The Simplest Collinear
Wherever there is a need for a vertically polarized omni-directional antenna with gain, a collinear is used. The vertically polarized collinear takes the form of two or more half waves in-line and in-phase, resulting in an omni-directional pattern with maximum radiation near the horizon. The techniques used to ensure that the half waves are in phase are:
1. Interspersing the in-phase half waves with quarter wave stubs. This is the method described in the ARRL Antenna Book.
2. Making the antenna out of half wavelengths of coax with the outside of the coax used as the radiating elements and the inside of the coax used as the phasing elements.
3. Interspersing the half waves with resonators to provide the necessary phase reversal.
Each of these techniques is designed to reduce or prevent radiation from the out-of-phase components of the antenna. The Simplest Collinear uses a different technique.
The Simplest Collinear makes use of the fact that large diameter elements within an antenna can radiate more readily than small diameter elements. Therefore if the large diameter elements are in phase and the small diameter elements are in phase but out of phase with the large diameter elements, radiation from the large diameter elements dominate.
Shown above is a schematic of a 5 element collinear—3 large diameter elements and two small diameter elements. Each element is about ½ wavelength long, and since the phase switches every half wavelength, the 3 large diameter elements are in phase and the 2 small diameter elements are in the opposite phase. The result is an antenna with about 2 dB less gain than if all the elements are in phase. Simulations show that this can be applied to very large arrays. A 45 element array (23 large diameter and 22 small diameter elements) is simulated to have 14.38 dBi gain.
The table above shows the gain and impedance of a 2 meter 45 element antenna, consisting of 23 elements, 1 inch diameter and 38.15 inches long and 22 elements 0.040 inch diameter and 40.95 inches long. The antenna has its feed point the center of the lowest element, although it could be fed at the center of any of the 45 elements. Also shown in is the performance of various versions of the same antenna, varying the number of elements between 1 and 45. Note that the gain increases by 3 dB for each doubling of the number of elements. I stopped the simulation at 45 elements because I can’t imagine building an antenna larger than this—at least at this time. Also shown in this table is the simulated gain of an ideal antenna assuming all 45 elements have 1 inch diameters and are fed in phase with voltage sources. Note that gain is only about 2 dB higher than the simulated real antenna. This is because the in phase currents are distributed quite uniformly across the antenna, an effect achieved by adjusting the lengths of the elements slightly from half wavelength. The 0.040 diameter is a compromise between making the diameter small to reduce the radiation, but not so small that ohmic losses become important. The 45 element array is simulated to have only 0.1 dB loss due to ohmic loss in the 0.040 diameter copper elements. All these values were determined using K6STI’s AO Antenna Optimizer MININEC program.
A 5 element test antenna was simulated and built. It was optimized for 5 elements and not created from a reduction of the 45 element array in the table; it’s gain was simulated to be 5.7 dBi and impedance about 100 ohms at 147 MHz. The small diameter elements were simulated to be 0.015 inch because for a small array, radiation from these elements is more of a factor and ohmic loss is less of a factor. All elements were 38 inches long. The 1 inch elements were made from aluminum, but any metal will have negligible loss—copper, aluminum, galvanized, or steel are all ok. The small diameter elements should be 0.010 to 0.020 inch in diameter (#30 to #24). There will be a slight reduction of gain if #24 wire is used. I used #26. Support of the antenna in the small diameter sections was provided by ¾ inch diameter schedule 40 PVC pipe. Guying was provided by 3/16 inch polyester cord. The antenna was fed through a quarter wave transformer made from RG59U. A few ferrite beads on the RG59U provide any needed current balun action. The antenna was installed on the roof in place of an old TV antenna. Simulated and measured SWR, with measurements taken at the end of the transformer, are shown.
This antenna has many attributes. It is easy to model. It is easy and inexpensive to build. A high gain version can be built in-situ, adding a couple of elements at a time, adding guying as elements are added. It is broadband. Even the 45 element version is simulated to have an SWR less than 1.7 across the 2 meter band (144-148 MHz). Current distributes well over the entire antenna making the aperture efficiency high.
Here are some possible applications.
1. Small versions like the 5 element antenna described can be roof mounted for repeater access.
2. Versions of this antenna should be good for applications in the field such as for contesting or DXpeditions, giving good gain and good height cheaply.
3. This antenna would make a good base station antenna for 2 meter and higher repeaters. Down tilt can be designed in if desired by slightly shortening the elements.
4. It should be a good antenna for communicating via Tropospheric propagation.
5. Several could be put together to make a phased array.
6. It could be used as a feed for a cylindrical parabolic antenna. Large cylindrical parabolic antennas should be easier to make than large paraboloid (parabola of revolution) antennas.
Ross Anderson W1HBQ September 7, 2002
E-mail me at email@example.com
Notes and References:
ARRL Antenna Book, 19th Edition, pp 8-36—8-38.
Collis, “Omni-Gain Vertical Collinear for VHF and UHF” http://www.repeater-builder.com/rbtip/wa6svt.html
Oblivion and Kaboom, “A 2.4Ghz Low-Power 5dBi Vertical Collinear Antenna for 802.11 Applications” http://www.guerrilla.net/reference/antennas/2ghz_collinear_omni_lowpwr/
Maxwell, “Some Aspects of the Balun Problem” http://home.iag.net/~w2du/Reflections2Chapter21SomeAspectsoftheBalunProblem.pdf
The beads I used were FB-43-5622. http://www.cwsbytemark.com/prices/largerBeads.asp or http://www.amidon-inductive.com/associates_prod_largerbeads.htm
A 928 MHz version of this antenna was made by K0EMT: http://www.dbbear.com/k0emt/projant/902/index.html
Link to “The Quadix” http://home.comcast.net/~ross_anderson/quadix.htm
Link to “Ross’s Antennas” http://home.comcast.net/~ross_anderson