Now that there are social distancing rules that have us mostly at home for the near future, I realised that this is an excellent time to do some upgrades to the VK5LA antenna farm.  I have wanted a low angle, dedicated and resonant radiator for 40M for some time, and recent reading had me thinking about trying a Half Square antenna for that band. The full size dimension would fit easily across my my back yard, and I had available the supports required to get it to the correct height.

An added bonus is that I would be utilising Mmana Gal antenna modeling software and my new nanoVNA to observe and adjust the antenna once erected.

Like many other VK hams, I really enjoy watching VK3YE’s You Tube videos, particularly when he heads to the beach and operates portable…it was this video that started me on the journey for this antenna…if Peter’s antenna worked ok on 20M why not try one one of these on 40M at my place?

For those unfamiliar with a Half Square, the easiest way to describe it, is to think of a single staple, like you would find holding a couple of sheets of paper together.

That hasn’t been used yet!

Now orientate it so the pointy points, point at the ground!

So as you can see in the above screen shot from the Mmana Gal software, here is a diagram of the  antenna. It is simply 2 1/4 wave verticals on 40M joined by a 1/2 wavelength wire at the top. This diagram show it being fed at the top corner (the small red circle) this is a 50 ohm direct feed point, great electrically, but it’s not very convenient to feed it here as *ideally* you have to get heavy coax to this point and lead it away at 90 degrees for a 1/4 wavelength…not very practical…Other more obscure feed methods are feasible, and you can find them on the internet or in good antenna books if you look hard enough. Perhaps the best way to feed it is at the bottom of one of the vertical wires, far more practical, but tricky electrically!  Nothings easy is it? Using the software, we can see in theory how the antenna will perform when fed with RF at our chosen frequency, at the place we want to feed it…

So let’s run model of the antenna in the software. I’ve told Mmana Gal to feed it in a top corner…

This image above shows the antenna radiation pattern. The diagram on the left shows the antenna, looking down directly on top of it. You can see that the long halfwave section going from left to right, the feedpoint (red dot) is shown for reference. The diagram on the right shows the results of the modeling and the elevation pattern. We can see here, that at our design frequency of 7.074 Mhz (for FT8), the Gain is 4.31dB over an isotropic radiator, there is a 0dB front to back ratio as it’s radiating equally well in 2 directions, the  impedance at the feedpoint is very close to 50 ohms and that the reactance is very nearly 0 Ohms. I made the lowest point of each the 2 vertical radiators 250 mm from the ground, as I wanted there to be sufficient room without having the feed point in the dirt.

Let look and see what happens if we feed the antenna at the bottom of one of the vertical radiators…this is far more convenient…

We can see here that comparing the 2 diagrams, the only real change at our 7.074 Mhz design frequency is the resistance at the feedpoint is now at 3250 Ohms with LOTS of reactance. Other important parameters like elevation angle, gain and vertical radiation remains virtually the same.

So there we have it, here are the plots of the radiation patterns, first, horizontal polarisation and then Vertical polarisation…this indicates that at least some of our vertically polarised radiated power should head towards the horizon at around 23.3 degrees. There seems to be very little Horizontally polarised radiation, and what little there is, is going straight up to warm clouds. Vertically polarised radiation, at a low angle towards the horizon, is what we want.

OK so now we have to put power into the antenna. How can we do this so that the maximum power is transmitted by the antenna?

What we have here is basically an End Fed antenna, in this case on our design frequency of 7.074Mhz, it’s a full wave (1/4,+1/2,+1/4 =1) End Fed. Steve, AA5TB explains all about the End Fed antenna in this link way better than I can…Another way to describe the antenna is an 80 Metre End Fed Half Wave.

So we need to transform the high impedance, in this case 3520 ohms (at the base of one of our 1/4 wave legs) to 50 ohms so our Transmitter is happy to deliver full power.

We could use a matching Transformer, like the popular 1:49 or 1:64 transformers that many hams are using with end fed antenna these days. See here for the lowdown on the transformer construction. A 3 turn Primary and 24 turn secondary version (roughly 1:64) works well with this antenna on the bands that are multiples of a half wave. In this case, 80 – 10M. This is convenient if you just want to push the tuner button on the rig when you change bands, but you sacrifice efficiency and the radiation pattern has many more sharp lobes and deep nulls, with increasing Horizontal radiation at high angles away from the design frequency. That’s not good for DX.

I chose to feed the Half Square as a single band antenna on 40M with a parallel resonant circuit. This is quite efficient and ensures the maximum power is transferred to the radiating element rather than heating a ferrite core in a broadband matching transformer.

John, M0UKD, has an excellent website on how he made a matching unit for his 15M 1/2 wave vertical antenna. I used the information from that site and the calculations to come up with my own matching unit, shown below. I used an old capacitor and coil section I had in my junk box and finished up with this unit. Not as pretty as his, but it certainly does the job…(hot melt glue for the win!)

Ok, so now we have a design, we have a way of matching it and we have plenty of time on our hands in self isolation to put it up…

I positioned the antenna in the back yard, the height of the two vertical supports are around 11 metres each. It fits nicely in between the two side boundaries of my average size block. Here is one of the vertical radiators,  both supports are lash ups of broken squid pole sections and ally/steel tube sections I had kept/salvaged for projects just like this…never throw anything out!

Here you can clearly see the 90 degree transition to the 1/2 wave phasing wire…

…to the other Vertical radiator at the other end of the antenna.

Here is the feed point, here I’ve temporarily strapped the matching unit and a 1:64 Transformer to the stake for testing until I come up with a more permanent setup. The capacitor in the matchbox can be adjusted so that X=0 on an Aerial Analyser or vector network Analyser.

So how do you adjust something like this? how do you know it’s doing its thing?

Ok, lets go with what we know. We want the antenna to be resonant at 7.074 (in this case for FT8 mode on 40M) so we want a low SWR reading on this frequency and we also want the resistance as close to 50 Ohms as possible with a reactance close to 0 as well, to make our transceiver sing, and develop full power to the antenna.

I was actually quite gobsmacked at what I was seeing on the VNA. This is one of the first readings I took,  I set it to have a centre frequency of 7.075 Mhz with a 1 Mhz span. (this was straight after adjusting the capacitor in the matching box to peak the noise on RX with my FT817 Transceiver). This is what the VNA showed first up. The SWR is 1 to 1.16, the Resistance is 58 ohms with 24.nF of reactance. The frequency is shown as 7.005 Mhz. Moving the marker (the little triangles on the VNA screen) so the reactance is very close to 0, (resonance) showed a frequency of around 7.050 Mhz. So if we’re splitting hairs, the antenna is a fraction too long. The Mmana Gal model prediction seemed pretty spot on. I would only need to make minor adjustments to the lengths of the vertical radiators to shorten the antenna to raise the resonant frequency to 7.074 Mhz

These numbers makes our Transmitter very happy to deliver full power to the antenna, no tuner (internal or external) needed…

OK so the $64.000 question, does it work? Well yes, it certainly seems to!

I have the antenna orientated north/south, so it is radiating pretty much east west.

The antenna went up on the 2nd of May 2020 at around 5:00pm local time here in South Australia, here are the first few contacts on the antenna after calling CQ on 7.075 Mhz. 1st up is VK3FAC, who was receiving me at +10dB

Then followed by the USA,  Robert, K9U0 in Portland Oregon with a -13 report…nice, looks promising…

Then the friggen DOMINICAN REPUBLIC answers my CQ call!!! Migue HI3MPC with a -22 report…very nice!

Next? CUBA! Eduardo C07EPP answers at -17…Then I think I had to go have a lie down!! Here is a snippet from my log…

The contact circled in green, with Slovenia is actually on 20M when I was testing the antenna with a 1:49 transformer to see if my radio would tune it on all bands from 80 to 10M with the rigs internal tuner, via the broadband 1:49 transformer…th 30, 17 and 12 Metre bands won’t match which isn’t surprising as they aren’t a multiple of 80, 40, 20 or 10…

So yes, it certainly seems to get out, and in the direction the modelling said it would.

I’ve worked 158 qso’s 19 DXCC entities so far in the 9 days as of 11/5/20 since the antenna has been up. I certainly do ok in the direction of the USA, Canada, the Carribean and Africa. An added bonus is the antenna is very quiet noise wise on RX, and the lack of horizontally polarised high angle qrm signals from Indonesia is noticeable. Also, I don’t get many JA stations returning my CQ calls as well, when before they were in plague proportions…this further reinforced the radiation pattern is favoring east/west.

So there you have it, I have described my journey with the Half Square antenna. I’m very impressed, I’m working stations I could only hope to before! I intend to try a portable version for 20M fairly soon, either near some salt water or a lake, as soon as these bloody social distancing rules are relaxed.

Happy Hamming, Andy, VK5LA