by Martin E. Meserve
Single Wire/End Fed Antenna
The Single Wire End Fed Antenna is probably the simplest antenna to use. But when you start looking around,
you find that no one can agree on what length to use. Some say any length will work, some stress odd multiples
of 1/4 wavelength, while others lean toward even multiples of 1/2 wavelengths. Some say design for the lowest frequency
you want to use, and the others say design for the highest. Some refer to the antenna as a
End-Fed Zep, a Inverted-L, and a few other names. Which one is right? Well... for the most part, they are all right.
You just have to dig a little deeper to determine the designer's reasoning.
Am I going to solve this dilemma here? Probably not. I will just present several ideas and you can choose which
one you might like to work with.
I will start with probably the simplest antenna you could think of. A random length of wire fed directly from the output
of your transmitter (Fig-1). The impedance of a end fed wire can be anywhere
from a few 10s of Ohms to a few thousand Ohms, depending on the frequency in use and the length of the wire.
Other factors that can affect the impedance are bends in the wire, height above ground, proximity to buildings
and wire diameter. Older tube transmitters generally have PI-Network as the output circuit. The PI-Network
output circuit is often capable of matching a pretty wide range of antenna feed impedances, without causing any issues
with the transmitter.
However, some of the PI-Network output circuits in transmitters my not have the range needed and sufficient loading
can not be obtained. Also, most of the newer transmitters have transistor output circuits and can not
handle the wide variation in load impedances, without possible damage to the output circuit components.
In that case, a L-Network (Fig-2) can be used to obtain a sufficient match between
the transmitter and the random length of wire.
The L-Network in Fig-2 should handle instances where the transmitter output impedance is less then the antenna
impedances. But what if you choose a wire length that produces a lower impedance than the transmitter output?
This is often the case when the random length of wire is less than 30 feet (10 Meters). In that case you should be able to
just switch the variable capacitor to the output of the transmitter. To cover all cases, it might be prudent
to include a switch that would allow you to switch the position of the variable capacitor (Fig-3).
Fig 4 - L-Network Tuner - Click for Larger Image
The inductor in Fig-4 is listed as being 20 to 25 uH. The coil has 30 turns of #12 bare
wire on a 2-1/2" form, spaced over 5 inches (~6 Turns Per Inch). Taps at 0 (coil shorted), 1, 2, 3, 5, 7, 9, 12, 16, 21,
and 26 turns would be a good choice to start with. A coil with those specifications could be used with a 250 Watt
transmitter. For a 250 Watt transmitter, the variable capacitor should also have sufficient spacing to handle the power.
If you are using a 50-100 watt transmitter, receiving type variable capacitors should be sufficient. The Knife Switch
in Fig-4 allows you to switch the position of the variable capacitor, depending on whether
you are working with a low (C-L) or high (L-C) impedance load.
If you need to change the coil dimensions because you don't have #12 wire, or your form is smaller, or any other reason,
you can use my Single-Layer Air-Core Inductor Design page. Start off with the "Initial Design"
section and enter 25 uH and #12 Bare Wire. Then skip down to the "Alternate Form Size section" and set the diameter for 2.5 inches.
That will set you up with the basic inductor design. Then you can adjust the wire gauge or form size and see what that does to the
number of turns required. As an example, if you were running QRP you could use a plastic film container (approximately 1-5/16"
or 3.3cm diameter) with a #22 enameled wire. For that, 29 close wound turns would provide the same inductance.
One big disadvantage of the End-Fed antenna is that the end of the antenna is very near, or inside, your shack. This
enhances the possibility that there will be RF in the shack. This can often be minimized by choosing a somewhat
less than random length of wire. Choosing a λ/4, at the
operating frequency, would put a current loop near the transmitter, avoiding a
voltage loop at the feed point. An effective
λ/4 can be calculated with the equation
Length(ft) = 234/F(MHz). So a good length for 3.6 MHz would
be 65 feet. For 7.1 MHz, 33 feet would work well. Odd multiples will also work. For example,
3λ/4 would be 195 feet for 3.6 MHz and 100 feet for 7.1 MHz.
Note: For the antenna element, insulated wire has a lower velocity factor then bare wire. This may cause
a shortening effect of about 3% to 5% (wire appears higher in frequency). So always
start with a little extra length and the prune it to get it near your operating frequency.
All of that is good, if the antenna is only meant for one band. What if you want to use the antenna
on multiple bands, like 40 to 10 meters? Well then you need to choose a wire length
that is at least 1/4 λ on 40 meters, and doesn't fall
on any of the 1/2 λ multiples for ALL
the bands between 40 to 10 meters While you could certainly calculate the lengths to avoid using the
standard formulas, and then make a chart, the section below may be what you are looking for.
The chart below is intended as a visual aide. It should help in understanding why certain lengths might
be good on some bands and terrible on others. The idea behind the chart is pretty simple.
First, for any band selected, lengths that are less than 1/4 λ
are considered undesirable and are colored RED.
I then calculate length pairs for that band's 1/2 λ
multiples (.eg. 1/2 λ, 1λ,
3λ/2, 2λ, etc..).
I calculate one length for the low end of the band and another for the high end
(Length Low = 468/FLOW, Length_High = 468/FHIGH). The entire
band, from FLOW to FHIGH
for each of the multiples, is then colored
Note: Because the WARC bands (30M, 17M, and 12M) have small
bandwidths, these bands, and their harmonics, appear on the chart as thin
This should then give you a chart showing the undesirable wire lengths in
RED. Since multiple bands can be selected,
it also shows lengths, that might be good for one band, that are bad for another. If you poke around at
the button a little bit, it should be self explanatory, but a more detailed explanation is listed
below the chart.
Chart and Controlls Details - The chart is split into three sections.
Example 1: Say you have a End-Fed wire that is
20 Meters (~66 Ft) long. If you select Metric
dimensions and enter 20 in the space provided, a green line will show on the chart.
Following the green line on the chart will show that, you should be able to load up your
transmitter, with a tuner, on 80, 30, 17, and 12 Meters. But the other bands may have too
high of a impedance to obtain a good match.
Example 2: I have a End-Fed wire that is
92 Feet (~28 m) long. If I were to select
US/Imperial dimension and enter 92 feet, a green line will show on the chart.
Following the green line down the chart would show that I should be able to load up my
transmitter, with a tuner, on 160, 80, 40, 20, 17, and 10 Meters. But the other bands,
30, 15, and 12 Meters, may be a problem obtaining a good match. If I shortened the antenna
by a few feet, I might be able to use 12 Meters, but I also might lose the ability to load
up on 40 Meters.
When trying to eliminate excessive RF in the shack, different grounding approaches may help. You may even need
to use different ground setups based on the band in use. A good ground to start with would be to return the antenna
ground to the transmitter. Or, attaching the antenna ground directly to close a metallic water pipe. If that doesn't help
you could try a λ/4 wire simply drooping down from the antenna feed point
or elevated slightly from the ground. Just don't tie it to an actual ground.
While random length wire antennas can be run without ground radials (counterpoise), some kind of ground system helps
stabilize the tuning. A single ground rod, or group of them bonded together, is seldom as effective as a collection
of random-length radial wires.
The diagram at the right shows some of the details of the Novice 90. C1 is a 350 µµf.
(or larger) midget TRF single-gang variable capacitor. For low power work, under 100 watts, almost any receiving type
variable capacitor will work. L1 is 6 turns of No. 14 wire approximately 2" in diameter.
The turns are spacing is approximately 1/8". You can wind the coil yourself or cut some turns from a piece of B&W
pre-wound stock. Actually, any coil that is close to those dimensions will work. The coil is approximately 2.182 uH, so
you can use my Single-Layer Air-Core Inductor Design page to work out your own
The antenna wire is 90' "L"-shaped length of wire. 30' vertical and 60' horizontal is optimum. But almost any other
combination will work. Do your best at keeping the vertical portion near 30'. The ground lead, however, needs to be kept
short. Around 10' should be the maximum. For every foot that the ground lead is over 10', subtract that amount from the