What you always wanted to know about Fox Hunt Antennas.
Thoughts about the 3 element Tape Measure Yagi.
There are many webpages describing how to calculate the basic design parameters of the
3 - element Yagi direction finding antenna and practical examples of how to make it using
The classic design page is by Joe Leggio, WB2HOL, on the webpage
and tape measure pieces as radiating
elements. The middle element, the
feeder dipole, is placed left from the
center, in vicinity to the reflector and
with a larger distance to the front
element, the director.
(Image from the Leggio page)
The design is based on non-conductive boom material. There are special publications which
deal with boom conductivity influences on Yagi antennas, but for the sole purpose of a hand-
held fox hunt antenna, those issues can be safely ignored. The boom sits in the electrically
neutral center and can be made of any material, conductive or not, as far as mobile direction
finding is concerned. What really matters, is the positioning of the center element along the
boom, because the place where the feeder sits, strongly affects the Front to Back F/B ratio.
The one-sided antenna directivity will help you avoiding the "Signals from all Directions"
situation which occurs when the equipment is not suited for the task. Either your antenna is
designed for a different operating frequency than the signal you are chasing at the very
moment, or you don't even know it's design frequency because you bought it on a flea market.
An antenna which has helped to collect many DX signals for your logs, does not automatically
qualify for direction finding. One's favourite Antenna may be gain optimized,
which causes side lobes to appear on the backwards direction instead of a zero hole there.
Furthermore the forward gain diagram is relatively broadband over the whole range of the
frequency band in use, while any F/B ratio is very narrowbanded only at the design frequency
of the antenna. To cover the whole 2 meter band from 144 MHz to 148 MHz with always near
optimum F/B ratio, one needs to build 2 or 3 antennas of the above design with individually
different length elements and spacings. This paper attempts to describe such a 3 element
2 meters direction finding Yagi with adjustable F/B ratio within a working range of 10 MHz
from 144 MHz up to 154 MHz. Simulation results from the "4nec2" software will be shown
and then compared to actually hand-made samples and results from the antenna analyzer.
The compromise taken is, that element lengths remain constant and only the center dipole
element will be moved front to back along the boom by mounting it on a slider.
The dipole position in between reflector and director both affects antenna resonance frequency
as well as the F/B ratio, which both together are thus adjustable within a >10 MHz range.
The CEBIK pages, W4RNL.
May those pages last forever: http://www.cebik.com
Hidden behind passwords I don't have access now, but have some copies from older times.
On those pages you will find many tables and diagrams which help building Yagi antennas.
We need to discuss the following comparison:
The High-Gain antenna is the longest and the three elements are very regularly tapered
in element length. The radiating element sits only a small distance below the half boom
length. We do not further care for that example, because we already know it is not suited
for direction finding, due to it's bad F/B ratio.
The Very-Wide-Band antenna has an outstanding short director element. That design trick
lifts the gain diagrams up in the frequency range and flattens their response. Also note the
relatively large distance between reflector and driver. Moving the driver towards the front,
increases the operating frequency also. The proof will be seen later with the simulations.
The design frequency of a Yagi antenna has to be always the lowest number. For example,
if we want to build an antenna for the 2 meter band, then we must design the antenna for
its lowest frequency, which is 144 MHz in that case. Automatically it will extend upwards
from that frequency, but never or only very little downwards. It would have been wrong to
design the antenna for the center frequency of the band. The reflector is the longest
element and nothing can resonate below its working frequency, but driver and director are
shorter and thus cover the higher frequency parts of the operating range. Same applies to
logarithmic periodic LPDA antennas, for example. I actually built that Wide-Band structure
and found that it behaves rather strange. I was not able to measure its resonance
frequency with the analyzer, because there was none, only a long deep ditch in the curve.
Furthermore the antenna reacted over-sensitive to everything in the vicinity. Theoretically
it has an excellent F/B ratio, as the 4nec2 simulation with the EVOLVER wants to make
me believe, and I believe it. Yes, the EVOLVER optimizes exactly for that design when
weighted on F/B during optimization, while the OPTIMIZER yields rather the
Max-Front-to-Back design shown in the image.
The Maximum-Front-to-Back design is what we are looking for. The boom is a little shorter
than the High-Gain, but the most important thing to watch is the element spacing. When you
encounter an antenna with a much different element spacing, you may assume a wrong
design. The feeder dipole must be placed just a little in front of the boom centre as shown in
the middle image. We will deviate from that and put the driver on a slider to move it front to
back. Remember exactly what you see in that Cebik image: This is the proper driver position
for the design frequency of the Yagi which is equal to its lowest working frequency. From
there on we can only move upwards in frequency by pushing the slider with the driver to the
front towards the director, but we will have only little success by moving the driver more to
the back towards the reflector.
Since we now know what we want to see, let us have a look at the work of others.
Most of them follow the same recipe, but not a single one is correct,
according to the results from antenna simulator programs.
They will work, though, more or less to the satisfaction of the owner,
because there are wide tolerances in design and manufacturing and operating,
and even a bad antenna is better than no antenna.
The Proof of the Pudding is in the Eating.
To the left imagine a wideband logperiodic LPDA antenna connected to a sweeper signal
generator. (I actually use broadband noise). The hairpin matching stub, that is the white
wire loop on the center element, is a still little too long. That would be changed later.
The height above ground on that tripod is not quite correct. As the simulation will show,
the optimum height above ground for such a 3 element direction finding Yagi is 2 meters,
with another only slightly smaller maximum at 1 meter height, which both are convenient
working heights. I just mention that because it is very peculiar. I will rather skip that
height optimization and keep to a more stringent line. The driver element sits on a sleigh.
the internet collection shown. Most of them use that particular configuration. They are
detuned below the actual design frequency. Reception from front will not be largely affected,
but such an antenna receives from the rear also because the front to back ratio is not optimal.
An optimally designed and constructed antenna looks like the image before with the tripod.
If you want to use a lower than design frequency, you will have to move the driver element
to the position shown here. But there is only a very small useable range, about 1 or 2 MHz
for that 2 meter yagi to be tuned low. Tuning upwards gives us more frequency space,
about 10 Mhz from the nominal frequency.
well enough without moving the driver position, the front to back ratio would be quite bad.
Moving the slider changes the "piece of wire" back to a direction finding antenna.
There is a max. 80 dB attenuator with switches, mounted on the Kenwood TH-F7.
That Kenwood is the best for direction finding, because it can work as an SSB receiver with
audible BFO tone generation. With an FM receiver you would only hear more or less noise.
And it has an additional 10dB attenuator built in.
Fox hunting is not the only reason why we need adjustable directional antennas.
Chasing unknown or disturbing signals in your neighbourhood may be necessary.
I once had to find 16 2/3 Hz interference from a nearby railway line high voltage isolator.
That insulator was sparking during moist weather and fog. The noise ranged from long wave
up to 70 cm all over the way. Short wave reception seemed virtually impossible.
During a 2 meter direction finding fox hunt I could spot the bad component and
notified the railway administration. They were happy to replace the insulator before
it actually failed and broke to pieces. Be prepared to do variable frequency fox hunting,
you never know in advance which frequency will be of interest next.
The Proof of the Pudding is in the Simulation.
Simulation Software used: 4nec2
Measuring tape is not a commonly used construction material in the antenna design business.
Maybe that explains why no software can deal with it. 4nec2 allows to select "stainless steel"
which sounds somewhat similar, but when looking into the specs, they only incorporated the
Ohmic losses, but no ferromagnetic effects. After running the numbers on the PC screen,
I have to shorten all elements calculated by 4nec2 for -3,8 % in total length. Sometimes
I ended up with my adaptive calculations at -4,2 %, but likely made some errors.
For the time being I cannot give a more precise number than the -3,8% which seemed to fit
best in most cases.
On the other hand, when entering a tape measure recipe into 4nec2 I have to stretch every
element for +3,8% and then enter those numbers into the program. I do not know how to
manipulate the Ohmic LD values with complex numbers to get the ferromagnetic effects right.
The program itself optimized the element wire diameter radius to 0,01 meter which suits
excellently. The tape I use is 19 mm in width. Luckily all those material factors do not affect
the optimization results strongly, there is not much of a difference. The assumed design
frequency for the following images will be 144 MHz.
Just a normal, standard 144 MHz antenna.
Dimensional numbers for element length and spacing can be found on various webpages.
Users of the 4nec2 software will also know where to find example files. Therefore I do not
want to give recipes here. I only want to demonstrate what can happen when antennas are
working off their design frequency and what we can do about.
for 144 MHz optimized with 4nec2 for maximum
Front-to-Back ratio. The center element sits higher
than half the distance from reflector to director.
The arrow tapering decrease in length is moderate,
but that depends on the wire diameter you choose.
We want to use that antenna from 142 to 150 MHz.
The question is: will we achieve a good F/B over
the whole range? And how good is the F/B at the
Next slide, please.
In numbers it is somewhere around 50 db or
above, depending on how long you let the
optimizer run. At some time then there will be
no more progress and the numbers won't
change anymore. The antenna will have a
decent null in the rear. The two earlike side
lobes are hard to recognize in reality, because
the diagram is easily distorted by nearby metal
objects and the perfect roundness gets pulled
and pushed a little. At 144 MHz we are fine.
No fox can escape our hunting.
What happens if we want to use that antenna
on 142 Mhz or 150 MHz?
Next slide, please.
Stray from the straight and narrow frequency band.
Watch the F/B ratio deteriorate offside the design frequency of 144 MHz.
First we look at 142 MHz and the next image is for 150 MHz, same antenna.
144 MHz Yagi working on 142 MHz:
144 MHz Yagi working on 150 MHz:
Sweeping the frequency over a range from 140 to 160 MHz reveals how narrow the usable
frequency range is actually. See image below.
This will erroneously lead to the assumption that the
antenna works fine and can be used for a broad
frequency range. Indeed it can be used and we will
always be able to achieve successful connections
with it. and quite good DX signals.
Only the Front-to-Back ratio is that sensitive.
Red dots depict how small the good range is.
(Ignore the green dots).
Any 3 element Yagi direction finder antenna has a very small range where it can yield
good results. Working other frequencies than the nominal design frequency,
144 (+/- a bit) in our example, will result in "Signals from all directions".
Riding on a sleigh.
The boom is made of 10x10 mm square aluminium tubing. The white bar is the cable clamp.
On top sits a 5,2 cm hairpin match. The hairpin size is not critical, a few millimeters up or down
do not matter. The screws in the slider are 3 mm with self-fixing nylstop nuts. The tape
screws and wing nuts are 4 mm. Friction can be adjusted by tightening the 3 mm screws
in the slider. I may even fix it in a position. Three and five millimeter plexiglass was used.
There is no need to connect the hairpin center to the boom for grounding.
This is a portable handheld antenna. Grounding would be advisable for a fixed antenna
mounted on a high tower for lightning protection and reducing static noise.
The passive elements are mounted with a wing nut.
Better design splits the elements in two parts which overlap at the mounting screw, so the antenna
can be folded. You may want to stiffen the center part of the radiators by fitting a small part
of tape, about 12 or 15 cm with liquid superglue.
A tape measure Yagi will flip and flap in the wind. It's bad when I sit in a helicopter or on the back
seat of a motorcycle, but it is good when hunting a fox in the corn field. Strenghten the center
of the dipoles by adding a second layer of tape. Liquid superglue sticks perfectly on the yellow paint.
The ends of the blades are razor sharp and must be protected somehow.
Changing the center element position, but nothing else.
Let us put the radiating driver element, the feeder, into positions ONE and TWO.
ONE is intended for 142 MHz and TWO is going to work on 150 MHz.
Position ONE, low frequency: Position TWO, high frequency:
The blue line is the forward gain, the red line is the F/B ratio and the green line
is not our business. This is the behaviour of a 3 element Yagi when the center
element is moved. The forward gain does not change much. Please note the different
scaling of the blue gain curves, something the 4nec2 software did on it's own.
The radiation diagrams taken in both slider positions.
Position ONE, low frequency: Position TWO, high frequency:
Both diagrams are a little worse than the original centered on the design frequency of 144 MHz.
But both diagrams are better than the unmodified antenna was on both frequencies,
and especially the higher frequency range profits from the slider adjustment. Therefore I think
the design frequency of an antenna equals it's lowest working frequency, within some tolerance,
but the antenna can be tuned to higher frequencies and still work as direction finder.
Improvements to the above shown diagrams may be possible.
What to do next?
element allows for better backwards dip adjustment.
Such a design will look like this.
For the moment I cannot report more data on that design.
This is work in progress.
The impedance is capacitive and requires a longer
than 4 cm hairpin match.
(Image from an unknown web page, thanks guys).
Balun transformers and other symmetry matches are fine, and will improve the antenna by cancelling
cable shield currents, but are not really necesary. The antenna still works when only connected directly
with a coax, and without any matching and symmetry transforming. But it is advisable to shorten all
unwanted lower frequency signals e.g. with a hairpin and keep them off the receiver input.
A portable antenna will continuosly be exposed to varying environmental influences when you
move around in the terrain with it. Therefore it is fruitless to demand precision matching.
A Direction Finding Coaxial Cable.
Since when can coaxial cables find a signal direction?
They can't of course, but the cable can spoil all your efforts, and again you will encounter
"Signals from all Directions". Imagine you are already near the source and there is a very strong signal.
Pushing the attenuator switches will bring down the level. Suddenly you notice that your antenna does
not pinpoint the direction as good as before. Your antenna cable may pick up signals and add them
to the signals coming from the antenna itself. There are cheap RG58 cables around the surplus amateur
radio shops which only have a 40% shield. The outer braid which is wrapped around the center insulator,
will have holes in millimeter size.
A good cable looks like this one:
Double wrapped silver coated shielding braid.
Other quality cables use copper or aluminium sheet metal as the second layer below the wire braid.
This results in 99% (or so) shielding effectiveness. Cable shielding is a critical factor when searching
signals in very strong EMF environments.
Measuring Measuring Tape Antennas.
Checking the Standing Wave Ratio (SWR) with a scalar antenna analyzer.
The optimal height above ground is 2 meters for that antenna, when used horizontally.
The upright position shown here is for demonstration only.
The Proof of the Pudding is in the Impedance Match.
Nominal frequency 144 MHz
Higher frequency 148 MHz
Still higher frequency 154 MHz
(On the SWR scale the number "1" is the first horizontal line from the bottom).
I could go a little further in each direction, down to 140 MHz, up to 160 MHz.
That is about the maximum range possible by moving the driver element alone
with no other changes on the antenna. Beyond that range I will have to use another set of
tape dipoles which I can optimize for the fixed reflector to director spacing. Of course I can
change the spacing too by drilling another 4 mm hole into the aluminium boom to reposition
the reflector element at the bottom. Alternatively this antenna can also be used for the 70 cm
band just by exchanging the elements and the single screw for the reflector mount.
Physical dimensions of the antenna shown in the above images:
Boom: 10 x 10 mm aluminium square tube, total length one meter.
Reflector length: 1010 mm
Driver dipole length: 961mm
Director length: 900 mm.
Distance from reflector to director: 621 mm
Tape measure material: 19mm width (18.8 mm exactly)
Driver dipole center gap: 29 mm. This equals the distance
of the two 4 mm diameter wing nut fasteners center to center.
The actual distance between the two inner tape ends is 15 mm,
but that is not critical and not important. The tape ends must extend
a few millimeters beyond the wing nut fasteners for mechanical reasons.
The matching stub is made of 2.5 mm copper wire.
Width is 29 mm center to center wire and the length from the wing nuts
center to the stub end (wire center) is 52 mm. Again, the exact dimensions
are not very critical, a few mm more or less don't matter.
The driver dipole size is measured over the total of both elements,
including the space in the center. The center space width is not critical.
With a total length of 961 mm the half dipole would be 480,5 mm.
Subtract 7.5 mm from both elements to get the 15 mm center gap, and the
actual physical length of the two dipole halves would be 473 mm each,
with the 4 mm diameter mounting hole drilled 7 mm from the inner end
of the dipole half-pieces.
With above dimensions the antenna will work from 143 MHz up to 153 MHz.
When the numbers as given are entered into a Yagi antenna simulation software,
the results will be wrong, because the software does not know how to correct
for the magnetic properties and the geometric dimensions of steel tapes.
The Antenna Workshop.
Work outside, keep the home clean. The boom actually needs only two drills:
two holes in 621 mm distance, diameter 4 mm.
Use Nylstop nuts instead. Look further down for an example foto.
indeed labeled "FOX". The ocillators were collected from old CPU boards
and run in parallel. Each oscillator output is connected to a 50 cm long wire.
That simple gadget produces frequencies all over the range from shortwave
up to 1 GHz. Best results are obtained with an SSB receiver.
The insulation is there rather for optical reasons. Blank wire would also work.
The boom is 10 x 10 mm size and one meter long.
direction finding antenna which I built for the tests are listed after the measurement
diagrams under "physical dimensions."
Mount the plastic parts with self-holding nuts, sometimes called Nylstop or
Nylon Locking Nut. The nut can be seen above the wing nuts which hold the
tape. The dark ring inside the nut is made of Nylon and prevents the nut from
getting loose. Then adjust the screw for the best operating friction of the slider.
By tightening more the slider can be fixed in a position.
The fox hunting antenna made small. Unfold it like an umbrella.
Many thanks to the 4nec2 people, the antenna gurus in the internet and to all the other folks.
I hope this report was a little entertaining.
Will add element size numbers for different frequencies some time.
If you need the dimension numbers for building such a yagi,
you can safely use the numbers from the many other sites
which mostly describe the PVC tube CEBIK or Joe Leggio 3 element tape measure Yagi.
The numbers are the same, only that this aluminium boom design with the slider allows
to adjust the frequency response of the antenna for a wide range.
Thank you for reading this.
Story by Helmut Wabnig, OE8UWW, August 2011.