that I could not resist
giving it a try.

The next morning  I made the rounds of the
local Radio Shack stores and found enough
screw in type insulators (intended to
support 300 ohm TV type ribbon line) to
complete the job.


















With a tape measure, a roll of wire and my
MFJ-269 I went to work. I spent the
remainder of that day measuring, installing
and tuning the antenna.

This is the first antenna that I have ever built
without first doing reams of calculations,
studying the antenna design books and
researching the internet. However, Steve’s
paper was very convincing and further I
would only be out a few hours effort if it
didn’t meet expectations.

I finished around 4 PM and connected the
antenna to the Alpha 9500. I fired five watts
into the Alpha from my Kenwood TS-590
and pressed auto tune button. I checked
the SWR and then increased the drive to
20 watts. I was now on the air with around
600 watts of SSB into my brand new loop
antenna. My first contact on my first call
was from the island of Hawaii. Wow, what
an antenna!

I was amazed at what I was hearing with
this antenna. As a receiving antenna, the
noise level is substantially lower than it is f
or my OFC inverted V with the apex at 45
feet. Preliminary readings on a few dozen
assorted signals (monitored on day one of
using the loop antenna) show that the signal
to noise ratio is improved by 3 db or more
when using the loop antenna. Signals that
are at the noise level on the inverted V and
barely readable jump right out of the noise
and are a pleasure to listen to using the
loop.

During the main part of the day, our skip
zone from San Diego for 17 meters is
virtually a dead zone with only weak copy
on most signals from California, Nevada,
Arizona, as well as the nearby parts of
Utah and New Mexico. It is commonplace
to listen to a booming signal from Hawaii
and not have a copy on the ham on the
other end of the line in nearby Phoenix.
This loop, with its improved signal to noise
ratio, helps immensely in copying signals
from within this first skip zone. Signals from
within the next skipped zone (generally
along a line from Minneapolis to New
Orleans for us in San Digo) are weak but
are usually good copy using the OFC
Windom antenna but are even better on
the loop. Throw in some Pederson effects
and anything can happen. It will take some
additional time to determine if the loop is
an equally superior transmitting antenna.  

Here are some of the construction details
about the antenna:

The antenna wire for this loop is perched
on the top of a 4 foot high wall that goes
around the perimeter of the deck over the
garage. This wall has a 2 x 14 inch red
cedar cap and two sides covered with
Hardy board siding (a concrete like
material). The dielectric constant for the
wood cap is around 2.0 and is about 4.5
for the siding. This will distort the pattern
slightly but probably not enough to matter.
Photo above: This photo shows the
antenna wire suspended three inches
above the deck cap strip. It is almost
invisible from the street.

__________________________________

The antenna wire is lifted about 3 inches
above the top surface of the wooden cap
strip so that the effects of rain soaked
wood will be minimal. My experience with
antenna wires that are nailed directly to
wooden support structures is that this
makes an antenna very difficult to keep
tuned as the weather changes, especially
when it rains.


















The maximum size of the loop is set by the
dimensions of the deck. Although a
triangular shape is shown by Steve to have
the most uniform azimuth pattern, my
garage rooftop has only enough room for a
rectangular layout (provided that I want an
exact two wavelengths and the ability to
feed the antenna with coaxial cable without
the use of a tuner). The longest side of the
rectangular loop is 31 feet and runs East
and West. The short sides are 20.5 feet
and run North and South. This provides a
maximum perimeter length without any
“folding or bending” of 103 feet. Although
simple calculations show that the wire
length could be as long 110 feet, my
experience with antennas that are covered
with a thick insulation, that are close to the
ground and are near high dielectric
materials is that they will be considerably
shorter than calculated for free space.

This antenna is only about 13 feet above
the ground (about ¼ wavelength). The view
North is “through” the house with a shallow
6 degree rise in the Northward direction.
The view East is also up a 6 degree rise
and the view Northeast is a more steep at
about 10 degrees. My site looks slightly
downward to both the West and South with
views that are clear of major obstructions
all the way to the Pacific Ocean.  Therefore
the broad takeoff angle lobe peaking at
around 45 degrees for this design (but with
lots of lower and higher angle radiation)
should experience no significant
deterioration due to the terrain (except
Northward where it passes through our two
story chicken wire and stucco wrapped
house).

After running the wire around the perimeter
of the deck, the loop was tuned to 18.135
megahertz. The final length of wire for this
antenna is 99 feet. The antenna was
shortened (by about 4 feet) from the
maximum perimeter around the deck by
cutting diagonally at 45 degrees across
one of the corners. The resonance null is
very sharp for this antenna and there are
no “false” resonances within a mile of the
desired 18.135 megahertz frequency. A
sharp null is good from both the point of
view of maximum effective capture area
and minimum response to off band signals.

The loop has a SWR of 1.8 or less across
the entire band. The SWR has been
confirmed within 10% by the metering on
both the Kenwood TS-590 transceiver and
the Alpha 9500 amplifier. Both the
Kenwood and the Alpha are happy tuning
this SWR value. In this initial version of the
loop, the antenna is being fed directly at
the center of one of the long sides with
RG-8 coaxial cable. Later, I intend to add
a 1:1 balun as well as few ferrite cores
around the cable in order to minimize RF
carried into the shack by the cable and to
keep the SWR from being effected by the
cable length.

This antenna can be constructed at the
very low cost of 100 feet of 12 gauge wire
plus a few support insulators and enough
coaxial cable to reach the shack. I am
adding a five kilowatt 1:1 balun and few
ferrite cores in order to keep any RF from
entering the shack via the coaxial
transmission line.

Because of the very low height of the
antenna there should be minimal lightning
strike risk. Therefore, the coaxial cable
transmission line has not been run down
to the “earth” ground stakes. However, the
equipment is protected from static
discharges by an Alpha-Delta 2 KW spark
gap surge protector which is located at the
output terminal of the Alpha amplifier.
Further, the Alpha chassis is connected to
the earth ground stakes by a wide copper
braid and by the braids of the six cables
the come up to the shack from the ground
stakes below. Maybe, at some later date, I
will route the coaxial cable down to the
ground stake before proceeding on to the
shack.

One wild card factor for this particular
antenna layout, that has yet to be fully
evaluated, is the effect that the RF radiated
from this nearby antenna might have on the
various systems in the shack. This
becomes a very important question since
a portion of the antenna runs along the wall
immediately outside of my second story
shack location. The distance from this
section of the antenna, through the outside
house wall to the equipment (transceivers,
computer and monitor, etc) is only about 3
feet!

I am considering elevating the antenna
wire above the top of the deck access
door and, at the same time, increasing the
distance from the antenna wire to the s
hack. This should fix the two issues that I
have with the present antenna wire
placement.

This antenna is very, very stealthy and is
less visible that a string of small Christmas
lights (see the photo at the top of this
page). This is in keeping with one of my
primary goals of having great antennas
without excessive visual impact on the
property.

Finally, I now have four antennas to choose
from on the 17 meter band. The antennas
that I can now select include a hex beam,
an inverted V off center feed Windom (a
long wire at this frequency) with good omni-
directional characteristics due to the
inverted V construction), my vertical dipole
to which I have added the 17 meter band
and this loop antenna. Each has its own
advantages and disadvantages and the
shoot out between them has been very
interesting.  The four antennas can be set
up to use the four antenna selector switch
buttons on the front of the Alpha amplifier
which makes A-B-C-D testing very easy
and a lot of fun!

The preliminary 17 meter observations are:

1. The hex beam is always quiet and has
the strongest signal in both transmit and
receive mode even though it is only 24 feet
above the ground.

2.  The loop has an equally quiet noise
level but delivers signal strengths that are
about 6 db less than the hex beam.  

3. The OCF inverted V has a noise level
that is 2 to 3 db higher than the hex beam
and produces signal levels that are
between equal and 4 db lower.

4. The vertical dipole usually has a 3 to
6 db higher noise level, compared to the
hex beam, and produces signal levels that
are typically 3 to 6 db less than the hex
beam. It does better on transmitting than it
does on receiving due to the high noise
level when it is used for receiving.



  


     Last Revised December 10, 2012