5
$\begingroup$

Consider a center-fed dipole antenna. It‘s well understood how the bottom element can be replaced by a ground plane (in practice, generally a few rods but in principle a conducting plate would work as well), creating a „groundplane antenna“.

From a dipole to a ground plane

I was wondering whether it would be possible to do the same transformation for directional antannas with either parasitic (Yagi-Uda) or active (HB9CV) directors.

Illustration of suggested Yagi transformation

Are there any publications out there discussing this idea? What consequences would this change have in terms of antenna properties, and how would antenna theory explain radiation?

Improve this question
$\endgroup$
8
  • $\begingroup$ Also, please excuse the quality of those sketches ;-) $\endgroup$
    – jstarek
    May 6, 2020 at 11:32
  • $\begingroup$ See my earlier answer to a similar question here: ham.stackexchange.com/a/14509/2988 $\endgroup$
    – tomnexus
    May 6, 2020 at 13:11
  • $\begingroup$ @jstarek I like your sketches -- very refreshing! $\endgroup$
    – Brian K1LI
    May 6, 2020 at 15:11
  • $\begingroup$ @jstarek If appropriate, please remember to mark a response as "Answered" so that it comes off the list of "Unanswered" questions. $\endgroup$
    – Brian K1LI
    May 6, 2020 at 22:40
  • 1
    $\begingroup$ @MikeWaters That’s just bad drawing, the line is supposed to be connected to the ground plane and the radiator. $\endgroup$
    – jstarek
    May 8, 2020 at 7:45

2 Answers 2

Reset to default
6
$\begingroup$

Yes, this is a time-honored practice, particularly at the longer wavelengths where horizontal antennas need to be mounted quite high to provide low-angle radiation for DX work. Callum, M0MCX, presents a nice gallery of information about his three-element parasitic array for 40-m:

enter image description here

The radiating elements are the three black verticals in the foreground behind the fence.

While a "vertical yagi" can be an effective solution for low-angle radiation, it's important to remember:

  • the ground screen consumes considerable area
  • you lose the reflection gain attendant upon horizontal antennas
  • ground conductivity near the antenna strongly affects efficiency and feedpoint impedance
  • ground conductivity out to several wavelengths from the antenna determines how low the radiation angle will be

M0MCX demonstrates the advantage he hopes to achieve:

enter image description here

but this comparison doesn't tell the whole story:

  • the text says the feedpoint shows a very low SWR across the entire 40-m band without any matching circuitry, indicating substantial losses in the system which may reduce radiation more for a vertical than for a horizontal antenna
  • the comparison puts a dipole at 20-ft, while the vertical elements are at least 50% taller

I point this out only to illustrate the complexity of an antenna system in the real world, not to criticize the efforts of M0MCX, who is obviously enjoying his fine antenna. All antennas require compromises, so carefully consider your goals and resources, then set your expectations accordingly.

Improve this answer
$\endgroup$
2
  • 2
    $\begingroup$ There are even examples of log-periodic arrays using monopole vertical elements driven against a ground screen. See, e.g. Vertical Log Periodic Antenna for 80m, 40m. $\endgroup$
    – Brian K1LI
    May 6, 2020 at 15:23
  • 2
    $\begingroup$ Good answer. I would also note that even if the losses are high, the losses don't matter so much as the improvement to directivity in the receive case. As an example, W8JI has a circular array for rx only. $\endgroup$
    – Phil Frost - W8II
    May 6, 2020 at 15:29
0
$\begingroup$

The answer to your second question is no, you will not receive the gains equal to, or even close to half of a Yagi-Uda or HB9CV, either of these designs requires proper phasing of of the signal which is not present on the other(virtual) half of the monopole.

Half of the signal is in the ground plane (which is phase variant with moisture and material).

The same is true with artificial grounds such as a radial system. Even with a radial ground system, the ground conductivity still causes some impedance, although greatly reduced, to signal flow. Thus reducing the velocity of the wave in a non-predicable way.

If you are asking if such a design can produce gain the answer is maybe sometimes; this depends on the phase of ground conductivity.

The length of the above-ground parasitic element is not properly represented in the ground reflection, due to the differing wave velocities between the two mediums. This phase difference mostly invalidates the advantages of element distance(phase)/length dependent designs.

Given a common ground(radials) and proper spacing/phasing between driven elements of the same size, gain can be achieved through phasing of the elements since their local ground plane reflections are close equivalent.

Adding a 90 degree corner reflector to a monopole antenna will still increase directivity and gain since phase is corrected and reflected by the reflector(including most of the image giving a ~2.7 db gain).

Since you did not specify a target frequency or ground plain height in your question, for a monopole antenna mounted at least 1/2 wavelength above the ground with the boom connected to a radial, a Yagi-Uda design antenna will improve directional gain by a small amount, but the gain will be restricted by the ground plane reflections(really poor f/b ratio).

As indicated by the the previous answer, driven elements of the same size over an artificial ground radial plane will give an improved dx pattern/gain.

As to the comment of the ground plane attached, most mag mount antennas rely on coupling to ground as opposed to a direct electrical connection.

While I have attempted to generalize this answer, the values given are only theoretical approximations based on empirical results in certain environments. The expansion of claims to any real world applications may result in different values.

I welcome any corrections or clarifications on any points presented.

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .