Elevated Radials

When initially installed, the 30 foot vertical antenna used two ground rods for the ground side of the antenna. Later, a copper pipe radial ring and 16 radials on the ground surface were added as described here. However, over time, these radials got ripped up by the crew doing yard work.

Many papers have been published on the benefits of elevated radials as compared to buried radials. For example, USING ELEVATED RADIALS IN CONJUNCTION WITH DETERIORATED BURIED RADIAL GROUND SYSTEMS describes the results of adding four elevated ground radials to an AM broadcast antenna system where the 120 buried radials had been disturbed. USING ELEVATED RADIALS WITH GROUND MOUNTED TOWERS does extensive analysis of the driving point impedance with radials at 5 and 10 meters above ground, with a 45 degree and "very steep" slope of the radial at the tower. All modeling was done at 1 MHz. Scaling into the amateur bands, the frequency of 5 MHz should yield similar results with the elevated radials 1 meter above ground.

Elevated Vertical Antennas on 80 Meters tests these concepts on the 80 meter amateur band. It is interesting to note that with 120 radials buried 2 inches, the maximum power gain is 0.06 dBi at an elevation of 25 degrees. As the radial depth is increased to 6 inches, the power gain decreases to 0.03 dBi, an insignificant difference. With four buried radials, the power gain decreases to -2.62 dBi at 25 degrees when buried 2 inches, and -2.66 dBi when buried 4 inches. The buried depth does not make much difference, but the number of radials (decreasing from 120 to 4) reduces the gain by about 2.7 dB. In addition, the driving point impedance changes substantially from about 40 +j22 ohns to about 74.5 + j34 ohms. If we assume the 120 radials had a ground loss of zero ohms, and the increase in driving point resistance was due to ground loss, the ground loss resistance was about 34.5 ohms. Making a voltqage divider with 34.5 ohms in series, then 40 ohms to ground yields a gain of 0.537 or -5.4 dB. So, the actual gain reduction is about half that predicted from the driving point resistance increase. Perhaps the radiation resistance also increased.

Table IV in Elevated Vertical Antennas on 80 Meters shows the power gain for a quarter wave monopole with four quarter wave radials at various heights above ground. Note that the masts holding up the radials are conductive, connected to the radials, and connected to ground rods. At a height of 30 feet, the highest gain is 0.01 dBi at 20 degrees. Decreasing the heigh to 10 feet yields a maximum gain of -0.21 dBi, a fairly low loss.

Based on this, I added four radials elevated four feet. The radials leave the "radial ring" at 45 degrees and reach the four foot elevation about four feet from the radial ring. The radials are held up with fiberglass insulating rods. The total length of each radial is 25 feet. Each radial uses @12 solid THHN wire. The wires are attached to the radial ring using ground clamps.

The top image at the right shows an impedance sweep before and after the elevated radials were attached. There are some discontinuities in the curves. The sweeps were run several time, and various discontinuities showed up. To ensure the issue was not a loose connecion between the antenna analyzer and the antenna J-plug, all the connections were sprayed with DeoxIT. The curves with the elevated radials were done after the DeoxIT was applied. There are still some discontinuities. This may be due to the slight breeze blowing causing variations in the resistance between tubing sections of the antenna. The joints are coated with JetLube SS-30, but the low power of the antenna analyzer may still allow some resistance to be present in the connections.

Ignoring the discontinuities, we see some interesting changes when the radials are added.

• At 5 MHz, the impedance changed from 134.6 - j326 ohms to 42.4 - j145 ohms. Both decreased substantially when the elevated radials were added.
• At 7.5 MHz, the impedance changed from 117 - j180 ohms to 46 - j52 ohms.
• At 9.525 MHz, the impedance chaged from 180.7 - j160 ohms to 57.8 + j0.7 ohms.
• Between 9 and 11 Mhz, there is an interesting resistance dip to 41 ohms, then a peak of 200 ohms before dropping again.
• Similarly, between 9 and 11 MHz, the reactance dips to -70 ohms, then peaks at +30 ohms before dropping again.

N6LF has written some excellent articles on elevated ground systems ( A Closer Look at Vertical Antennas With Elevated Ground Systems and A Closer Look at Vertical Antennas With Elevated Ground Systems - Part 2). A few interesting points out of the article include:

• With few radials (such as four), even small elevations above ground substantially improve the power gain. For example, see Figure 17.
• As radial length increases approaches 1/2 wavelength, gain drops substantially, especially with good ground conductivity. See Figure 15. Figure 11 shows that radials near a half wavelength result in very high field intensidies at ground level, increasing ground losses.

Based on the N6LF articles, it appears the 30 foot vertical with 25 foot radials should perform best on 40 meters (where each is about 1/4 wavelength). The 25 foot radial length was chosen as the largest that would fit in the yard. The four foot elevation was chosen based on the length of the fiberglass rods. N6LF suggests that the antenna be isolated from ground through used of a choke on the transmission line. Having an additional path to ground (such as through the coax shield or ground rods) can decrease the gain by up to 0.5 dB. Here, however, the antenna tuner is mounted on two ground rods, and there is not a common mode choke on the coax. So, there are additional ground losses due to currents not going through the radials. However, we MIGHT assume this loss resistance is in parallel with the resistance presented by the radials, which should be low when the radials are a quarter wavelenght long (40 meters). It's a different story on 20 meters where they are about a half wavelength long.

Previous tests included the change in relative field strength as measured by the SHR at the KO7SS RBN receiver, a short distance away. However, that receiver is not currently online. When it returns, I will try comparing the SNR to previous values.