This antenna is based on a design for a portable 2 element 6m Quad by VE7CA, which is featured in the 19th Edition of the ARRL Antenna Handbook. Click here for the original construction article. While the basic 2 element design was OK, I felt it worth trying 4 elements.
Unfortunately, there are several different sets of dimensions for quads. The main differences seem to revolve around different spacing between the elements, some suggest that even spacing for all the elements is best, others move the driven element nearer the first director while others move it towards the reflector. The boom length I chose was limited by the size of available "off the shelf" timber and the maximum size I could transport by car, longer boom lengths may result in higher forward gain. The element spacing I used makes for a symmetrical boom, so it doesn't matter which way round I assemble the antenna, except to ensure director 1 is next to the driven element, etc. All the spreaders and plastic corners for the elements are marked with indelible marker pen as "R", "Driven", "D1" and "D2".
The final dimensions for the four element quad I chose were based on the "on-line" calculations on the Signal Engineering web site . The dimensions for a centre frequency of 50.150 MHz are:
Reflector circumference: 247 inches, driven element: 240.5 inches, director 1: 233.3 inches and director 2: 226.3 inches.
Spacing: Reflector to driven: 35.3 inches, driven to director 1: 30.6 inches and director 1 to director 2: 35.3 inches. Note these dimensions are for bare or enamelled copper wire, not PVC covered wire.
Note that Imperial and American wire gauges are not the same, 18g Imperial is approximately 16g American, which in "real measurements" are 0.049" or 1.25mm. I used this size enamelled copper wire for both the 2 element and 4 element quads shown here.
Materials required to build the antenna are :
1 off 44mm square section x 2700mm pine for the boom
1 off 150mm x 200mm x 20mm thick plank "off cut" for the mast to boom clamp (see image of 2 ele quad below)
8 off 12mm diameter x 2400mm long wooden dowel for the spreaders *
8 off "L" shaped screw hooks to secure the dowels into the boom
Spool of enamelled copper wire - 16 or 18 gauge, (see above)
Length of plastic waste water pipe (or similar) to form the "corner" pieces, threaded onto the wire
Short length of 1.5" to 2" diameter plastic waste water pipe (diameter not critical)
9 foot length of good quality RG58 coax.
2 off "U" bolts and nuts for the mast to boom clamp
4 wood screws and water proof glue to fix wooden plate/clamp assembly onto boom
SO239 free socket
Small tin of outdoor grade varnish
Small number of cable ties, both re-usable and permanent use. The re-usable ones are used to secure the feeder to the boom and mast, two small permanent use ties are to attach the coax cable to the plastic pipe for the choke balun.
* Beware of dowel that is joined - the joint may not be water proof and can fail in damp conditions! See image below for the method of joining cheap dowel. The "zig-zag" dovetail joint can be seen on careful inspection. Pine dowel from B&Q and Homebase both feature these joints.
Assembly
The wires are measured and cut to one inch longer than needed to allow an overlap for joining, for example cut the reflector wire to 248 inches. Thread the corner pieces onto the wire, then scrape the varnish from the end inch of each end of the wire. Overlap the wires by one inch, then grip the overlapped wires with two pairs of pliers and twist the wires together. Secure with solder. Mark the corner pieces "D1", "D2" and "R" to avoid mixing them up when assembling the antenna out in the field.
In the case of the driven element, I overlapped the wires through holes drilled in the end of a 6" x 3/4" diameter piece of plastic water pipe leaving a gap between the wires of 1/2". The ends of the wires were then bent over and soldered to a 9 foot length of RG58 coax. Over the 3/4" pipe I hot melt glued a shorter length of 1.6" diameter plastic waste pipe and wound the coax around the pipe to form a coaxial choke balun. The coax was secured to the waste pipe with a couple of cable ties (Ty-Raps) and terminated in an SO239 socket. The coax should be sealed with either hot melt glue or silicon rubber to keep water out. In order to obtain a better fit for the corner pieces on the spreader dowels, a short length of plastic waste pipe (the same as used for the corner pieces themselves) was cut lengthwise to remove a section, squeezed and inserted into the corner piece. The corner pieces are approx. 3/4" diameter x 1" long, the inserts are 3/4" long.
Carefully mark out and drill the boom, DO NOT make the distance between each horizontal and vertical hole pair too close otherwise the "L" shaped screw hooks will foul the spreaders. Use a drill stand/press to ensure the holes go through the boom at 90 degrees. Drill pilot holes for screw hooks. Screw the hooks into the boom, then remove them and grind off the point to minimise damage to the spreaders when used to secure them in the boom.
Cut the spreader dowels slightly too long for the wires, fit them into the boom and trim for nice tight wires, without excessive "bowing" of the spreaders. The boom can be planed to an octagonal section in order to reduce wind resistance and weight. Both the dowels and boom should be varnished to protect them from the weather.
Tuning and performance
The antenna was assembled using the dimensions shown above, the SWR was at a minimum around 200 KHz lower than the design frequency, removing one inch from the driven element, and re-soldering it, moved the tuning sufficiently higher in the band for my purposes. Note the tuning will vary slightly with height, if possible test the SWR at the intended operating height. The minimum SWR was approximately 1.2:1, with quite sharp increases as the frequency is changed, which at least indicates it isn't acting as a dummy load. This 4 element quad gives a good match at resonance, measurements on an MFJ 259B antenna analyser read close to 50 Ohms at 50.2 MHz with a 1.2:1 SWR. Built as above, the antenna doesn't cover the whole 6 metre band. As I only use CW and SSB over a small portion of the band it's not an issue to me.
On air performance of the 4 element quad is very impressive, not having an antenna test range I can really only report my findings on a distant beacon. In the forward direction the signal was S7, off the rear S zero but audible. At 90 degrees to the wanted direction the signal could not be detected. While I could quote the signal level required to obtain S7 and speculate what S zero might have been, for all practical purposes the antenna does what I expect it should. Book figures for gain give 10 dB over an isotropic radiator for a 4 element quad. These are theoretical figures and don't necessarily equate to reality. The "free space" figures in EZNEC 4 give a gain of 9.6 dBi and a 26 dB front to back ratio for this antenna.
To pack the wires for transport, twist and loosely fold them, secure with re-usable ties, those ties found in garden centres which are used to secure plants to stakes in the garden are ideal, as they are soft and pliable.
If you have EZNEC 4, click here for a zipped file of this antenna, alternatively click here for an azimuth plot, or here for the SWR plot (both in free space, not over real ground). EZNEC 4 can be downloaded from here: http://www.eznec.com
Since originally building the quad, I purchased EZNEC and modelled the antenna. The plots and files in the paragraph above are based on the reflector having a circumference of 247 inches. Several other models of quad antennas appear to have a significant notch in the rear lobe when modelled with EZNEC, by increasing the reflector cirumference to 248 inches the front to back ratio should in theory increase significantly. When I built the antenna, attempts to tune a smaller sized reflector by adding a stub didn't appear to offer any advantage over the 247 inch reflector. I've added this comment for the benefit of those who like to experiment. The forward gain doesn't change significantly when one inch is added to the reflector. For a free space plot of this version of the antenna, click here
If you really want to lengthen the boom to 15 foot, the gain will increase by approx 1 dB. By increasing the diameter of the reflector loop the front to back ratio can be maintained. The driven element will need to be slightly shorter for an optimum match. For an EZNEC 4 file of this longer antenna click here. Note I have only modelled it, not built it.
2 element version - built to VE7CA's design
My original quad started life built to the exact dimensions in VE7CA's design, however there was feeder radiation due to not seeming to use enough turns in the coax feedline choke and both a poorer than expected SWR and front to back figures. Going back to basics, based on previous experience with HF quads, I felt that using a loop size of 1005/f, where f = frequency in MHz and the result is in feet, together with a gamma match in the driven element and a shorted stub in the reflector would "simply just work". Both the reflector and driven element are the same size (20' 1" loop), the spreaders are 7' 1" long overall (not boom to corner). Reflector to driven element spacing is the same as the original article (24 inches).
There was feeder radiation with the original coax choke, this manifested itself by a difference in SWR as seen by the transceiver and an external SWR meter near the antenna feed point. When the SWR was at a minimum on one meter it wasn't on the other, and vice versa. The number of turns of coax in the choke balun didn't seem sufficient, so I increased the amount of coax (9 foot of coax wound on a 1.6" diameter piece of plastic waste pipe), while this cured the feedline radiation, the lowest SWR I could obtain was 2:1, which was probably due to trying to directly feed a 2 element quad which is not a 50 Ohm resistive feed.
By using a gamma match instead of direct feed, which is a mini ATU at the feed point, it is fairly easy to obtain a really good SWR. Avoid running more than 5 Watts when standing near the antenna to adjust the gamma, as the RF field strength can be hazardous at close quarters. The ideal method of tuning is to use an antenna analyser if you have one. The distance from the shorting bar on the gamma match to the ground point of the loop is approximately 14", the parallel gamma rods are spaced 1" apart and the series capacitor is roughly 30 pF (after tuning). I would suggest using 18" rod lengths and a 50 or 75 pF trimmer capacitor.
The reflector stub started at 18" with a 1" spacing between the rods. There is a sliding shorting bar made from brass tube and copper wire, this can be tuned for minimum signal off the back of the antenna by running a length of cord over the boom and adjusting the bar from ground level. There didn't seem to be a peak of forward gain but there was quite a sharp dip in signal off the rear of the antenna. A signal source for these adjustments was an old Heathkit solid state dip meter placed some distance away. Once the correct position for the reflector shorting bar has been determined, carefully lower the antenna and solder the brass tubes in place. The excess rods can then be cut off, in the case of piano wire a mini drill with a cutting disc makes a quick job of the hard steel wire. The original oversized reflector gave a poor front to back ratio, where the front signal was S8 reducing to S3 off the back. Using a stub tuned reflector reduced the signal from the rear to S zero, a significant improvement. In practice the tuned stub reflector version seems much more directional than the previous version, both antennas were tested and used at 25 feet above ground.
Click here for an azimuth plot of the antenna using VE7CA's dimensions, plotted with EZNEC 4 in free space. The gain figure is 7.2 dBi with a front to back ratio of 18 dB (figures in free space).
Note, there is no need to use a choke balun with a gamma match as the outer of the coax is at zero RF potential (centre of the loop). There is a minor disadvantage to the gamma match as the direction of maximum signal is slightly offset to the plane of the loop, i.e. the antenna "squints". In practice this is not an issue as the beam width of a 2 element Quad is very broad. There may be additional losses involved in using a gamma match compared to alternative feed methods, given the relative ease of constructing a gamma match I don't feel it's worth worrying about the issue.
Extensive use was made of hot melt glue, without such glue construction would have been much harder. White plastic electrical conduit tubing was used for the various corner pieces. Short pieces of tube had a 3/8" slot cut in them, they were then squeezed and inserted into the various corner pieces to give a better fit for the dowels. There was no need to glue them in place.
Wood and plastic parts were obtained from B & Q, who stock 12mm pine dowel in 2.4 metre lengths. 44mm square section pine was used for the boom. B & Q also stocked the U bolts. 1.25mm enamelled copper wire and the "plastic project box" came from Maplin. The capacitor was a surplus item. Brass tube and 14g piano wire came from a model shop. Brass rod would be a better choice than piano wire, unfortunately I could only find it in short lengths.
The following individuals are thanked for their patience with my many questions on 4 element quads.
Brian Hummerstone, G3HBR, LB Cebik, W4RNL (sadly now both SK) and Dale Hunt, WB6BYU
And of course thanks to Markus, VE7CA, for the original idea. |
