The loop is very simple and is fed with a toroid transformer coupler.
We purchased two diameters of Pexal, "Gas-TITE", polyethylene cross linked aluminium, 20mm and 16mm. The idea is that by shaving down the thinner tube, used for most of the loop, the thicker pipe can be slipped over both ends and works as the capacitor.
Pexal is easy to cut and Ross recommends using a box cutter as shown here. (No blood was spilled in this video).
Here's how the pipe looks. It is insulated on the inner and outer surfaces and has aluminium in it. This works up to hundreds of watts if required as the breakdown voltage for this material is 19KV.
To get the smaller tube to slip in to the larger tube we had to do a considerable amount of shaving with the knife. We also tried heating the outer tube but it didn't seem to help much.
There are some excellent calculators for computing the size of magnetic loops and figuring out the capacitance of the run of tube within tube. Here's the windows app called "Magnet Loop".
The cylindrical capacitor calculator is here.
The loop for 20m is now up and receiving WSPR.
We ran a WSPR receiver with a Raspberry Pi decoding over night and here are the received spots.
We also made a larger loop for 30m but it's a little unwieldy.
Some tuning can be achieved by sliding the smaller tube in and out of the larger linking part. Note that because this section is two capacitors in series the value is halved so you need to double the amount of capacitance needed to resonate.
Another way to tune is to use aluminium flashing tape over the ends of the thicker tube which can be trimmed to adjust.
We used between 2 and 4 turns on the toroid and this affects the SWR you can get. You'll need to use a minimum of 850 initial permeability. We used a toroid from Jaycar marked as L15 but we're skeptical.
Thanks Ross for sharing this interesting technique. Ross is well known as the "loop meister" on wsprnet.org, here's a few of his past loops.