Why Are Antennas Built
to Look Like They Do?

Published by The Tech Bench Elmers Amateur Radio Society

By: John Wendt WA6BFH

We come to recognize the proportionate shape and appearance of antennas. If we see a half wavelength dipole we recognize it for the antenna it is. When we see a Ground Plane antenna we know what it is. Its just the same as when we see a Ford automobile next to a Volkswagen we know which is which. It is possible though for Ford to build a car that looks like a Volkswagen but, it's not possible to build a dipole that looks like a Ground Plane, or a "J" antenna that does not look like the letter J! Lets investigate this, and in fact we can start with the "J" antenna as our object model.

Observe that the vertical portion of the letter J is about two times higher than the portion that forms the crook of the J, or we could say that the height of the J is three times the height of the crook. It is for this reason that the J antenna got its name.

The crook portion of a J antenna forms a "Linear Impedance Matching Transformer" or "Q-Line" transformer because of these two parallel conductors that are 1/4 wavelength long. Above this Q-Line is the radiating portion or "radiating element" that is 1/2 wavelength long.

At the bottom end of this quarter wavelength Q-Line which is electrically shorted together, there is a dead short zero Ohm impedance. One quarter wavelength above this dead short is an infinitely high impedance of thousands of Ohms. This is how any Q-Line device such as a "Bazooka Balun" works.

Now some who have read this article so far might be scratching their chins about now thinking, he said the radiating element is 1/2 wavelength long. Gee, a dipole is one half wavelength long! That's right, a "J" antenna is merely an "end-fed" dipole! Another name for an end-fed dipole is a "Zepp", because this form of dipole was first used on Zeppelins. So how is the more common version of a dipole different?

In the J antenna we feed the dipole on its end at the high voltage point of the antenna. If we feed it at the center at its high current point, we will see a much lower impedance or alternating current (AC) resistance. In fact the characteristic "radiation resistance" of a center fed dipole in free space is 72 Ohms. Free space by the way means that the antenna is several wavelengths above the ground, or any other conductive object. Usually free space means at least 10 wavelengths but, for practical design considerations 3 to 5 wavelengths is often times hard enough to achieve!

What happens if we feed a dipole not at the center, and not at its end but, half way in between. This sort of dipole we call a "Windom" named after the antenna's originator. This type of dipole has a characteristic impedance or radiation resistance of 600 Ohms. This feature allows this sort of dipole to be operated on almost any frequency within several octaves of its design frequency, and always present a relatively moderate impedance and consequently a decent "SWR".

Next let's take a look at "Ground Plane" antennas, afterall, aren't they just another variation on a dipole? Well, it's certainly true that they are "current-fed" at the center of one half wavelength. If you have ever seen a Ground Plane fabricated on a chassis mount coax connector you can see how this antenna works.

You start by cutting five quarter wavelength metal rods, I have always used Brazing rod. If we were going to make such a Ground Plane for the 2 meter wavelength band we would cut these rods to about 19.25 inches. If we start by just soldering on two of them, one to the center connection, and one to one of the flange holes, we have sort of a dipole. Actually this probably looks closer to an "Inverted V" type dipole but, I think you get the picture! So, now we have one of these 1/4 wave rods connected to the center conductor of our coaxial transmission line, and one of them connected to the shield. So, why should we solder on the other three, won't the antenna work with just these two? It would work as far as the transmitter is concerned. It would have a characteristic impedance pretty close to 50 Ohms, so the transmitter would be happy! The trouble is that without the other "radials" to form a uniform "counterpoise" , the antenna is not the "omni-directional" antenna we were seeking! If we left it looking like an Inverted-V, it would have a figure-eight radiation pattern broad-side to the two rods. If we provide three radials 120 degrees from one another, or four radials 90 degrees from one another, the antenna will have an omni-directional radiation pattern. By the way, the radials really should be about 5% longer than the radiating element. Also, if the antenna has 3 radials, they will have to be bent down at a lower more acute angle to achieve a 50 Ohm impedance match to the transmission line.

So, what's the bottom line to all this palaver? Simply this, all antennas, any antenna can be analyzed as to its design by analyzing its current and voltage distribution. The end or tip of the antenna is always going to represent a high impedance and high voltage point. If we measure down 1/4 wavelength we will find a high current point and a relatively low impedance. If we follow this process all the way back to the feed-point we can determine all aspects of the antenna including the antenna's aperture size, and the aperture size will tell us the antenna's approximate gain. Every time you double the aperture size of an antenna you double its gain, which means you pick up 3 decibels of gain.

Lets check this out by looking at one last J antenna which has come to be called a "Super J". A Super J starts with a normal looking J just like we see so many of nowadays. At the tip top of this J a quarter wavelength phase de-coupling stub is added, and then another half wavelength dipole is placed on top of the phase decoupler. Guess what happens next, we gain 3 "dBd", or 3 dB's above a dipole reference! In "dBi" this would be 5.2 dB's compared to an "Isotropic" reference.

Q-Line, Bazooka Balun, or linear impedance matching transformer: All of these are electrically speaking the same thing. A Bazooka Balun only differs in that it is fabricated from two lengths of tubing, as well as a central coaxial inner conductor. These are all one quarter wavelength long!

Radiating Element: This term is both hard to closely define, and in fact is a bit of a misnomer. The vertical element in a Ground Plane is sometimes called the radiator or radiating element but, it really radiates in conjunction with other associated elements that form part of a half wavelength.

End-fed, and center fed: These terms are closely associated with the terms, "Voltage Fed and Current Fed".. At the end of a half wavelength there is an infinitely high impedance and consequently an infinitely high voltage. At the exact center of a half wavelength is an infinitely high current and virtually by contrast, no voltage and a very low impedance.

Characteristic Impedance: All conductors or wires have both some amount of inductance and some distributed capacitance, this in itself provides a "lumped constant" derived impedance. In various configurations such as two wires parallel to one another, a characteristic impedance will result. Wires that are brought more closely together will have a lower impedance as parallel capacitance rises, or if they are farther apart this impedance will rise as capacitance is reduced.

Radiation resistance: All antennas have a characteristic radiation resistance because of the comparative effects of their distributed inductance and capacitance. This can also be expressed as a current to voltage ratio. Whatever this ratio is, a characteristic impedance will result. For a dipole this is 72 Ohms, for a 1/4 wave Ground Plane with radials at 90 degrees to the radiating element this is about 34 Ohms, and for a 5/8 wavelength Ground Plane its about 90 Ohms.

SWR: Standing Wave Ratio is the term given to the measurement of current or voltage distribution as imposed within the antenna. It is usually measured as a voltage and therefore the term often used is "VSWR".. If an antenna has a radiation resistance of 72 Ohms and we feed it with 50 Ohm coaxial cable, the SWR will be 1.44:1. If we fed a 90 Ohm antenna directly with 50 Ohm cable the SWR would be 1.8 to 1 (1.8:1 or 90/50 = 1.8).

Phase de-coupling: When ever the aperture size of an antenna is increased we have to make provision for the additional antenna elements to work in phase with the other elements. On vertical omni-directional antennas this is done by phase de-coupling half wavelength radiators with quarter wavelength phase de-couplers.

Gain: Antenna gain is often times a controversial subject. It really shouldn't be, for the following reason. Every time an antenna's aperture size is doubled, its gain will double. If I properly stack one beam antenna of equal size over its predecessor I will have doubled its aperture size. If I ignore the losses imposed by the feed line and phasing network, I will have added 3 dB's of signal gain. Don't forget though, there's no free lunch. If I put a 10 dB gain antenna on a 100 foot tall tower and use poor or cheap coax cable to feed it, it may well turn out that I have less signal gain than I would have had by putting a unity gain "J" up at 30 feet with good coax.