Antennas/Antennas_CD-ROM_Walnut_Creek_September_1996.iso/articles/longwire/longwire

Text File | 1996-06-30 | 15KB | 364 lines

THE UNI-DIRECTIONAL LONG WIRE ANTENNA
R. P. Haviland, W4MB

When the frequency of a signal fed to a half-wave dipole is
increased above it's resonance value, the antenna lobe at right
angles to the antenna pattern first develops a dimple at the 90
degree point. With further increase, the lobe splits into two
lobes, narrower than the original, each at about 45 degrees to
the original lobe.

With further increase in frequency, additional lobes are formed.
A step in this process is shown in Fig. 1. This is for a
frequency at which the antenna is essentially 3 wavelengths long.
The general rule is that the number of lobes between 0 and 90
degrees is equal to the antenna length in wavelengths, for
integer lengths. If there are fractional parts of a wavelength,
there is also a lobe at 90 degrees. This is fully formed if the
fraction is 1/2, and shows the dimple leading to two lobes for
fractions between 1/2 and unity.

This pattern appears to be somewhat different from the construct
usually shown in textbooks. These show a line parallel to the
0-180 degree axis, tangent to the end lobes. This is intended to
show that the intensity of the minor lobes is equal to gain of
the end lobes in the direction at right angles to the antenna.
The difference is due to the presentation, which is usually
linear in textbooks, and is logarithmic here.

EFEECT OF END FEED

When the feed is off center, the pattern changes, becoming
asymmetric. An example is shown in Fig. 2, which is for a 6
wavelength antenna, fed 1/4 wavelength from the left end. Note
that there are six lobes between 0 and 90 degrees, and also
between 90 and 180 degrees. However the two halves are different,
the lobes away from the feed being the largest.

The reason for the asymmetry lies in radiation from the antenna.
A wave starting from the feed point and traveling towards the far
end is losing energy by radiation. As a result, the wave
reflected from the far end back to the source is smaller than the
exciting wave. The antenna is acting as a combination of a
travelling wave antenna and a normal antenna with a standing wave
present. The currents are considered later.

This feed location is convenient, since it gives a low impedance
feed point. On the other hand, nominally, it can be used only at
one frequency, losing the multiple frequency capability of a true
end fed antenna such as the Zepp. It is also convenient from an
analyitical view. The program used for analysis, Mininec, has no
provision for direct voltage feed. This could be modeled by
adding a transmission line section to form a Zepp, but at the
expense of more complexity and time. The simple approach is used
here.

EFFECT OF ANTENNA RESISTANCE

The above patterns are for ideal conditions, for antenna
conductors so large that their resistance can be neglected. When
conductor resistance is appreciable, there are further pattern
changes.

The magnitude of these is shown in Fig. 3. One direct effect is a
decrease in antenna efficiency, due to the current squared loss.
Another is changes in lobe structure. The main effects of this
are a reduction in intensity of the main lobe, the one off the
far end of the antenna. At the same time the F/B ratio degrades.
These effects are the reason very fine wire "invisible" do not
always perform well. And, of course, iron and stainless steel
wire is poorer than copper.

A "feel" for wire size effect can be obtained from Fig. 4. This
shows the calculated RF resistance as a function of wire size, at
a frequency of 20 MHz. Multiply the radius in meters by 80,000 to
get the wire diameter in mils. For example, #12 wire has a radius
of 0.001 meters, very closely.

RESISTANCE LOADING

Instead of using small wire to make the lobe changes, we can
obtain superior results by placing a concentrated resistance at a
point where it will affect the wave reflected from the far end of
the antenna. A convenient place is the first far-end current
maximum, at one quarter wavelength from the end.

Fig. 5 shows the pattern with a 1000 ohm load resistance at this
point. The far lobe gain is essentially unchanged, but the lobe
at the feed end has decreased, giving a F/B of about 10 DB.

Instead of a single loading resistor, several can be used. Fig 6
shows the efect of three 333 ohm resistors, one at each high
current point from the far end. Gain has decreased slightly, and
the F/B has increased further, to about 14 DB.

The value of resistance giving best F/B is a function of antenna
length and diameter, and also varies with height. The best ratio
obtained in many trials is a F/B of 28 DB, with a resistance of
400 ohm, for a free space antenna.

It is not necessary to place the resistance at the antenna. It
can be at the end of a transmission line, which allows placing
the resistance at ground level for convenient adjustment. For
example, such a line can be placed at both ends of the antenna,
giving voltage feed at one end, and resistance loading at the
other. The far end line can be made of resistance wire,
cancelling the reflected wave as is common in rhombic
installations. (This has not been studied, but would seem to
eliminate the need for resistance adjustment when maximum F/B is
desired.)

CURRENT ON THE ANTENNA

Fig. 7 shows the current along the 6 wavelength end fed antenna
with no loading, and Fig. 8 the phase of the current. There are
essentially 12 half-waves of current along the antenna,
alternating in phase by 180 degrees, as shown in Fig. 8. Note
that the antenna is not exactly at the 6 wavelength resonant
point. This is shown by the fact that the current at the feed
point is not exactly at 0 degrees phase, and by the changes in
current magnitude along the antenna. The apparent "spike" in the
phase plot comes from the plot routine's interpretation of
slightly more than 180 degree phase shift as compared to
slightly less.

Figs. 9 and 10 show the corresponding plots for the same length
antenna with a loading resistance at the far end. Now the current
along the antenna is a combination of a major component, the wave
traveling down the antenna, plus a standing wave component due to
the interaction of the small component reflected from the far end
of the wire. The reduction in current along the wire is due to
loss by radiation and resistance. As is true in terminated
transmission lines, the phase of the current increases linearly
along the antenna, to a good approximation. The antenna is acting
as a traveling wave antenna, akin the Beverage and others of this
class.

EFFECT OF GROUND

Fig. 11 shows the vertical plane pattern for a 6 wavelength long
antenna at 20 meters or 66 feet above ideal earth. Since there is
no loading, the F/B ratio is poor, about 3 DB. As a result of the
height, the main lobes are relatively strong. The pattern at
lower hights is similar, but with the low angle lobes redduced in
size.

Fig. 12 shows the vertical plane pattern for this antenna
installed as a sloper, with the high end at 35.8 meters or 117.4
feet, and the low end at 8 meters or 26.25 feet. While this
antenna is not loaded, the F/B ratio in the plane of the antenna
has improved, to about 9 DB. This improvement is also shown in
Fig. 12, a horizontal cut at 4 degrees above the horizon. Note
that the reference angle is 180 degrees from that of the previous
plot.

However, this apparent improvement in F/B does not occur for
other vertical planes or elevation angles. This is shown by Fig.
14, the same antenna but for the horizontal cut at 8 degrees
above the horizon. The maximum lobe intensity is at 40 degrees to
the line of the antenna.

It is likely that other combinations of length and slope angle
will give better patterns, especially if two such slopers are
used back-to-back. Such an antenna becomes a half rhombic, in the
vertical plane, and rhombic design factors apply to it. These are
not explored further here.

BROADBAND OPERATION

The matter of use at other frequencies has been mentioned above.
Fig. 15 summarizes the main pattern features for a 90 meter long
wire in free space, nominally designed for 20 MHz. Feed is a
quarter wavelength from one end (at 20 meter wavelenth
execitation), with a 400 ohm load resistance a quarterwave from
the other end.

Both maximum gain and maximum F/B occur for a frequency about 10
percent below the 20 MHz design value. Forward lobe gain equals
or exceeds that of a dipole over the range 4 to 32 MHz. The lobe
in the opposite direction is from 3 to 16 DB smaller.

Fig. 16 shows the drive point impedance for the range of
frequency. Drive resistance is close to 300 ohms for the range 8
to 30 MHz, and becomes large only above 30 MHz. Drive reactance
varies less than 200 ohms from zero over the range from 11.5 to
30.5 MHz, becoming large below 8 MHz. The values are within range
of a open wire line antenna matcher of normal design.

Figs. 17 through 25 show the free space patterns at frequencies
of 4, 6, 8, 10, 12, 14, 16, 18 and 20 MHz. As frequency
increases, the number of lobes increases and the main lobes move
closer to the axis of the antenna. At the same time the F/B
increases from about 4 DB to about 14 DB.

This arrangement makes a useful multi-frequency antenna. It is
probably not quite as good as a true end fed, end loaded design,
the double ended Zepp, but this design has not been investigated
in this study.

OTHER DESIGNS

The data shown here is for a single antenna dimension, 90 meters
or nominally 6 wavelengths in length. To a close approximation,
the performance of other lengths can be obtained by scaling. For
example, for a 3 wavelength design, patterns and drive
resistances are those for a frequency just twice the value for
Figs. 15-25. This approximation neglects the effect of wire size,
but this will be small unless "invisible antenna"  wire sizes are
involved.

For more detail, calculate the performance of the specific design
with Mininec. Since no small angle wire intersections are
involved, any version should give adequate answers.

CONCLUDING NOTE

There is little practical experience with distibuted resistance
loading to increase F/B, and as far as a literature search goes,
none with lumped far end loading. Perhaps some farm / ranch
owning amateur will undertake practical trials. It appears that
performance approaw


THE UNI-DIRECTIONAL LONG WIRE ANTENNA
R. P. Haviland, W4MB

CAPTIONS


Fig. 1. Free space pattern of a center fed antenna 3 wavelengths
long.


Fig. 2. Free space pattern of an end current fed antenna 6
wavelengths long. The feed is 1/4 wavelength from the left end.


Fig. 3. Effect of radiator resistance on a long wire antenna.
Efficiency, forward gain and F/B ratio all decrease as wire
resistance increases. Calculated by Mininec with resistance load
at each wire segment.


Fig. 4. RF resistance of a wire as a function of wire size, at 20
MHz. Multiply size in meters by 80,000 for size in mils.
Resistance varies as the square root of frequency.


Fig. 5. Free space pattern for the antenna of Fig. 2, but with a
1000 ohm resistance located 1/4 wavelength from the end opposite
the feed. Note the improvement in F/B. There is an 0.8 DB loss in
gain of the main lobe.


Fig. 6. Same as Fig. 5, but with 333 ohm resistors at 1/4, 1/2
and 3/4 wavelengths from the end opposite the feed. Note the
Further improvement in F/B, and the added 0.2 DB loss in gain.


Fig. 7. Current distribution of the antenna of Fig. 2, with
current value shown for each Mininec calculating segment. The
actual curve would have more rounded points, as for a sine wave.


Fig. 8. Phase of the current of Fig. 7, showning that the
standing wave predominates. The apparent spike is a phase of just
over 180 degrees at segment 21. Compare to that at segment 44.


Fig. 9. Current distribution for the antenna of Fig. 5. The
dominant component is now that of a travelling wave progressing
from the feed to the right. There is a small standing wave
component impressed on this.


Fig. 10. Phase of the current of Fig. 9. Primarily, the phase
steadily increases along the antenna. The small standing wave
component shows in the detail variation.


Fig. 11. Vertical plane pattern for the antenna of Fig 2,
installed relatively high above earth. The azmuith choice makes
the feed at the right.


Fig. 12. Same as Fig. 11, but with the antenna installed as a
sloper. See text.


Fig. 13. Horizontal plane cut pattern at 4 degrees elevation for
the sloper antenna. See Fig. 14.


Fig. 14. Same as Fig. 13, but for the cut 8 degrees above the
horizon. Note that maximum gain at this elevation is at 40
degrees to the antenna axis.


Fig. 15. Lobe gain versus frequency for the 6 wavelenth at 20
meters antenna of Fig. 5.


Fig. 16. Drive point resistance and reactance as in Fig. 15. The
antenna is easily fed with 300 ohm twinlead or open wire line.


Fig. 17. Free space pattern for the antenna of Fig. 5 at 2 MHz.


Fig. 18. Free space pattern for the antenna of Fig. 5 at 4 MHz.


Fig. 19. Free space pattern for the antenna of Fig. 5 at 6 MHz.


Fig. 20. Free space pattern for the antenna of Fig. 5 at 8 MHz.


Fig. 21. Free space pattern for the antenna of Fig. 5 at 10 MHz.


Fig. 22. Free space pattern for the antenna of Fig. 5 at 12 MHz.


Fig. 23. Free space pattern for the antenna of Fig. 5 at 14 MHz.


Fig. 24. Free space pattern for the antenna of Fig. 5 at 16 MHz.


Fig. 25. Free space pattern for the antenna of Fig. 5 at 18
MHz.ching that of a rhombic is possible.


Sidebar for
The Uni-directional Longwire Antenna
R. P. Haviland, W4MB



It is well known that an endfed antenna has pattern distortion,
with the lobe in the direction of the far end being stronger than
the lobe in the direction of the feed end. The effect is most
noticable in antennas which are several wavelengths long, the
long wire antennas.

W4MB examines this phenomena in detail, with a view of
determining the amount of directivity available. He then shows
how the directivity can be increased, by introducing resistance
loading at the end away from the feed point. Such antennas are
members of the class called traveling-wave antennas, The Beverage
being the most common in Amateur useage.

Since the antenna analysis program, Mininec, is not designed for
true end or voltage feed, the analysis is based on current point
feed, at one-quarter wavelength from the end. W4MB also shows how
this feed technique can be used to give low impedance feed while
retaining the broadband capability of the end-fed long wire.

Because if the space needed, the long wire is not an antenna for
everyone. For those with space, it is an useful alternate to the
Yagi, giving good performance in a simple antenna. It is also a
useful antenna for portable or emergency use.