# 160 meter inverted L radials

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by KG4DYN, Jan 8, 2021.

1. ### KA0HCPXML SubscriberQRZ Page

Methinks you are banging the drum of "Perfect modeling" vice more than adequate 'approximations".

We often hear the claim that "This mystery antenna can't be modeled".

Close second is the "Your modeling software can't produce usable results because 1. It isn't the right version/powerful enough or 2. No real world antenna site can be sufficiently defined."

Sigh.

AK5B likes this.

3. ### WA7ARKHam MemberQRZ Page

I have been working on your issue, and have come up with what may be a novel way of using your existing tower as the main "radiator".

When folk connect a wire to a tower (usually called a "sloper"), they are thinking that the wire is the "antenna" and that the tower is just a convenient way of holding up the wire.

I just learned that maybe we should be thinking of the grounded tower as being the "antenna", and think of the sloping wire as being the "counterpoise or radial" that provides a method of driving RF current into the tower. Certainly, if we can establish an RF current that flows in the highest part of the tower, then that produces the desired low-angle radiation. Most are not interested in the radiation created by the sloping wire; it is just incidental to create something to push against.

I created the following model:

I am using 6ea radial wires (radius = turning radius) to represent the two Yagis at the tower top. The large set of radials under the simulated tower is not physical. I place them about two wire diameters above ground and then add an estimated lumped loss resistance (20 Ohms) in the bottom of the tower to simulate your eight, parallel ground rods. The length of those artificial radials is chosen so they couple to earth without reflecting a reactive term into the tower...

I discovered that there are two significant variables (everything else remaining constant) that control the feedpoint impedance, the radiation efficiency, and pattern shaping of this "antenna", namely the height of the horizontal wire1 (variable K) and the length of the horizontal wire1 (variable L). The discovery is that the antenna works best when Wire1 is relatively low (~15ft), so for the purpose of the simulation, it might as well be horizontal.

I used the optimizer in AutoEz to maximize the current in Wire5 and Wire6 (Tower above Wire1) to create maximum low-angle radiation while at the same time minimizing the current in Wire4 (because that current flows in the "lumped loss resistor"). K controls that and also creates the feedpoint R.

Changing L is primarily responsible for bringing the system to resonance (jX=0), so the goal for the Optimizer was to make the Swr50 at 1.86MHz be less than 1.5 while maximizing the gain and minimizing the power lost in the lumped loss resistor.

Here is what I came up with:

-1.26dBi is not as good as a 138ft vertical on 16 on-ground radials (+1.56dBi), but it is not bad for such little investment.

To feed the horizontal wire with 50 Ohm coax, connect the center conductor to the horiz wire, and connect the shield to the tower at 14ft height. You might have to tweak the length of the horizontal wire to make the frequency where Swr50 hits minimum.

If you go ahead and try this, please come back with some results and measurements, because I have never seen this way of "shunt feeding" a tower.

4. ### W9XMTHam MemberQRZ Page

Below is a start at modeling the 160-m inverted L supported by a 60-ft tower, based on the description of WA9GON in this thread. NEC4.2 was used in order to model the buried conductors (interconnected ground rods) around the tower base, which NEC2 cannot do. Earth conductivity was set for 4 mS/m, d.c. 10 — hopefully to approximate that at the antenna installation site in Wisconsin. Aluminum conductors were used, a network was included to get the match ~suitable for a 50 Ω source, and both of those losses are included in the analysis.

The graphic below shows the result so far, but at least the following info is needed to proceed...
• end-end length of slant wire leading from the top of the vertical conductor of the inverted L
• exact offset of the vertical wire of the inverted L from the supporting tower (I guessed it to be 8 inches, at this point)
• physical orientations of the two HF arrays at the top of the tower, with respect to each other
• any other input desired

5. ### WA7ARKHam MemberQRZ Page

Using the Nec4.2 data posted by Richard, the lumped ground loss for @WA9GON 's ground rods I guessed at in post #53 (20 Ohms) is actually a bit higher at 32 Ohms. If I simulate the Inverted L (58ft tall, slanted wire 135ft long), fed at the bottom, using 32 Ohms of lumped loss, I match Richard's -6dBi for the peak radiation for the no radial, only ground rods version of the inverted L.

If I then go back to the "single horizontal wire shunt feeding the tower" model of post #53, but I use 32 Ohms for the lumped loss instead of 20 Ohms (no other changes), the peak radiation (-2.31 dBi) occurs at 28degrees elevation on the side of the tower with the horizontal wire.
The Radiat-eff (Average gain) is 16.4%, or about double the 7.7% Richard shows for the inverted-L. With 32 Ohms of loss instead of 20, the total radiation is reduced due to a higher percentage of the transmitter power being lost in the ground rods, but is about twice Richard's grounded inverted-L.

Notice that back when I was showing the performance of inverted-Ls with 4ea radials elevated 8ft (divorced from earth) in Post #28, the peak radiation at ~25 deg elevation from the inverted-L was about -0.5dBi, which is much better than what Richard showed driving an inverted-L against @WA9GON 's ground rods. My direction here is not to place an inverted-L up against the existing grounded (radial-less) tower; rather it is to figure out how to shunt feed the existing tower as it stands without having to augment the existing ground rods with radials. Obviously, adding radials at the base of the existing tower would improve things...

Last edited: Jan 21, 2021
6. ### WA9GONHam MemberQRZ Page

Richard, the vertical wire is 15 feet away from the tower going up to the tower about 50 feet high then down towards the ground with the end of the wire about 15 feet off the ground with about 85 feet of wire on the slant. Both antennas on top on the tower face in the same direction . Hope this helps. I appreciate any help I receive.

7. ### W9XMTHam MemberQRZ Page

Thank you. I'll modify my NEC model and post the results — probably early tomorrow morning.

8. ### WA9GONHam MemberQRZ Page

Mike, your previous post is intriguing. I would like to try this when the weather gets warmer. What I am understanding is the coax connection goes at 14 feet with the braid going to the tower and a horizontal wire 14 feet off the earth for about 138 feet. Is this information correct?

9. ### WA7ARKHam MemberQRZ Page

Yep. Adjust the 138ft length to move the resonant frequency to your favorite 160m frequency.

Presumably, there are existing coax runs (and a rotor control cable) running from the shack to the tower base. Are those cables presently bonded to the tower base as is required for lightning protection?

The new 160m coax should run parallel (bundled with) to those existing cables...

10. ### WA9GONHam MemberQRZ Page

Both Mike and Richard, I really appreciate both of your efforts!

My goal is to reach DXCC on both 80 M and 160 M. To date, I have used an inverted V for 80 M with apex at 50 feet on my tower and have a total of 34 countries confirmed in two years of casual operating. Not great but I feel there is a need for a better antenna for DXing. On 160 M, I only have USA and Canada confirmed countries. I know of local hams using verticals with 12-16 short radials and having limited success thus my search for a better solution.