Marconi and other unbalanced antennas

. Marconi antennas are those that are fed unbalanced with respect to ground ( FIG. 1). In the case of a transmitter, one side of the output network is grounded, while the other side is connected to the radiator element of the antenna. The length of the radiator element varies with type, and determines many of the operating properties of the antenna. The direction of the radiator element also varies. It may be vertical as shown (indeed, vertical antennas are a species of Marconi antenna), or at a slant. In some cases, the Marconi antenna is connected directly to the rig or receiver, but in other cases an antenna tuning network is needed.

FIG. 1

BASIC RECEIVING 'TEE' ANTENNA

Although most of the antennas in this Section can be used on either receive or transmit, this one ( FIG. 2) is recommended for receive use only unless an antenna tuner is provided. This antenna is not particularly well thought of, but is the antenna of choice for many first-time receiver owners, or those who live in a difficult antenna erection venue.

The radiator element is a 10-40m long (more or less) piece of wire, preferably in the 12 AWG (= 14 SWG), 14 AWG (= 16 SWG), or 16 AWG (= 19 SWG) size. Smaller sizes can be used, but they are not recommended for structural reasons (they break and fall down too easily).

FIG. 2

FIG. 3

THE HALF-LAMBDA-'TEE' ANTENNA

The antenna in FIG. 3 is called the half-lambda-'tee' because it’s one half wavelength long (i.e. j/2) and is a 'tee' antenna. The 'tee' is seen by the fact that a downlead section is provided, and it comes away from the radiator element at a 90-degree angle. The length of this downlead section is quarter wavelength. This length differs from the top-hat tee antenna used by many short-wave listeners where the radiator is half wavelength or longer and the downlead is a random length of wire. The lengths of these elements are

2L = 143 F_MHz meters and L = 71:5 F_MHz meters

These lengths make the antenna resonant. Examples of the proper lengths for various bands are:

Band L 2L

60m (5MHz) 14.30m 28.60m 31m (9.75 MHz) 7.33m 14.66m 20m (14.2 MHz) 5.03m 10.06m

An antenna tuner is required for the half-lambda-'tee' antenna. Both the inductor and the capacitor should have reactances of 450 ohm, i.e.

XL = 450 ohms and XC = 450 ohms. Typical values for these components are:

Band CL 60m (5 MHz) 31m (9.75 MHz) 20m (14.2 MHz) 65 pF 33 pF 23 pF 16.0 mH 8.2 mH 5.6 mH

For operation over a wider band, try using variable capacitor and/or inductor elements.

FIG. 4

THE 'QUICK 'N' DIRTY' TWIN-LEAD MARCONI ANTENNA

The antenna shown in FIG. 4 is one of the first that I ever used (when a mentor, MacIvor Parker, W4II, gave me a roll of twin-lead wire). I ran this antenna out of a basement window, and attached the top end to an old pine tree, and the middle supported by a 5 _ 10 cm piece of construction lumber nailed rather unceremoniously to a tool shed on the back of the house.

Basically a quarter-wavelength 'round robin' Marconi, it consists of a 300 or 450 ohm twin-lead transmission line with a length (in meters) of 71.5/ F_MHz. Thus, for the 60m (5 MHz) band the length would be about 14.3m long overall. The length of the coaxial cable feedline is not critical. The 52 ohm coaxial cable feedline is connected such that the center conductor goes to one conductor of the twin-lead, and the shield goes to the other.

This antenna is basically a species of Marconi antenna, so requires a good ground in order to work properly. At the minimum, one of the longer copper or copper-clad steel ground rods should be driven into the earth at or very near the feedpoint. The best method is to use both a ground rod and a system of two to eight quarter-wavelength radials as a counterpoise ground.

The 'quick 'n' dirty' (or 'round robin' as my mentor called it) can be installed at an angle (as shown), although I suspect best performance occurs when the wire is straight. In the version that I had, there was only a small voltage standing wave ratio (VSWR) with one bend that put the vertical and horizontal segments at about a 120-deg. angle.

Note the far end of the twin-lead wire (i.e. the end away from the coaxial cable). The two conductors of the twin lead are shorted together. I have encountered some confusion over this matter, after all a short is a short, isn't it? No, because this antenna is a quarter wavelength long, so the impedance reflected to the feedpoint by a short at the end is very high.

This antenna works better than single-conductor Marconi antennas because regular Marconi antennas suffer from ground losses. By supplying a return line, the radiation resistance is raised from 10 or 15 ohms to some thing of the order of 40-50 ohms - which makes it a good match to 52 ohm coaxial cable.

THE SWALLOW TAIL ANTENNA

FIG. 5 shows two variants of the swallow tail multiband antenna. The radiator elements in both FIGs 5A and 5B are cut to specific frequency bands, and are quarter wavelength (L_meters = 75=F_MHz). As many as needed to cover the bands of interest may be used, provided that they don’t bear a 3:1 frequency ratio. The reason for the constraint is that for all antennas but the resonant one the impedances are so high that connecting them in parallel with each other does not affect the overall feedpoint impedance. However, at the third harmonic, the impedance again drops low and will load down the impedance seen by the transmission line. This is another situation where the results are more profound for amateur radio transmitters than for short wave receivers, but it’s a good idea to avoid it if possible. Besides, the antenna is actually resonant on its third harmonic, so you lose nothing.

The version in FIG. 5A uses a pair of insulating masts, or support structures (roof of a house, tree, etc.) with a rope stretched between them.

The quarter-wavelength resonant wires are spread out along the length of the rope, spaced approximately evenly.

FIG. 5

The version shown in FIG. 5B uses a large wooden cross-like structure. The antenna wires are connected to the cross-piece at the top end, and to the coaxial cable at the bottom end. This antenna apparently worked well for a fellow who wrote to me recently. He lives in a townhouse community that has a 'homeowners' association' of nit-picking little dictators who like to tell people what they may do with their houses. One of the rules is 'no outdoor antennas' of any sort. He erected a mast in his rear garden shaped like a ship's mast and yardarm, and then convinced the busy-bodies snooping for 'the committee' (dread!) that it was somehow nautical, not radio.

RANDOM LENGTH MARCONI

There are several antennas that qualify as 'all-time favorites' with both ham operators and short-wave listeners. In the top three or four is the random length wire antenna. This form of antenna consists of whatever length of wire that you happen to have (although there are reasonable limits!) run from the rig or receiver to a convenient elevated support (e.g. a tree, mast, or roof of a building). FIG. 6 shows this most basic form of Marconi antenna.

Although a receiver owner can connect the random-length radiator wire to the antenna input of the receiver, the transmitter operator most certainly requires an antenna tuning unit (ATU). Even receiver operators will find this antenna a better performer when an ATU is used between the coaxial cable to the receiver and the radiator wire.

FIG. 6

One requirement of this antenna is that it requires a good ground. It’s essential that either an earth ground or a counterpoise ground (i.e. a system of radials) be provided. Methods of providing a good grounding system are described in Section 13.

There are several conditions for this form of antenna: an antenna that is shorter than a quarter wavelength, an antenna that is a quarter wavelength, and longer than a quarter wavelength. The inset in FIG. 6 shows several options for the antenna tuning unit. At position 1 is a variable inductor.

This is used when the antenna is quite short compared to a quarter wave length. At position 2 is a variable capacitor. It’s used when the antenna is longer than a quarter wavelength. When the antenna is exactly a quarter wavelength, then it may be possible to eliminate the antenna tuner. It’s also possible to use the L-section couplers at positions 3 and 4. These ATUs combine the inductor and capacitor. These are used to match the feedpoint impedance to the 52 ohm coaxial cable.

Many commercial L-section couplers are designed with either switching or shorting straps that allow any of the combinations in the inset to FIG. 6 to be realized by making some simple changes.

Receiver operators who use this antenna are well advised to use an ATU, just like transmitter operators. But there is at least one difference: you should be able to switch the ATU out of the system when you are receiving a frequency outside the limits of the tuner.

THE WINDOM ANTENNA

One of the first ham radio transmitting antennas that I ever used was not my own, but rather the antenna at my high school station in Arlington, Virginia (K4BGA, Washington-Lee High School). Although I had owned receive antennas, the K4BGA club station was one of the first I had operated. It used an HF Windom antenna running across the roof of the industrial arts building. The Windom antenna ( FIG. 7) has been around since the 1920s. Although Mr Loren Windom is credited with the design, there were actually a number of contributors. Co-workers with Windom at the University of Illinois were John Byrne, E. F. Brooke, and W. L. Everett, and they are properly co-credited. The designation of Windom as the inventor was probably due to the publication of the idea (credited to Windom) in the July 1926 issue of QST magazine. Additional (later) contributions were rendered by G2BI and GM3IAA (Jim MacIntosh). We will continue the tradition of crediting Loren Windom, with the understanding that others also contributed to this antenna design.

The Windom is a roughly half-wavelength antenna that will also work on even harmonics of the fundamental frequency. The basic premise is that the antenna radiation resistance varies from about 50 ohms to about 5000 ohms, depending upon the selected feedpoint. When fed in the exact center, a current node, the feedpoint impedance will be 50 ohms; similarly, end feeding the antenna finds a feedpoint impedance of about 5000 ohms. In FIG. 7 the feedpoint is tapped away from the center at a point that is about one-third (0.36L) the way from one end, at a point where the impedance is about 600 ohms.

The feedline for the basic Windom of FIG. 7 is an insulated length of wire. Of course, the size of the wire depends on the power level, but I suspect that No. 14 insulated stranded wire will do for most people who run less than 200W of power (indeed, I would not like to use a Windom at high power levels because of the 'radio frequency (RF) in the shack' problem).

The Windom antenna works well - but with some serious caveats. For example, the antenna has a tendency to put 'RF in the shack' because of the fact that it’s voltage fed. This is why Windom antenna users (along with those using the random-length Marconi) get little 'nips' of RF shock when touching the transmitter chassis, or putting a lip on the microphone.

Second, there is some significant radiation loss from the feedline. Finally, the antenna works poorly on odd harmonics of the fundamental frequency.

The antenna tuning unit can be either a parallel resonant, link-coupled, LC tank circuit or a _-network (i.e. high-impedance end toward the antenna).

Note that a good ground should be used with this antenna (note the ground connection at the output of the antenna tuning unit). This basically means (for most people) a 2-2.5m ground rod, or a system of radials (see earlier discussion for random-length wire antennas).

FIG. 7

A reasonable compromise Windom, which reduces feedline radiation losses, is shown in FIG. 8. In this antenna a 4:1 BALUN transformer is placed at the feedpoint, and this in turn is connected to 75 ohm coaxial cable. It’s necessary to ensure that the feedpoint is at a current node, so it’s placed a quarter wavelength from one end. Also, the antenna on the other side of the BALUN transformer should have a length that is an odd multiple of a quarter wavelength. In the case shown it’s three-quarter wavelength long.

FIG. 8

QUARTER WAVELENGTH VERTICAL

Perhaps the most classic form of Marconi antenna is the quarter wavelength vertical ( FIG. 9). It consists of a vertical radiator element that is quarter wavelength long, and a system of quarter wavelength radials for the ground system. The length of the radiator element is found from:

L_meters = 75/F_MHz

…where L_meters is the length in meters and F_MHz is the frequency in megahertz.

The radiator element can be made of pipe or wire. The former case, the pipe can be copper or aluminum, although I suspect that aluminum will be the choice of most readers. In the case of a wire radiator, it will be necessary to support the wire up top, so some form of mast or support is necessary.

Make the mast or support out of an insulating material such as wood.

The radials are somewhat optional if you use it with a 'good ground'. But in the absence of a 'good ground', I would use at least four radials and as many as sixteen. Above sixteen the benefit of adding further radials drops off substantially.

Radials should be buried if the vertical is ground mounted (that is to prevent lawsuits over pedestrians tripping over them). In the case where the vertical is mounted on a mast, off ground, the radials are not optional - they are mandatory. In that case, use four to sixteen radials arranged in a circular pattern around the base of the radiator element.

Connect the system so that the radiator element is connected to the coaxial cable center conductor, and the radials are connected to the shield of the coaxial cable. Because the feedpoint impedance of the vertical antenna is 37 ohms, use of 52-ohm coax will result in a VSWR of only 50/37 = 1.35:1. In the case where the vertical has a substantially lower impedance (as low as 2 ohms!), use a broadband transformer between the coaxial cable and the antenna.

FIG. 9

FOLDED MARCONI 'TEE' ANTENNA

The antenna shown in FIG. 10 is popular for receiving and transmitting on the lower bands (<7 MHz) when space for antennas is a bit limited.

Although it can still take a lot of space at the lowest frequencies, it’s quite a bit shorter than a half-wavelength dipole for the same frequencies.

Two pieces of 300 or 450 ohm twin-lead are used to make this antenna (A and B in FIG. 10). The A-section is the main radiator element, and it has a length of

A meters = 82 F_MHz meters

Note that section A is built like a folded dipole. At the ends of the twin lead the two conductors are shorted together. At the center point of section A one of the two conductors is cut to accept the two conductors of the twin lead used for section B.

Section B is vertical, and should come away from section A at a right angle. It has a length that is similar to that of A except that it’s reduced by the velocity factor (VF) of the transmission line:

B meters = 82(VF) meters / F_MHz

For ordinary 300 ohm twin-lead, the length of B is 0.82A.

FIG. 10

Note that the conductors making up A and B form a continuous loop of wire. The coaxial cable is connected such that the shield goes to one lead of B and the inner conductor goes to the other conductor. At the top of B the two conductors are connected to either side of the cut in section A.

As with the other Marconi antennas in this Section, it’s imperative that a good ground be used with the 'tee' Marconi. Otherwise, losses will be too great, and performance will suffer considerably.

FIG. 11 shows the end-fed Zepp antenna. This antenna has a radiator wire that is half wavelength long at the lowest frequency of operation. It will work on harmonics of that frequency as well as the frequency itself. It will also work on other frequencies if a high VSWR can be tolerated. It’s fed by 600-ohm parallel feedline and an antenna tuning unit (ATU). The ATU is used to tune the VSWR to 1:1 at the frequency of operation.

ANTENNA TUNER COAX TO RIG ROPE; ROPE END INSULATOR END INSULATORS FEEDLINE RADIATOR WIRE

FIG. 11

EWE ANTENNA

The EWE antenna emerged recently as one solution to the low-noise low band antenna problem. FIG. 12 shows the basic EWE antenna. It consists of two vertical sections (labeled L1) and a horizontal section (L2). The EWE looks superficially like a Beverage antenna, but it isn't.

Like the Beverage it’s erected about L1 = 3 meters above the Earth's surface. Unlike the Beverage, it’s only L2 = 6.5 meters long at 3.5 MHz.

Those dimensions make it affordable for most people.

The far end segment is terminated in an 850 ohm resistor. This resistor should be a carbon composition or metal film resistor, and never a wire wound resistor.

The receiver end must be matched to the receiver's 50 _ antenna input impedance. Transformer T1 is provided for this purpose. It has a turns ratio of 3:1 to provide the 9:1 impedance ratio required to match the 450 _ antenna impedance to the 50 _ receiver impedance. A powdered iron toroid core made of -2, -6 or -15 material will be sufficient. A suitable transformer can be made using a T-50-15 (red/white) core. Use about 20 turns of any size enameled wire.

The azimuthal and elevation patterns for the Koontz EWE antenna are shown in FIGs 13, 14, 15 and 16. These patterns were simulated from the Nec-WIN Basic software available from Nittany-Scientific. The patterns in FIGs 13 and 14 are based on the Sommerfield-Norton standard ground model, with the azimuth being seen in FIG. 13 and the elevation in FIG. 14. The same types of pattern are seen for a 'real' ground based on suburban soil and are shown in FIGs 15 (azimuth) and 16 (elevation).

FIG. 12

FIG. 13 Low-Noise RX Antenna

FIG. 14 Low-Noise RX Antenna

FIG. 15 Low-Noise RX Antenna Rocky Soil

FIG. 16 Low-Noise RX Antenna Rocky Soil

REVERSIBLE EWE

The EWE antenna can be made reversible by using a system such as FIG. 17. The feedpoint and termination circuits are co-located at the receiver.

Transformer T1, coil L1, resistor R1 and DPDT switch S1 are installed in a shielded metal box. The outputs of the box (i.e. center terminals of the DPDT switch) are connected to the bases of the vertical (L1) sections of the EWE antenna. According to one source, the simple resistive termination was not sufficient, so they added an inductive reactance in series with a resistance. This is the method used on Beverage antennas to make a steer able null, and that effect is seen on the EWE as well.

FIG. 17

DUAL EWE ANTENNA

FIG. 18A shows a modification of the EWE antenna that permits switch able bi-directionality. Four EWE antennas are arranged in north-south (N-S) and East-West (E-W) directions. The feedpoints (A, B, C and D) are connected to a switch circuit such as shown in FIG. 18B. The directivity of the antenna is controlled by opening and closing the four switches (S1-S4).

FIG. 18A

FIG. 18B

Prev. | Next

top of page   Home