Fullscreen HTML5Med Res (??MB)HTML4

2m Elevated Groundplane Antenna An antenna designed to exactly match the impedance of the feed cable has much to recommend it. The transmitter will develop its maximum power, losses in the feed cable will be minimised and any risk of damage due to mismatch is avoided. By PHILIP WATSON, VK2ZPW 62  Silicon iliconCChip hip This view shows all the pieces of the antenna just before the final assembly. The copper tube forms part of the matching section. The materials used are all ready available and you should be able to scrounge most of the parts for little cost. T HIS VHF ANTENNA was originally constructed as part of the author’s research into the impedance of an elevated groundplane antenna, as set out in the June 1999 issue of SILICON CHIP. In particular, the author wanted to establish that an impedance matching section (or “Q” section) could be constructed, to match the 52Ω impedance of the feed cable to the 18Ω antenna impedance. In fact, the finished device has proved to be a completely practical antenna. It is simple to construct, easy to mount and because it provides the correct load, it allows the transmitter to generate maximum power. This is important because not every transmitting device is completely safe from mismatch damage. Typical commercial power amplifiers (“afterburners”) frequently carry a warning that an SWR above a specified figure will void the warranty, for example. A feature of the unit is that the “Q” section is of solid construction. It makes a substantial “handle” which can be lashed or clamped to a mast or, if the mast is in tubular form, the “Q” section can sit inside the mast, along with the coax cable. Another feature of the unit is the use of screw-on radials which can be easily detached for transport. In fact, this antenna has proved extremely useful as a temporary base antenna during WICEN exercises. Alternatively, as a permanent base station antenna, it would suit any situation requiring an omnidirectional VHF antenna for 2-metres (144-148MHz). Useful background In the June 1999 issue of SILICON CHIP, the author present­ed an article entitled “What Is A Groundplane Antenna?”. This article sought to clarify the difference between two different types of groundplane antenna – the earthed variety and the ele­vated type. Having established that the elevated version has a theoretical impedance of 18Ω, the article went on to discuss the problem of matching the antenna to the feed impedance (52Ω) and briefly described a practical antenna. However, the arrangement described in that article is not the only approach. In that case, the idea was to design the antenna itself to provide the required 52Ω feed impedance. In the current approach, the antenna is left in its simple basic form, compatibility between the two impedances being achieved by in­serting a matching device between the cable and the antenna. One of the simpler forms of matching device is what is commonly called a “Q” section; a quarter wavelength coax section having an impedance value intermediate between the two impedances (ie, between 18Ω and 52Ω). A simple formula (1) is used to cal­culate this value: (1). Zq = √(Za * Zb) where Zq = Required Q section impedance; Za = Cable Impedance; and Zb = Antenna Impedance In this case, the value for Zq comes out at 30.6Ω. And so it all appears to be delightfully simple; just insert a quarter wavelength of Zq impedance coax between the cable and the anten­ na. It’s all too easy. But of course, there’s a catch – just where do you find 30.6Ω coax? You certainly can’t get it from any of the regular electronic outlets. In fact no such material exists – all that is readily available are the (nominal) 52Ω and 75Ω varieties. Granted, there are some tricks available – eg, two lengths of coax connected in parallel will provide half the impedance. From this, the best choice would seem to be two 52Ω parallel lengths to produce an impedance of 26Ω. That’s much better than the gross mismatch of 2.88/1 using a straight connection but still short of the ideal. The error is similar using two parallel 75Ω lengths. A possible solution to this problem might be to use a parallel arrangement made up of one length of 52Ω cable and one length of 75Ω cable. This would produce an impedance of 30.7Ω (which is very close to the required value of 30.6Ω) – assuming that the simple resistances-in-parallel law holds true. The author hasn’t tried using this technique, however, and so is unable to vouch for its authenticity. Transmission line tricks Fortunately, some of the techniques FEBRUARY 2001  63 How it goes together – the 6.35mm (OD) brass tube is pushed down the 12.7mm copper tube to form the matching section. It is then soldered to the centre pin of a PL259 plug via a short length of tinned copper wire – see Fig.1. employed by users of open wire transmission lines can be used to solve our impedance matching problem. An open wire transmission line can be made with any impedance value (over a wide range) simply by selecting a suitable wire gauge and spacer dimensions. So, if a “Q” section is required, it is easily made to the required value. Coming back to the coax scene, could the same trick be pulled there? Could a length of “coax” be constructed to have any required impedance? The answer is yes and a formula is available to design it. Taken from the ARRL Antenna Handbook, it is as follows: (2). Zo = 138(log D/d) where Zo = Characteristic Impedance; D = Inside Diameter of Outer Conductor; and d = Outside Diameter of Inner Conductor This, of course, is for an air-spaced device. In theory, the use of any spacers would alter both the impedance and the velocity factor. In practice, this can be ignored – at least in the context of this article and the antenna described here. Practical considerations So much for the theory. Putting this idea into practice is another matter, since we are no longer thinking of a flexible cable. Instead, we are talking about a rigid device which must somehow be mounted. And, of course, suitable materials with the appropriate dimensions must be found to build the matching sec­tion. Strangely enough, finding the materials turned out to be the least of our problems. A hunt through my scrap­metal box soon yielded an odd length of 12.7mm OD copper water pipe plus a length of 6.35mm (0.25inch) OD brass tube. Well, that was as good a place as any from which to start. The water pipe would serve as the outer conductor, while the brass tube would become the inner conductor. As it turned out, I was lucky – when the appropriate meas­urements were The PL259 plug is connected to the “Yorkshire” fitting at the end of the copper tube using 1/8-inch Whitworth screws, as shown here and in the photo at right. 64  Silicon Chip fed into Eqn.(2), the result came out within a whisker. More exactly, it worked out as follows. The diameter (d) of the brass tube inner conductor (6.35mm) was already known but the inside dia­ meter of the outer conductor had to be measured. Since I didn’t have an inside micrometer or callipers, the best I could come up with was a finely calibrated steel rule and this gave a figure of 10.5mm (D). When these two figures were fed into Eqn.(2), the charac­ teristic impedance (Zo) came out as 30.14Ω – not quite the 30.6Ω being sought but probably “within acceptable tolerance” as an engineer might say, or “near enough” in layman’s terms. Mechanical design Now it was a matter of deciding on a suitable mechanical design and the physical construction. Originally, the idea was to build the matching section as a separate unit which could be coupled to the antenna base using an appropriate plug and socket combination. However, while assembling a rough mock-up of the inner and outer conductors, a much simpler approach suddenly suggested itself. If the inner conductor was extended beyond the antenna end of the “Q” (matching) section, it would form the beginning of the antenna. And by further extending this to an appropriate length, it would form the antenna itself. This changed the whole approach; the tail was starting to wag the dog. Instead of starting with an antenna and making a “Q” section to attach to it, we are now making a “Q” section which also becomes the antenna. So the logical approach is to combine the two items into one structure. Not only is it simpler and cheaper to build, obviating the need to supply and fit a plug and socket assembly, but also rather more elegant technically. In theory, the presence of conventional plug and socket assemblies – or any junction – in a coax cable creates a discontinuity which increases losses. Just how serious this is in practice may be debatable but, anyway, every little helps. So much for the theory. The first construction step is to ensure that the ends of the copper tube are cut square. Ideally, this should be done using a tube cutter, as used by plumbers, if one can be begged or borrowed. If a hacksaw is to be used, take care in marking and cut­ ting. Use the straight edge of a piece of paper wrapped around the tube to mark out a cutting guide, then cut a shallow groove right around the tube. Deepen this cut progressively by rotating the tube a little at a time until the operation is complete. Be sure to cut slightly to one side of the cutting guide, so that the end can later be cleaned up with a file. Don’t try to cut straight through the tube in one opera­tion. It will almost certainly come out crooked if you do. The antenna end of the tube is fitted with a small metal plate which supports the four radial elements. In the writer’s case, this was made from a piece of scrap brass, cut to about 110mm square (although this isn’t critical) and drilled with a central hole to match the OD of the brass tube. The plate is simply flush-mounted with the end of the tube and secured by soldering (eg, using flux, a solder Fig.1: this exploded diagram shows how the antenna is assembled. Note that the 6.35mm OD brass tube is used as both the radiator and as part of the matching (Q) section (ie, the brass tube is 1004mm long). An insulating grommet isolates the radiator from the copper tube at the brass plate end. FEBRUARY 2001  65 The radials are tapped at one end to 4BA x 10mm to match the spacers on the brass plate and fitted with a soldered “stopper” nut. This makes it easy to dismantle the antenna for transport. Alternatively, for a fixed installation, the radials can be soldered directly to the brass plate. stick and a gas flame to provide the heat). An alternative form of plate is a press-on lid as used on large coffee tins or similar containers, painted for protection from the weather. The radial elements can be made from any convenient size and type of material. The writer used 3.175mm (0.125in) brass rod but larger dia­meter tubing could also be used. The radials are each about 450mm long and can be directly soldered to the four corners of the metal plate. Alternatively, the radials can each be tapped at one end to 4BA x 10mm. A brass nut is then threaded onto each radial to act as an end stop (and soldered in position). The radials are then screwed into 4BA brass spacers soldered to the corners of the metal plate (see photo & Fig.2). The prototype used round brass spacers but hexagonal spacers would be much easier to position during soldering. You can buy a pack of six from any of the parts re­tailers for around $3.00. The advantage of this latter scheme is that it allows the antenna to be easily dismantled and transported, if necessary. plumber’s “Yorkshire” fitting. (Note: the metric dimen­sions are rounded to 12mm in hardware literature). The “Yorkshire” and “Yorkway” unions are designed to join (ie, buttjoin) two lengths of 12.7mm OD copper tube. In this case, the unit used should be specified as a “slip fitting” which has no stop in the centre (as normally used to simplify correct This close-up view shows how the end of the brass rod is plugged and drilled to accept the short length of 1.3mm tinned copper wire which connects to the PL259 plug. Termination The cable end of the tube is terminated with a PL259 plug. The plug body is the same diameter as the OD of the tube (ie, 12.7mm) and is buttjoined to the tube. It is secured using a brass sleeve consisting of a 12.7mm (0.5-inch) ID union – a standard 66  Silicon Chip The insulating grommet should be a tight fit over the inner brass tube. It is pushed down into the copper tube at the end of the matching Q section during the final assembly. positioning over a junction). Both fittings are designed to be soldered to the copper tube. The “Yorkshire” fitting is supplied with two internal rings of solder. The copper tubes should first be cleaned and fluxed, after which everything is fitted together and hit with a gas flame until solder flows right around the end of the union. A “Yorkway” fitting is treated similarly, except that the solder has to be applied externally to the ends of the union. In the writer’s case, a “Yorkshire” fitting was used simply because it was on hand but it would probably be the preferred device. The PL259 plug was secured into the sleeve using two 1/8-inch Whit­ worth screws. Matching holes are drilled through the sleeve and the plug body, initially to suit a 1/8-inch Whit­worth tap. The holes in the plug are then tapped, while the holes in the sleeve are enlarged to provide clearance. The same arrangement can be used to secure the sleeve to the tube or, as in this case, the sleeve can be secured by sol­dering. The distance from the end of the PL259 plug to the start of the pin is about 20mm (as measured inside the plug) and the internal diameter of its body is similar to, but not identical with, that of the tube. But although not exact, it really is close enough considering that only 20mm is involved. As a result, the plug can be treated as an extension of the tube. This means that the tube must be cut 20mm shorter than the calculated section length (ie, to 494mm instead of to 514mm, as quoted later on). The cable end of the inner conductor (ie, the 6.35mm tube) has to be joined to the pin of the PL259 plug. This was done by first plugging the end of the tube using a 3/16-inch brass ma­chine screw. This screw was soldered into place with its head cut off and filed square with the end of the tube. The screw was then drilled longitudinally to accept 16g (1.3mm) tinned copper wire (about 25mm long), using a No.55 or 5/64-inch drill, to a depth of about 6mm. The 16g wire was then sol­dered in place. When the antenna is later assembled, this 16g tinned copper wire slides into the plug pin and is soldered. Longitudinal drilling can be a tricky operation unless one has access to a lathe. However, provided care is taken, it can be done using a hand drill (eggbeater). Just be sure to accurately centre-punch the end before drilling. Small off-centre errors are easily corrected simply by bending the wire. Larger errors can be corrected by drilling an oversize hole and accurately positioning the wire prior to sol­dering. The inner conductor is secured where it emerges from the outer conductor (ie, at the ground­ plane) using a simple plastic grommet. This insulator should have a bore size of 6.35mm (0.25-inch) to accept the inner conductor and should be a tight fit into the 12.7mm outer brass tube. Scrounging the copper tube Obtaining a suitable length of 12.7mm copper water pipe for the outer conductor shouldn’t present any problems. Normally, plumbers buy it in standard 6-metre lengths but most hardware stores will sell it to you by the metre. There are also two other likely sources of scrap lengths: (1) a local plumber and (2) a scrap metal yard. A scrap metal yard will also usually have brass rod and tubing in a range of sizes and this can be cheaper than buying commercial lengths from a hardware store. Fig.2: the four tapped brass-rod radials screw into threaded brass spacers (or standoffs) which are soldered to the four corners of the mounting plate. Antenna dimensions The exact dimensions of the antenna assembly will depend on the particular frequency to be favoured. The antenna described here was designed for 146MHz which equates to a freespace quar­ter-wavelength dimension of 514mm. This means that the outer copper tube in the “Q” (matching) section had to be cut to 514 - 20 = 494mm, as mentioned previously. The calculated length of the radiator, after allowing for end effect or “K” factor, is 490mm (ie, 514/1.05) and so the 6.35mm brass rod is cut to 514 + 490 = 1004mm. And how did all this work out in practice? Extremely well, as indi- This photo shows the finished antenna with the four radials screwed into position. Also visible is the insulating grommet (red) at the end of the matching section. Use silicone sealant to seal around this grommet. cated by the following SWR figures, which speak for them­selves. 144MHz ..................... 1.02/1 145MHz ..................... 1.02/1 146MHz *.................... see note 147MHz *.................... see note 148MHz ..................... 1.02/1 * Too Low To Measure So that’s it; a near perfect antenna - well, impedance-wise anyway. SC FEBRUARY 2001  67