The Double Zepp The W3EDP Longwire with MTFT

Feeding the Double Zepp with Baluns

Fan-Dipoles Vertical-L

Vertical-L-Antenna for portable operation 10-40 m

The starting point of the considerations was the question of how much "counterpoints" a vertical antenna actually needs. The design of the horizontal 90 ° angle dipole described here is also very popular as an "upper-and-outer" especially for portable fans and only comes with such a counterpoise. This is one reason to take a closer look at this type of antenna and to develop a version covering 10-40 m bands.

If simple solutions are chosen as the 1: 9 broadband Un-Un ("Magnetic Balun"), so you still need an antenna tuner and may have significant losses but usually only as a monoband version uncomplicated build.

The most well-known vertical radiator is probably the lambda/ 4 Marconi antenna, which requires a ground network, which must be very expensive for a good efficiency. Now the Marconi antenna is comparable to a dipole, the second half of which is replaced by the reflecting earth surface. It is easier if you actually build 2 x lambda/ 4. However, then the question arises, where to go with the second "leg" if the total height for a true vertical dipole with a half-wavelength is not reached. The classic solution is called "groundplane antenna", in which the counterpoise consists of 4 lambda/4 long sections at right angles. But does that really have to be so complicated?

Theoretical considerations

The advantage of a vertical antenna is that, assuming one builds up and has a free environment, it provides a 360° radiation at a relatively flat elevation angle. This can sometimes be better for DX than a horizontal dipole, but in which - especially due to lack of structure height - the elevation angle is larger and thus less favorable. The minimum for a perfect round radiation are two "counter poises" in opposite with 180 ° distance arrangement. This version, developed as "inverted T-antenna", was described by me in [1]. But it is even easier, if one accepts that the all-round radiation is not perfect.

The vertical angle dipole is obtained by bending the classic half-wave dipole in the middle and leading one section vertically upwards and the other parallel to the ground. In free space, this angle dipole has an impedance of 45 , the length must increase compared to the stretched normal form by about 3%.

If you build them strictly symmetrical with equal lengths for the vertical and horizontal parts, an interesting effect occurs. The capacitive load occurring through the ground leads to an asymmetry of the two mechanically equal antenna sections and the impedance decreases. As a result, the 50-Ohm-point on the counterweight moves electrically further outward. This applies to the case where e.g. such an antenna for the 20-m-band is configured so that the feed point is 1 m above the ground and also at a corresponding height of the horizontal part is arranged.

Analyzation with EZNEC

If a symmetrical antenna has an impedance <50 Ohm, then one can optimize for a higher radiation resistance by arranging the feed point off-center. This is known from the FD-4, in which the feed takes place at a point of about 300  and thus the antenna is divided into two unequal length sections.

Here, the vertical part is made longer and shortened the horizontal section accordingly. This results in a ratio of about 55% to 45% for 20 m to reach the required 50 Ohm. This results in addition to a slightly flatter beam angle. The diagram of the antenna with the current distribution can be seen in Figure 1. Depending on the wire and insulation, the length of a normal half-wave dipole should be approx. 0.92-0.97 * lambda / 2. The vertical L-dipole is pretty much longer than lambda / 2, which has to be taken into account during the adjustment. For the different bands, the ground clearance (relative to lambda) is correspondingly different, which causes new factors and makes the adjustment not easy. Especially at 30 m the different conditions surprised me.

An alternative option is to lower the counterweight slightly closer to the ground. Due to the larger angle you can increase the impedance within certain limits. Systematic experimentation is important here, as near-earth simulated wires are problematic in NEC-II with regard to the calculated impedance. The dimensions given here were roughly determined with EZNEC, the fine adjustment was carried out experimentally.

Such antennas, which are fed off-center, are referred to in the English-speaking world as "OFC" antennas (off-center-fed). Since they (such as the classic Windom antenna also) tend to common waves, they must necessarily be connected via a balun.

The azimuth diagram (i.e. viewed from above) in Figure 2 is not circular, resulting in a loss of 3.5 dB on the opposite side of the horizontal part. If you have a preferred direction in mind, the counterweight should be aligned in this. The "gain" in this area is -1.5 dBd. This appears, as with other vertical antennas too, as very little. This is compensated by the flat elevation angle. In comparison, a horizontal half-wave dipole produces a lot of useless sky wave radiation at a low installation height, which is counterproductive, especially on the longer-wave bands.

The elevation diagram is therefore not symmetrical. At the counterpoise, this looks like in Figure 3, at 90 ° rotation you can see the slight directivity in the longitudinal direction of the horizontal wire in Figure 4. The radiation pattern is quite DX suitable despite the low height.

Based on the vertical angle dipole described in [3], for which, however, a significantly greater installation height is required, the up-and-outer can also be fed with a two-wire line. In this case, both sections should be 7 m long. This operation is 10-40 m possible, but only with a Antennenanpassgerät.

In practice, I started with the calculation values of EZNEC and then optimized each band by Hand. The effort was considerable, but it was it worth.


Practical structure

If you want to work from 10-30 m, a 9 m fiberglass mast (no carbon fiber material!) is sufficient as a support for the vertical wire. If 40 m is also to be used, a construction height of 12.5 m would be required. By inserting an inductance in the vertical part of the 9-m mast can also be sufficient for the 40-m band. It is then only the counterpoise further extended. As expected, much had to be snipped during the comparison, I have used low-priced electric 1.5 mm²-wire with PVC insulation. In fact, I used up a complete 100m roll. Since this is easily available, the builder can resort directly to the specified lengths [4]. With other wires result in other dimensions that require a readjustment!

In the run-up I have made experiments with monoband antennas and then used the dimensions determined there for the repluggable version. For operation on several bands, therefore, the two sections must be changed in length when changing the band. The simplest solution, which however requires a folding of the mast, is a plug-in system for the wires. It offers this plug and couplings for automotive wire connections. Depending on the band, the connections are closed or interrupted as bridges.

I had made 50 mm long sections with epoxy plates, through which two holes are drilled for strain relief on each side. The wire is pulled through as a loop. Figure 5 shows this method with the original "jumpers". Surprisingly, the resonances, especially on the higher bands, were well below the previously determined lengths and also the SWV minimum was significantly worse than before.

Obviously, the freely hanging ends play a serious role and negatively affect the properties on the tape used. Then I made the insulation much longer and enlarged to 120x20 mm (Figure 6). As a material, I then used a 2.5 mm thick kitchen cutting plate made of PE, this can be cut with a stable pair of scissors. When opening the bridges, the ends of the wires must be bent away from each other as much as possible to prevent the mentioned unwanted coupling. Now the result looked much better, even if the lengths are still a bit shorter than with the pure monoband setup.

I have summarized the finally determined dimensions of the wire sections in Table 1. The shorting inductance for 7 Mhz is approximately 10 uH and is made from the same installation wire as the band sections. For this purpose, 42 windings are applied to a PVC pipe with 25 mm outer diameter (Figure 7).

Table 1: Lengths of the sections for the different bands

This coil is inserted with the jumpers between the 10 m and 12 m sections of the vertical section (Figure 8). I have put the resonance in the lower half of the band at 7 Mhz, so that CW is still possible with good SWR. If you shorten this section by 15 cm, the best fit is between 7.0 and 7.1 Mhz. Alternatively, you can make this shortened piece even pluggable and let it hang down in the end. Since changing the frequency to 40 m requires only the tail of the horizontal part, which is easily accessible, this is a good option.


Because the two antenna halves have different environmental conditions, in addition to the above-described effect of the off-center feeding common waves can be expected. For this reason, a balun should be provided in the feeding point. This consists of 2x4 turns Aircell-5 on a toroid FT240-43. So you are up to a power of 1 KW on the safe side. Figure 9 shows the junction box with the built-in toroid. It makes sense to connect the vertical conductor on the inner conductor of the coaxial cable and the horizontal part on the shielding. If you only want to work with smaller powers up to 200 watts, then RG316U teflon cable with 2x5 turns on a FT140-43 offers. If necessary, you can also use the inexpensive and easily procurable RG174 up to 100 watts.

Adjustment and operation

Basically, when using other wires, the ends should first be left slightly longer than indicated in Table 1. In Table 2, I gave the approximate correction values ​​per 100 KHz. By this amount, both ends must be corrected if the frequency is wrong. This should make the adjustment of the entire antenna no difficulties. An exception is the 40 m band, in which the adjustment is made exclusively on the horizontal section. For this reason, the double value must be set!

First, you connect the individual sections with luster terminals, so you can easily make length changes. Start from 10 m and then shorten the ends again slightly, when the metal sleeves of the couplings and connectors are soldered. Then you go to the 12-m band and proceed accordingly with all other bands. So you can put the best fit in the preferred areas. Up to 10 m, where I have selected 28.0-29 MHz, a balance is just below the mid band. The measured SWR gradients on the different bands are documented in Figures 10-16. Feeder cable was 10 m H-155.

With a ground socket the GRP mast can be set up, for the second leg one uses a tree or fence. If necessary, there are also other attachment points. If necessary, a short auxiliary mast is used for bracing. It is important that this half of the antenna, as shown in Fig. 17, is parallel to the ground at a height of 1 m and does not descend obliquely, because this results in shifting the resonance again. To change the band, the mast is folded over and the corresponding connectors are opened or closed. This is easier with the horizontal part.

However, one thing has to be pointed out: As with all antennas that are set up near the ground (the counterpoise is only 1 m high!), the impedance and resonance are strongly dependent on the ground conductivities. On dry sandy soil can be compared to wet meadow floor quite different. That's the price for the simple construction. On the other hand, a correction within certain limits by height change or ground angle of the horizontal portion is possible. It should be noted that with a smaller distance, the impedance is shifted upwards, with a larger downwards. The resonance is exactly the opposite.


[1] Program EZNEC+ Ver. 6.0.13 by Roy Lewallen (W7EL), P.O.Box 6658, Beaverton, OR 97007, USA (e-Mail [email protected]),

[2] Steyer, M. (DK7ZB): Auf dem Weg von der Vertikal zur Inverted-T-Antenne, FUNKAMATEUR 62 (2016), H. 9, S. 853-855

[3] Steyer, M. (DK7ZB): Vertikaler Winkeldipol für KW, FUNKAMATEUR 53 (2007), H. 10, S. 1090-1091

[4] Pollin Electronic GmbH, Max-Pollin-Straße, D-85104 Pförring,, Best.-Nr. 560292 (Litze H07V-K, 1,5 mm², 100 m, schwarz)