Doublets, dipoles, and other Hertzian antennas

A doublet antenna is any of several forms of dipoles (di - 'two') that are balanced with respect to ground. In other words, neither side of the antenna feedline is grounded. These antennas are fed in the center, which point will be either a current node or voltage node depending on the design of the particular antenna. Perhaps the most common form of dipole is the half wavelength (j/2) center-fed dipole, which can be fed with ordinary 75 ohm coaxial cable. It’s so popular that when people use the word 'dipole' in an unqualified sense, it’s the j/2 center-fed version that is meant. Although dipoles can be made of either wire or tubing, the most common practice is to make them out of wire, especially antennas at the lower end of the high frequency (HF) spectrum. Above about 14MHz, however, it becomes increasingly practical to make these antennas of aluminum tubing.

DIPOLE RADIATION PATTERNS

When one uses the term 'dipole' in the common sense it’s usually under stood that the antenna being discussed is the half-wavelength doublet. One reason for this practice is the widespread popularity of that particular dipole. But the practice is actually erroneous because the dipole can be any size up to about two wavelengths (although some would argue that any balanced antenna, even longer than two wavelengths, is a dipole, I demur on grounds that longer antennas begin acting like the long-wire antenna and its relatives). FIG. 1 shows the effect of different wave length dipoles on the azimuthal radiation pattern of the antenna.

At j/4, 3j/8, and j/2 the azimuthal pattern is the classic figure '8' that is commonly associated with dipoles. The principal difference in these patterns is the particular shape (width, etc.) of the pattern, which varies with antenna size. At 5j/8 wavelength, the figure '8' persists, but becomes narrower and minor lobes begin to appear. To a receiver, these side lobes represent responses in directions that may not be desired, and for transmitters it represents wasted energy. At 3j/4, the pattern blossoms into a four-lobe 'clover leaf' with significant minor lobes. This pattern is seen in ham radio when a half-wavelength 40m (7 MHz) dipole is used at 15m (21 MHz). The same physical length that is j/2 at 7 MHz is 3j/4 at 21 Mhz. At 7j/8 wavelength the minor lobes disappear, and at 1j the pattern thins out a bit but retains the clover leaf shape.

FIG. 1

HALF-WAVELENGTH DIPOLE

It’s often the case that a half-wavelength dipole is the first antenna that a new ham operator or short-wave listener will put up. These antennas are low cost, easy to build, and have a delightful tendency to work well with only a little muss and fuss. They are thus well suited to the newcomer.

FIG. 2A shows the basic half-wavelength dipole. It consists of a half wavelength radiator ('B') that is cut into two sections ('A') which are each a quarter-wavelength long. The feedpoint is the middle of the half-wave length, and for this size dipole the feedpoint is a current node. The feedpoint impedance is therefore a minimum, and the dipole makes a good match to 75 ohm coaxial cable.

It’s commonly assumed that the feedpoint impedance of the dipole is 73 ohms, but that is a nominal value which is only true at certain heights above the ground. As can be seen in FIG. 2B, the actual impedance is a function of the height of the dipole, and at a certain height it converges to 73 ohms. At other heights it can vary from a very low impedance (near zero ohms) up to about 110 or 120 ohms.

The reason for the impedance situation is shown in FIG. 2C. This graph shows the distribution of voltage and current along the length of the antenna element. Note that the current peaks at the feedpoint and drops to a zero at the ends. The voltage, on the other hand, peaks at the two ends and drops to a minimum at the feedpoint. One of the principal differences between different dipole designs is the issue of whether they are voltage fed or current fed.

The element lengths (in meters) of the half-wavelength dipole would normally be found from 150/F_MHz for the overall length, and 75/F_MHz for each quarter-wavelength segment. However, because of the velocity factor effects of the length-diameter ratio (which is high for wire HF antennas), and the capacitive end effects, a small foreshortening occurs, making the actual lengths closer to

B_meters = 143 meters F_MHz

A_meters = 71:5 meters F_MHz

FIG. 2

These lengths are still only approximate, although in many cases they will be right on the money. There is nonetheless a possibility of needing to lengthen or shorten the antenna if an exact resonant frequency is needed.

This is done by either noting the voltage standing wave ratio (VSWR) for the minima being at the desired resonant frequency, or by the feedpoint impedance being close to the design value.

The standard dipole is fed in the center with coaxial cable. For installations really on the cheap, ordinary lamp cord (alternating current wiring) can be used, as it looks like 75-100 ohm parallel transmission line to the antenna. However, except for receiver and the lowest power transmit operations, it’s definitely not recommended. Even receiver operators will fare better with coaxial cable because it has superior noise immunity.

In the usual installation, the dipole will have a center insulator to support the middle ends of the two sections, and to provide the connection spot for the transmission line. The outer ends of the two sections will be supported by end insulators and rope or heavy twine tied to a mast, roof top, or convenient tree. Antenna and radio shops sell special center insulators that provide strain relieved coaxial connections for the transmission line and solder points for the antenna elements.

To BALUN or not to BALUN, that is the question

The BALUN transformer is used to convert between balanced and unbalanced loads. In the context of the dipole, the feedpoint is balanced with respect to the ground, while the transmitter output is unbalanced with respect to the ground. What this means in plain language is that an unbalanced load has one side grounded, while a balanced load is not grounded. Some BALUN transformers also provide impedance transformation, with 4:1 being the most commonly seen. This means that a load of R across the output ports of the transformer is reflected back to the input so that 'looking into' the transformer it will seem like a load of R/4. This is the transformer used to reduce 300 ohms from television-style twin-lead to 75 ohms for coaxial cable.

It’s common practice to use a 1:1 BALUN transformer at the feedpoint of the half-wavelength dipole antenna ( FIG. 3), even though it provides no impedance transformation (and none is needed). The BALUN transformer balances the load, causing equal but opposite phase currents to flow in the two conductors, and thereby reduces radiation from the transmission line. The effect of the feedline radiation is to distort the radiation pattern.

FIG. 4 shows the effect on the pattern, at least in simplified form. FIG. 4A shows the 'normal' pattern, which is what we saw in FIG. 1C. But radiation from the feedline, and unbalanced current flow, tends to cause the nulls to fill in and the main lobes to reduce a bit ( FIG. 4B). In some cases, the afflicted pattern is a lot more distorted than shown in FIG. 4B, which is a highly simplified case. When a BALUN transformer is used, however, the resultant pattern will look a lot nearer the ideal than the distorted version, and the dipole will perform as expected.

The use of the BALUN in the half-wavelength dipole is one of those controversial things that radio enthusiasts debate seemingly endlessly, but there is no real controversy. The simple fact is that half-wavelength dipoles work better in practical situations when a 1:1 BALUN is used to interface the coaxial cable transmission line to the antenna feedpoint. The only excuse to not use a BALUN, in my not so humble opinion, is the economic one.

There were, after all, periods in my life when even so low a priced item as an HF BALUN was very dear indeed, so I used an end insulator at the feed point instead (once I even used a discarded toothbrush as a center insulator - and it even lasted a month or two before it broke).

FIG. 3

FIG. 4

MULTIBAND AND WIDE-BAND HALF-WAVELENGTH DIPOLES

The half-wavelength dipole is a resonant antenna, so it works best only on and near the center design frequency. The classic figure '8' pattern is found only at this frequency and certain lower frequencies (for which the antenna is j/4 or 3j/8). The antenna also provides performance at odd integer harmonics of the design frequency, but with a multi-lobed pattern (see FIG. 1 again). But there are some things that can be done to accommodate the multiband interests of typical radio hobbyists. For one thing, we could put up a different dipole for each band. But even if you have the space, the installation would look like an aerial rat's nest of wires, and it requires switching the feed lines at the rig or receiver to change bands. Not a good idea, actually.

The other approach is shown in FIG. 5: connect dipoles for different bands to the same transmission line. One can get away with this because the feedpoint impedances of the dipoles are very high except at their resonant frequency, so don’t materially affect the feedpoint impedance of whichever dipole is in use at any given time. In FIG. 5, dipole A1-A2 works at the highest band, B1-B2 works at a middle band, and C1-C2 works at the lowest-frequency band. One precaution is to ensure that none of the dipoles is cut to a frequency that is an odd integer multiple of the resonant frequency of one of the other dipoles. If, For example, you make C1-C2 cut to 7 MHz, and A1-A2 cut for 21 MHz, then the half-wavelength 21 MHz dipole A1-A2 will be in parallel with the 3j/4 wavelength 21MHz operation of A1-A2. The low impedances will be in conflict and the pattern will be bizarre. Note, however, that 3j/4 operation means that proper selection will yield six-band operation from this 'three-band' antenna.

FIG. 5

A certain precaution is needed when the multiband dipole is connected to a radio transmitter. Many of the HF amateur radio bands are harmonically related. This was done, I am told by some old timers, because in the 'golden oldie' days, trapping harmonics was not as easily done as today, and amateur transmitter harmonics would fall into another amateur band, rather than onto a commercial band or public safety band. If one of the other dipoles is cut to a harmonically related band, then that harmonic looks into a resonant antenna and will radiate efficiently. If you use this type of antenna, then be certain to use a low-pass filter with a cut-off between the two bands, or use an antenna tuning unit that is not a high-pass filter (as some are), or otherwise make sure that your transmitter is harmonic free.

FIG. 6 shows a method for producing a very wide-band dipole. This particular version was designed for use by radio science observers who listened for radio emissions from the planet Jupiter in the 16-26MHz short-wave band. These sounds are rather strong, and can appear anytime Jupiter is above the horizon, day or night. Although a beam antenna would have been more useful in this project, the small college physics department who needed it was unable to afford a proper array, so had to make do with a simple dipole.

FIG. 6

The idea in this design is to parallel connect three half-wavelength dipoles with overlapping frequency responses. In this particular case, the antennas were cut for 18, 21, and 24MHz. Some amateur radio operators will use this type of antenna, but with only two dipoles, with one cut for a frequency one third the way through the band and the other at two-thirds the way through the band. For example, in the 15m band (21.0-21.450 MHz), the low-end antenna is cut for 21.15 MHz and the high-end antenna for 21.3 MHz.

FIG. 7 shows how this antenna works. In FIG. 7A we see the frequency response of a single-dipole antenna. This curve plots VSWR versus frequency. The lowest point on the VSWR curve is the resonant frequency of the antenna, and is hopefully where you designed it. Note that the VSWR climbs fairly rapidly as you get away from the resonant frequency. In FIG. 7B we see the effect of overlapping the VSWR versus frequency curves of three antennas designed for adjacent segments of the band. The overall VSWR seen by the antenna is basically the lowest point on the composite curve (although there is some interaction in real antennas!).

FIG. 7

FIG. 8

FIG. 9

FOLDED DIPOLES

The folded dipole antenna ( FIG. 8) has a wider bandwidth than the common dipole, although at the expense of a little constructional complexity. This antenna consists of a loop of wire made from a half-wavelength section of twin-lead or parallel transmission line. The most common form of folded dipole uses 300 ohm television twin-lead for the antenna element.

Note that the two conductors of the twin-lead are shorted together at both ends of the radiator. The feedpoint is formed by breaking one of the two conductors at the midpoint. The feedpoint impedance of this antenna is just under 300 ohms, so it makes a good match to another piece of 300 ohm twin-lead used as a transmission line.

The overall length of the folded dipole is found from the same equation (see above) as a regular half-wavelength dipole. There may be a small amount of additional foreshortening due to the velocity factor of the twin-lead, so this length may be a tad long for the desired frequency. You can adjust to the desired frequency by trimming the same amount off of each end while monitoring the VSWR to determine the resonant frequency.

If you want to use coaxial cable to feed the folded dipole, then place a 4:1 BALUN transformer at the feedpoint. This transformer reduces the feed point impedance from near 300 ohms to near 75 ohms, and so makes a good match to 75 ohm coaxial cable (e.g. RG-59 or RG-11).

The folded dipole made of twin-lead can be used with moderate amounts of radio frequency (RF) power in transmitting stations, and all receiving stations, but one has to be a bit careful as power levels increase. I recall one 40m folded dipole excited with around 300Wof RF that got uncomfortably hot to the touch after a few minutes of CW operation.

Folded dipole construction always looks easier in books than it’s in real life. The problem is that the conductors used in 300 ohm twin-lead are small, and not intended to support any weight or sustain under wind forces (which can be larger than most of us imagine because an antenna has a large 'sail area' in wind). As a result, folded dipoles usually break and come down either at the feedpoint or one of the two ends. The solution is to make fittings such as shown in FIG. 9 for the center ( FIG. 9A) and ends ( FIG. 9B).

The folded dipole fittings of FIG. 9 are made of an insulating material. I have seen similar devices made of hardwood salvaged from flooring, and then coated in polyurethane or varnish to protect against rain. Others use Plexiglass, Lexan, or other materials. I have even seen one commercial folded dipole center insulator made of nylon, but it seems to have disappeared from the marketplace. In FIG. 9A, the wire connections are soldered to solder lugs attached to brass machine screws. All other screws and nuts are made of nylon to minimize interaction with the antenna. Note in FIGs 9A and 9B that holes are cut into the insulation of the twin lead, and nylon machine screws are passed through both these holes and mating holes cut into the mounting device; matching hex nuts on the other side secure the screw (see side view inset to FIG. 9A). This keeps the twin-lead from slipping out of the block, or at least delays the day when failure might occur.

The screw holes in the twin-lead (see inset to FIG. 9A) can be made with an ordinary paper hole punch, although one has to squeeze a bit hard on the better grades of twin-lead.

INVERTED-'VEE' DIPOLE

The inverted-'vee' antenna ( FIG. 10) is a dipole that is supported in the center from a mast or other support, with the ends drooping toward the ground. As with other forms of dipole, the ends are supported using end insulators and rope. But instead of finding a convenient high spot to support the ends, the rope is tied off to a ground level support, or a stake driven into the ground if none exists. Keep in mind, however, that only the type of tent pegs that are intended for high-wind areas (i.e. the type with prongs) should be used. Otherwise, drive a long stake into the ground. The wires of this antenna have a rather large sail area, and the forces could easily pull out the ground supports over time.

The inverted-'vee' antenna is fed in the center with 75 ohm coaxial cable and a 1:1 BALUN transformer. In some cases, you will find that 52 ohm coaxial cable is a better match. The angle between the radiator elements at the top of the mast should be 90-1208. The directivity is based on the figure '8' of a dipole, but with considerable filling in of the nulls off the ends.

The length of the wire elements can be a quarter wavelength each (a half wavelength overall), or any odd integer multiple of a quarter wavelength.

Because of the sloping of the elements, the actual length is of the order of 5-6%longer than for an equivalent horizontal dipole at the same frequency.

The quarter-wavelength element lengths are found from

L_meters = 76 F_MHz meters

As with all antennas, the actual lengths required will be found experimentally using the calculated figure as a starting point.

FIG. 10

SLOPING DIPOLE

The sloping dipole ( FIG. 11) is a half wavelength dipole. Like the inverted-'vee', this dipole slants from a support (mast, tree, rooftop) to the ground. Unlike the 'vee', however, the sloping dipole has the entire half wavelength in one sloping element. The length of each quarter-wave length element is found from the same equation as for the inverted-'vee'.

VERTICAL DIPOLE

A refinement of the sloping dipole is the vertical dipole shown in FIG. 12. Like all verticals, the azimuthal pattern of this antenna is omnidirectional if nothing distorts it. The vertical dipole antenna works especially well in locations where there is limited space, and the desired frequencies are in the higher end of the HF spectrum. Lower-frequency vertical dipoles can be accommodated if a support of sufficient height is available. The lengths of the quarter-wavelength elements are found from 150/F_MHz, although this length will almost certainly need trimming to find the actual length.

The feed shown in FIG. 12 is a direct feed with 75 ohm coaxial cable.

If you wish to substitute a 1:1 BALUN, however, feel free. Indeed, it’s recommended.

FIG. 11

DELTA-FED DIPOLE

FIG. 12

Most of the dipoles shown thus far are fed with coaxial cable. The delta-fed half-wavelength dipole ( FIG. 13) uses parallel feedline (either open line or 450 ohm twin-lead) and a delta match scheme. The feedline is non-resonant, and must be connected to an antenna tuning unit that has a balanced out put. The lengths are found from:

A meters = 142 F_MHz meters

B meters = 54 F_MHz meters

C meters = 45 F_MHz meters

The delta-fed dipole was very popular prior to World War II, and still finds some adherents.

FIG. 13

BOW-TIE DIPOLE

FIG. 14 shows the 'bow-tie dipole' wide-band antenna. It’s not quite like the wide-band antenna shown in FIG. 6 because both wires in both elements are cut to the same frequency. Wide-band operation is achieved by spreading the ends of the two wires in each element approximately 11% of the total length. Length L (in meters) is found from 127/F_MHz, and the spread width W = 0:11L. For a frequency of 4.5MHz, therefore, the overall length of the antenna is 127/4.5 = 28.22m, so W = 0:11 _ 28:22 = 3:1m.

The bandwidth of the antenna is achieved because spreading the ends supposedly controls impedance excursions that would normally occur when the frequency departs from resonance.

FIG. 14

FIG. 15

WIDER BANDWIDTH FOLDED DIPOLE

FIG. 15 shows a wider bandwidth folded dipole antenna. This antenna is made from 300 ohm twin-lead. Like all folded dipoles, the parallel conductors are shorted at the ends, and one conductor is broken at its center to accommodate either a 300 ohm twin-lead feedline to the receiver, or a 75 ohm coaxial cable coupled to the antenna through a 4:1 BALUN transformer. In the antenna of FIG. 15, however, there is a difference from the normal folded dipole format: shorts are placed between the two conductors at very critical distances (B) from the feedpoint. The overall length (A) of the folded dipole is given by 142/F_MHz, while the distance between the inner shorts (B-B) is 122/F_MHz, i.e. the distance from the feedpoint to each inner short is 61/F_MHz.

FIG. 16

DOUBLE EXTENDED ZEPP

The double extended Zepp antenna ( FIG. 16A) is a dipole in which the overall length is 1.28j, and each element is 0.64j. It provides about 3 dB gain over a dipole. The azimuthal pattern ( FIG. 16B) is a narrowed figure '8' with side-lobes appearing at _358 from the line along the antenna length.

The length of each 0.64j element is found from

L_meters = 163 F_MHz meters

This antenna must be fed with parallel line connected to an antenna tuning unit with a balanced output. However, if you build a one-eighth wavelength matching section from 450 ohm twin-lead, then you can reduce the impedance to 150 ohms. The length of the 450 ohm twin-lead matching section is

L_meters = 31:4 F_MHz meters

The impedance looking into the matching section is 150 ohms, so using 75 ohm coaxial cable results in a 2:1 VSWR. However, if a 4:1 BALUN transformer is provided, then the impedance is transformed to 150=4 = 37:5 ohms. If 52 ohm coaxial cable is used with this scheme, then the

VSWR = 52=37:5 = 1:4:1.

MULTIBAND TUNED DOUBLETS

An antenna that was quite popular before World War II with ham operators and short-wave listeners is still a strong contender today for those who want a multiband antenna, provided that the bands are harmonically related.

FIG. 17 shows the multiband tuned doublet antenna. This antenna consists of a nominal half-wavelength wire radiator that has a physical length (in meters) of 145/F_MHz, where F_MHz is the lowest frequency (expressed in megahertz) of operation. Note this difference: many antennas are designed for the center frequency of the band of interest, while this one is designed for the lowest frequency. The antenna will perform well on harmonics of this frequency, and will perform at least somewhat on other frequencies as well if a higher VSWR can be tolerated. This situation is seen for receiver operators, but is often a limiting case for transmitter operators.

The feed of this antenna is through a quarter-wavelength matching section made of 450 ohm twin-lead transmission line. The length of the quarter-wavelength section is 73/F_MHz. Although one may get away with replacing the heavier 450 ohm line with lighter 300 ohm line, it’s not appropriate to replace the line with coaxial cable and a 4:1 BALUN as is done in other antennas. The reason for this is that the twin-lead forms a set of tuned feeders, so is not easily replaced with an untuned form of line.

The lengths for typical antennas of this sort are given in TBL.1.

FIG. 17

TBL.1 Lower frequency (MHz) Radiator element (m) Matching section (m)

One consequence of this configuration is that a balanced antenna tuner is needed at the feed end of the matching section. The version shown here is a parallel resonant LC tank circuit that is a transformer coupled to a low impedance link to the receiver or transmitter. The inductor and capacitor for this tuner should be designed so that they have a reactance of about 600 ohms at every frequency required. This may mean a tapped or variable inductor, or band-switched inductor, as well as a variable capacitor.

Commercial antenna tuners can also be used for this application.

THE G5RV ANTENNA

FIG. 18

The G5RV antenna ( FIG. 18) is controversial for several reasons. One reason is the originator of the antenna. The name 'G5RV' is the call-sign of the claimed originator, Louis Varney, a British ham operator. Others claim that the G5RV antenna is nothing but a 1930s or 1940s vintage design by Collins Radio for the US military. However, the similarities between the Collins antenna and the G5RV are, it seems to me, at best a case of 'further development' or 'co-invention,' rather than something more sinister.

Because of the obvious differences between the two antennas, I prefer to continue the credit due to G5RV.

Another reason for the seeming controversy over the G5RV is perhaps the 'NIH syndrome' (not-invented-here). The G5RV is more popular in Europe than in the USA. In my own experience, the G5RV tends to be built in the US by antenna experimenters, and most of those who I have talked to are happy with the results.

Still another controversy is over whether, or how well, the G5RV works. One reviewer of one of my other antenna books (Joe Carr's Receiving Antenna Handbook, Universal Radio) stated that he 'wish(es) the G5RV would just go away.' The same reviewer stated that it would be better to just put up a 'dipole of the same size.' Wrong! The dipole is a single-band resonant antenna, whereas the G5RV will work on several harmonically related bands. The G5RV has two poles, but it does not exactly fit into the same category as the half-wavelength dipole.

PHYSICAL STRUCTURE OF THE G5RV

The G5RV antenna looks like a dipole, to be sure, but its length is considerably longer. Unlike many multiband antennas, the G5RV is not cut to the lowest frequency of operation, but rather to the middle frequency. For an HF ham band antenna, one designs it for 20m (14 MHz).

Like the dipole, the G5RV is fed in the center. Unlike the dipole, a matching section made of 450 or 300 ohm twin-lead transmission line (450 ohm is preferred) is connected between the antenna feedpoint and the 75 ohm coaxial cable. The length of each radiator element (A)is

A = 220 F_MHz meters

While the length of the matching section (B)is

B =146V F_MHz meters

A is the length of each radiator element, B is the length of the 450 ohm matching section, F_MHz is the middle frequency of operation, and V is the velocity factor of the twin-lead (typically 0.82 for twin-lead and 0.99 for open-wire parallel line).

In case you don’t like to do arithmetic, the calculations have already been done for the HF ham bands: A = 15:55 meters (2A = 31:1m overall), and B = 10:37m for open-wire transmission line and 8.38m for 300 or 450 ohm twin-lead. There is some argument over these figure s, but they are regarded by many hams who have actually used the antenna as a good trade-off.

If you want more technical details on the G5RV, Louis Varney, the inventor, has written about the antenna in a number of publications.

Varney gives the basis for operation of the G5RV antenna at 3.5, 7, 10, 14, 18, 21, 24, and 28 MHz. The VSWR on each band is a bit different, and Varney recommends the use of a trans-match or similar coaxial-to-coaxial antenna tuning unit between the transmitter and the input to the coaxial cable transmission line.

There is a possibility of an unbalanced line condition existing that causes some radiation from the feedline - and for transmitters that can cause television interference (TVI) and other forms of unpleasantness. The solution to this problem is to wind the coaxial cable into an in-line choke at the point where the coaxial cable connects to the twin-lead or parallel line matching section. This is done by winding the coaxial cable into a 15 cm diameter coil of 10 turns right at the feedpoint. The coiled coaxial cable can be secured to the center insulator by tape, string, or some other mechanism.

BALUNS ON RECEIVER ANTENNAS?

Above I mentioned a reviewer who did not like the G5RV antenna. The same reviewer stated that he did not believe that 1:1 BALUN transformers are useful on half-wavelength dipole receiving antennas. My calm, professional, well-considered response is: rubbish! The purpose of using a 1:1 BALUN (which after all provides no impedance transformation) is twofold.

First, as in the transmit case, the BALUN prevents radiation from the transmission line. Perhaps the reviewer was thinking of TVI as the reason for avoiding feedline radiation. But, as ample test chamber evidence shows, the radiation from the feedline tends to distort the figure '8' azimuthal pattern. When a BALUN is used, the currents are balanced, and the radiation pattern is restored. And guess what? Antennas are reciprocal in nature - they work the same on receive as on transmit.

The second reason is that the receiver antenna feedline may pick up strong signals from powerful local stations. Other hams and AM broadcast band stations are particular problems. Any large signal at the input may challenge even the best receiver front-end, but if the receiver design is in any way mediocre in the dynamic range department (and many are!), then the signals picked up on the dipole transmission line shield can overload the front-end of the receiver. Using a 1:1 BALUN transformer between the antenna feedline and the radiator elements balances out the cur rents and seriously reduces the amount of signal seen by the receiver input.

LINEAR LOADED WIDE-BAND DOUBLET

FIG. 19 shows an antenna that superficially resembles both the folded dipole and the G5RV, but is actually a wide-band antenna with a linear loading section between the radiators. The overall length of the main radiator elements is 146/F_MHz meters, while the linear loading element is 133/ F_MHz meters long. The radiator elements are separated about 30.5 cm, with the linear loading element placed in the middle, 15 cm from each element.

The linear loading element is placed horizontally about the mid-point of the antenna. The spreaders can be ceramic, or made of any insulating material.

In one variant that I built, ordinary wooden dowels (1 cm diameter) were used, although I am not too sure it will weather well unless treated with some kind of coating.

This antenna is fed through 'any' length of 450 ohm twin-lead, but practical advice from several people who have actually built this antenna suggest that 'any' should be read as 'one-fifth to one-half the length of the radiator elements.' A 9:1 BALUN transformer is connected to the feedpoint, and coaxial cable routed to the receiver.

One practical problem with this antenna is that it tends to flop over because it’s top heavy. Two solutions present themselves. First, don’t tie off the ends in a single rope, but rather attach two ropes to the supporting structure at each end. Second, tie rope or fishing line from the bottom radiator at each end to a supporting structure below, or to stakes in the ground. One fellow used fishing weights for this purpose. The weights were attached to the bottom radiator element at the far ends.

FIG. 19

THE AUSSIE BROAD-BAND HF DOUBLET FOR RECEIVERS

One of the problems with most of the antennas in this Section is that they work on a single resonant frequency, and a narrow range of frequencies around it. Some of them will work over relatively wide bandwidths, but the coverage is nonetheless limited with respect to the entire HF spectrum. The Aussie doublet shown in FIG. 20 is a receive antenna with a VSWR less than 2.5:1 over the entire 3-30 MHz range of the HF band.

This antenna is rather large, being 40m long and 1.8m wide, so requires some space to erect. Also, the two conductors of each segment are separated by insulating spacers. One approach to making these very long (1.8m) spacers is to use PVC plumbing pipe or small size lumber. Even then, the antenna will tend to flop over and lay flat unless rope moorings to the ground are provided along the bottom conductor.

Part of the broad-banding effect is provided by the two-wire design and their spacing, but much of it’s also provided by a pair of RL networks ('RLN' in FIG. 20). Each network is a 390 ohm resistor in parallel with an 18 mH inductor. It’s these components that limit the use of this antenna for transmitting, although moderate powers can be accommodated if a suitable non-inductive, high-power resistor and a suitably designed inductor are provided.

FIG. 20

THE CAPACITOR-TUNED WIDE-BAND DIPOLE

Dipole antennas are resonant and so will operate effectively over a relatively narrow band, as well as the third harmonic of that band. It’s possible to broaden the response of the antenna by inserting a variable capacitor at the feedpoint ( FIG. 21). This capacitor should have a maximum value of 500 pF, although 730 and 1100 pF have been successfully used. The latter two values represent the values obtained when two or three sections of a 365 pF broadcast variable capacitor are used (receive only). For transmitting, a wide air gap capacitor ('transmitting variables') or vacuum capacitor should be used.

The best solution for using this antenna is to mount the capacitor at the feedpoint inside of a weatherproof housing, and then use a low-voltage, low RPM direct-current (DC) motor to drive it. Voltage to the motor can be supplied through a separate wire, using the coaxial cable shield as the DC return line. In another scheme, it’s also possible to use the coaxial cable center conductor to carry the DC (with the shield acting as the return), but it’s necessary to separate the motor and signal line with an RF choke at both ends of the transmission line. IT IS VERY IMPORTANT TO USE LOW VOLTAGE DC FOR THIS JOB. DO NOT USE AC POWER MAINS VOLTAGE!

FIG. 21

THE INDUCTOR-LOADED SHORTENED DIPOLE

At low frequencies, dipole antennas can take up a large amount of space. At 75/80m ham bands, For example, a half-wavelength dipole is 41m long; at 5 MHz, the length is still 28.5m. This fact of physics can ruin the plans of many radio enthusiasts because of space limitations. There is, however, a method for making a dipole one-half the normal size. FIG. 22 shows a dipole that is one-half the size of the normal half-wavelength dipole, even though it simulates the electrical half wavelength.

FIG. 22

An antenna that is too short for its operating frequency offers a capacitive reactance component to the impedance at the feedpoint. In order to tune out this impedance the opposite form of reactance, inductive reactance, is placed in series with each radiator element. Although the inductors can be placed anywhere along the line, and the overall length can be any fraction of a 'full- size' half-wavelength, the design gets a bit messy. A compromise is found by limiting our selections: (1) the overall length is one-half the normal length; and (2) the inductors are placed at the center of each element. If these rules are followed, then we have a shot at making this antenna with only a small amount of discomfort in the math zone. In this limited case, the reactance of each coil is 950 ohms in the middle of the band of interest. The procedure is as follows:

(a) Calculate the overall length of the elements: L_meters = 37:5=F_MHz. This is the overall length of the half-size antenna.

(b) Divide the number obtained in step (a) by 2. This is the length of each segment of wire (A or B). In our limited case, the assumption is that A = B, but in some other cases this assumption might not be true.

(c) Calculate the inductance required to produce a 950 ohm reactance at the design frequency:

L mH = XL _ 106 2_F_MHz

Antennas don’t give up convenience for free, however. The price of this antenna design is narrower bandwidth.

MORE MULTIBAND DIPOLES

The shortened inductively loaded dipole of FIG. 22 can be used on two bands by increasing the reactance of the coil to 5500 ohms at the middle frequency of the highest band. For example, you can make a 75/80m plus 40m dipole. The inductor in this case serves not only to foreshorten the antenna's overall length, but also acts as an RF choke to the higher frequency band. In that case, the sections labeled A in FIG. 22 are for the higher-frequency band, while A ) B is for the lower-frequency band.

Approximations of the lengths are found from

A = 74/FHI(MHz) meters

B = 4:64 /FLO(MHz) meters

In this antenna, the sections A act like a regular half-wavelength dipole at the higher frequency, while the sections A ) B act like a shortened inductively loaded dipole at the lower frequency. You are well advised to leave extra space on the B-sections for adjustments. This antenna tends to be narrower in bandwidth on the lower band, so may require some fiddling with lengths to achieve the desired resonance.

Another approach to multi-banding is shown in FIG. 23. This antenna is called a trap dipole because it uses a pair of parallel resonant LC networks to trap the higher frequency. A parallel resonant network presents a high impedance to signals at its resonant frequency, but a low impedance to frequencies removed from resonance. Like the antenna above, the sections marked A are a half-wavelength dipole at the higher frequency, while the A ) B sections form a near-half wavelength at the lower frequency. These lengths are approximated by the normal half-wavelength equations, although some adjustment is in order due to the effects of the trap. The traps consist of an inductor (L) and capacitor (C) in parallel. At resonance, XL = XC = 370 ohms.

FIG. 23

MULTIBAND RESONANT ANTENNAS: MULTIBAND TUNED DOUBLET

An antenna that was quite popular before World War II is still a strong contender today for those who want a multiband antenna, provided that the bands are harmonically related. FIG. 24 shows the multiband tuned doublet antenna. It consists of a nominal half wavelength wire radiator.

The antenna will perform well on harmonics of this frequency, and will perform at least somewhat on other frequencies as well if a higher VSWR can be tolerated.

The feed of this antenna is through a matching section of 450 ohm twin lead transmission line. Although one may get away with replacing the heavier 450 ohm line with lighter 300 ohm line, it's not appropriate to replace the line with coaxial cable and a 4:1 BALUN as is done in other antennas. The reason for this is that the twin-lead forms a set of tuned feeders, so is not easily replaced with an untuned form of line. One con sequence of this configuration is that a balanced antenna tuner is needed at the feed end of the matching section. The version shown here is a parallel resonant L-C tank circuit that is transformer coupled to a low impedance link to the receiver.

The antenna is a half wavelength long at the lowest frequency of operation, and the twin-lead is a quarter wavelength long, or an odd integer multiple of quarter wavelengths. The lengths are:

A = 144 F_MHz B = 73 F_MHz

This antenna depends on an antenna tuning unit (ATU) to be effective, unless the transmitter has a balanced output that can handle the high impedance. The ATU needs to be the type that has a balanced output, rather than the coaxial cable 'line flattener' variety.

FIG. 24

THE TILTED, CENTER-FED TERMINATED, FOLDED DIPOLE

FIG. 25 and the following figure s show the tilted, center-fed terminated, folded dipole (TCFTFD) antenna, which is a special case of a loop antenna and a folded dipole antenna. The inventor, Navy captain G.L. Countryman (W3HH), once called it a 'squashed rhombic' antenna. The antenna is a widely spread folded dipole, and is shorter than a conventional folded dipole. It must be mounted as a sloper, with an angle from its upper vertical support of 20 to 408.

The feeding of the antenna is conventional, with a feedpoint impedance close to 300 ohm.A75_ coaxial cable is connected to the bottom half of the antenna through a BALUN transformer that has a 4:1 impedance ratio.

At the top side of the antenna, the 'feedpoint' is occupied with a termination resistor of 370 to 430 ohm (390 ohm, 1 or 2 watts, makes a good compromise for receiving antennas).

The spread (W) of the antenna wire elements is found from:

W = 2:99 F_MHz

…where W is the width of the antenna in meters (m) and F_MHz is the frequency in megahertz.

The spreaders are preferably ceramic, strong plastic, or thick-walled PVC pipe. The spreaders can be made of wood (1 _ 2 stock or 1-inch dowels) for receive antennas if the wood is properly varnished against the weather.

The overall length of the antenna is calculated a little differently from most antennas. We need to calculate the lengths from the feedpoint to the middle of the spreaders, which is also the length from the middle of the spreaders and the terminating resistor. These lengths are found from:

L = 54:3 F_MHz

…where L is the length in meters (m) and F_MHz is the frequency in megahertz.

Four sections of wire, each with a length defined by the equation above, are needed to make this antenna.

The height of the upper antenna support is determined by trigonometry from the length of the antenna from end-to-end (not the length calculated for D, but approximately twice that length), and the angle. For example, at 7 MHz the lengths are 7.76 meters, and the spreaders are 0.47 cm. Thus, the overall physical length, counting the two element lengths and half of both spreader lengths, is [2 _ 7:76 meters - (2 _ 0:5)] meters, or 16.52 meters. If the angle of mounting is 30 degree, then the antenna forms the hypotenuse of a 60=30 right angled triangle. If we allow two meters for the lower support, then the upper support is:

FIG. 25

height = 2 ) LcosQ= 2 )( 16:52 cos 60) = 10:26 meters

This antenna has a low angle of radiation, and at a tilt angle of 308 (considered ideal) it’s nearly omnidirectional.

The termination resistor can be mounted on a small piece of plastic, or alternatively, as shown in FIG. 26, it can be stretched across the end insulator in the manner of the inductors in the previous section. Use a 390_, 2 watt resistor for this application.

FIG. 26

FIG. 27

FIG. 28

FIG. 29

FIG. 27 shows the installation of the antenna. The upper support and lower support are made of wood, or some other insulating material. There are ropes tying the ends of the antenna to the supports. Detail of the spreaders and end insulators are shown in FIGs 28 and 29, respectively.

Prev. | Next

top of page   Home