This section of my web site will offer you the results of the years I have spent modeling and optimizing antenna designs to fit my modest means and yet still fulfill my big-signal ambitions. The designs here are computer-aided designs, based on modeling software that enables the designer to easily modify and optimize a design for a particular parameter objective. In my case, that has mostly meant raw, old fashioned gain. In constructing most of these models, I have deliberately designed as possible variables the various parameters needed to adjust for various design objectives including gain, F/B ratios, feedpoint impedance, etc. Since QRM is not a serious issue in the part of the world I live in, but distance to other stations is, gain is the primary design objective. But you are free, of course, to download these designs and optimize for other parameters you may need, including front-to-back, front-to-side, accuracy of match or whatever. If you have come up with MultiNEC designs yourself that you think I ought to look at, click on the email link in the navigation bar and tell me about it. If I like your design, I will be happy to put it here and give you credit for it.
These files are all in .weq format, and are intended to be utilized in antenna simulation software. They were created and optimized using MultiNEC, v.2.2, a really nifty software package created by Dan Maguire, AC6LA, who has created a truly addictive piece of ham radio software. If you are into homebrewing antennas, the $40 it will cost you to register and continue to use it, is one of the best bargains around. I am a big advocate of this package, because it can utilize the freeware 4NEC2 as a plugin computation engine, which enables one to visualize in rotatable 3D the antenna model being defined in the wires tab (extraordinarily helpful in troubleshooting a difficult model), and visualize in rotatable 3D the radiation pattern the design will actually generate, in the computation tab. Then using the freeware VOACAP as a plugin, MultiNEC generate propagation predictions of how it will actually perform on-air, using either signal-strength maps of the coverage the antenna will produce, or a path profile of signal levels along a signal path you define, or what signal levels the antenna will produce at a given receive site across a 24-hour period. MultiNEC allows the antenna designer to set up to three variables (four if one is the test frequency), and then automatically run dozens, hundreds, even thousands of test cases to find the absolutely optimum design. One can then copy and paste the results into another part of the spreadsheet, sort them by the parameter being optimized, and in seconds, find the absolute optimum design among the variables being tested to a hundredth of a decibel. There is a caution, however; the program is not without its bugs, and sometimes it produces erroneous results without warning you that the results are unreliable - you can check for this in two ways: first, look at the wire model in the 4NEC2 viewer, and if the pop-up window shows a cell with a red background, that means that parameter is in error, and the MultiNEC result cannot be considered reliable. I have found that the second reliability bug usually appears in a series of test cases. If you get a test case that you suspect is wrong or simply find hard to believe, or is inconsistent with previous calculations, the best way to check it is to set the parameters of the suspect test case manually in the variables, save the weq file, close the program, reopen MultiNEC and run that test case again, without any other test cases being run ahead of it. If you get the same result, and there is no red cell in the 4NEC2 viewer parameters window, you can rely on it. But this assumes, of course, that all the other parameters are defined properly - wire size, metal type, ground type, height above ground, etc. Remember, the old adage applies: garbage in, garbage out.
As an example of what I have been able to do with MultiNEC, I needed to create a wire dipole antenna that would produce the optimum signal level at the net control station in Portabello, Panamá of a net I was fond of working in the early morning hours on 40 meters from my home in Costa Rica. I also wanted to use the same antenna in the opposite direction for working other stations in Central America during the day. Using MultiNEC, I created a double-extended zepp in the software model, and then varied its height above ground to find the best overall height for my soil conditions and for the propagation conditions for the path and time of day for which I wanted to optimize it. I then went back and optimized the antenna gain by length for that particular height. When this was all done, using the MultiNEC results plugged into the "Scott's Tools" transformer-section calculator, I designed a transmission-line transformer to allow me to properly match the resulting complex feedpoint impedance to ordinary 50-ohm coax. The result, when constructed, astounded everyone on the net with its performance. When everyone else on the net was barely over the noise at S6 or S7, I was 10dB over S9 - everyone wanted to know what I was using. This software package is an extremely powerful tool, and I recommend it highly. It takes all the guesswork out of designing your station antenna facilities, saving you lots of time, lots of money, and eliminating most all of that frustrating trial-and-error of antenna construction.
If your browser tries to open these .weq files as a web page, simply click the back button, then right click on the link, and choose "save link as" or "save target as" from the pop-up menu to save these files to your disk.
TI5N Ultimate Quad This antenna is a 5-band quad on an 18 foot boom that sports three elements on 20, 17, and 15, and four elements on 12 and 10. It is optimized from the design of the main antenna in use for many years at the contesting station at TI5N. That antenna is a remarkable performer - signals consistently an S-unit stronger, both transmit and receive, and much quieter on receive, than a TH6 on a nearby tower at nearly the same height. It produces usable signals much earlier and later in the band opening that does the TH6 - typically as much as an hour at each end of the opening. I have used this particular quad a great deal with my barefoot TS-430, and have rarely ever called anyone on it and not had them come back on the first call. The contesting results for this antenna speak for themselves - there are over 30 plaques and certificates for major contests (such as the ARRL DX contest or CQ WorldWide), mostly first or second place winner worldwide, hanging on the wall at TI5N - and this was the principal antenna used for most all of them. All that with the modest height above ground at TI5N of only 65 feet, and the antenna is an ordinary single-bay antenna - not stacked. For such a modest-looking antenna, it is truly amazing. This is my dream antenna, and the one I intend to build for use at the new station I am planning.
7 element Open Sleeve Tribander If you are into yagis rather than quads and want to go for the gain in a compact design, here is an open sleeve (trapless) tribander that you can build using readily available aluminum tubing stock. Requires a 17 foot boom and split driven element with a single feedline. The design also happens to produce quite respectable performance on both 17 and 12 meters as well, but requires the use of an antenna tuner on those bands as the VSWR is unacceptably high. Roughly based on a Force-12 design, and the calculated gain figures are approximately the same.
Optimized Skywire Loop The ARRL Antenna Book (19th ed., pp5-19 thru 5-21) claims that the optimum design circumference in feet for a horizonal quad loop (what they are calling a "Skywire Loop") for use on 40, 80 and 160 is based on the simple quad driven-element rule of 1005 divided by the design frequency in Mhz. Checking this on MultiNEC, I have found that this is clearly not correct - the optimum design rule is closer to 1200 (or 365 to get the optimum circumference results in meters). The optimum height above ground for long distance is above 1/4 wavelength, but even below that, you will get very respectable results. This file will allow you to vary both the circumference and height to optimize the design for your particular piece of real estate, and if necessary, the shape as well. This antenna is always a good performer - its roughly circular azimuthal pattern makes it a good choice for those stations fond of those late-night roundtables on 75 meters. On-air experience with using this design at TI5N has confirmed my suspicions that the Antenna Book does not offer even close to the optimum design for this antenna. For those blessed with the room, the antenna could be optimized even further by erecting it as a circle, using Dacron guy ropes from three or more supports, to pull the antenna wire into a more or less circular configuration.
Double-Extended Inverted Vee It is this antenna that, with just one QSO, completely changed my thinking about large wire antennas for 40 and 80 meters, at least for stations blessed with the room to erect it. Late one night, reading the mail on a QSO between a station in Texas using this antenna (S9+20 running 900 watts), and a (much closer) station in Florida (S7 using 700 watts and an ordinary inverted V), the Texas station gave a demonstration that astounded everyone listening. He simply called called CQ - and generated a pileup of at least 10 stations trying to work him - all trying to work a station in Texas on 40 meters! No rare DX there, but certainly a signal that got lots of attention. The Florida station's reaction was the same as mine - "holy s**t!" I just had to optimize it in MultiNEC, and based on his description, the file here is the result.
This antenna is basically a pair of conventional inverted v's, but with each leg of each "v" just a bit over 1 wavelength long and hung from the tower at right angles to each other - one on a north-south axis, the other east-west, and fed in-phase, 90 degrees out of phase or 180 degrees out of phase as needed. A major advantage of this antenna is that it can be erected from a single support. The result is a very high feedpoint impedance, so the "v's" are each fed with a separate 600 ohm open wire line (or even better, a 4-wire open line could be used), and are fed in phase for northwest-southeast, out of phase for northeast-southwest, or fed 90 degrees out of phase for a "fattened box" pattern for roundtables (which was what I was hearing that night). The optimum apex height for 40 meters appears to be about 85 feet (the station in Texas that night was using it with remarkable success at only 55 feet - apparently about the minimum useful height, according to MultiNEC), with the end height about 20% below the apex height. The model here includes all those as variables which you can change at will and play with for optimizing for your particular situation. I would love to see how this antenna performs when optimized and constructed for 80 meters. It must truly be remarkable. Even the 40 meter version, used on 80, should produce excellent results, according to MultiNEC.
Log Periodic Quad Here in Latin America, one often sees log periodic quads used as TV antennas, especially at UHF. This got me to thinking - could this be a useful design at HF? So I designed this antenna based on the criteria in the ARRL Antenna Book, and optimized it for a 26 foot boom to see if it would produce decent results from 20 through 10 meters. Indeed it does. But is it practical? Using wire for the elements, which would be suspended from Dacron concantenary ropes strung between quad spreaders at each end of the boom, it looks to me like it would indeed be practical to actually build. Gain and front/rear ratios is nowhere near that of the TI5N Ultimate Quad above, but respectable nonetheless, and comparable to yagis of similar boom length. It offers the additional advantage of requiring only a single feedline, rather than five separate feedlines for the Ultimate Quad, and resonant everywhere between the design limits, so it should be great for shorwave listening as well. Being constructed of wire and with only eight spreaders, it should be fairly light-weight - considerably less weight than the Ultimate Quad.
5-band Open Sleeve Vertical Here is an antenna that will do everything some of the Butternet and Hustler verticals do, but you can build it yourself using absolutely nothing but materials you can buy at the corner hardware store (ever try to buy a doorknob capacitor in a third world country?). It works well on 40, 20, 17, 15 and 10 meters, and can be ground-mounted or roof-mounted, but, like all 1/4 verticals, requires an adequate radial system for good performance. Because it is full sized and has no coils or traps, it should perform about as well as a single-band 1/4 wave vertical and survive in any weather for many years without maintenance. The parasitic sleeve elements should be connected to ground at their base. The 33-foot long driven element is defined in the model as a 6-inch diameter cylinder, but this is based on the assumption that a light-weight TV tower will be used as the driven element. The model can be adjusted as needed for whatever material for a driven element you have at hand. If you shrink down the driven element to a single piece of pipe, you will probably have to reduce the spacing to the parasitics to bring them back to a low VSWR.
2m.-70cm. Dual-Band 4-Bay Colinear The advent of dual-band FM radios on 2m. and 440 Mhz. that use a single antenna port have created the need for a gain antenna that can be used on both 2 meters and 440 Mhz., and has decent gain and acceptable VSWR on both bands, using a single feedline. To this end, I have modeled a dual-band 4-bay colinear dipole stack that uses a single phasing harness and feedline, but resonates and has good gain on both bands at the same time. The free-space gain of this antenna is 8 dBi on 2 meters and 11 dBi on the 70 cm. FM band - and has a 1.3:1 SWR or better on both.
The key to the construction of this antenna is that the driven elements are split-fed dipoles, with driven elements constructed from 3/4" tubing, and the open-sleeve parasitics from 1/2" tubing. The antenna must be mounted 19" from the supporting tower, so that it is 1/4 wave away on 2 meters and 3/4 wave away on 70 cm. This ensures that the SWR effects of the tower is nullified on both bands at the same time, and while there will be a distortion of the azimuthal pattern, particularly at 440 Mhz., it will be acceptable and the effect on the SWR will be minimized. The phasing harness is constructed as a conventional 1/4 wave phasing harness, as described in the ARRL Antenna Book, dimensioned for two-meters, and it will act as a 3/4 wave phasing harness on 440. All dimensions for this antenna should be calculated for the two-meter frequency of interest, and recognize that for the VSWR to be acceptable on 440, the resonant frequency on that band will be precisely three times the resonant frequency on 2 meters, like it or not. This means that if the antenna is designed for 146.0 Mhz., it will have the best gain and lowest VSWR on 70 cm. at exactly 438 Mhz., whether that is what you like or not. Because this antenna is dependent entirely on the harmonic relationship of the two bands, there is no way to change that significantly and still maintain an adequate VSWR. Also, this antenna cannot be built with downtilt, except on the 440 frequency. If you desire downtilt, design it for 438 Mhz., and you'll just have to accept that the 2-meter downtilt will simply be 1/3 of the 438 Mhz. value. If you design the downtilt for the 2 meter frequency, the downtilt on 440 mhz. will be three times that and will result in an unacceptable pattern. No other way to do it, so I don't recommend that this antenna be used for sites requiring downtilt. If you absolutely insist on downtilting this antenna, the downtilt calculator in "Scott's Tools" will calculate the incremental offset for you.
Monster-Gain Colinear For stations needing all the omnidirectional VHF or UHF gain they can manage to get, this design is the ultimate. This antenna produces enough gain, at least in theory, to actually make possible an EME contact for a minute or two as the moon is rising or setting over the horizon through the pattern! This antenna is basically a collinear stack of four double-extended Zepps, all fed in phase and carefully dimensioned and spaced for absolute maximum gain. With the bottom end of the array at 50 feet above average soil, it produces more than 20 dB gain! If you have the tower space to accomodate it (it requires nearly 50 feet of tower space at 2 meters), and the tower is low enough to the average terrain that the "dead zone" it will produce below the antenna pattern would not be a problem, then you might want to consider constructing this antenna. The feedpoint impedance for each dipole is complex and inconvenient, but using the "Scott's Tools" calculator, a 450-ohm ladder line transformer section can be easily designed and constructed that will get the feedpoint impedance transformed to a convenient 50-ohm non-reactive match. From there, a conventional phasing harness as described in the Antenna Book will get it all combined into a single feedline to go to the shack.
One word of caution: should you decide to build this antenna, it is absolutely vital that the feedlines (including the transformer sections) between the driven elements and the phasing harness be exactly the same electrical length to prevent distortion or uptilt or downtilt of the pattern - this is especially critical, more so than the usual dipole colinears, because of the uniquely high gain. Additionally, the phasing harness must be exactly symmetrical for the same reason. If you actually need some downtilt, the downtilt calculator in "Scott's Tools" can calculate the required incremental difference in feedline lengths for you. But just be sure to cut the feedlines to the exact length required. Also, you should consider the effects of spacing from the tower on the azimuthal pattern of the antenna, and orient it carefully accordingly. With this much gain, the difference in dB between the null and the main lobe center can be substantial, and will make a big difference out at the far horizon. This antenna should not normally be used for mountaintop repeaters, because the high gain will result in a rather considerable dead zone near the mountain - it is best used on sites at low heights above average terrain - downtown buildings of modest height, for example.
For stations not needing all that gain, for mountaintop sites, or sites lacking the tower space to accomodate the full sized antenna, the antenna can be simply cut in half - two dipoles instead of four - which greatly simplifies the design (and reduces the losses) of the phasing harness. I have actually built and used this two-dipole configuration on both 2 meters and 70 cm. quite successfully, and have found that in the real world, it performs at least as well or better than the best conventional four-bay half-wave dipole collinear stacks. It also has some larger sidelobes that help fill in the dead zone under the antenna, making it more useful for hilltop repeaters. With that antenna mounted only a few feet above ground on a modest hilltop, I have been heard throughout nearly a fourth of the national territory of Costa Rica on 70cm. I can only imagine what the results would have been with a full-sized, four bay 2-meter antenna on a respectably high tower.
© 2008, Scott Bidstrup, all rights reserved. Todos derechos reservados.