antenna-tx

IV3PRK Pierluigi “Luis” Mansutti

160 Meters: DXing on the Edge

My new transmitting antenna «T Vertical».

Structurally the antenna is the same I had in Ecuador, but changed from “Inverted L” to “Vertical T” in order to improve low angle radiation as preferable at high latitude locations.

The vertical section is made of 8 aluminium tubes starting from 65 mm. to 18 mm diameter, for 20.5 m. high.

The top wires are 19 m. long, sloping from 20.5 to 8.7 m. height (1 mm. diam.).

Despite the success of the K2AV FCP system in the HC1PF setup, I changed to the classic “on ground radials”, as I have here the necessary space. The antenna has been modeled with Eznec 5 and these are the screenshots of the “Elevation Plot” and “View Antenna” with 32 radials and wire currents:

EZNEC source data indicates:

Impedance 20,72 - J 1,872 ohms.

Than, entering those data into TLW to design the L-Network, it requires:

C – parallel input: 2064 pF

L – series coil: 2,30 µH

The following is an SWR run after the insertion of this L-Network:

Raising the antenna was quite easy by means of a gear wrench and a gin pole: just me and my wife Luisa.

The radial system.

After reading the great QEX articles on this subject by Rudy Severns, N6LF, I learnt that there is no reason to exceed the number and the length of radials: very small improvements going above the number of 32, and even worse by making them too long. In direct correspondence with the author of EZNEC, Roy Lewallen, W7EL himself wrote to me:

«Luis, You can see by the current distribution on the radial wires that they are electrically longer than a quarter wavelength. At a quarter wavelength, the current should just reach maximum at the junction. I think you'll find the best efficiency to occur when the radials are an electrical quarter wavelength or shorter, and increasingly poor as they get longer than a quarter electrical wavelength. The electrical lengthening of the radials is due to the low propagation constant of the ground. The amount of lengthening (the effective velocity factor) that happens depends strongly on the height above ground as well as the ground characteristics. Calculation of the physical length needed isn't simple -- I recommend just shortening them until the current maximum occurs at the junction. 73, Roy, W7EL».

EZNEC reports all the detailed currents calculated along the single wires in a tabular form; it is much easier to see their whole development in the graphics using AutoEZ by AC6LA.

This is the current on the first two radials of 40 meters length ===>

There are 10 segments for each radial. Current is max (0,07) at wire junction (seg. 1 and 11) and min. on the tip

There is too much current on the first segments (1 and 11) and this causes a non-uniform distribution in the whole ground system as shown in the following graph ===>

which includes all the 32 radials with their 320 segments

I found that the best radial length, which accommodates a full quarter wave sinusoidal current, is around 28 meters.

Here we see that the maximum current on a single radial is reduced from 0.07 to 0.05 ===>

Thus, reducing the current magnitude on the first radials from 0,07 to 0,05, produces a perfect uniform and efficent current distribution around 360 degrees in the full radial system - with 32 radials - all 28 meters long.

====>>

There is no reason to waste more wire!

Actually, I made some measurements with my AEA antenna analyzer on the sample dipole - 76 m. long - I was using in Ecuador to determine the ground resistance and thus the velocity factor “VF” for my BOG studies.

On the Andes, on a poor desertic soil, that dipole resonated around 1.200 KHz, while over my garden grass in Italy it resonates around 800 KHz: thus, quite a difference in the velocity factor.

Then I tested it on different heights and it is clear that lowering a wire to ground level its resonance changes a lot:

Dipole heigth cm. - Resonance Freq. KHz - Velocity Factor VF

180 1.810 0,92

90 1.750 0,89

30 1.640 0,83

15 1.250 0,63

5 1.070 0,54

0,5 780 0,40

This is the reason of such a great difference between phisical and electrical lenght of a wire near the ground.

My practical installation.

I covered the concrete base of the vertical with a copper sheet for easy soldering. Than, four 1.80 m. long copper rods have been hammered down at each angle and connected through a solid bus wire. For lightning discharge, I used two iron 50 mm. balls, with 1 mm. separation, as suggested by Tom, W8JI.

As static drain, I inserted in the matching box a series-parallel combination of several 10 Mohm resistors.

For radials, I used a roll of enameled wire 1.32 mm. diameter (12 Kg.) and to be double safe with their connections, I soldered them first on the bus wire, and then on the copper sheet after tightening there with bolts. I wrapped with electrical tape the tinned joints before covering them with sand.

Impedance measurements made with AEA antenna analyzer.

BEFORE installing the radials was: R = 27 Ohms; X = 13 Ohms; Z = 30 Ohms. Resonance point on 1.830 KHz with two top loading wires 20 meters long.

AFTER completing the installation of 32 radials, these were the readings:

R = 12.5 Ohms; X = 0; Z = 12.5 Ohms, with resonance point still on 1.830 KHz.

Wow, almost 15 Ohms of ground resistance removed!

Now, entering these real impedance data in the ARRL TLW program, it actually resulted that the L-Network requires an input parallel cap. of 3.000 pF and a series inductance of 1.9 µH.

Thus, I just added a further 780 pF - 2500 V. mica capacitor in parallel to the string already in the box with the vacuum Jennings (7-1000 pF) variable, to reach a value ranging from 2400 to 3400 pF.

The matching box includes also two relays for detuning the vertical Tx antenna while listening on a separate receiving one; Rly1 is a vacuum “Kilovac HC1” for high-speed switching.

Final results.

After installing the L-Network, the 50 Ohms impedance point was soon found on the antenna analyser, but with the resonance frequency shifted down on 1.800 KHz. That’s probably due to the excessive inductance of the coil. It was easier for me to shorten the top loading wires from 20 to 19 meters (as designed) and the final results are given on this antenna analyzer graph:

December 12, 2016 Luis IV3PRK

Finally, I made my Xyl Luisa - who always supported me - happy when she saw all the wires disappeared in the garden lawn by means of 300 “Gardenmate” steel pickets bought on Amazon.