This gallery contains 3 photos.
Bookmark on Delicious Digg this post Recommend on Facebook Buzz it up Share on Linkedin share via Reddit Share with Stumblers Tweet about it Buzz it up Subscribe to the comments on this post Tell a friend
This gallery contains 9 photos.
Bookmark on Delicious Digg this post Recommend on Facebook Buzz it up Share on Linkedin share via Reddit Share with Stumblers Tweet about it Buzz it up Subscribe to the comments on this post Tell a friend
This gallery contains 10 photos.
Bookmark on Delicious Digg this post Recommend on Facebook Buzz it up Share on Linkedin share via Reddit Share with Stumblers Tweet about it Buzz it up Subscribe to the comments on this post Tell a friend
I have always been fascinated by antennas and how they work as have most amateur radio ops. A few years ago we built this house in the woods and just to get on the air I sling shot up a simple long wire V- beam in some of our Ga. pine trees. I thought it would be a simple antenna using open wire feeders and have a little gain. It is bi-directional with major lobes to ZS and JA. Each leg is 210 feet long and has an internal angle of about 67 degrees with both wires horizontal up about 70 feet above ground. I figured it would be up not longer than 6 months while I put up the towers, stacked yagis and log periodics. Now eight years later, no towers or beams yet and that pine tree supported v-beam is still up, cracking pile ups and working DX like crazy. Why? With all due respect to my location, it makes no logical sense, or does it?
I think I understand how two long wires forming a V, non-terminated, combines their along the wire, in-phase lobes together to form that bi-directional major lobe. How can such an antenna perform so well in pile ups against yagis and large arrays on 20 and 40 meters?
Then one day I met Roy, W1VDE on 17 meters. This guy is a long wire specialist and an excellent electrical engineer. He uses nine rhombic arrays. He reminds me of Don Wallace, W6AM. Roy said it was simple: Volumetric Efficiency. I am familiar with the term capture area and figured that was what Roy was referring to. However, he explained that he defines a term not limited in two dimensions, such as “Area”. Volumetric or “Volume” is at least a three dimensional concept. Applying a fourth variable, time, it gets even more descriptive. But before we try to understand Roy’s concept I will refer back to a brief review of what is known as capture area.
A good refresher for the concept of capture area can be found in an excellent article by Peter A. Stark, K2OAW, 73 Magazine, March 1975, pp. 39-46. This may be available on the net. Other terms such as isotropic radiator and path losses are also discussed but I will only look at what actually is capture area. By the way, it is NOT the physical cross-sectional area of an antenna (a parabolic dish antenna is close though). The definition is as follows: “It is the amount of power (power density) received over an area of one square meter located on a sphere surrounding the antenna (the antenna is at the center of the sphere)”. It is measured in watts/square meter, or in real life, micro-watts/ square meter. The following relationship can determine power density based on a measured field strength (volts/meter) and free space wave (vacuum) impedance (very close to air by the way) 377 ohms:
Power density (p) = (FS x FS)/377
It is assumed that the field strength (volts/meter) measured at the receiving antenna (FS), is measured with both receiving and transmitting antennas pointed at each other at their maximum gain directions.
A half wave dipole placed at the center of a sphere will concentrate power in two directions that are broadside to its length. Assume the sphere to be in space, no earth ground effects. The power density would therefore be stronger at two opposing directions, broad side to the dipole, on the sphere’s surface. The omnidirectional antenna would have its power density divided evenly over the entire surface of the sphere. The surface area of a sphere 4 pi (3.1416) times its radius squared.
As an example, an omnidirectional type antenna, radiating a total of 100 watts, at the center of a sphere whose radius is 1000 meters would have a power density of 7.958 micro-watts per square meter.
Power density = Total power radiated power divided by the sphere’s surface area = 100 watts/4 x 3.1416 x 1000 meters = .000007958 watts/meter. (7.958 micro watts per meter)
If we substitute a half wave dipole for the omnidirectional antenna at the center of the same sphere, we now have focused or concentrated some of the 100 watts in two spots on the sphere’s surface. That concentration of energy/power is 1.64 times that of the omnidirectional antenna. If we call the power density of the omnidirectional antenna 1.0 watts/meter, the dipole focuses its power density 2.14db higher. Using the above example, the power density of the dipole would be 1.64 x 7.958 = 12.446 micro watts per meter in two spots on the sphere broadside to the dipole. The TOTAL power density over the entire sphere is still 7.958 micro watts per meter. Power has not been created; it was simply focused and concentrated on parts of the sphere’s surface.
Therefore, the concentrated or focused power density AREA on the sphere’s surface, captures (or transmits) the electrons, represents what is called the capture area. Notice, that I have not mentioned the effects of external forces such as; propagation, ground effects, impedance, reactance and etc. These factors can add or subtract from the focused power density. Also the type of antenna; yagi, logperiodic, V-beam and etc. is not necessary for an understanding of capture area.
Capture area is proportional to the power density in watts per meter on a portion of the surface of a sphere. Also, if we replace the half wave dipole with a higher gain antenna, the concentration of power or electrons will increase. However, the area covered by a higher concentration of electrons on the sphere’s surface will decrease due to the antenna’s directivity. As was stated earlier, the total power over the entire sphere’s surface has not changed. The concentration is usually defined as antenna gain and from that statement, is derived a relationship between antenna gain and capture area. I get a kick out of those “term defining folks”. The computer folks are full of terms. It’s almost as if another language has been developed to keep us in the dark. Invent new terms and build mathematical models around them! You can’t lose!
So, it is back to square one, the definition itself. The capture area is the area illuminated by electrons on one square foot of sphere surface area at a given distance, measured in watts per meter. Notice that the area illuminated is not uniform. The center of the antenna’s main lobe (power maximum) is in direct line of sight with the center of that area. The power drops off on either side of the maximum. The 3db or half power points on either side of the maximum may or may not represent the limit of the illuminated area on the sphere. Also, there is no clear definition for the area’s geometry. The area geometry must be determined by the lobe itself. It could be elliptical, circular or any other combination. What about illuminated area beyond the half power points? How is it defined? Field strength is usually measured at maximum points on the lobe (see paragraph above), but the entire lobe must be analyzed using accurate volumetric field strength measurements.
What this means is that the relationship between gain and capture area is more complex than a simple two dimension relationship: Power received or captured (transmitted) = power density times capture area. It is a stretch to go a step further and let power density (P) = gain (G). Since our illuminated area is on the surface of a sphere, Area = 4 pi times radius squared. Then using this line of reasoning: capture area (Ac) = gain (G) x wavelength squared (l x l) divided by 4 pi. (Ac = G x l x l / 4 pi) Under this reasoning, capture area is simply defined in terms of gain. If the gain is known, we can calculate the capture area.
This simple approach does give us a rational way to arrive at a conclusion, but is it accurate? Does it reflect answers to questions for antennas that are special or volumetric, such as long wire, curtain, quad and stacked arrays? As we noted, the sphere is illuminated at irregular points on its surface. Many lobes both major and minor are possible. My experience using linear, single boom arrays, always provided less receiving power than my stacked arrays though the gains were identical. The atmosphere is full of RF energy and electrons. Shouldn’t the probability that more will be pulled from air by more resonant metal than less?
Of course ground conditions, propagation angles and electrical connections play an important role and probably propagation conditions are paramount. However, something is missing. Roy, W1VDE, probably has it right. Volumetric Efficiency is the reason “mega metal” arrays win. In the early days of radio the saying was: “The more wire the better”. At VLF this definitely was true. But with HF and VHF, yagis ruled with their high gains and small metal type construction. It is interesting to note that Dr. Yagi did his groundbreaking work a few years prior to WW2. The yagi is a low impedance antenna and required a low impedance feed system. Open wire feeder systems are too difficult to use and restrict rotation of the antenna unless unusual feed systems are employed. Large, open wire fed antennas, seem to have been moth balled after his work and the production of reliable coaxial cable. Television, invented in the 1930’s also was moth balled!
Earlier I said that the capture area of a parabolic dish array was almost equal to the area of the dish itself. If a dish has a surface length (not radius) of ½ wavelength, the gain is close to 10db over our omnidirectional antenna when placed at the sphere’s center. The gain of such an antenna increases with each increase in frequency and jumps in gain can be achieved at each harmonically rated frequency. As the frequency is raised, the surface area of the dish does not follow in a linear relationship to its increase in gain. In fact, the gain increases at a much slower rate than its surface area as the frequency is raised. In other words, the capture area of the dish (the dish surface area) increases faster than its gain as the frequency is raised. So a relationship has to be developed that ties wavelength, structural dimension (length, area) and volumetric dimension of the major lobe(s).
I will use my V- beam as an example. Like the dish, the long wires forming the V-structure are usually many wavelengths long. A 3 wavelength per leg V-beam has a gain of about the same as a 2 element yagi whose elements are only ½ wavelengths long. The two structures are about as dissimilar as you could imagine. However, accepted practice would say that their capture areas are identical. Similar arguments were made when the quad verses yagi war raged. Jim Lawson,W2PV, put that one to bed with his description of the 2 element quad being a bent element 2 over 2 stacked array in reality. When comparing the gain of a 2 element quad to a 3 element yagi, they are almost equal. However their major lobes were drastically different due to the quads stacking factor. Lawson calls it “stacking gain”. This is just another name for volumetric efficiency.
73′s Pete, K4EWG
Due to circumstances beyond our control, it was necessary to transfer to a new server once again. In doing so, we lost the previous database. Please be patient with us as we rebuild the site.
Recently I have been asked to provide a high gain LPDA design for Rick Doughtery’s championship contest station NQ4I. His primary concern is 80 meter performance verses the New England big gun contest station runs to Europe on that band. I met with Rick at his antenna site and evaluated it’s suitability for such an array. Fixed, point to point communication is desired for both Europe (NE) and Japanese (NW) paths. Obviously the structure requires a large land area and suitable supports for the multiple wire elements. Prior to the site visit I had done some preliminary calculations assuming the land areas available were not a restriction. Rick had mentioned his 4 square arrays were in the trees and I was concerned that the area(s) available may be restricted by woods.
The site analysis is critical prior to any type of high gain array location and design. Rick’s plan was to extend an existing tower to 200’ and use that for one end of the boom. A catenary cable would then support the boom from longest element end of a mono band log periodic dipole array (LPDA). The other end of the boom could be supported by an existing 80 foot tall tree (10 to12″ DBH) located across a stream and about 210 feet NE of the tower. I was not surprised to find the proposed site to be heavily over grown with many 6 to 8 inch DBH trees scattered throughout. Two to three acres must be cleared prior to the construction of an array this size. The clearing limits must be accurately determined and staked (flagged) prior to clearing. The site in a NE direction contained a slight rise in elevation and was heavily wooded with hardwood trees (10 to 12 inch DBH). Rick said the land was flat. However, since a small stream exists a few feet from the base of the tree, that indicated a slight rise on its opposite side. The rise in elevation (from the stream) and dense hardwoods would reduce the gain and efficiency of a vertical four square or monopole LPDA. Existing beverage wires cover the area and could interfere with a vertical elevated LPDA counterpoise. Rick said his 4 and 6 square arrays perform poorly to Europe on 80 meters. They seem to perform well on 160 meters, however. Both vertical arrays are beaming into dense hardwoods that may impede higher frequency efficiencies. The decision was to design and build a mono band, horizontally polarized LPDA.
The LPDA boom would slope along its boom from a height of about 150’ (later 200’) to 80’ at the feed point. The elements would also slope in an inverted –v configuration depending on the element end support system. The final design will include either Rohn 25 or telephone pole supports for the end element cantenaries. Four will be required and the clearing limits must include room for these supports.
Art Collins in his “Fundamentals of Single Sideband” book contains interesting information concerning log periodic dipole arrays. Since such an array is frequency independent depending on various design parameters, many different combinations of element assemblies are possible. The term “dipole” is a little misleading. The actual array is an area or surface excited system. It is triangular in shape and one triangle area is excited 180 degrees out of phase with its counterpart. The half wavelength of the triangle’s base determines the low frequency cut-off and the apex determines its high frequency cut-off. Since at HF, the wavelengths are large relative to array structure, pipes or wires can be used to approximate the triangular surfaces. Art chose two planar, end fire arrays on two separate booms joined together at their common feed point. At UHF, surface arrays are practical and are used in many high gain cavity or slot array designs. Collins labs measured gain, front to back ratio and VSWR as the physical dimensions of the array were changed. Since the charts reflect direct antenna range field measurements, they can be considered to be accurate. His book was published in 1959 long before computer modeling was available. Gains are typical 8 to 17DBi depending on design parameters. The gain, front to back ratio and feed point impedance follow the phase center of the array. That is the distance from the feed point to the phase center measured in wavelengths.
Rick’s array will be designed covering a frequency range of 3.456 to 3.950 MHZ. As the frequency is varied from say CW to SSB, the phase center will move in a direction along the boom closer to the feed point and the shorter elements. For optimum performance the phase center height above ground must also change. Collins discovered that this required tilting the array, along the boom, toward the ground at its feed point. Or, placing the array at an angle to the ground! This seems to be counter intuitive since most horizontally polarized arrays are USUELY horizontal to the ground surface. My good friend Ron Lowrance, K4SX has given me a lot of grief concerning this phenomenon. Ron is correct for frequency DEPENDANT arrays like Yagis. However, this is a frequency INDEPENDANT array. Another way of looking at this is to consider the concept of “optimum gain height” (approx. 1.5 wavelengths above poor ground). The concept predicts an optimum height for a horizontally polarized antenna array based on its frequency. Multi band arrays face this problem since they are usually located at a specific height. An excellent work on the subject can be found in the June issue of QST by Kai Siwiak, KE4PT. Collins discovered the phenomenon while field testing the LPDA designs. As the higher frequencies are approached, the optimum gain height is reduced. Since Rick’s design is a mono band array tilt is not a factor. Even at 200’, optimum gain height cannot be realized. Gain the will be determined solely by the design parameters and the physical installation.
For example, it is understood (almost intuitive)that there exists an optimum height above ground for a mono band Yagi depending on the MUF and desired wave angle for the strongest signal into Europe. Other than MUF perturbations or multiple hop conditions, that angle is relatively constant. My good friend, Jim Lawson (SK), W2PV, provides an excellent description of this phenomenon in his book “The Yagi Antenna Handbook” (ARRL publication). Rick, NQ4I, needs good runs into Europe to compete with the New England contesters and he does not have enough room for a 6 wavelength per leg Rhombic on 80 meters. Art Collins shows that the vertical plane pattern of the LPDA array and ground system (to the same degree that the ground conductivity and dielectric constant are independent of frequency) is independent of frequency and the maximum beam occurs at an elevation angle at which the array is inclined to the ground. In other words, the LPDA should be tilted toward the ground at the desired angle for optimum signal into Europe on 80 meters. To accomplish this, Collins stacked two LPDA arrays and fed each array 180 degree out of phase. I will review this technique later under the section I call “Physical Construction Techniques”.
So the design must optimize taper (tau), spacing of elements (sigma), boom length (a function of tau and sigma) and array tilt angle. Tau determines the number of elements and sigma determines their spacing. For example, if the “run station” uses 3530KHZ during a contest, the OPTIMUM phase center must be at that frequency. If it is desired that other frequencies are to be used, other phase centers must be able to shift without any reduction in performance. To accomplish this, more elements are needed and tau must approach 1.00. In a well-designed LPDA, only one reflecting element is active while the major end fire current distributions (end fire power directivity) are contained in the shorter elements within the log cell. Since this design is essentially a mono band LPDA array, all elements SHORTER than ½ wavelength will contribute to the overall array gain. Jim Lawson, W2PV, also found that 2 reflectors on a conventional yagi array showed little or no additional gain. Cheong, et al, in his monumental work and experiments with LPDA element to element current distributions also found that the shorter elements (as short as 3/8 wavelength) were the major contributors of array gain. Elements longer than an equivalent yagi reflector contained very small current distributions and therefore did not contribute to LPDA end fire directivity and array gain. As it is with a well-designed Yagi, maximum gain is also dependent on boom length. The same is true with a log periodic array. Rick measured the distance from the tower to the tree at approximately 210 feet. I will use 200 feet as a starting point with an optimum tau and sigma.
Let: tau = 0.9868; L1 = 135.11’; F1 = 3.456MHZ; Fn = 3.950MHZ; Boom = 200.00’
Then: L2 = 133.32’; L3 = 131.57’; L4 = 129.83’; L5 = 128.12’; L6 = 126.43’; L7 = 124.76’; L8= 123.11’ L9 = 121.48’; L10 = 119.88’; L11 = 118.30’
There are no multiple reflectors and the log cell gain is primarily dependent on the shorter elements including L11. L1 is designed to perform as a reflector for the entire array band pass. The array bandwidth from 3450 to 3950KHZ is essentially the same due to the array’s periodicity controlled by log(1/tau ) + 1= 1.00577. The antenna gets its name “Log Periodic” as a result of this phenomenon. That is, for this design the array parameters of gain, vswr, patterns (E andH) and front to back ratio repeat by a factor of 1.00577 along the array’s band pass. This particular design will contain the following periodic phase centers: 3500,3516, 3536, 3556, 3577, 3598, 3618, 3639, 3660, 3681, 3703, 3724, 3745, 3767, 3789, 3811, 3833, 3855, 3877, 3899, 3922, and 3945KHZ. The run station(s) should choose frequencies on or close to these for optimum array performance.
The maximum spacing between elements L1 and L2 is fixed based on the following physical constraints: Boom length = 200’; total number of elements = 11; tau = 0.9868. Since it is desired to maximize array gain for the space available, the relative spacing constant sigma and subsequent spacing between L1 and L2 must be determined with a fixed boom length of approximately 200 feet.
Let Alpha = the horizontal array apex half angle; Alpha = arc tan (L1 – L11)/400 = 2.4064 degrees
Let d12 = spacing between elements L1 and L2; d23 = spacing between elements L2 and L3; d34 = etc.
Using algebraic ratios for similar triangles: d12 = (L1 – L2)x200/(l1 – L11) = 21.30’
Then: d23 = tau x(d12) = 21.02’; d34 = 20.74’; d45 = 20.47’; d56 = 20.20’; d67 = 19.93’; d78 = 19.67’; d89 = 19.41’; d9,10 = 19.15’; d10,11 = 18.90’. Total design boom length = 200.79’.
The relative spacing constant Sigma = 0.0912
Physical Construction Techniques:
Since the element number, length and spacing have been determined, the physical structure can be evaluated. The proposed tower height at L1 is approximately 150 feet (after July, hopefully before CQWW, 200 feet). The tower (tree) height at L11 is approximately 80 feet.
(A) Single LPDA:
1. Assuming a single LPDA, a cantenary line of Kevlar will be strung between both supports and fixed at each support by a pulley and tied off at the base of each support (tower and tree).The center insulator of each element will be attached to a pulley that will travel freely along the cantenary line. Each element’s center insulator pulley will be located at the insulator’s apex and bolted in place. Each element center insulator shall be triangular in shape, 8″ base width, 12″ altitude height and fabricated from ½” thick sheet fiberglass. The array feeder shall be #12 or #14 copper weld and shall contain fiberglass spreaders 3’O.C. The spacing between each feeder wire shall be 6″. There shall be NO crossing of the feeder wires. Transposition will only take place at the insulator itself using soldered in place jumper wiring. I will prepare shop drawings showing fabrication dimensions, hole size and cutting required. All feeder wires between each element shall remain parallel to the ground and be spring loaded tight at the base of each tower (tree) support. The entire central feeder line system shall float on its cantenary line and remain in position only by spring loading at each tower (tree) base. The cantenary line will NOT be spring loaded.
2. Each element end shall be tied to a fiberglass insulator with a hole spacing of 8″. The end insulators shall be 8″x2″x1/2″. There shall be a Kevlar cantenary support system for the element ends and each element end shall be tied in place. The element end cantenary shall be supported by four supports. The supports should be constructed by suitable towers (Rohn 20 or 25) or masts no shorter than 75 feet each. The cantenary lines will be fixed at the top of each support. Guy wires for the supports shall not be longer than 100’.
(B) Dual or stacked LPDA:
1. This type array removes the hassle of a parallel feeder system and increases gain by an additional 3DB. It does not require a parallel two wire feeder system between each element. It does require 22 elements in lieu of 11. It also increases volumetric efficiency (capture area) since it approaches that of a stacked array in a similar way to that of a Quad (close vertical spacing and bent yagi elements). Two identical wire arrays are constructed on two separate central WIRE cantenary lines. Each central cantenary line is the boom for each array. The wire elements are then soldered in place along the boom at the element centers. At L1, the cantenary is grounded to the tower at the 150’ height (top array, LPDA1). At the LEFT end of L1 an electrical connection is made to the LEFT end of L2. The RIGHT end of L1 is left open. Then the right end of L2 is connected to the right end of L3. This process of alternating end connections is repeated for all elements, ending at L11. At the boom feed point, the LEFT end of L11 will remain open. End cantenaries of Kevlar and wire connections will tie the right and left sides of the array.
2. The second array (LPDA2) is dimensionally identical to the one described in paragraph 1 except that the RIGHT end of L1 is connected to the RIGHT end of L2. The LEFT end of L1 is open. The alternate end connections are repeated to L11. As it was with the top array, the bottom array will continue the alternating end connections to the boom feed point at L11. The RIGHT end of L11 will remain open. The second array’s central WIRE cantenary will be grounded to the tower at L1 and at the 75’ height.
3. LPDA 1, the top array, will run from a height of 150’ at L1 and 80’ at L11. LPDA 2, the lower array, will run from a height of 75’ at L1 and 75’ at L11. So you can get a picture of the two arrays fed 180 degree out of phase relative to each other by stacking and transposition taking place at the ENDS of each element than at the element centers. The angle formed by the two booms (cantenaries) sloping toward each other is defined as “psi”. The psi angle in this array is 20 degrees. E plane (horizontal) half power beam width is almost constant at 63 degrees at psi from 0 to 60 degrees. H plane (vertical) half power beam width varies from 60 to 100 degrees at psi from 0 to 60 degrees. Front to back ratio approaches 30DB at a 20 degree psi. When the tower is raised to 200’ and the top array (LPDA1) is raised to that height at L1, psi will increase slightly even if the 80’ support at L11 remains with the lower array (LPDA2) is raised to 125’at its L1.
4. There is a small physical problem if the element end support cantenaries are very low. Both arrays will be an inverted-v. The end element cantenaries should be at a height no less than 75’ at L11 and 100’ at L1 (the central cantenaries are 150’ and 75’ at L1). That is, the two front towers would be 75’ in height, the rear towers would be 100’ in height. The vertical separation between the element ends for the top array at L11 (at a height of 75’) and the lower array at L11 (at a height of 70’) will be 5’. The vertical separation between the two arrays at L1 is 100’-75’ = 25’. The top array element end cantenary slopes from a height of 100’ to a height of 75’ at its feed point. The bottom array element end cantenary slopes from a height of 75’ to a height of 70’ at its feed point.
5. The feed point is at L11 and a two wire open line feeder is connected to each boom or central WIRE cantenary line. One feed line wire is connected to the top array boom at L11 and one feed line wire to the bottom array boom at L11. Since the top array boom is 5’ above the lower array boom, a vertical section of wire 2.25’ long is connected to each boom and supported by a central insulator that will provide the support for the open wire transmission feed line.
6. So we have developed a “plumber’s delight” type of construction for an LPDA! Structurally, this is a great system for a wire, point-to-point LPDA. It rivals a VOA Sterba Curtain and if Rick is going to put this kind of effort and money into such an undertaking, it should be done right. For such a massive 80 meter VOA point to point wide band array it is my choice. Such a system utilizing tubular elements and rotating, however, presented structural challenges as Collins discovered. I may build one for HF at K4EWG. I will not use high tau in the design like Rick’s monster, but it may be interesting to measure the gain verses a single LPDA. I like “Volumetric Efficiency” as my rhombic buddy, W1VDE, calls “Capture Area”.
This completes the K4EWG design for NQ4I. Rick should model both arrays and choose the best for gain and capture area. Cost should be only a secondary consideration if he wants to hang tough with N1 land to Europe on 80 meters during contest time. Other than buying more land or moving to N1 land there just isn’t much left to do, gain wise that is. 6000 Q’s during the last contest is not too shabby from Georgia! I think NQ4I can be on top if he will clear a few trees and put up some super gain point to point antenna arrays. VOA did.
73′s Peter Rhodes, P.E., K4EWG
Need an antenna design? Contact me K4EWG@K4EWG.com.
As a Registered Professional Engineer I feel obligated to warn the radio amateur community regarding the use of guy wire rope clips made in China. I purchased twenty five ¼” guy wire rope clips from a local Home Depot store. Their packaging contained the required country of manufacture: “Made in China”. I decided to try these clips anyway. While tightening one clip u bolt against a ¼” cable splice, its thimble broke in two pieces. Thinking this was a single event, I tried to tighten another rope clip only to find a similar break. In fact over half clips purchased, the thimbles broke in half.
I was tightening the clips using a simple nut driver for a snug fit on ¼” steel guy wire prior to the final tightening with a ratchet wrench. I was surprised to see the thimbles break apart with the light torque of a nut driver. A rope clip with a broken thimble will allow the guy wire to slip. During one of my continuing education classes, I remembered the instructor stating that structural steel made in China must not be used in construction. China’s steel production is poorly controlled and does not meet U.S. standards for structural steel. Also, project construction plans should contain a note that steel made in China is not acceptable, approved or to be used in the construction of the project or any portion thereof. The instructor’s words rang true as I tightened these rope clips with only a nut driver.
I have checked other stores such as Lowes, Ace Hardware and found that most of their fastener and hardware products are made in China. Perhaps for light duty “home use” it’s O.K. but for structural use it is not safe and should not be used. It should not be used for a tower guy wire support system.
K4EWG
Now to the tuner. Some time ago I prepared a fairly detailed description of why perfect reflections are the best transfer of energy from the generator to the antenna and back in multi-band antenna systems, provided there are no losses between antenna and generator.
Coils, coax, baluns, taps, gamma matches, RF chokes, tuners and any other
device within the RF circuit contain losses. I have attached that paper for
your reading, when u can’t sleep. The simple explanation is this: for each
traveling RF wave front along a feed line there is a RF traveling back along the
same feed line if there is any miss-match between the generator, antenna, feed
line impedance and resistance losses. Resistance losses increase inside the
feed line as the number of wave fronts contained within the line increases. The
terms forward and reflected power are related. The forward wave front will be
totally accepted by the antenna in a perfectly matched, no loss condition. In
reality, this is never the case, since in a perfectly matched condition there
is always resistance loss. In our ham world, we need reality. A miss-matched
condition, usually exists. The forward wave front meets the reflected wave
front and both are carried by the feed line. Depending on the degree of
miss-match, the feed line can contain multiple wave fronts at the same time,
one on top of the other. This process of
forward-to-reflected-to-forward-to-reflected continues until all the RF energy
is radiated, converted to heat or both by the RF circuit. Since we don’t want
our RF energy to heat our room or the birds outside, we need to design a RF
circuit that uses the reflected wave front. A multiple wave front circuit is
not a problem provided circuit losses can be minimized. That is: no coax, no RF
choke, no ferrite balun coil, no gamma match, no hair pin match with 1:1 balun,
no tee match and 2:1 coax balun, no so called infinite match with RF choke
(Mosely CL-33,TA-36,etc. system) or any other lumped reactance within the RF
circuit. To eliminate all of these is a tough task. However, if we use open
wire line, we can eliminate most of these losses from the circuit.
Look at the tuner as a mirror. It
should bounce the reflected wave front back to the feed line. An aluminized
mirror will reflect almost 99% of the light it receives, depending on the
smoothness of its surface and its polish. A well designed tuner can do the
same, provided loss elements are removed from the circuit. Then the multiple
forward-to-reflected processes can continue indefinitely until almost all the
RF energy is radiated. Since resistance losses are dependent on the SQUARE of
the RF amps, a 1.2:1 VSWR on a coax feed line can convert many RF amps to heat.
It is an exponential loss, not a simple 1-to-1 loss. RF amps inside the coax
are the result of multiple forward-to-reflected RF wave fronts. The coax actually
will contain more RF current than the generator (transmitter) can deliver. Coax
line loss tables in our handbooks have no use when miss-matched circuits are
used. The losses are exponential! A tuner at the transmitter means nothing at
all in a miss-matched coaxial RF circuit. A coax-to-coax trans-match provides a
resistive load for the transmitter ONLY. And if it uses the so called “current
balun”, it simply adds an additional loss to the RF circuit. I get a kick of
the tuner manufacturers using the term “current balun”. They seem to think they
can disguise the RF amp losses in the circuit by inserting a current balun
between the tuner and the transmitter (they apparently got tired of hearing
complaints about 4-to-1 voltage baluns blowing up when located on the antenna
side of the tuner). A current balun is no more than a RF choke! No different
than a coil of coax between the transmitter and the tuner! Coax is an
un-balanced RF conductor with the RF currents carried by the center conductor
and the INSIDE surface of the braid. A matched, un-balanced coaxial circuit
will contain the RF wave fronts inside the coax cable and the outside surface
of the braid will act as an RF shield. However, once the un-balanced RF coaxial
circuit is connected to a balanced RF circuit, miss-match occurs at the
connection, and the RF wave front travels on both the inside and the outside
surfaces of the braid. Hey, that is energy that is lost! No worries, we will
put an RF choke in the circuit and convert that reflected energy to HEAT! No
one will know the difference and our coax line will be cool as a cucumber. It
costs a lot to produce RF energy. Even our best Alpha designed amps are 50 to
60% efficient! What’s a few more lost RF watts anyway? We exponentially loose
these watts due the I square (RF Watts=IxIxR) law when multiple wave front
loads exist on the feed line even with a coax VSWR as low as 1.2 to 1!
One more note regarding the infamous
balun coil. Broad band balun coils have been around for a long time. Originally
they were a pair of 3.5” diameter coils, air wound, 30 bifflar turns for each
coil, # 22 tinned copper wire, 1/8” separation between the bifflar windings
(mfg. by B&W). The coils could be connected for balanced to un-balanced,
balanced to balanced, un-balanced to un-balanced in ratios of 1:1 and 4:1 only.
They were extremely power limited and were rarely used. With the invention of
iron ferrite materials, the number of turns decreased and wire gages increased,
but the principles were the same. Two coils tightly coupled by bifflar windings
could provide an almost RF frequency independent transformer similar to
transformers used at audio frequencies (ie. most audio transformers transfer
all ear detectable audio frequencies). The catch is that these RF balun transformers
(including audio transformers) MUST terminate into a pure resistance on both
ends (in-put and out-put) of the transformer. The “current” balun is no balun
at all. It is a simple RF choke that is terminated in the characteristic
impedance of a coaxial feed line. It does not transform impedances, it simply
prevents the travel of an RF wave front due to its high resistance (impedance)
to the wave front’s RF current. When a balun transformer is not terminated in a
pure resistance a portion of RF current is stored within the coil and its core
and the balun has reactance that not only impedes RF transfer but creates heat
(ie. loss). Early manufactured tuners incorporated a 4 to 1 balun at the
antenna connection for balanced feed lines. They contained multiple stacked
ferrite toroid cores wrapped with many layers of fiberglass and Teflon tape.
Stacking was to prevent RF current saturation of the core. Teflon and
fiberglass was to prevent RF high voltage arcing and breakdown. Most amateur
installations, especially multi-band dipoles, zepps and G5RV types came nowhere
close to a 4 to 1 impedance transformation ratios. Resistive terminations were
non-existent. The industry had to come up with a simple way to prevent the
solid state transmitters from shutting down in the presence of such reactive
terminations. Today they seem to have replaced the 4 to 1 in favor of the RF
choke (“current balun”) between the transmitter and the tuner. We don’t want to
turn any of our transmitted RF energy into heat! We want to use it all,
including reflected energy as a result of miss-matched antennas and assorted
reactances within the circuit. We want an RF mirror!
Look at the entire picture, the
whole RF circuit flow path: generator to tuner to half the feed line to half of
the antenna to the other half of the antenna to the feed line wire from the
antenna to the tuner and back to the starting point at the generator. The
process reverses for a receiver. The RF travels from the generator into a 50
ohm resistance, no reactance at the tuner input. The tuner must be able to
transform a 50 ohm termination into a complex reactive impedance from the
antenna and feed line connected to its output. The tuner must provide an equal
and opposite reactance that will cancel this antenna and feed line reactance.
The feeder wire to the antenna contains a traveling wave front to the antenna
and has a standing wave along its entire length. Half of the antenna receives a
reactive RF current that induces (creates) an equal, 180 degree out of phase RF
current in the opposite half of the antenna. The return feed line wire carries
this out-of-phase RF current back to the tuner. Since both feed line wires are
parallel and close together (6” to 8”) and if the antenna is physically
balanced (equal half lengths), the reactive RF current traveling in the return
wire will be equal and out of phase to the opposite feed line wire and there
will be a cancellation of RF fields (minimum radiation from the feed line).
Note that this is true regardless of the physical length of the antenna, the
physical length of the two feed line wires, the frequency, the characteristic
surge impedance of the two wires, the antenna resonant frequency, or the value
of the reactance and resistance at the tuner. The entire antenna length plus
the TOTAL length of the feed line wires is said to be a RESONANT antenna
system. If the TOTAL physical length of the system is shorter than one-half
wavelength, the reactance will be capacitive. If longer, it could be inductive
OR capacitive. The design of our pure reflection tuner must use coils and
capacitors that will send that return wave front back up the feed line with no
lossy gimmicks or foolishness along the way!
The design I drew up for you at
Dayton will do that. I will send the schematic with some photos. The tuner on
the web page is the same circuit with plug in coils. The differential variable
capacitor is a combination on one shaft. It can be made using bone yard
variable capacitors with some difficulty. The rotors are not grounded and are insulated
from each other by a ceramic insulated coupling. There are 4 stators insulated
from and mounted on top of each other. The rotors rotate between top and bottom
stators. Ill try to give a photo of how its done. It will require two split
stator variable caps to build the differential variable cap.
I am still building my second
transmitter. I had some photos and video of a transmitter on my web page a few
years ago, but that design, well let’s just say I had a few expletives, was
SCRAPPED! Should have new pictures soon.
So, 73′s and good DX!
Pete, K4EWG
Article written by KI4KJV with
excerpts from K4EWG correspondence.
From K4EWG Correspondence files
It was great to hear from you! I look for you on 40M CW but I guess we’ve been zig-zagging. Glad you’re not using one of those so called balun RF chokes in your antenna system. It’s too bad they became an easy fix for the tuner manufacturing industry and won approval of the un-knowledgeable amateur community ( My friend, Lew McCoy, W1ICP, may have been responsible because he put it in his original QST article: “The Ultimate Trans Match”). It is indeed difficult to design and manufacture a tuner that is capable of matching a 50 ohm unbalanced output to a balanced, multiband antenna using open wire feed line containing reactance (Resonant system). Johnson did a fairly good job with their trans match using the dual differential variable capacitor. Their limitation was a switch controlled inductor. Lew’s trans match “T” bridge auto transformer design works great for un-balanced to un-balanced systems. However, huge circulating ground currents occur when a balanced, resonant system is used with a T bridge auto transformer design.
I am of the opinion that thirty feet is just too low for the low bands, however, I read some interesting info contained in a 1953 antenna Army Technical Manual. It dealt with Beverage antennas used for transmitting. The manual is TM-11-666, “Antennas and Radio Propagation”, pp. 148-151. I quote from page 150: “Experiments in the range of 100KC to 200KC with wave antennas (Beverages) of 0.6 to 1.5 wavelengths, compared to standard flat top (1/2 wave dipoles) mounted on 180 foot towers, showed gains of 10db for the wave antenna. Highest gains were noted for longer wave antennas erected over poor ground. The efficiency of the wave antenna increases rapidly as the height is increased from 0 to the range of 12 to 15 feet. Above 15 feet, there is little increase in efficiency.” They go on to say a wave antenna of 2 wavelengths compares favorably in its favored direction (along the wire axis) to a full sized quarter wave vertical. A 4 wavelength compares favorably to a half wave vertical.
They have test data to substantiate the claim, but I had never heard such claims before. I think it would be neat to test such a system against your vertical. I guess what I am saying is that your Vee Beam height may be O.K.
I will be looking for you on 20 or 40M CW.
Best 73, Pete, K4EWG
follow: Follow me on Facebook RSS Tweet with me