At our 16 February meeting, James ZL2ET gave us an interesting talk on WSPR (pronounced “whisper”) and its applications.
Doug ZL2AOV summarised it as:
“James ZL2ET discussed the challenges and possibilities involved in low-power, very low-speed propagation testing. WSPR (Weak Signal Propagation Reporter) is a computer-driven way to find out which bands are open, when and between what sites; it correlates this data and provides it on-screen.
“Using a presentation prepared by Steve Nichols G0KYA, James took us through the equipment needed: a computer with a sound-card to generate the slow-speed FSK used, a transceiver with accurate and stable frequency control, a source of accurate time (propagation paths are found by comparing accurate time slots of data), and an internet link to upload the data to a central repository for automatic analysis and dissemination. Viewers can see what’s going on by pointing their web browsers at www.wsprnet.org. The files can be analysed by band, time, and even by transmitter source … allowing insights into amateur radio DX prospects.
“Transmitters don’t want to be high-powered (maximum five watts is fine) and milliwatts is possible. A Raspberry Pi model B can be programmed to provide FSK on a preferred frequency at about 10 milliwatts out; an amp can push that to 100milliwatts and a tiny bandpass filter does the rest. Timing accuracy can be guaranteed using a wi-fi usb stick to synch to external clocks.
“The protocol provides for transmission of callsign, locator square, power output (in dBm) for 100 seconds, then listening for 120 seconds. Transmission is FSK, just six Hertz.
“The designer of it all? Professor Joe Taylor K1JT, well known for other innovations in digital modes. The system’s sensitivity is stated to be between 11dB and 15dB better than the human ear. That’s way better than using PSK31 and searching for the characteristic double-lines in the waterfall display!!!!”
Ralph Sutton ZL2AOH, our oldest member, mentioned during his talk on CW and FISTS at the November 2015 branch meeting that he conducted regular skeds with three hams, the Three Morseketeers! (Mack JE1CLH, David G3KMG and Jonathan W0XO/7J1AWL.).
In 2009, all three of them happened to be in Mack’s shack in Kimitsu, a city just across Tokyo Bay from Yokohama. Jonathan recorded their QSO with Ralph ZL2AOH and posted it on YouTube. It’s a fine example of an enjoyable rag-chew in very good conditions.
A four square antenna has now been built and assembled, assisted by a number of local amateurs (Doug ZL2AOV is photographer, Bernard ZL2BD, James ZL2ET, flatmate Josh, and Malcolm ZL2UDF and myself are in assembly mode).
Full testing wasn’t possible because 20m at the test location in Grenada Village was full of local QRN at S6+ and peaking over S9. It did, however, pull in stations on 15m, even if it was not designed for that band!
The team are busy assembling the portable four square array for 20m.
The array takes about an hour to assemble: four antennas, four feedlines, a control box, coax and an ethernet line back to the rig and remote control, and 32 radials!
Bernard ZL2BD measuring the impedance at the feedpoint with his antenna analyser. Note that we found in this first test that the initial feed arrangement wasn’t satisfactory, creating too much coupling to ground. Doug reworked the “feet” giving them nice painted toenails (in British racing green).
I started this project with the intention of engendering interest in phased vertical arrays. I imagine that the club will get to play with this antenna (and hopefully others like it) in our annual efforts in Oceania DX Contest and CQ WW contests.
20m was picked as a starting band because element lengths are manageable (approx. 5m). If there is a band where your signal needs to stand out in a contest then it would be 20m! A four square has around 5.5dB of gain over a monopole, similar to a horizontal beam antenna. However, all assembly is done on the ground, there is no rotator and it is very quick to switch directions.
What a tiny box! James ZL2ET and Mike ZL1AXG can be seen pointing at the controller.
You can spot the 1/4 wave current forcing 75 ohm feed-lines from the controller box to each monopole.
The control box and remote unit
The controller is not frequency dependent – containing only switching equipment (3 DPDT relays) and a broadband transformer.
To get the 90 degree offset for the centre radiators you use two pieces of coax, in a Christman matching arrangement. The lengths are calculated using a computer program downloadable from W7EL’s website at EZNEC.com. You simply input the impedances of the component monopole antennas measured using an antenna analyser, the configuration of the array, frequency and velocity factor of the coax. Out pops the lengths of coax to be cut in metres. The program solves two complex number calculations behind the scenes. Simple!
The two pieces of 75 ohm coax (one is 2.395m long, the other is 7.799m long) are joined together with a Tee connector creating a common voltage feedpoint where they are fed with 50 ohm coax (grey RG8U) from the rig. These feeders plug into the controller, giving 0 (rear element) and 90 degrees (centre elements) of phase difference. The 180 degree shift (front element) is achieved using a broadband transformer constructed on a FET240 type 43 ferrite toroid sourced from SICOM Ltd.
I am working on constructing a hybrid coupler as well, in order to compare the different feed arrangements. Hybrid couplers are frequency specific as they contain both inductance and capacitance. They are popular in commercial units as they are not as sensitive to the antenna feedpoint impedance. Some say they are more lossy than Christman feed. We shall see!
A special thanks to Malcolm ZL2UDF who provided the aluminium tubing for the four antennas (ex Quartz Hill 20m beam), Bernard for assistance in calculating the 1/4 wavelength feeder lengths and velocity factor of the 75 ohm coax and supplying the antenna analyser, and Doug ZL2AOV who reworked the “feet” and connection mechanisms for the antennas, and resolved the resonance issue so that they tuned up properly.
At the next club meeting on Wednesday 19 November at 7.30pm, you will hear me talk a bit more about phased vertical arrays in general. I will discuss why these arrays make sense for contesting, basic configurations (2, 3, 4 or more elements), switching and feed arrangements. I will also cover off practical design ideas for both this portable four square and my as yet unfinished two element multi-band array under construction.
