Through the courtesy of the Teleconference Radio Net and net manager Rick Whiting, W0TN, we've received a transcript of the June 2nd Teleconference by Joe Reisert, W1JR. Joe spoke on the subject "Antennas and Antenna Systems, Where is the State of the Art Going?"
PART I - Overall Summary and Definitions
Good evening, my fellow amateurs. It is a great pleasure to be here
tonight. I feel honored to be selected to speak at this very large and possibly largest ever teleconference. Many thanks go to the Honeywell Amateur Radio Clubs and, in particular, Dave Meldrum, KA1MI, and Rick Whiting, W0TN, for the confidence they have placed in me and also for their helpful hints and suggestions to make this presentation a success.
Tonight my talk will be about "Antennas and Antenna Systems, Where is
the State of the Art Going?" I will divide the talk into four separate segments. The first part will deal with general terms and definitions which will set the stage for the rest of the talk. The other three segments will be the "low HF" (40 through 160 meters), the "middle HF" (10 through 30 meters) and finally VHF/UHF and EME antennas. In most cases I will be talking about the top of the line, state of the art or future antennas and antenna systems. There is probably no other amateur radio topic that inspires such a
vigorous line of conversation as the subject of antennas. Virtually every amateur has some interesting story to tell about his or her favorite antenna or antenna system or one yet to be fully tested. In reality there is no better place to spend you time improving the performance of your station since if the same antenna is used for transmitting and receiving, every dB of improvement in gain yields a two dB overall station improvement, one dB on transmit and one dB on receive. The old saw still stands... "If you can't hear them you can't work them."
If there is one big complaint that can be leveled against the amateur
community it is that they almost never use true gain standards when evaluating antenna performance below 30 MHz. Typical performance is measured in dBs above my last antenna or above a long wire in the trees of how many guys I beat out in the pile up on XZ2XX. Seldom regarded are changes in radio propagation, power of the competition, or operator savvy not to mention good luck in being at the right frequency or timing of the call. Another problem is the sheep following the wrong leader copying an antenna that Joe Blow uses because he works more DX than I do without regard for his physical setup, power or operating ability. The trouble with this kind of approach is that you never really know what you have. You even may have built a real good antenna and replaced it with a poorer one. The situation I've described is not hopeless or beyond even the novice doing antenna tests if you understand your limitations and make a few basic tests.
Typical amateurs can only measure a few antenna parameters such as
VSWR and in the case of a rotary beam, front to back ratio. Therefore, the commercial manufacturers make sure that these parameters are good. Let me be more specific about these parameters and first look at
VSWR. All kinds of myths have evolved such as the one that says a 1:1 VSWR is necessary for an antenna to be working properly. This is entirely false. More recently, there has been an upsurge in the belief since many of the modern solid state rigs will only put out power if the VSWR is below 1.5:1. The lowly antenna tuner is now enjoying a big comeback. We have gone to solid state rigs that don't require tuning but now we have to add an antenna tuner that takes more time to adjust than the old pi-networks we all used to use that would literally load into any VSWR.
To further muddy the water, the typical VSWR measuring gear used by
amateurs is patterned after the monimatch, a breakthrough in its time but an instrument that is sadly lacking in accuracy when compared to a good directional coupler like the Bird Model 43 thru-line wattmeter and its equivalent toroid directional hybrid VSWR meter. Whenever I hear someone tell me their dipole covers the full 80 meter band with a 3:1 VSWR I just laugh to myself and wonder if they have a lossy feedline, a dummy load or just another incorrect reading VSWR bridge.
Probably the most important antenna parameter is gain. Gain is an
antenna property dealing with its ability to radiate power in a desired direction or conversely to receive energy preferably from a desired direction. It is a relative quantity, not measured in watts or ohms, etc. Hence gain must be referenced to something such as a dipole or an isotropic radiator, a theoretical antenna that radiates power equally well in every direction. HF'ers rarely measure this very important parameter and usually blindly accept the manufacturers claims especially if the front to back looks good. Many persons wrongly feel that if an antenna has good front to back it has good gain. This is not always true and will be discussed later in this teleconference. In many cases the antenna manufacturer can't even measure gain or compares his antenna to the competition and inflates his gain figures by adding a single finagle factor to the competition's claims. Now don't misinterpret my remarks as a slap at the manufacturers. I have noticed a significant improvement in this area in recent years but we still have a long way to go and I hope to give some guidance on this subject later in this teleconference.
VHF'ers realized this problem years ago when they organized a antenna
measuring parties. Some of the setups have gotten very exotic but the accuracy can be quite good. The main things that must be taken into consideration are an accurate gain reference antenna, a well illuminated source antenna, accurate measuring instrumentation and a reality as to what to expect. Two excellent articles on the subject are "Antenna Performance Measurements" by Dick Turrin, W2IMU in November '74 QST pgs 35 thru 41 and "UHF Antenna Ratiometry" by Dick Knadle, K2RIW in February '76 QST pgs 22 thru 25.
The National Bureau of Standards in Boulder, Colorado has done alot
of work in gain measurements and NBS Report 5539 entitled "Methods of Accurate Measurement of Antenna Gain" by H. V. Cottony is well worth reading. NBS developed an antenna gain standard which consists of 2 full wave dipoles mounted 1/4 wavelength in front of a 1.6 by 2 wavelength reflector that yields an accurate gain of 9.31 dB over a dipole. This standard was later redesigned for the EIA (Electronic Industries Association) by Richard Yang, a consultant to the Andrew Corporation to a simpler and smaller reference which consists of 2 half wave dipoles space 1/4 wavelength in front of a 1 by 1 wavelength reflector and yields 7.7 dB gain over a dipole. The EIA adopted this smaller reference and incorporated it into EIA Standard 329 entitled "Minimum Standards for Land Mobile Communications Antennas, Part 1, Base and Fixed Station Antennas." This reference antenna standard is the one most commonly used by amateurs and is often referred to incorrectly as the NBS Standard when it is actually the EIA Standard.
While on this subject, the EIA has issued another standard, RS-409
entitled "Minimum Standards for Amateur Radio Antennas, Part 1, Base and Fixed Station Antennas." It uses a half wave dipole for reference and is more apropos to HF antennas. This standard is very specific about the range itself including the reference antenna height, the source antenna height, the minimum distance between antennas and the minimum gain of the antenna under test. It should be noted that the the formula for minimum separation distance between the source and reference antennas is often quoted in the popular literature as 2 D squared/lamda where D is the largest aperture dimension of the antenna under test. This minimum can introduce an error of up to 1 dB. A better antenna separation standard and the one used by the EIA is 10 D squared/lamda which is accurate to 0.2 dB.
There is a crude but informative method of gain measurement that can
be done on rotary beams using simple amateur techniques. It relies on the fact that gain results by redirecting the power radiated in many directions into a single direction or directions. If the half power bandwidth of the radiated signal can be measured, the gain can be calculated using the formula 41253 divided by the product of the beamwidth in the horizontal and the beamwidth in the vertical plane. Using a locally generated low power signal (such as a local amateur), you first measure the half power bandwidth of your antenna (the points where the signal is down 3 dB from the direct heading). The vertical beamwidth can be estimated to be 5 to 15 percent greater than the horizontal beamwidth. To give a numerical answer, a typical well designed 3 element Yagi has a 60 to 65 degree beamwidth in the horizontal or "E" plane and 70 to 75 degrees in the vertical of "H" plane. Dividing 41253 first by 65 and then by 75 yields a gain of 8.5 or approximately 9.25 dB. This is isotropic gain which is approximately 2.15 dB above the gain of a dipole. Therefore the gain of a typical 3 element Yagi is roughly 7.1 dB over a dipole. By measuring your beamwidth over the frequency bands of interest you can estimate the gain. The wider the beamwidth, the lower the gain and vice cersa. The only restriction to this formula is that all side lobes and read lobes should be at least 15 dB below the main lobe. If not, the gain may be lower than calculated. For further information on this subject, I refer you to "Antennas" by John Kraus.
Always be aware of the beamwidths quoted for a specific antenna. This
parameter can usually be accurately measured and tells you if the gain is true gain or specsmanship. Also, some manufacturers list half beamwidth. To convert, just multiply by 2 and proceed. Check the gain reference carefully. Some sources quote isotropic gain which is 2.15 dB above the gain of a dipole.
Transmission Lines and Baluns:
No antenna talk would be complete without at least a few words on
transmission lines and baluns. Time does not permit a long segment on this subject. A few general rules apply. Use good low loss non-contaminating types of coax cable such as RG 213. Most RG-8 is poorly shielded, contaminating and deteriorates rapidly meaning your feed line becomes a big attenuator after a few years. The CATV foam coax is low loss, inexpensive and 75 ohms. It will require special connectors and matching transformers (such as synchronous transformers) if you want to go between 50 ohm sources. Make sure no water gets inside as some of the older types of this line will draw water through like a sponge and then cause discontinuities. I prefer Andrew Corp. Heliax or its equivalent at VHF and UHF. It costs more but has long life and is very low loss making it less costly in the long run. Oper wire lines are great and low loss but require special handling techniques and are particularly vulnerable when humidity or rain is present.
Baluns are a subject that invokes strong arguments. Suffice it is to
say that a balun probably doesn't help much on wire antennas and dipoles. On directional antennas it can prevent re-radiation which will reduce front to back ratio. I prefer the balun types that do not require extra wires or windings that interrupt the feedline. My article in September 1978 Ham Radio and the one by Walt Maxwell, W2DU in March 1983 QST discuss this balun type in detail.
Part 1 Summary:
To summarize this portion of the teleconference, we need some
accurate antenna gain references. EIA Standard RS 409 may be a step in the right direction. A good directional coupler type of VSWR indicator is a must for the serious amateur. As a side benefit, it may be used to measure output power if the FCC changes the amateur regulations to PEP output power as discussed in the recent Notice of Proposed Rulemaking. Amateurs can determine gain if they make the effort to measure or study the beamwidths and antenna patterns on their antennas and calculate gain as I have described. Hopefully in the not to distant future there will be general agreement on amateur antenna standards so we can objectively compare antennas. [End of Part 1]
Lower HF Frequencies
In the past few years we have enjoyed some of the greatest radio propagation ever. Now the sun spots are declining and the fervent DX'ers and those looking for a challenge are heading for the lower frequencies. All kinds of new or improved systems are evolving and I will now attempt to cover this frequency range and development.
Simple Antennas:
1. 1/2 wave dipole is hard to beat. It has good directivity, very efficient and the ground reflection in the far field is the only real loss (and that we have no control over!). The biggest problem is broadbanding especially on 80 meters. The open sleeve dipole invented by H. E. King and J. L. Wong (IEEE PGAP, pg 201-204, March 1972) is now being explored for HF. If it can be successfully scaled down from 225-400 MHz., it could improve bandwidth by a 2 to 5 factor.
2. Inverted Vee - radiates equally poor in all directions. Not really my favorite antenna!
3. Verticals: There are many articles on this antenna type by Jerry Sevick, W2FMI, Paul Lee, ex W3JM and now K6TS (?) etc. There are several popular lengths - 1/4, 3/8, 1/2 and 5/8 wavelength. See Ham Radio September 1981 for an interesting article on the 1/2 wavelength vertical by VE2CV. The main problem is ground losses. The ground plane is an exception since it has 3 or 4 resonant radials and hence is very efficient. Typical resistance for the conventional vertical 1/4 monopole is 30 to 36 ohms. Top loading, especially with a top hat is recommended to improve efficiency especially on shortened verticals. Also bandwidth can be very narrow especially on shortened verticals since they are highly reactive. I am somewhat against verticals for QTH's where ground conductivity is poor or where there are lots of local obstructions. A good vertical has most of its radiation near the current point which is usually the base! Absorbtion by trees, local objects, houses, etc. is very detrimental. Also we have very little control over the far field unless we live on or near a salt marsh or alkaline flat in the prairie. 4. Loops, Quad, Delta, side-fed Delta loop and Bi-Square. Great antennas if you have the space. The most popular seems to be the delta loop apex up fed on the lower corner up part way up the side.
5. Slopers: This is typically a 1/4 or 1/2 wavelength antenna that hangs off a tower and in a semi-vertical fashion and therefore may have some directivity (due to the tower acting as reflector) and a low angle of radiation. I prefer the G5RV antenna (June 1977 Ham Radio Horizons) since it is shorter than a 1/2 wavelength dipole. It consists of 51 feet of wire each side of the center insulator fed with 30 feet of 300 ohm feed line which then connects directly to a 50 ohm coax line. It does have poor VSWR over most of the band but never infinity. Advantages are multiple band operation (eg. 80/40/20/10) and it acts like a collinear (with gain) on harmonic bands. At my station I use three G5RV antennas as slopers spaced equally around a 97 foot tower and hence get good coverage over most of the world on multiple HF bands with fair directivity.
6. Beverage or traveling wave antenna is especially good for receiving despite its low efficiency. This is true because the outside or ambient noise is very high and hence compensates for the loss. Use a trifilar wound transformer and a low noise high dynamic range preamp to make up for the losses. Keep the height up at say 10 feet so no one walks into antenna and files a law suit against you. This happened locally when a horseback rider was knocked off a horse by a local's beverage antenna! The length should be greater than a wavelength at the operating frequency but 2 wavelengths is probably the maximum recommended length. To keep noise down, use a wire with at least 30% copper and is PVC coated. I have used beverages for transmitting and John Belrose, VE2CV, has recently written an article on same in a recent QST. Guys and Guy Wires: They must be tested for resonances especially if they are not broken up with insulators. The difficulty is testing. One test is to monitor VSWR carefully and remove or change a guy. Any changes indicate problems. Likewise, the front to back ratio carefully monitored on a local controlled station can give a feel for the problem. In some rare cases such as sloper arrays, etc., they can actually be part of the array such as working like reflectors, etc.
ARRAYS:
1. Yagi: Very large at HF, especially if full size! Bandwidth can be a big problem. One 75 meter fan (W2HCW) had problems hearing the Russian SSB stations operating on 3640 KHz, despite the fact that he was very strong over there when transmitting in the US phone band at 3800 KHz. When he turned his beam 180 degrees he could hear them but now they couldn't hear him. It turns out that the front to back ratio flipped over below 3700 KHz!
Many stations on 75/80 meters are using wire Yagi beams quite successfully even at low heights (30 to 50 feet). They do work but there is much tuning needed to determine correct lengths, etc. The problem of narrow bandwidth mentioned above must be considered. Loaded Yagi antennas have even narrower bandwidth.
2. W8JK: This antenna has been around a long time and is very successful at HF but it is bi-directional.
3. The ZL Special and KB9CV modern version of same is seldom considered but I think a worthwhile antenna. It is essentially a 2 element log periodic invented over 10 years before the log periodic! It has excellent gain (like the W8JK), directivity and is uni-directional. The feed system forces the pattern so it does not have the limited bandwidth and pattern reversal problems as severely as the Yagi does. See Ham Radio, May 1976, "Understanding the ZL Special."
4. LPA (log periodic array): It is essentially a wide-band uni-directional antenna. It has a sort of cardiod pattern at its lower frequency end so a reflector is worthwhile. Make the low frequency cutoff a few % below the lowest frequency of interest to enhance the lower frequencies. The best references are George Smith's articles is 73, Ham Radio and QST. Other good HF articles of interest on the subject are:
- "Log Periodic Antenna Design," Ham Radio, Dec. '79 by P. Scholz
W6PYK and G. Smith W4AEO.
- "Vertical Monopole Log-Periodic Antennas for 40 & 80 Meters," Ham
Radio, Sept. '73 by G. Smith.
- "Feed System for Log Periodic Antennas," Ham Radio, Oct. '74, G.
Smith W4AEO.
5. The bobtail array: This simple array has recently enjoyed a comeback. It consists of three 1/4 wavelength verticals spaced 1/2 wavelength joined at their tops by a single wire. Usually a high impedance antenna tuner is used at the base of the middle vertical to match the high impedance to coax. This antenna has 3 to 5 dB gain and is bi-directional. Recently articles have appeared in 73 magazine on how to feed the array directly with coax at the top of the array.
6. Vertical Arrays: In the last decade or so, many amateur radio state of the art advances have been made in vertical arrays by the late Jim Lawson, W2PV ( QST, March and May 1971), Dana Atchley, W1CF et al (QST April 1976), "Updating Phased-Array Technology," W1CF (QST August 1978) and Richard Fenwick, K5RR and R. Schell, PhD (QST April 1977). They have used computer aided techniques to design optimum 2, 3 and 4 element arrays using triangles, squares and lines of verticals. Their work has considerably improved not only the gain but also the front to back and patterns of arrays.
More recently, Roy Lewellen, W7EL (QST, Aug. 1979 pgs. 42/43) and Forrest Gehrke (Ham Radio, May, June, July 1983 and other articles to follow which will tell all!) have shown how to improve the feed systems of such arrays to guarantee that the mutual coupling between elements will not deteriorate the gain and patterns in the real world. This work and computer aided work in the future will have a big effect on operations in the lower HF region. 7. Other Arrays: Don't forget "V" beams and Rhombics. They can yield high gain. The principle problem is patterns which are not always very good (side lobes, etc.). These types of antennas are particularly good if you have lots of real estate and only are interested in one or two directions. I think the sloping terminated "V" beam is particularly worthwhile.
8. The active antenna array: Last but not least let us explore the active array. This usually consists of a small (0.5 to 1.5 meters) vertical monopole feeding the high input impedance of a low noise high dynamic range FET preamp. Arrays of these are in commercial service and can provide extremely high directivity. I am presently working on one for myself for solving some HF receiving problems. The chief advantages of such a scheme are that it is small and doesn't need an elaborate grounding system. Phasing is easy since the outputs are not reactive and mutual impedance affects are low compared to a conventional full-sized array. Also don't overlook ferrite loaded antennas and loops. A good reference for HF DXing and antennas is ON4UN's book on 80 meter DXing.
Summary: There is lots to be done. Computer aided design will help. We must explore optimum topology for vertical arrays (2, 3 and 4 elements etc.) to find best layout. Maybe we should look at the Mill's Cross! The sloper system used today can probably be improved. The software just emerging in the last few years will greatly help in the design of high performance arrays. Don't overlook the log periodic or the ZL Special. The biggest problem to solve may be the wideband feed system. Only now is the open sleeve dipole by Howard King and J. Wong (IEEE PGAP March 1972) being explored. If it can be successfully scaled from the 225 to 400 MHz spectrum, it could potentially yield a 2 to 5 times bandwidth increase over the present half wave HF dipole!
Upper HF Frequencies:
This is probably the frequency region where most amateurs are really concerned about their antennas and probably where the most $$$ are spent. Here the rotary antenna is very popular and being competitive in the pileups is very important. Other than the usual ground planes, dipoles and long wires, the most often used antennas in the 10 to 30 meter frequency range are the Yagi and the cubical quad.
First my thoughts on the quad. It may be a fine antenna for some amateurs. It is surely an inexpensive antenna but is difficult to keep in the air. It really has high Q in that it's front to back ratio detunes rapidly versus frequency. I strongly doubt that it has greater than 1 dB gain over a properly designed Yagi (more on this later). The usual way a quad is tuned is to maximize the front to back ratio. This does not necessarily mean maximum gain. I personally feel the quad is popular because its construction is simple and low cost. A quad using aluminum tubing would probably do much better but would obviously be unwieldly. One big plus for the quad, and I may add its original invention was for this reason, is its lower static reception level during rain and snow storms. This is unquestionably true. I've gone the quad route twice. Despite 2 years of work on a 3 element quad, it never could compare with a well designed 3 element Yagi and hence was finally scrapped in favor of the Yagi. Tests at VHF and UHF on scaled quads have never successfully shown the gains claimed except on the loop Yagi which I will discuss later in this teleconference. In this frequency range, the Yagi is King especially among the DXCC Honor Roll members. This antenna has been around the amateur community since the late 1930's. Many people have used Yagi antennas but few have really paid any attention to proper element lengths. Recent work on scaling and especially on element tapering have been thoroughly discussed by the late Jim Lawson, W2PV, in a series of articles in Ham Radio from August 1979 thru December 1980. These articles show that on 20 meters the elements may have to be lengthened as much as 12 inches and more to equal the free space length of an equivalent untapered element. The results of not performing this extension are lower gain and poorer pattern than expected! This same series of articles is probably the best collection of references on Yagi design to date. W2PV meticulously explored all details. Of greatest interest are his charts and patterns showing what can be done and how to do it. This is must reading for the serious HF'er.
W2PV also shows how to use computer aided optimization, a technique that is presently beyond those without access to a large computer, but surely something that will be within the realm of home computers in the not too distant future. The principle advantage to computer-aided Yagi design is the ability to optimize gain, front-to-back ratio or side lobes. One big problem is bandwidth, typically 250 to 300 KHz. maximum at 20 meters. This is true because of the feed systems we amateurs use and the cutoff of the first director causing the pattern to break up at the top of the band. As a word to the wise, design your Yagi antenna on the high side of your favorite operating frequency since this type of antenna cuts off rapidly above the design center but drops performance slowly as the frequency is decreased. All things being equal and optimum, the boom length, not the number of elements, is the important parameter when determining the gain of a Yagi antenna. A larger number of elements than required insures a good pattern over a wider bandwidth but more elements can also be a negative since there are more things to go wrong both electrically and mechanically! Maximum gain on a one wavelength boom is about 10 dBd! Compare this with the high gains you hear amateurs bragging about on the HF bands. Another interesting phenomenon on Yagi's is the improved pattern at certain boom lengths. This was first discovered by Peter Viezbicke and reported in NBS Technical Note #688 and later confirmed by W2PV. The NBS data showed slightly higher gain at certain boomlengths which is probably due to some pattern cleanup. Suffice it to say that for best pattern and gain, the boomlength of a Yagi antenna should be an odd number of quarter wavelengths ( eg .25, .75, 1.25, etc.) long. (The only known exception is the famous W2PV 0.575 wavelength boom published in the Yankee Clipper Contest Club Bulletin. However, this design is asymmetrical about the axis and uses very close reflector spacings and is believed to be a special case). We will discuss the NBS Technical Note in depth in the VHF/UHF portion of this talk because the antennas in that report are mostly longer than the typical designs used on HF. Boom resonances can be a problem especially at HF and where mono-band Yagi's are often stacked Christmas tree fashion for multiband operation. Again, computer optimization has shown that these effects are real. Gain and front-to-back ratio can be significantly decreased when one Yagi is placed close to another one even though they are on different frequency bands. Computer techniques have been used to reduce these effects by re-tweaking the element lengths to offset the detuning effect but even then the results show bandwidth may be decreased by up to 50% of the original design. Some amateurs have used insulated boom mounting clamps in an attempt to offset this effect. Another technique but an ungamely one is to rotate the offending antennas at right angles to the lower antenna. If you place one antenna in close proximity (1 to 2 meters at HF) to another, check the VSWR carefully before and after the change. If the pattern or the VSWR shifts or changes, it is a possible sign of an interaction problem.
Let's not forget the log periodic array! "The Log-Periodic Dipole Array" by Peter Rhodes, K4EWG, QST, Nov. '73, "The Log-Yag Array" by K4EWG and J. Painter, W4BPP, QST, Dec. '76 and "The Log-Periodic V Array" by K4EWG, QST, Oct. '79 articles are must reading. The addition of the new WARC bands in the future will make log-periodic antennas much more practical. Their main andvantage is good gain, VSWR and pattern over a very wide frequency range rather than the usual narrow bandwidth of the conventional Yagi antenna. One amateur antenna manufacturer presently employs a log-periodic feed system to some of their antennas to increase bandwidth. The log-periodic structure forces current and therefore pattern by its unique feed system and I think we will see more antennas of this design in the not to distant future. Summary: We've come a long way in the HF region. There will be a swing towards wider bandwidth and perhaps LPA's will find their way into the amateurs bag of tricks as more spectrum and bands become available (eg. 18 and 24 MHz.). We are getting more discriminating and will demand good patterns and gain at the same time! In the future I see the use of computer aided design to improve patterns and gain as well as bandwidth. Wider bandwidth feed systems are needed. The LPA is one example, the use of the open sleeve dipole is another. Amateurs have notoriously ignored the feed systems and consistently used narrow band feed systems. We must develop wider feed systems and consistently used narrow band feed systems. We must develop wider bandwidth feed systems in the future.
VHF and UHF Antennas:
The spectrum above 50 MHz has special significance to the development of antennas and antenna arrays. This is the frequency range where you can build a really high gain antenna without owning a large piece of real estate. It is also the region where antennas can be tested easily in preparation for scaling them to the HF region. At the upper end of our frequency spectrum the antennas are more aking to optics. I'll divide this segment of the spectrum into two parts, the VHF and UHF regions.
The two major types of antennas used in the VHF spectrum (50 to 225 MHz) are the collinear array and the Yagi structure. The collinear array usually consists of a group of 1/2 wavelength dipoles in front of a screen or set of half wave reflectors. In the later case, it technically could be called an array of two element Yagis. The unique thing about the collinear is the simplicity of the feed system which usually is an open wire line. The collinear is usually quite broadband, unlike most high gain antennas, and efficiency and gain can be quite high. The extended expanded collinear is a stretched out version that has less elements and was described in an article I wrote in Dec. '74 QST. Both the conventional and the extended expanded collinears were widely used in the days before good Yagi designs were available and are still in use by some 144 and 432 EME operators. This type of antenna has two main drawbacks: 1. It is large and hence it can be large enough, an expression the late Sam Harris, ex W1FZJ, used to use for antenna that couldn't stay up under adverse weather and 2. Its size usually prevents mounting other antennas on the same mast. The workhorse in the VHF spectrum is truly the Yagi antenna. The first high gain VHF Yagi designs were published by Carl Greenblum (QST, Aug/Sept. '56), J. Kmosko, W2NLY and H. Johnson, W6QKI (QST, Jan. '56) and Dr. Hermann Ehrenspeck and H. Poehler (IEEE, PGAP, Oct. '59, pp 379-386). Unfortunately, these Yagis weren't always as good as claimed and had only fair cleanliness in the side lobe and front-to-back ratio. In Jan. '72 (QST pg 96 and March pg 101 corrections), Don Hilliard, W0EYE, now W0PW, published his 4.2 wavelength 15 element Yagi based on the unpublished works of Peter Biezbicke at NBS. Don and I urged Pete to publish his work and he finally did so in Dec. '77 in NBS Technical Note #688, now out of print. This publication was the result of extensive studies done by the NBS in the 1950's to develop high gain arrays for ionospheric scatter and included models with boomlengths of 0.4 to 4.2 wavelengths plus new information on scaling and boom corrections. In August 1977 "Ham Radio" I published a full length article on the NBS report including all the necessary details to build your own Yagis and sketched several models for 50 thru 432 MHz. There are some errors in the NBS publication which are corrected in my article. Not correct was the gain of the 2 element Yagi which should be approximately 5.0 dBd, not 2.6 as reported by NBS (they must have had some measurement errors). The NBS Yagis are not the only Yagi designs available but they are easily duplicated and near the maximum gain attainable for the appropriate boom lengths. They have excellent patterns and are easily stacked for additional gain.
One more point in passing. The trigonal reflector system in NBS 688 definitely is no good on the 3.2 wavelength and shorter booms. It actually reduces gain by up to 1.5 dB! By lengthening all three elements in this reflector system, I have been able to recover all the gain but no real gain improvement over a single reflector. I have not tested the trigonal reflector on the 4.2 wavelength designs.
In Feb. 1978 QST, Wayne Overbeck, N6NB, published an antenna he named the Quagi. It is basically a Yagi using a quad driven element and reflector. It is low in cost using a wooden boom and fed directly with coax cable. DL9KR and others have done further optimization on the Quagi and have used arrays of 16 to do 432 MHz EME. This design could still use some optimization in gain and only a limited number of designs are available.
Other versions of the Yagi have also been used including the log- periodic fed Yagi developed by the late Oliver Swan and now manufactured by KLM (See Ham Radio, Jan '76, pg 46). The log periodic antenna discussed earlier in this talk has never found much favor with amateurs since there is no need for the bandwidth and it has less gain than a well designed Yagi. Along these lines, we can now make high gain Yagis with clean patterns using the NBS designs. These antennas seem to stack well in larger arrays yielding the 20 plus dBs required for 144 and 220 MHz EME. One EMEer, Dave Olean, K1WHS, is using an array of 24 of the 2.2 wavelength NBS type Yagis stacked 8 feet apart for EME and he has worked stations all over the world who are only using single Yagis and moderate power. Most recently, with the help of a large computer, a special program and a local person interested in the design of VHF antennas, we were able to develop a very unique Yagi, an 8 element one on a 12 foot boom for 144 MHz that had extremely high gain (greater than 11.5 dBd true gain) with excellent pattern (all lobes down 20 dB). It worked so well that I made 8 copies and first tested them on a 144 MHz EME DXpedition to Rhose Island where 25 stations were worked off the Moon in two nights of operation. Computers will undoubtedly be useful in the future as this work continues.
UHF:
The 420 MHz and up area is in a transisition region. Long Yagi antennas can be made with high gain such as the NBS and Guenter Hoch, DL6WU, types. The later designs are an extension of the Greenblum designs mentioned earlier and can be designed up to 20 wavelengths (see VHF Communications, #3 and #4, 1977, and #3, 1982). These designs show an increasing gain of approximately 2.2 dB for every doubling of the boom length which is about the maximum so far reported. Indeed I built a 9.25 wavelength (21 foot) 432 MHz Yagi using this design material and achieved a verified gain of almost 17 dBd at the 1981 Central States VHF Conference in Sioux Falls, SD. Long backfire ("A New Method For Obtaining Maximum Gain from Yagi Antennas", IEEE, PGAP, Vol 7, Oct. '59) antennas have been tried by the EMEers but gains have failed to live up to claims. The short backfire ("The Short- Backfire Antenna", H. W. Ehrenspeck, Proc IEEE, Vol 53, Aug '65) has been duplicated by myself and others and gains of approximately 15 dBi have been achieved. Perhaps more work should be done in this area as an array of short backfire antennas has the potential of higher gain without the problems of the surface tolerances on the parabolic reflector.
Loop Yagi: Another popular UHF antenna is the loop Yagi developed in 1974 by Mike Walters, G3JVL (Radio Communications, RSGB, Jan '75 and Sept '78). Although it looks like a quad, it is distinctly different in that it uses wide but thin metal scraps for elements. Mike started out with wires but could never achieve high gains (like discussed earlier on quads). He recons that the wide but thin strap improves bandwidth and hence gain. The loops are bolted directly to a metallic boom thus solving the mechanical problems of mounting elements at UHF. It is a very practical antenna for 902 MHz and above and has worked well for me on 902, 1296 and 2304 MHz. G3JVL has even designed and tested to specifications a 10 GHz model. The principle designs use 26, 38 and 45 elements. The gain on the 45 element model (which is 16 wavelengths long) is 21 dBi! G3JVL has also published correction factors so that the loop width thickness and boom size can be scaled.
Dishes: There is something esoteric about the parabolic dish antenna. It just has to work but the typical dish only has a 55% efficiency at best. Furthermore, it has a large wind surface. Therefore, it is not too popular except at frequencies where loop Yagis are no longer economical and for EME where it can often be mounted close to the ground. More on this subject later.
High Performance Arrays: I'd now like to turn to the subject of high performance arrays and more specifically EME (Earth-Moon-Earth) antennas. EME affords a unique property, viz. that due to the approximately 2-1/2 seconds it takes a radio wave to traverse the 450,000 mile path to the Moon and back, the EMEer can make improvements to his antenna system and actually hear the difference by listening for his own echos. Furthermore, EME antennas have such high gain (typically greater that 20 dBi) that you can listen to the noise generated by the sun to measure beamwidth, patterns and hence determine actual antenna gain (see "Requirements and Recommendations for 70-cm EME", J. Reisert, W1JR, Ham Radio, June '82) as well as system noise figure.
Large Yagi arrays are becoming increasingly popular especially for EME. WB0TEM has 24 5.75 wavelength 19 element Yagis on 432 while K1WHS has 24 14 element 2.2 wavelength Yagis on 144 MHz. Both stations have big signals and are able to work small (1 or 2 Yagi) stations off the Moon.
However, the really big EME stations use parabolic dishes up to 40 feet in diameter! The advantages of a dish for EME operation are numberous despite the low (55%) efficiency (some commercial antenna manufacturers have claimed up to 80% efficiency but use cassegranian feed systems that are quite complex). First off, the feed system can be changed to permit multiband EME. Circular polarization is also possible by using dual dipole feeds or the W2IMU multimode horn. Dish type antennas are usually much quieter on reception because of low side lobes and hence are very desireable with the low sky temperatures experienced on 432 MHz and above. On 432 MHz where linear polarization is still predominant, the most efficient dishes are using the EIA symmetrical "E" and "H" plans and works well with a dish with a 0.45 to 0.5 F/ D ratio. VE7BBG has such a feed with a W2IMU horn built into the center and has made cross band (23 to 70 cm) EME QSO's. A single dipole in front of a splasher plate is definitely not recommended due to its unequal "E" and "H" beamwidths! We still have a long way to go to improve efficiency and the offset parabola recently introduced to EMEers by W2IMU from Bell Labs has considerable advantages if the construction can become feasible for amateurs.
Summary: The VHF/UHF frequency region is a good test bed for developing and improving antennas. Recent developments in the Yagi and loop Yagi have greatly advanced the state of the art in VHF/UHF communications. Antenna patterns have improved and hence the noise temperature of the antennas used is now more compatible with the state of the art preamplifiers. The NBS Yagi data now gives everyone interested a recipe for a suitable antenna without guesswork. EME antennas have taken a big leap forward in performance and made EME operation almost commonplace. We still need to do more work in the area of low loss feed systems especially for Yagi arrays.