It is the current in the antenna wire that creates the radiation. A larger current means higher radiation and so increased signal strength at the receiver. And a larger current can be acheived with final amplifiers. This is however costly. Another way of increasing the signal strength is to let several smaller radiators combine and positively reinforce with each other, thus creating lobes of maximum radiation at the expense of lowering the emission in other directions.
A common way of doing this is to put two radiators at a distance between them and feeding them both. This is exactly what happens in the half-square (HSQ) antenna for example. In the case of the HSQ the distance between the vertical wires, each 0.25 wavelengths long, is typically 0.5 wavelength. While this is (presumably) originally selected in order to get the proper phase relationship between the two verticals it also is a near optimum distance for maximum gain. Due to symmetry the antenna is bidirectional and has two lobes, a forward one and an equally large rear one.
Another simple vertically polarised wire antenna with bi-directional properties is the Bruce array. Apparently it stems from astrophysics research in the 1920’s. I have heard from older hams that this antenna used to be popular, but it has not survived the competition from todays plethora of other antennas. I find it interesting since it is:
- simple and lightweight
- does not need any particular tuning procedures,
- and seems to get my signal out in a first rate fashion!
Another interesting feature of the Bruce array is its theoretical simplicity:
The current distribution is what makes the Bruce array tick. Beginning with a wire in resonance, it is folded in such a way that the high current parts are all in the same direction. See the text. (The ends of the array can be bent according to the dotted lines in d), but the antenna becomes somewhat shorter if bent inwards as depicted.)
Starting off with the figure a) we have a 3.5 wl long wire with a generator to the left. The generator’s frequency is selected to give a number of standing waves on the wire as shown. As the wave travels to the right along the line from the generator, it finally bounces at the far end and returns and mixes with the right-going wave, resulting in the current distribution shown by the approximate sine-wave. In each half-wave the current is opposed in direction to its neighbours.
Now, the Bruce array wrinkles the wire so that the opposing currents actually combine, i.e. do not oppose each other any longer. The figure b) shows how one can imagine that creating corners on the wire actually separates the different current directions. In the same vein we can create more corners at points along the wire so that we select the points of maximum current going mostly upwards. This is depicted in c), where we have squashed the corners in b) and inserted two new corners. The complete antenna, with measurements, is found in d). In fact some sources say that the lengths should be 1.05 times the indicated figures. Using normal isolated wire with a velocity factor of about 0.95 this adds up to a length equal 0.99, which is close enough to 1.00 to be of no practical importance. This antenna is fairly wide in bandwidth terms, so lets not be overly fuzzy!
The end effect of all this is that we have a number of currents in the same direction, and a number of opposing currents in the horisontal parts of the wire. Seen from a long distance away the currents opposing each other will more or less consume each other and give a net contribution to the radiation of almost nothing. The all-vertical currents do, on the other hand, combine into a larger current and a distinct vertical polarisation.
Not every Bruce array has to be built according to the figure. One could use as little as two verticals, but the directionality is not that much then. I suggest 4 or 5 verticals according to the MMANA-simulations below (the bulge in the waist when using 5 verticals does not go away when using more verticals, even though the maximum gain increases in the main direction):
A four element Bruce array is a good compromise between gain, directionality and size.
The 5 element Bruce array has about 1 dB extra gain but also less forward/side-ratio compared to the 4 element one. Its length is only 1/4 λ longer.
I made a 4 element Bruce array for 20 metres. It was designed for 14050 kHz and the exact lengths used was 1/4 λ = 5.76 metres. This is the length MMANA suggested for resonance in free space. I used isolated flex and assumed a velocity factor of 0.95 so the actual lengths cut were 5.47 metres. In fact I didn’t cut at all, I only made corners at these points. I used thin (3mm) polyester wire as a catenary and used sellotape to tie everything together. An upper polyester wire held the it in place between the highest points in my garden, and a second lower polyester wire made sure that the verticals were, well, vertical. When erected its lowest points are about 2 metres above ground level. Such a thin all-wire antenna doesn’t lend itself well to photography, so I had no photos of this contraption until one winter morning when large amounts of frost had settled on the wires.
The 4-element Brucearray one calm winter morning. It is fed in the vertical leg closest to the viewer. Height is lower than normal as it weighs quite a lot more with frost on it. Below is our cat, out on morning inspection for mice.
The usual way of feeding this contraption is in the middle of one of the center verticals. However not much happens with the radiation if we feed in one of the outer verticals instead. So this is what I did. I then ran some 20 metres of 300 ohm ladder line to the shack into my balanced antenna tuner. That tuned the assembly ok. But in order to be able to compare the antenna with my other balanced antenna I had to use a coax input on the tuner for fast switching. So, in the end, I made a quick 4:1 balun out of stuff I had in the junk box and connected the unbalanced end of that into the coax input of the tuner. That worked very well: SWR 1.5-1.6 all over the 20 metre band. No tuning required!
Measured SWR of the finished Bruce array. The measurements are made at the unbalanced end of the 4:1-balun mentioned in the text. Some might argue that the antenna is a trifle too large, I agree on that. I’d shorten it by 13.9/14.1=0.985 to get the exact resonance at 14.1 MHz, but this has no discernable effect on its performance. Getting the SWR down 0.1 is not worth the hassle.
Results! We want results! Ok, here they are: I am positively impressed by this antenna. Pointing towards Spain, the only strong stations I get are in the line of Spain e.g. France etc. I hear some others too, mainly french (which are in the same direction from my QTH), some British stations (slightly off the main lobe) and germans (slightly off the other side of the main lobe). Strangely I also got some bulgarians which are not in the lobes at all, I wonder why? Perhaps they were running with high output power? But apart from these the band is mostly quiet. I think the lobes are where they should be according to simulations. On my second day of operation I nailed my first 7P8 (Leshoto, Africa) and a russian base in Antarctica! I’ve got a feeling I heard these before “mayhem” broke out, perhaps due to the low lobe?
I had the ability to switch instantly between this antenna and my main antenna, the vertical dipole. The latter is always 2-5 S units lower. A result which doesn’t make sense as it is such a large difference. I imagine this is an effect not only due to the directionality of the Bruce, but also different take-off angles – or the vertical dipole’s gone blink suddenly… It is however absolutely clear that the Bruce is a far superior antenna to the VHD – in these take-off angles anyway, and if you are willing to accept the directionality. The VHD was however always the quieter antenna, not only due to low signals but also due to less noise and so better signal to noise ratio. In the radiation diagram for the Bruce array one can see that there is some upwards radiation too. I think this is the source of interference. The vertical dipole does not have this bulge in its diagram.
I made this antenna out of junk wire. I had to buy the polyester line though. But that is cheap. So my cost was virtually nil. The resulting antenna is not, however, at all cheap in performance. With a gain of circa 5 dB it replaces a final amplifier of 3 times my output power, i.e. 375 watts. A 375 watt amp is bound to cost more than the parts of the antenna!