In the midst of a ham-radio 2-day “bootcamp” license exam course, I wanted to help the group recognize just how easily and cheaply they could make much of their own ham radio station. People today are so used to BUYING everything, that they have lost touch with their ability to MAKE small pieces of equipment for their own use. Licensed ham radio operators are unique in that they are specifically authorized to construct their own radio equipment. I decided to make a simple antenna tuner mostly out of homemade components – literally making the components, not just the circuitry.
While a resonant half-wavelength dipole (468/F(Mhz) = ½ wave dipole in feet) fed with RG-58 50-ohm coaxial cable is probably one of the simplest and most common antennas, many experienced ham radio operators want an antenna that can be tuned to any frequency on any band, and will still present pretty close to a 50-ohm impedance to their precious transmitter’s output stage. A dipole antenna, or even a single random length wire (with “ground” as the other conductor), as long as it is longer than a small fraction of a wavelength, can be made to work by using an appropriate matching system (“antenna tuner”). (When the total length of the antenna gets down to 0.1 wavelength or so, there are likely to be inefficiencies because of the low radiation resistance, but that is complicated). In practice, most people can throw up a 30-100 foot dipole, a dozen or more feet off the ground. The insulators for the ends & center can be made from commercial insulators, or even segments of PVC pipe with holes drilled. Electrical pvc conduit is sunlight-resistant. Instead of using expensive coaxial cable (which can have very high losses (and even arc-over) when used with a non-resonant dipole like this) it is simpler to purchase or even make “ladder line” open (air insulated) parallel wire transmission line. One way to do this is to cut sunlight-resistant electrical conduit PVC pipe into 5-inch segments, drill a hole just inside each end, and pass two regular old insulated 14-gauge household wires (either stranded or solid) one through each of the holes. Spacing the PVC pipe spreaders every 12-18 inches or so should keep the two wires nicely separated and bingo, you have made a “ladder-line” transmission line that will have extremely low loss, no matter what is the SWR on the line. What you care about is not the SWR on the ladder-line, but instead what is presented to the transmitter. (I’ve written about this before, suggesting a simple L network that can also be homemade if need be.  ) If you choose to purchase transmission line, you can use inexpensive 300-ohm “TV ribbon line” for perhaps several dozen watts, or 450-ohm line just as well.
The real education for the group was building the homemade T network. The T network is an extremely versatile matching system that can generally match almost anything your antenna can do, and make it look just like a resistive 50 ohm load to your transmitter. The ARRL has a nice write-up that will explain a lot more about it.  In my case, I found some 2” water pipe PVC, drilled a start hole at one end and an end hole at another, and had one of the energetic teens wrap about 60 turns of insulated 14 gauge house wire, tightly wrapped, from the start hole to the end hole, passed the wires through them so it wouldn’t unravel. (This created 55 microhenries or so of inductance.) I then use a kitchen knife to cut off the surface insulation across many of the wires, and soldered in “taps” every ½” or so – or about every 3-4 turns. An alligator clip on a wire from one end of the inductor allows you to short out a variable amount of this inductor. While a nice roller inductor would give infinite variability, I was trying to show how cheaply you could do things. If you want maximum variability, at one end put taps every single turn for 5 turns or so and use a second alligator clip from a wire connected to that end of the inductor to allow “fine tuning”.
We had one nice high-voltage air tuning capacitor with 1/16” gaps between the plates, probably 200 pF or more, but I decided to simply MAKE the other one. Taking a large manilla envelope, I taped aluminum foil to the outside of one side, and then taped aluminum foil to a throw-away advertising magazine that would slide nicely in and out of the envelope. These two aluminum surfaces, separated by one thickness of the manilla envelope, make a variable capacitor. Alligator clip leads connected to the aluminum foil. I attached an “ear” of duct tape so I could slide the inner magazine in and out without having to touch the foil. It turned out that the capacitance was more than I needed, and I ended up cutting away half the envelope, leaving a surface of about 60 square inches. I measured the maximum capacitance of about 260 pF. The breakdown voltage of paper is about 16kV/mm , so a .003” thick paper has a breakdown voltage > 1000 volts. Both variable capacitors could easily have been made by this same technique and then all three components could have been homemade.
I wired the commercial variable capacitor as my input capacitor, the inductor to ground, and the homemade paper capacitor as the output capacitor in the traditional T-network format. (See accompanying schematic). If you are going to use up to 100 watts of RF, I would suggest using 16 gauge or better wiring (lamp wire works). IMPORTANT: My quick-built example has the components spread way out and very long wires between them: it works on 80 meters, but you should make the wiring much shorter than my quickie-built example, and the components closer together, if you want to have this work well on 7MHz or above.
Such a simple T network should be able to easily drive an end-fed random length dipole, with the ground wire of the T network sent to an earth ground. Commercially built T networks use a balun (“balance to unbalanced”) toroidial transformer to provide two completely symmetrical connections to a ladder line transmission line. I have such a commercially built system (ancient, and procured through Ebay) but I also found that I could skip the balun transformer and simply wire the ladder line to the two outputs of my homemade T network and it works! Use standard coax wiring from your HF transmitter to your SWR meter, and then from the SWR meter on to the input of this T network.
It takes trial and error to find the best adjustment points for each band for your T network. The first time can be a bit frustrating, so write down or mark the approximate positions after you find them for each ham band. The ARRL article  provides a strategy that is likely to more quickly reach an optimum solution – there are actually an infinite number of combinations that work with a reasonable SWR presented to your transmitter, but you prefer to arrive at one that causes the least circulating currents through that homemade inductor, so that you have the lowest heating effects and losses. Typically, this occurs with maximum output capacitance.
 PrepperDoc. Inexpensive and Simple Mono- or Multi-Band HF Ham Radio Antennas.
 Griffith, AS. Getting the Most Out of Your T-Network Antenna Tuner. Accessible at: http://www.arrl.org/files/file/Technology/tis/info/pdf/9501046.pdf
 Breakdown voltage of paper: http://hypertextbook.com/facts/2007/VashtiPrasad.shtml