Introduction
End fed wire antennas have been around since the early days of radio, recently half wave end fed wires with a 49:1 matching transformer have become fashionable, as has using an ATU to create a low SWR at the transmitter.

Typical end fed wire installations are mostly horizontal, which is fine for contacts out to around 500 miles with sky wave, but they are not optimum for DX.

An easy option for those of us with relatively small gardens it to use a fiberglass pole to support a vertical wire, a quarter wave vertical for 40m is roughly 32 foot 6 inches with PVC insulated wire. When used with a reasonable number of shallow buried radials, it will work well for DX on the 40m band. Unfortunately, a ground mounted quarter wave vertical for the 20 to 10 metre bands has a less competitive performance relative to other stations, even though a good SWR is easily obtained.

Using a 49:1 transformer does allow a half wave vertical to be matched, for example a quarter wave 40m vertical wire is a half wave on 20m and a full wave on 10m. Unfortunately, the base loading effect of the 49:1 transformer results in needing to shorten the wire for optimum matching on those higher bands and that creates an unacceptable SWR on 40m when used as a 40m quarter wave.

Another option could be to use a remote ATU at the base of the vertical. The limitations are twofold, most will not match a half wave wire and few are rated more than 100 Watts. While a simple L-Match is only a single band solution, it is easy to make one that handles high power levels and will match a very high impedance.

Radiation pattern
On 7.1, 14.1 and 18.1 MHz, a ground mounted vertical wire around 32' 6” long will have a low angle of radiation on the two higher bands and around 26 degrees on 40m The angle will be 19 degrees on 20m and 16 degrees on 17m. The low angles of radiation on 20 and 17 are particularly useful for DX. By way of comparison, a 20m half wave horizontal dipole at a height of 30 feet has a main lobe at an angle of 34 degrees. A 17m horizontal dipole at 30 feet has a 26 degree main lobe.

Feeding the vertical
The source impedance of our vertical wire is around 2450 Ohms on 20m, which sounds ideal for a 49:1 transformer, but as it also introduces base loading, is not so good. Martin, K1FQL, produced an excellent paper on building single band L-Matching units to feed a half wave antenna (link at bottom of page). I will not go over the same ground as Martin as his paper covers the issues well. Taking the idea a stage further, courtesy of EZNEC and “on-line” calculators, it is relatively easy to compute the values of inductance and capacitance to match the same wire on 18.1 MHz. While we could in theory obtain the resistive and reactive figures using an antenna analyser, these instruments are usually designed for 50 Ohms and may not give accurate results on end fed wires. Note that the 1/4 wave vertical for 40m is a 5/8 wave on 17m and could be fed with an electrical quarter wave of 75 Ohm coax to transform the 100 Ohm antenna feed down to 50 Ohms.

Practical L-Match results
Using the same 32 foot 6 inch wire, this time on 18.1 MHz, by inputting the wire fed against ground into EZNEC, clicking the source data button gives the following:

         --------------- SOURCE DATA ---------------

Frequency = 18.1 MHz
Source 1      Voltage = 463.6 V at -76.68 deg.
              Current = 1 A at 0.0 deg.
              Impedance = 106.8 - J 451.2 ohms

Take the impedance figures and enter the following into the Analog Devices web page calculator.
Frequency 18.1 MHz, Source 50 Ohms (0j), Output 106.8 Ohms (-451.2j)

The result is shown in the first diagram
L1 = 2.755 uH and C1 = 8.916 pF

You could make the alternative version where the coil is across the output, I personally prefer the easier layout where the coil is in series. Also, the choice may depend on the capacitor values you can source.

Construction and measured results
The image at the top of the page shows the typical construction of a single band L-Match built in a Geros GR17004 case, these are 100 x 100 x 60mm and IP65 rated. Drill a small vent hold in each lower corner, seal the inside of the sockets and ground screw with 704 silicon sealant. The inductor is wound on a piece of 40mm diameter plastic tube.

The static drain resistor is a 1/2 Watt high Voltage type. Any value of a few K Ohms upwards will be fine, obviously not needed if you opt for the serial capacitor and shunt inductor configuration.

For the 18.1 MHz example, wind 7 turns of 16g enameled copper wire on a 40mm diameter former (in the UK we use that size for kitchen and bathroom waste water pipe), a 15pF and a 22pF cap in series (equals 9pF), plus a static drain resistor across the input and you are QRV.

The calculated values are not guaranteed, some stretching and squeezing of the coil may be needed, once a satisfactory match is obtained, secure the coil with a couple of blobs of hot melt glue. The resultant SWR plot close to the L-Match on 18.1 MHz is excellent, see below.

The capacitors are surplus "door knob" types from Eastern Europe, with a specification of 8 Amps and 10 KV. Sadly the supply of these capacitors has more or less dried up in recent times. You could use a piece of RG400 PTFE coax as a capacitor, it is 94pF per metre (96mm is about 9pF).

Power rating
Using the above surplus capacitors with my 500 Watt linear has not caused any issues. Using the same capacitors at higher power levels should be OK, they have been previously tested at 1 KW in the trap on a dual band vertical. RG400 coax seems OK when used as a capacitor at 500 Watts, but it is not rated to the same high Voltage as surplus door knob capacitors, in any event if using coax as a capacitor, trim back the braid at the open end and cover the end with hot melt glue.

Links
Analog Devices L-Network Calc
Original L-Match article by Martin K1FQL
EZNEC