One of the most efficient multi-band atennas is one that has been around almost as long as Amateur Radio has existed. It has been referred to by many different names over the years; dipole, doublet, double- Zepp, etc. Basically, it is a somewhat random length dipole using a balanced feedline and a balanced tuner.

     In this artcle I will discuss a few anectdotal information based on my experience in using this antenna, supported by some technical facts I have gleaned over the years.

The Basics of the Doublet

     For the purpose of this article I will define the terminology I will use to decribe the Doublet antenna system. This is due to the fact that many different terms are used by Hams to decribe this antenna and its components. The following are the terms I will use:

• Dipole - A two element radiator typically 1/2w in length commonly fed with coaxial feeder.
• Doublet - An open-wire fed antenna of random length.
• Feedline - Some cable that connects the transmitter to the feedpoint of the dipole.
• Matchbox - An antenna tuner designed as a balanced symetrical tuner.
• Ladderline - meaning two independant wires separated by insulated spacers and usually having an impedance in the range of 400 to 600 Ohms.
• Windowline - meaning the commercial feedline with a nominal impedance of 450 Ohms. (Be aware that it is often closer to 400 Ohms).

     Standard resonant dipoles fed with coax have many disadvantages. Usually the are only efficient on one band and exhibit narrow bandwidth on that band. Another issue faced with coaxial fed dipoles is coaxial losses when SWR ration rises above the lowest ratio. SWR ratio has a direct and significant effect on coaxial losses, which is different from the regular attenuation loss. Another issue is Common Mode Currents on the sheild which has to be adresses by the use of a current balun at the feedpoint.

     Usually it is a simply a Dipole, fed in the middle with openwire feedline. Initially these antennas were one half wavelength long on the lowest operating frequency. Later, shorter versions were found to work successfully, with a few concessions that we will discuss below.

     "Doublet" was the common name for this antenna when it was not a half wavelength long on the lowest operating frequency.

     Originally these antennas were called "half wavelength Hertz Aerials" and were fed with a single wire. Later they were fed with two wires (which I generically call openwire) and called a Double-Zepp Aerial (or Antenna) because they could be viewed as being two single-wire Zeppelin antennas, fed with a common balanced feedline.

     These antennas REQUIRE the use of an Antenna Matchbox on all bands. They not only require an antenna matchbox, they also place some fairly significant requirements on the match box (e.g., very broad impedance matching range, from single digit Ohms up to multiple thousand Ohms, must be capable of coping with very high voltage as well as very high RF current).

     Doublets are a multi-band antenna, its primary advantage. Due to its fully balanced design, from the matchbox to the radiators, this antenna system is very efficient due to its inherant low loss characteristics. SWR on the feedline is of little consequence and on HF frequencies goes almost unnoticible in practice.

     On its fundamental frequency, the doublet antenna should have a feedpoint impedance near 73Ω if configured as a flat top. However, since the feedline is typically not 73Ω, specifically on the order of 300Ω to 600Ω, it will transform the feedpoint impedance to some other value. Depending on the length of the feedline, it may be either very low, very high or anywhere in between. Indeed, with this antenna, the impedance in the shack may vary greatly! High; Low; Anything.

     For instance, on the second harmonic the feedpoint impedance is roughly 4000 Ohms, well beyond the matching range of all automatic antenna tuners and even beyond the range of most manual matchboxes. Of course the feedline will transform this, but to what? Again it depends on the length of the feedline.

     Most people believe that this problem is solved by inserting a 4:1 balun. WRONG. If you do that, and are you running higher power, that's when the real trouble begins. (More on that later). There are numerous better ways to overcome this, including changing the length of the dipole to various non-resonant lengths, using specific lengths of feedline, etc.

     So you probably can see a pattern developing here. It is important to think of a doublet as a "system". It not just about the radiating elements. It's about the radiators, feedlines and matchbox working together as a system. While this should be the case for any antenna, with a doublet it is the essence of how it functions. Let's go further into this system and more details of things I have already mentioned ...

Doublet Size

Most people believe that the antenna should be a resonant half wavelength. This length has the advantage that when resonant, it is (theoretically) purely resistive. However it also has disadvantages, specifically the impedance on the harmonic bands is VERY high; many thousand ohms. This can be quite difficult to match.

Common sizes for the 80m Doublet:

• 132 ft. (plus/minus - depending on where the user wants it to resonate within the 80m band)
• 111 ft.
• 100 ft.
• 88 ft.
• 66 ft.

Of these, I have personally used 132 & 111 ft. The longest one, the full half wavelength was used primarily at my restoration work shop. At my home I tended to use the 111 ft. since it fit the size of the property space. While I know the 88 and 66 foot length would work, even on 80m, I have not had a desire to use them even though my matchbox could tune them.

Fundamentally and HIGHLY SUBJECTIVE, I can say the following about performance:

• Using the full size (132') as a base, I will give it a 10.
• The 111 ft. version seemed to work just as well. I will also give it a 10, although in theory it's probably more like 9.5. You won't tell a difference.
• I have no experience with the 88 and 66 ft. versions.

Doublet Matching

The Openwire Fed Doublet is simply a dipole fed in the middle with any kind of openwire or balanced feedline. It would normally be a half wavelength long on the lowest band you wish to use it on.

As I mentioned earlier, the impedance of the dipole is around 70Ω (flat top) or 50Ω (inverted-V) at the feedpoint. If we feed it with 450Ω or any other impedance not exactly equal to the feedpoint impedance, there is a transformation of impedance at the shack end of the feedline.

In order to predict the impedance seen at the end of the transmission line in the shack, we must know the following things:

• The Frequency of operation
• The Impedance of the Dipole at the feedpoint
• The Impedance of the Feedline
• The Length of the Feedline
• The Velocity Factor (VF) of the feedline

It is not always easy to know the exact values of of of this so often we can only make a rough estimate. Of course it is possible to exactly measure lengths, impedances and even the VF, but we need an antenna analyzer and a bit of math.

Although the impedance on the fundamental frequency is 70 (or 50) ohms, it is quite different on the upper harmonic bands. The impedance on the 2nd harmonic is about 4000 Ohms. It will vary with height above ground, but it is always very high.

If the length of the feedline is one half wavelength (after factoring in the VF), or any multiple thereof (e.g., 1 wl, 1.5 wl, 2 wl, 2.5 wl, etc.) the impedance in the shack will be exactly the same as the impedance at the feedpoint (i.e. about 4KΩ). This high impedance is impossible for most antenna tuners to match.

On the other hand, if the length of the feedline happens to be an odd multiple of one quarter wavelength (e.g., 1/4, 3/4, 5/4, etc.) the impedance in the shack may be quite close to 50Ω.

If the length of feedline happens to be anywhere inbetween a half wavelength (or multiple there of) and an odd multiple of a quarter wavelength, the impedance can be anywhere (all over the map).

Why not simply use a 4:1 Balun?

Mathematically, a 4:1 balun would reduce the 4000Ω impedance down to 1000Ω, a value nearly every matchbox can match efficiently.

BUT THERE IS A PROBLEM ...

The little known reality is, a 4:1 balun only performs well when it is closely matched on both ends. In other words, it likes to see 50Ω on one side and 200Ω on the other side. If the SWR on the antenna side rises, the performance of the balun, specifically, its ability to reject common mode current, deteriorates rapidly.

In general, once the SWR rises above 4:1, the balun is no longer functioning properly. In the case of the center-fed full wave doublet (i.e., 4000 Ohms), this represents an SWR of 20:1. Not good! It is imperative that we try and find a length of feedline that will transform the impedance down to under 800Ω. Generally this will keep the impedances comfortably in the range of most tuners; therfore, a balun is not truly necessary and in fact using one makes the overall Doublet system less efficient.

Conversly, we may actually incur very low impedance at the shack end of the feedline. Often this can be as low as 20 Ohms or less. Using a 4:1 balun in this case would transform the impedance even lower, down to 5 Ohms (or less). VERY BAD! Most matchboxes cannot match this low impedance. The few matchboxes that can actually match it, generally have very high internal losses when doing so. Therefore a 4:1 balun is a disaster with low impedances.

Instead of the balun being the limiting factor in the matching range of our setup, the matchbox itself becomes the limiting factor. We are able to use the full matching range of the matchbox. In most matchboxes, this is 50Ω to 2000Ω. So it is always better to design the doublet/feedline combo to operate withing these impedance ranges. It follows the old saying, "Less is more".

About Balanced Feedlines

Before we go any further let's take a look at balanced feddlines. Which type is best? As I mentioned earlier, there are different kinds of balanced feedline. How do we determine which one is best?

ANSWER: There is no such thing as "best".

Each type has advantages and disadvantages. The differences are found in size, weight, appearance, loss in dB, cost, ease of working with, and if home-brew, ease (or difficulty) of construction. The application and user preferences will determine which type to use.

Advantages and Disadvantages of Each

LADDERLINE:

Ladderline consists of two wires held parallel by spacers. There are only a few companies making and selling ready-built ladderline. In most cases, the operator must build the ladderline himself (or herself).

Advantage:

It is possible to vary the impedance by adjusting the spacing between the wires. A wider spacing means higher impedance. Some people believe a 600Ω line will have less loss than a 450Ω line. It is questionable whether the difference in theory will make any tangible difference at the other end of the QSO.

More important, ladderline maintains its impedance and velocity factor much better than other types of openwire feedlines, even when it is wet or covered with snow or ice. It is also less susceptable to wind than Windowline or 300Ω Twinlead.

If built using thin wire and clear Plexiglas spacers, it can be made nearly invisible, or at least much less conspicable than the brown or black Windowline.

Disadvantage:

Usually you must build this yourself. This involves a bit of work, but it is not difficult if you sellect the right spacers. (See: D-i-Y Ladderline). Openwire is a bit more difficult to work with, especially for making bends or going around corners.

WINDOWLINE:

Windowline is commercially available from many sources, though the most common source is THE WIREMAN or one of its resellers. This feedline comes in several different sizes, all of which are commonly said to have 450 Ohms impedance. In reality the impedance depends on which model of Windowline you choose. (See table below).

Windowline is usually made of CopperWeld (copper-coated steel wire) and as such is very strong. Never-The-Less it DOES BREAK. It often breaks at the point it connects to the antenna. You can buy it with stranded wires or with solid wires. Regardless, they both break over time unless you take very good care to strain-relieve the feedline, such that there is no strain on the solder connections or bare wires. The strain-relief must include the insulation.

Warning: you will hear arguements for both types, claiming one is better than the other. Forget it. Rely on proper strain-relief, not hear-say from hams.

There are severeal kinds of insulators providing good strain-relief in one form or another. It is also very simple to build your own out of Plexiglas or Epoxy Board (FR4). See below.

Advantage:

You can buy it ready-built and simply begin constructing your antenna. Easier to work with than ladderline. It holds its shape even when bending or going around corners. (AVOID 90 DEGREE BENDS).

You can buy center insulators (including good strain-relief) that exactly fit the 450 Ohm feedline. (One size fits all).

You can buy more later, in case you ever need to.

Contrary to popular belief, it does NOT lose significant efficiency when it gets wet or covered with snow.

Disadvantage:

Windowline has slightly more loss than ladderline. This is because it has much more dielectric between the two wires than ladderline has.

When it gets wet, especially covered with snow, its velocity factor changes, which requres a re-tuning of the matchbox. For most people this is of no consequence. If operating a contest and desiring rapid band change, this can become annoying because there are different matchbox settings for each band, depending on whether it is wet or dry.

This feedline is very conspicuous and can be an eye-sore to neighbors. It also tends to blow around in the wind more than ladderline because its insulation has more surface area to catch wind.