Contacter l'auteur / Contact the author

Recherche dans ce site / Search in this site



From Long wire to Yagi

Dipole (III)

A dipole is nothing else than a long wire cut in the middle and fed by the center. Its overall length is 1/2l or about 10 m for the very popular 20 m band. One segment is directed to the ground, the other one to your transceiver. Tight "flat-top" (horizontally over ground) it is horizontally polarized and both segments are symmetrical. There are however many models of dipoles. We will review the most well-know of them, from which most others are inspired.

At left, the famous G5RV 31m long (102 ft) multi-band dipole tight in my garden. The insert shows a close-up on the copper ladder line constituting the matching section. A balun 4:1 or an external antenna tuner stays mandatory to adapt the impedance to the 50W of the transceiver. At center, the radiation patterns of the G5RV tight 1/4l (5m high for the 20m band) above ground. To get many more lobes, the dipole should be placed above 1/2l high. You can also reduce the "thickness of the donut" in using either a colinear version, superposing two dipoles working in phase or increasing its length from 1/2l to more than 2l. But the fire angle will not change much, excepted that the azimuthal radiation pattern will show still more secundary lobes. The fire angle will only change if you work at higher frequency, if you increase its height above ground (over 1/2l) or modifying its design (inverted-V, V-beam, etc). At right, the coverage at short distance of this G5RV placed 1/2l high in the E-W direction calculated with MultiProp taking into account propagation conditions. Documents T.Lombry.

The LÚvy or half-wave dipole

Name after the French engineer Lucien LÚvy, the half-wave dipole is the mother of all dipoles. This classic antenna is the most common aerial usually find to amateurs who wish to preserve their budget, the visibility of their installation as well as the ease of the building.

The half-wave dipole is an harmonic antenna that prevents well interferences when it is tight horizontally. Its resonance length depends on the ratio of the length of the conductor to its diameter; the smaller is this ratio (using thicker wire), the shorter the antenna for a given electrical length.

The aerial length calculated using the formula given in the first page is for the centre of the band and insulators are attached at the end of segments.

The impedance of a dipole depends on several factors. If you modify its shape, tightening it in zig-zag or in slope, if you modify its height above ground and the conductor diameter, you will affect its impedance at the feed point and thus the SWR, all the more if you want to feed it with a coaxial. So, thicker is the conductor, lower will be the Q-factor and larger the bandwidth.

By design a dipole is a monoband antenna, cut for a specific frequency and you cannot use it on other bands without losing much power. Most amateurs wishing to work on more than one band, one developped the multi-band dipole.

Multi-band dipole

Theoretically, a 40 m long LÚvy is a multi-band dipole able to work with efficiency from... 2 to 160 m ! In the field you only loose 3 dB on 160 m using a quality antenna tuner. Of course such an antenna is not really your best choice to work on VHF.

You can also connect by their center several dipoles and fed them with a common feeder. This way your antenna system can work on several bands using a minimum of space. In this version of the half-wave dipole, each insulated end can be tight to any convenient support and the dipoles need not all be in the same plane.

Using the properties of electromagnetism, as we told before a dipole cut to be at resonance at a certain frequency will also be at resonance on its harmonics, for example at three half-waves higher, eliminating the need for an additional aerial. So if you built a dipole for the 40 m band it will be tuned for the 15 m band too. This is the way the famous G5RV multi-band dipole works as well as all harmonic antennas. 

If it is well cut and placed high enough above the ground (1/2l - 1l or at least 8 m high) to preserve the antenna takeoff angle, such a dipole provides an excellence signal. Some amateurs say even that it performs better than any vertical. I rather should say that it works another way with a different radiation pattern, it captures surely less QRM due to its horizontal polarization and constitutes an excellent complementary antenna, specially for the low bands.

If dipoles are required for optimum performance on several frequency bands they can be connected in parallel at their centres and fed with a common feeder thus providing multi-band facilities in a minimum of space. However the G5RV uses only 2 segments.

Another way to build a multi-band dipole when space is limited is to use traps, consisting in parallel tuned circuits inserted in the two dipole segments. Imagine that you want to build an aerial suited for the 20 m and 15 m bands. On each segment (the ones running from the isolator to the feeder) insert a trap. The full length of your segment is at resonance on the 20 m band while the length going from the feeder to the trap is at resonance for the 15 m band.

Traps should be designed to resonate at 21 MHz., isolating practically the end sections of the dipole from the feeder at that frequency. On 14 MHz, the traps would have a low impedance, and the whole come into use. Some trimming of the segments is however necessary compared to the calculated lengths to compensate for the effects of the traps.

Folded Dipole

In this version, the flat-top dipole is folded on itself in shape of inverted-V. The length of the wire is 1/2l as expressed in the previous formula but the end are folded to the center up to get an opening angle of about 110-120░ wide, and the center part is connected to the feeder. If a dipole displays an impedance of roughly 70W, the fact to fold the dipole transformsthe input impedance by a factor of 22x70W = 280W so that it can be fed with a 300W ribbon feeder (open wire or "ladder line"), provided the wires forming the folded dipole are the same diameter.

At left commonly used dipoles. Note that the feedline always leaves the antenna at 90░ to create the less perturbative effects as induced currents by RF energy which couple into the feedline and cause RFI. Keeping the feedline perpendicular to the dipole also help in limiting the amount of RF appearing on the coaxial feedline, it helps to keep the open-wire feedline balanced and most important for SWLs, minimizes the degree to which the feedline radiation messes up the antenna pattern. The dipole height is determined by the length of the open-wire portion of the feed that should not lay on the ground or have abrupt bends in it as this changes its impedance matching characteristics. For a receive only antenna this may not have too great an affect on what you will hear, but probably will have some affect. As long as the open-wire (or Twin-Lead) portion of the feedline is not touching the ground, it is not very critical how to coaxial portion of the feedline is run to the transceiver. It can be lay over the ground, run under ground, or however you have to do it. Watch only for the coax when you cut the grass ! At right specifications of an inverted-V. Documents T.Lombry.

Another solution is to use a 300W ribbon feeder for both the aerial and the feeder. Only one conductor of the aerial is cut at the center, the feeder is inserted and the joints soldered. The junction should be clamped between pieces of isolating material and properly proofed. The ends of the aerial are shorted to close the circuit.

In a three-fold dipole you need a third wire at the center that will be attached to the feeder and the ends linked to the external loop. Here the input impedance increases to 32x70W = 630W which provides a good match to an open line feeder or a double ribbon using insolating spreaders.

The center part contained the feeder and the open parts of the dipole can be connected together using a home-made insulating block or better, using a T-connector that provides mechanical anchorage and watertight termination for the feed line.

Usually the isolators of an inverted-V or OCF dipole are fixed ~30 to 50 cm above ground, the feeder being hanged at about 10 m high (what requires a trick to suspend the balun if there is one in placing it e.g. into a small container equipped with a ring). However, nothing prevent you to attach the ends a few meters above ground, or even against a wall, trees or a fence. I even should say that the higher the weaker the ground effects. Try only to avoid to fix your dipole over an object made of reinforced concrete or metal to avoid QRM or interferences. The best do to is to set it up over a lawn or even a salty water surface.

Feeders and insulators : the good practices

On dipoles the wires of the feeder are connected either side of the insulator. With dipoles of low impedance (50-80 ohms) you can use either a flat twin feeder constituted of 2 parallel wires (at left) or a coaxial cable (at right). Feeders requiring higher impedances (240 to 600W) need a two wires feeder spaced apart by a larger insulating material. A balun (4:1 for a G5RV of 31 m long) or a choke can be used between the feedline and the antenna to adapt the high impedance of the open wire to the coaxial feeder of 50 ohm. When properly balanced, you should not have any significant amount of signal on the outside of the coax shield; it will all be inside the coax, and won't couple into the ground. In all cases one of the two wires connected to one of the segments (the wire at left or the braided conductor at right) should go to the ground before entering the house for lightning protection. Think also to include drip loops to keep rainwater from getting into the house, or into the connectors. At center a hardware insulator kit Delta C sold by Universal Radio. This accessory must be very resistant : to warm and cold temperatures, to solar UV radiations, moisture and mechanical pressure (tension, wind gusts, etc).

W3DZZ, a trap dipole

At last C.L. Buchanan, W3DZZ, was the first ham to be able to create a trap antenna for the first five-pre WARC amateur bands from 3.5 to 30 MHz. This is a dipole 32.4 m long (108 ft) fed with a 70-ohm Twin-Lead. The dipole contains one trap on each segment at a distance of 9.75m (32 ft) from center. Traps are made of a parallel circuit constituted of a coil of 8.2 mH and a capacitor of 60 pF. 

After more than 60 years of trials and errors, multi-band antennas count today by tens of models and became true competitors. Of course as we explained in other pages, each trap inserted on an antenna reduces accordingly the efficiency of the system : the loss per trap is ranging between 0.2 and 10.5% (0.006 to 0.5 dB) depending on the frequency, the lowest the highest loss. 

OCF, Windom, FD4, and Carolina Windom

Historically all began with the Windom, a multi-band antenna first described in 1929 by Loren G. Windom, W8GZ in QST for general coverage of HF bands. It is an hybrid antenna mixing properties of a dipole and a vertical. It is sometimes called Conrad-Windom or FD4.

As shown below, the Windom uses a single wire feeder tapped on to an horizontal electric wire between 14-38% off-center, hence its name of off-center-fed dipole or OCF. Don't be surprise of the variable offset. It depends only on the required impedance, e.g. 300W or 450W on the feedline ended with a 52W coaxial.

Since the impedance will also depends on the frequency, using the Windom across a wide frequency range becomes difficult and, on some models, WARC frequencies can be impossible to tune (SWR > 10:1). Other drawback, both wire segments being unbalanced, the coaxial feeding the wire radiates RF up to the shack, and amateurs using this design must be careful using high power.

Today the impedance matching is often achieved using directly a balun without feeder. This is the solution provided by WiMo for their 40 m long Windom displayed below that comes with a 6:1 ordinary balun in T-shape placed at 1/3 of the length and ended with a PL connector (SO-239). This model can be tuned on all bands from 80 to 10 m including WARC.

To read : Wire antennas (incl. for Top band), M0MTJ

Off center fed dipole (OCF).

The Windom design is as simple as a dipole. The difference is that the feeder is off-centered and the ladder line is replaced with a coaxial. At right its installation 7 m high in my backyard. Tight exactly in the same conditions as a G5RV, it gave better results than the dipole and I was able to work JA and VK stations bare foot (100 W in SSB) from Belgium. Documents T.Lombry.

How does it perform ? I used a G5RV dipole 31 m long (102') and a Windom 40 m long (133'), both tight 7m high. In fact the Windom radiation pattern is much more omnidirectional than the dipole one and I don't have noticed more QRM. From what I have experimented, I received better signal reports with the Windom and I confirmed countries I was never able to reach with the dipole. Tight in the E-W direction, from Europe on 20, 17 and 15m bands I reached without much difficulties and often almost at first call North America (VE7, W6, 8P), South America (HK, YV, PY), Africa (SU, 9J, Z2), Middle-east (9K, A4, YI), Asia (VU, 4S, JA), South Asia (YB, XX9, VR) and many VK stations. Globally working bare foot with 100 W PEP, DX signal reports were between 55 and 57. Of course with 100 W in a fixed wire antenna there were many DX (D2, HC, VR, XW, XU, etc) I was unable to work due to pileups. But results exceeded my expectations. I kept it, and sold my G5RV, HI !

The Window is completed with a new version called the Carolina Window designed by Edgard Lambert, WA4LVB, and Joe Wright, W4UEB, that turns this potential disadvantage - the feed line radiation in the original design - into a potential advantage.The original design belongs today to Radioworks.

The radiation pattern in the H-plane of a Carolina Windom (red) in which 6.6m or 20' of the coaxial, thus a vertical segment, is source of RF radiation. In this design the antenna is tight 15m over saltwater compared to a dipole (blue) of the same length used in the same conditions.

Along the feed line, the Carolina uses on top a matching unit 4:1 or 1:1 completed with a vertical radiator 6.6 m long (20') for a 40 m long antenna made of RG-8X or RG-58C followed with a current choke balun acting as line isolator, preventing RF radiation along the coaxial to reach the shack as in the original design. Then an ordinary coaxial goes to the antenna tuner. The coaxial segment inserted between the two baluns acts like a small vertical and gives to the Carolina Windom some more gain at low elevation and a more omnidirectional pattern than a dipole of the same length. Indeed, it tends to fill in the deep nulls displayed at the four points of the compass as displayed at right.

Displaying usually an overall length of 40 m, all OCF and mainly the Carolina Windom work especially well between 3.5 and 30 MHz, including on WARC bands if well tuned. Its best performance are on 15 and 10 m bands.

When tight horizontally in straight line at good height (over 8 m high, like a dipole), users of the Carolina Windom are often very satisfied of its performance as they can work both local stations (thanks to the vertical segment) and DX, as good or better than using a dipole. Compared to a G5RV multi-band dipole used in the same conditions with the same length, users noticed that the Carolina Windom offered a signal strenght up to 10 dB stronger or 1.5 S-point. Little drawback, this vertical segment picks up a little more QRM that a true dipole tight horizontally.

About feed lines and RF field

We always say that we need two wires to power a bulb, the same for an antenna. That means that the best antenna should use a balanced feed line in order that all RF energy is transmitted to the radiative element. The RF field is balanced between the two conductors and the feed line does not radiate. So works a dipole using a ladder line or any other parallel-wire feed line.

As we will see about transmission lines, many amateurs replaced the parallel-wire feed line with a coaxial, much more convenient. However this feeder is unbalanced. If a part of the RF field flows between the central conductor and the inside of the shield braid, the outer surface of the braid is free to radiate RF energy up to the shack. We saw that this is particularly true using a Windom or any other long wire antenna.

A way to prevent this RF radiation, mainly when working with high power, is to insert a 1:1 current balun at the center of the dipole in order to isolate the unbalanced coaxial feed line. The current balun will suppress the common-mode current. You can easily made it in wounding 16 turns of side-by-side #20 AWG teflon-coated wire (thus a double wire) onto a ferrite core FT114-43 or similar. Check on the Internet for more detail; there are tens of pages about its design.

Advantages and drawbacks of dipoles and OCF antennas

A dipole displays a horizontal polarization and picks up much less QRM than a vertical antenna. By design it is also very light, all the weight being almost supported by the central matching section or the balun. It is also easy to build, very cheap and give rapidly excellent results. Many amateurs use this design, with or without traps, during portable or maritime activities. If you are searching for a master choice, for what it costs, you cannot beat a dipole or an OCF antenna.

A dipole requests some place left if you tight it flat-top. An alternative is to tight it in slope, in inverted-V or inverted-L, what improves its directivity.

But a dipole is usually fixed between its supports, and if it is tight in the N-S direction you will have difficulties to work stations located in these same directions (N and S) as up to now a dipole has never radiated by its ends, HI ! The solution is to install a second dipole perpendicular to the first or a rotary dipole, but it is almost impossible to work this way on the lower bands.

Inverted-V and inverted-L half-waves antennas can solve this problem and display properties of both vertical and dipole antennas. An inverted-L is close to a inverted-V shifted of 45░ but their respective radiation patterns are different. 

Globally the Windom or the Carolina Windom counts among the OCF antennas displaying the best efficiency due to its double polarization with a radiation pattern much more omnidirectional than the classic dipole. Unfortunately these designs are so-called outmoded and thus more difficult to find.

About dipoles directivity

Drawback of dipoles, tight too close to the ground (below 1/4l high), a dipole has a tendency to fire vertically and thus to work easily near stations, closer to about 3000 km away, at the expense of DX stations. In this configuration you will experiment much difficulties to work far stations, say over 6000 km away.

Can we cover all directions using a single dipole ? In the page dealing with basics of antennas, we have explained that when a dipole is placed over 1/2l high, it becomes directive and displays a main lobe in the vertical plane close to 26░ of elevation. Placed higher, at 3/4l the main lobe still decreases of 10░ compared to the same antenna placed 1/2l high. Tight 1l high, we get a F/S ratio that can reach 18 dB, this is the typical "8-shape", with a main takeoff angle close to 15░. Now, to cover all directions, we need to turn the dipole to work DX stations with efficiency. However, for the ease or due to space restrictions, many amateurs place their dipole much lower and often at 1/4l high over the ground or about 5 m high for the 20m band. In this configuration don't be surprised to get a radiation pattern showing a main lobe close to 90░ (max at about 60░ but very similar between 40-90░).

What happens then in the horizontal plane ? Up to 1/4l high, the F/B ratio of a dipole is not measurable and it displays a F/S ratio less than 5 dB in the best case; its radiation pattern is thus almost omnidirectional. It is far to display the theoretical 8-shape ! If your dipole is tight only 1/4l high, it is thus useless to cross two dipoles at 90░ to cover all directions as it already displays an omnidirectional pattern. But if you can get a very low takeoff angle, for example close to 15░ placing your dipole 1l high, in this case you can take advantage of crossed dipoles. This solution can be successfully used on low bands (40 and 80 m) in installing crossed dipoles 40 m high, if you can.

A magnetic loop antenna works like a directional dipole. At left the impressive model from WSPLC. At right the AOR LA350 model sold by Universal Radio.

 Magnetic Loop

This aerial is a compact antenna, mainly used to work with the longest wavelength between 0.5-2 MHz. Nowadays however we find magnetic loops for HF frequencies too.

This aerial consists of an aluminium ring of about 50 mm in diameter or made of 5 to 7 turns of wire depending on the frequency (7 turns below 3 MHz, 5 turns for HF bands) tight around a framework 10 cm wide and about 1m of diameter (1/10l).

The WiMo BABY Loop tuned from 7 and 30 MHz. It performs as a dipole but it is directional, offering a F/B gain falling down from 4 to 0.3 dB from 7 to 28 MHz while the F/S gain exceeds 23 dB.

A cheap solution consists in manufacturing the frame from a 6 mm thick plywood or any plate. The wires should be wound very tight and should be kept that way (under tension the wire tends to stretch slightly). The wood blocks merely act as bracers and as support for the broom handle. The ends of the wire are connected to a 500 pF tuning capacitor fixed in the middle of the framework. A second wire, wound around the center broom connects the coaxial cable with the aerial on one side and to the transceiver and the ground on the other side and preferably to a balanced input.

The magnetic loop forms a tuned circuit which capacitor provides a low impedance fed to the receiver. Among its disadvantages, the capacitor has to be tuned for each band but you can easily fit a slow motion drive to it or to wire a small value variable trimmer in parallel with it (10-20 pF). Then its bandwidth is shorter than the one of any other antenna.

Among its advantages, the magnetic loop is highly directional and by rotating it you can virtually eliminate any interfering station. The tuning is also very sharp and its selectivity is excellent. This is a good choice if your outdoor space is limited to a balconery or a vasistas. But this antenna works also fine indoors, stand near a wall in your shack or in the attic. Due to its great mobility this kind of aerial is often used by the army too. But as we told in the first page, do not forget the potential radiation hazard if you place the loop too close from you. A good distance is over 6m.

The gain of a magnetic loop is not as high, similar to the one of a good dipole, but it is more than outweighed by the much improved signal-to-noise ratio and the directional characteristics. The direction of a station can be determined within a few degrees by nulling it out to take its bearing. In that perspective it is useful to fix below the joystick (broom-handle) a recessed slot to prevent it slipping.

Note that magnetic loop antennas also exist for receive purposes, like active models (powered in 12V) manufactured by Wellbrook in the U.K.

Next chapter

Yagi, Quad and Log-Periodic

Page 1 - 2 - 3 - 4 - 5 -

Back to:


Copyright & FAQ