On another thread, a discussion of small home built antennas for UHF TV broke out. (h/t to Jim2) The traditional TV antenna is the Log Periodic. It has dipole like elements spaced on a log scale for broad frequency range and good gain. The problem with it is that at the longer wave lengths, the antenna can become large. For example, a VHF low channel number can have each element about 1.2 M long, or about 7 foot tip to tip at the wide end of the wedge.
Now it turns out that drooping the ends of a dipole does not weaken it much, nor does shaping it in other ways, even into a semi-circle. As a sidebar, I wondered if these could be made fractal elements, the answer is “yes” and it makes a smaller antenna, but without much change of performance. At the extreme of this drooping, you could end up with a circle. (There are “loop” antennas that are in fact made of circular loops). I got to wondering: “Can you make a log periodic loop antenna?”.
Well, turns out I’m only 17 years late in inventing it ;-)
This is a Masters Thesis from Virginia Tech from 1999:
Update: This first link is now dead (well, leads to the generic top page:
Searching that site lead me to the pdf article again here:
LPLA is log periodic loop array
LPDA is log periodic dipole array
GR is Ground Reflector – a wire or metal sheet at the back
6.1 Summary of Results
Compared to the LPDAs, the LPLAs provide the advantage of size reduction in transverse dimensions by a factor of 2/π.
The radiation characteristics of the loop element in the axial direction are preserved in both LPLAs with and without GR, whose fields in the broadside direction are cancelled due to 180 alternating phase shift between the adjacent loop elements. The Gθ far-field pattern is dominant in the E-plane, while the Gφ pattern is dominant in the H-plane. Hence, the LPLAs have linear polarization and unidirectional patterns.
Improvement of several decibels in gain can be achieved by using ground reflectors in the apex region of LPLAs. For example, the gain (or directivity) of a 6-turn 30° LPLA-GR can be about 12 dBi. Also, the smaller is the subtended angle (α) and the closer to unity is the scale factor (τ), the larger will be the directivity of the antenna.
The ground reflector seems to increase the bandwidth. For example, the 6-turn 30° LPLA with GR yields 47.62 % gain bandwidth.
LPLAs with GR are expected to find applications as feeds for reflector antennas, as mobile communication antennas, and as detectors in EMC scattering range.
Those interested ought to read the whole paper.
The key bits are that this looks somewhat like a conical birdcage, yet it works very well. It is a 12 dB gain antenna (that is quite good) and that is using only 6 ‘turns’ or circles of wire. It is also smaller than the regular log periodic dipole antenna at 2/Pi or about 2/3 the size. That’s a big improvement. You can form it, as the author did, by gluing wire onto a conical form. Heavy paper cones are VERY easy to make.