Long / Random / End-fed Antennas
An end-fed wire often seems like the simplest type of antenna, but getting it to work well is actually rather complex. The primary issues are:
- The impedance varies over a very wide range, depending on the length in wavelengths.
- It requires some sort of ground connection to work.
- The radiation pattern also varies with length in wavelengths, and in some configurations may not be suitable for the desired coverage.
Before we go any further, however, let’s try to clear up some common points of confusion. Many people seem to think that “end-fed antenna” implies a particular configuration and matching method, and they don’t always agree on what that is. So let’s start with some examples what types of antennas we are discussing here.
sample types of end-fed antennas
- A wire stuck in the antenna connector of a radio and tossed over a tree branch.
- A quarter wave or similar vertical antenna, fed against a ground plane or ground rod. They may use traps or other methods for multiband operation. We’ll discuss those more in the Vertical Antennas pages, but it is important to realize that they really are part of this group.
- A VHF rubber ducky or whip antenna used directly on a hand-held transmitter with no coax cable.
- A “random length” wire, the length of which is sometimes carefully selected to be non-resonant on any of the bands of operation, so it really isn’t very random. The current fad is to feed them with some sort of matching transformer at the feedpoint, but there are other options.
- An “End-Fed Half-Wave (EFHW)” antenna. These may actually be multiples of a half wave: the point is to have a high impedance at the feedpoint to reduce ground losses (although other losses may occur instead). There are any number of designs using step-up transformers (another new fad), some of which can be lossy. However, they have been used for over a century without such transformers.
- An HF mobile whip antenna.
There are other examples, but these should give a sense of the wide range of types we are including.
The basic theory and requirements are the same for all of them, but there are some implementation details, particularly in how they are matched to the transmitter. However, let’s look at the group similarities before we start dividing them into pieces.
Perhaps the most important principle is that an end-fed conductor, by itself, is not a complete antenna.
At the feedpoint, where the current flows into the wire, current of the same magnitude must flow into some other conductor, and that other conductor is an important part of the antenna, even if we don’t always pay much attention to it. But sometimes that other conductor can radiate more power than the part we consider the antenna. So we need to be explicit where the rest of our current is going, otherwise it may cause high losses, distort our radiation pattern, or cause other problems.
Let’s consider an example: a quarter wave whip antenna on a VHF or UHF handheld radio. This is a common application of an end-fed antenna (although they are often a helical wound “Rubber Ducky” instead of a straight quarter wave, but the same principles apply). It is easy to imagine the current into the antenna, but where is the other current?
Typically the output of the transmitter is connected to the case of the radio, and/or the common circuit ground. If the radio has a metal case, then the other current is on the outside of the case. As hand-held radios have gotten smaller over the years, this has caused more of a problem trying to provide enough of a ground system. Adding a quarter wave ground radial hanging down from the shield on the antenna jack can make a significant difference in performance. I’ve measured over 6 dB of improvement, and that was using an older HT that was at least twice the size of many modern ones.
Years ago, Tandy / Radio Shack advertised 27 MHz “Citizens’ Band” handheld radios with “Exclusive Range Boost Circuitry”: this turned out to be a metal trim on the case, positioned so it made contact with the operator’s hand, turning their body into part of the ground system!
A small lump of metal, such as the case of a radio, can work somewhat for the “other half” of the antenna: the effect is similar when using the chassis of a car as the ground for an HF mobile antenna. The smaller it is, the less effective, and overall performance of the antenna decreases.
So, while it might not be apparent, there is always an “other half” of the antenna, into which current must flow if it is to flow on the antenna. The impedance of the “other half” affects the feedpoint impedance of the antenna, and any losses will reduce the overall efficiency.
If a suitable conductor is not provided for the “other half”, then either the current must flow as common mode current on the feedline (and possibly on any other metal object connected to the transmitter, including the AC power wiring, computers, etc.), or the antenna can’t take power.
We can divide end-fed antennas into 4 general groups, based on the required impedance matching:
- quarter wave antennas (or similar) that don’t need additional matching
- short antennas, less than a quarter wave
- half-wave antennas, or multiples thereof.
- other, including multiband operation
Note that the radiators can be wires (vertical, bent, or sloping), guyed masts, rigid whip antennas, or any combination of these.
quarter wave wires
One of the simplest antennas is a quarter wave (or 3/4 wave) radiator plugged directly into the output of transmitter (similar to using a VHF/UHF quarter wave whip on a handheld). When the radiator is properly adjusted, the impedance generally is close enough to 50 ohms that no additional matching is needed. However, performance depends critically on the ground system, especially at HF and low VHF where the radio case is far too small to provide an effective ground. (Of course, these antennas can be used with elevated radials as a ground plane antenna.)
I used such an antenna when I first got on 80m SSB: a wire running out my window and over a tree behind the house. The ground was a wire to a water faucet (back when we had metal water pipes). The wire was terminated in a banana plug, which fit the SO-238 socket on the SWR meter. The ground wire was twisted around the outside of the socket. Later I ran coax out the window and attached a coax fitting to the water faucet for the feedpoint.
Such an antenna is not ideal – the ground losses reduce the efficiency – but it was adequate to get started on the band. The ground losses brought the feedpoint impedance up to where I could get a good SWR. If I had added more ground radial wires to reduce the ground losses, then the feedpoint impedance would be lower, and the SWR would have been higher.
A 3/4 wavelength wire was popular for 40m, particularly with tube transmitters, as the impedance was commonly somewhat higher than 50 ohms.
With either length, antenna current is maximum at the feedpoint, so ground loss resistance has more of an impact on efficiency than with wires having a higher feedpoint impedance.
With a good ground, the impedance of a quarter wave wire can be as low as 20 ohms (especially if the wire is sloping or bent). One common solution is to make the wire longer than 1/4 wavelength to raise the resistance closer to 50 ohms, then add a series capacitor to tune out the additional reactance.
Radiators shorter than about 1/4 wavelength are capacitive, and generally use a series loading coil. The feedpoint resistance will be less than 50 ohms, unless ground losses are high, or some other form of matching is used.
One simple matching network is a tapped coil: the bottom of the coil is grounded, the coax is attached to a tap point a little ways up from ground, and the radiator is attached further up on the coil, where the inductance resonates the antenna. In practice it may be easier to adjust if the coax tap is fixed and the ground and radiator taps are adjusted, as they will be more independent. This is equivalent to a beta match, where a shunt reactance is connected across the feedpoint to raise the impedance: either a coil or a capacitor can be used, but a coil is often easier to adjust, and can provide a DC ground for the radiator.
For low impedances, especially on the lower frequency HF bands, tuner efficiency can be an important factor with short radiators, as the resistance is low and the reactance is high. Many tuner designs are more efficient matching higher impedance antennas than very low ones. Ground loss resistance also becomes more significant when the radiation resistance of the antenna is low, so providing a good ground system becomes more important.
An end-fed half-wave (“EFHW”) antenna has a very high feedpoint impedance, often 1000 to 5000 ohms (depending on the diameter). The radiation pattern on the fundamental frequency (where the antenna is 1/2 wavelength) will be similar to the same wire fed in the center as a dipole (but this is not the case on harmonics). Longer wires also have high feedpoint impedances when they are multiples of 1/2 wavelengths, making this a popular multiband antenna.
The high impedance has several impacts:
- the current at the feedpoint is reduced, and the voltage is higher (both by a factor of 10 for a 5000 ohm impedance)
- losses due to ground resistance are lower, due to the lower current (but there can still be other ground losses)
- matching to a 50 ohm is more difficult
- some antenna tuners won’t match such a high impedance, due to component values or voltage ratings.
End-fed half-wave antennas are popular because of the potential for lower ground losses, but without careful attention to detail, the losses due to a poorly designed matching network or lack of attention to the “other half” of the antenna can result in worse performance overall.
Common matching methods include the “L” network, parallel-tuned circuits, high ratio step-up transformers, and the “Zepp feed” (originally designed to keep the high voltage ends of the antenna away from airships filled with flammable hydrogen gas).