A Pyramidal Antenna for 14-30 MHz
This article is about a type of log periodic antenna that surprisingly little has been written about. You have probably seen one on your neighbor’s roof [1,2]. The V shaped part on the left side of the antenna below used for the VHF channels is the pyramidal log periodic array.

This antenna was invented by Van Osdol [3,4], Canadian patent 713169. See also the July 1969 cover of Popular Electronics [5,6].
How it works
It looks and works in much the same way that a standard lp antenna works, with one big difference: the two halves of the transmission line are separated and positioned as a V, so each half of the transmission line is in effect a single wire transmission line. Despite the fact that the two halves are separated, radiation from the transmission lines is negligible, contributing a small cross-polarization component to the pattern. Unlike traditional lp antennas, this antenna is easily modeled in nec. Closely spaced wires needed for transmission lines are difficult to model in nec, so usually idealized transmission lines are used. Since the pyramidal antenna uses widely spaced wires for the transmission lines, modeling is not a problem. I find that I get nearly the same results if I model in nec2 or nec4, and AGT is always close to 1. This gives me confidence that the model is accurate. The nec program I use is 4nec2, available for free [7].
The model
I decided to model a version of this antenna to cover the 5 ham bands in the 14-30 MHz region of the spectrum. I rather arbitrarily picked a boom length of 285 inches and 8 elements (ie, 16 half elements). The booms double as transmission lines. I assumed that all wires are 1 inch diameter aluminum. I then let the 4nec2 genetic evolver pick the element lengths and spacings and the spacing and angle of the booms/transmission lines with the following conditions: frequency sweep 13.9 to 29.9 step 1, segmentation 20 per half wavelength, impedance, 200 ohms; SWR 100, Gain 50, and f/b 4. The schematic diagram of the antenna, its SWR, gain, and f/b plots are below as well as the 4nec2 program that created the data. The antenna dimensions can be found in the program. Elevation and azimuth plots are shown for 21.2 MHz. AGT is about 1.06 for most of the data, so gain is optimistic by about 0.2 dB. Higher accuracy can be obtained by using more segments per half wave and/or by using nec4. Even though the antenna was modeled with a 1 inch diameter boom/transmission line, it should work with a larger diameter. 4nec2 has a stepped radius correction option if you wish to model tapered elements.
If you plan to build this antenna,
I recommend you first become familiar with existing lp antennas and their advantages, disadvantages, and problems. Here are some links [8,9,10,11]. Also you should be familiar with modeling in nec. 4nec2 is recommended because it allows you to find the optimum dimensions for this antenna which deviate somewhat from true log-periodic form. If you build one, let me know how you like it.
Another antenna based on a single log-cell of the pyramidal log periodic dipole array, a wire antenna for 3.5 to 4.0 MHz, can be found here [12].
Ross Anderson W1HBQ ross_anderson@comcast.net Jan 8, 2008 Jan 10, 2008
[1] NZ4K, personal communication
[2] http://www.starkelectronic.com/wca1.htm
[3] N8TT, personal communication
[4] http://patents.ic.gc.ca/cipo/cpd/en/patent/713169/summary.html
[5] W9GB, personal communication
[6] http://www.swtpc.com/mholley/PopularElectronics/Jul1969/PE_Jul1969.htm
[7] http://home.ict.nl/~arivoors/
[8] http://www.realhamradio.com/Log_Periodic.htm
[9] http://www.cebik.com/lpda/lpnec.html
[10] http://www.cebik.com/amod/amod23.html
[11] http://www.universal-radio.com/catalog/hamants/1540.html
[12] http://home.comcast.net/~ross_anderson/80Wire.htm
keyword: LPDA
My homepage “Ross’s Antennas”, with links to my other pages, is http://home.comcast.net/~ross_anderson





CM A Pyramidal Antenna for 14-30 MHz by W1HBQ.
CM
CM This program is written in 4nec2.
CM The dimensions of this antenna were determined using 4nec2's
CM genetic evolver with SWR 100, Gain 50, F/B 4, 13.9 to 29.9 step 1.
CM Theta = 90, Phi = 0. Characteristic impedance 200.
CM Auto segmentation 20 per half wave.
CM
SY alpha = 70.87357 'Angle of rotation of the booms.
SY so = 15.87721 'Antenna is fed in center of this section.
SY sp1 = 27.39122 'All dimensions are in inches.
SY sp2 = 38.76605 'sp is the spacing between adjacent elements.
SY sp3 = 43.37986
SY sp4 = 44.97633
SY sp5 = 57.635
SY sp6 = 53.72762
SY l1 = 75.97424 'l is the length of the elements.
SY l2 = 94.21755
SY l3 = 111.3901
SY l4 = 132.6472
SY l5 = 160.856
SY l6 = 182.9957
SY l7 = 202.2512
SY l8 = 222.1566
SY sz1 = sp1 'sz is the distance along the boom.
SY sz2 = sz1 + sp2
SY sz3 = sz2 + sp3
SY sz4 = sz3 + sp4
SY sz5 = sz4 + sp5
SY sz6 = sz5 + sp6
SY sz7 = 285 'Boom length is 285 inches.
SY freq = 28.9
SY wrad = 0.5 'Diameter of the elements is 1 inch.
CM
GW 1 1 0 0 so 0 0 -so wrad
CM
GW 2 1 0 0 0 0 0 sz1 wrad
GW 3 1 0 0 sz1 0 0 sz2 wrad
GW 4 1 0 0 sz2 0 0 sz3 wrad
GW 5 1 0 0 sz3 0 0 sz4 wrad
GW 6 1 0 0 sz4 0 0 sz5 wrad
GW 7 1 0 0 sz5 0 0 sz6 wrad
GW 8 1 0 0 sz6 0 0 sz7 wrad
CM
GW 9 1 0 0 0 0 l1 0 wrad
GW 10 1 0 0 sz1 0 -l2 sz1 wrad
GW 11 1 0 0 sz2 0 l3 sz2 wrad
GW 12 1 0 0 sz3 0 -l4 sz3 wrad
GW 13 1 0 0 sz4 0 l5 sz4 wrad
GW 14 1 0 0 sz5 0 -l6 sz5 wrad
GW 15 1 0 0 sz6 0 l7 sz6 wrad
GW 16 1 0 0 sz7 0 -l8 sz7 wrad
CM
GM 0 0 0 -alpha 0 0 0 so 2
CM
GW 17 1 0 0 0 0 0 -sz1 wrad
GW 18 1 0 0 -sz1 0 0 -sz2 wrad
GW 19 1 0 0 -sz2 0 0 -sz3 wrad
GW 20 1 0 0 -sz3 0 0 -sz4 wrad
GW 21 1 0 0 -sz4 0 0 -sz5 wrad
GW 22 1 0 0 -sz5 0 0 -sz6 wrad
GW 23 1 0 0 -sz6 0 0 -sz7 wrad
CM
GW 24 1 0 0 0 0 -l1 0 wrad
GW 25 1 0 0 -sz1 0 l2 -sz1 wrad
GW 26 1 0 0 -sz2 0 -l3 -sz2 wrad
GW 27 1 0 0 -sz3 0 l4 -sz3 wrad
GW 28 1 0 0 -sz4 0 -l5 -sz4 wrad
GW 29 1 0 0 -sz5 0 l6 -sz5 wrad
GW 30 1 0 0 -sz6 0 -l7 -sz6 wrad
GW 31 1 0 0 -sz7 0 l8 -sz7 wrad
CM
GM 0 0 0 alpha 0 0 0 -so 17
GS 0 0 0.0254
GE 0
LD 5 0 0 0 30800000 'Aluminum
GN -1 'Free space
EX 0 1 1 0 1 0
FR 0 0 0 0 freq 0
EN