Build instructions for the Figure-of-Eight magnetic loop.

Following a number of requests, here are some simplified instructions for the 'Meight', or figure-of-eight magloop.

This design will work from 13MHz and probably up to about 18MHz, maybe higher, I haven't yet tried it out beyond 14MHz, where the match is 1.1:1.  Efficiency drops as you go down in frequency.  At 14MHz, efficiency will be around 95%.  Remember that there is still active debate about how accurate 'traditional' magloop efficiency formulas actually are.

(1) Buy a roll of 10mm copper microbore tube, the kind often used to plumb central heating radiators.  Alternatively, you can make a square loop array from 15mm copper tubing, except the build is slightly different.

10mm microbore, as usually found in hardware stores.

(2) Assuming we're using 10mm tube, this is quite easily bendable by hand, more so if it's warmed up.  It usually comes in a plastic bag as a coil.  You should be able to open up a decent circle of tube such that the diameter is about 75cm.

(3) Use pipe cutters or a hacksaw to cut two full circles from the roll, each ~75cm in diameter.

(4) You can either solder the tube together using straight couplers, or you can do what I do, for which there are photos.  This is a little tricky, but it does tend to yield quite a stiff loop array that is useful to maintain good matching.

(5) I took two lengths of 15mm copper tube and hammered it between two pieces of wood to make a flat bar.  Clean these very well with wire wool, coarse sandpaper or a rotary tool with sander attachment.  Cleaning is essential, otherwise you won't solder successfully.

(5a) Number 5's method works OK, but effectively makes a plate capacitor.  Even with an insulator (such as wood) between them, the intervening material provides a wettable surface that will allow a water bridge to form, and change the matching of the loop in rain.  You need to either cut the excess copper once it's all be soldered up and support the loop so that there is only a good air gap between the connections, or, preferably, just use soldered straight connectors to make the figure eight, rather than my initial method.

(6) Also clean all four ends of the copper tube rings to at least 2" from the ends.  A good guide to general soldering principles can be found in this video.

(7) Place one end of the bar in a vice or somehow otherwise make sure it's secure and can't move.  Heat the end of the bar with a butane torch strongly, and do not remove the heat (because it will allow contaminating oxides to develop) until you have applied a good amount of solder.  Use a fairly thick gauge, flux-cored solder.  When you are at the right temperature, the solder will run 'thin'.  You can remove the heat now.

(8) Apply heat to the end of the rings and apply a good amount of solder to each end.

(9) Bring the pre-soldered end of one ring to the soldered end of one of the bars and apply heat to both the bar and ring together.  After a short while, you will see the solder run thin again, and you can remove the heat but ensuring the ring and bar remain firmly in contact. You may need or want to add a little extra solder between the elements to make a stronger joint.  Having someone blow on the joint helps shorten the period you have to keep everything still.  REMEMBER THAT COPPER STAYS HOT FOR A LONG TIME AFTER REMOVING THE HEAT!

(10) You should end up with something like this:

The bars provide a way to easily make a figure of eight, and also provide a strap attachment point to wood, etc. BUT SEE NOTE 5(a)

On the left, you see one loop, with the ends offset by a couple inches.  On the right, you have loop two, also offset.  Soldered to the bars like this, they yield a figure of eight.  AGAIN, SEE NOTE 5(a)

This should give a clearer picture of how it goes together. The wine is 2018 Damson.  SEE NOTE 5(a)

(11) As you can see from the photos, you also need to solder either some thick gauge wire (something like Flexweave is OK), or copper/brass braid or shim metal to make the connections to the capacitor.  I solder spade connectors to the ends of these.  Always clean everything with wire wool or an abrasive tool to ensure good quality soldered connections.

(12) I made my own loop mounting stands out of old timber.  You can use a fence post with a piece of 2x1" timber screwed into it for added height, or some other arrangement that provides enough support and resistance to wind.  You can also add guy rope, provided it is not connected to the loop itself (which will lead to stray capacitances and detuning in damp weather).

(13) All you need to do now is add a suitable capacitor.  You need a wide, air-spaced variable of around 10 or 15pF to 275 or 300pF, preferably one that is of the butterfly type that has infinite rotation (i.e. no stops).  If you have one with stops, that's fine, but not so clever if you want to motorise your capacitor later, because you need to incorporate motor stopping switches.  You can also use vaccum capacitors, but these are heavy and fairly expensive.   A suitable capacitor is typically about 3-5 inches in length.  Be careful not to inadvertently buy a heavy, 12" long cap!

(14) It is fairly easy to just tie-wrap (cable tie) an air variable capacitor to a timber or plastic backing of some sort.  I recommend you make it weather tight if possible/necessary.  Sticking it in a plastic drain pipe with a cover is often a good housing.  A vaccum capacitor is more challenging to mount securely.

(15) Once you have soldered connections to the capacitor and secured it on your magloop stand of some sort, you then need to feed the loop.  Most people follow the herd and use a primary, or 'Faraday' loop.  I never use these, because they are often mechanically unstable, and can make it difficult to reach the best match without a lot of shape-bending and moving things around.  It's better to use a 'pseudo-gamma' match, which goes like this:

(a) Take a 4:1 balun.  Attach a short wire from one output to somewhere along the bottom of your lower loop.  Then take an insulated wire that's about the length of one side of one of your loops, and attach this to the other output of your balun.  Wrap the wire around one side of the lower loop, and use a crocodile clip to attach it roughly 3/4 the way up.  Once you find the correct spots to attach the wires, you can (and should!) remove the clips and solder the wires on.

(a1) UPDATE: The above crossed-through method works perfectly fine, but I have found it susceptible to minor but irritating SWR changes during wet weather, as water gathers between the wire and the loop, introducing capacitance changes.  If you look at the third image below, this provides a perfect alternative matching point (~1.09:1) and, at the time of writing, because the entire wire is well outside the loop and not in contact with it as previously, such that water can't accumulate, the SWR drift seems to be eradicated.

General view of the loop array, initial test.

Following day, with soldered connections!

Final iteration - modified feed.

(b) You can now just tune the loop by hand.  This is not easy, due to the very high Q of the antenna, but you will get a feel for it and so it can be done.  You can achieve reasonable tuning accuracy by ear, as the noise will get noticeably louder as you reach resonance.  Then use your rig on low power and a CW or digital tuning signal to test the SWR.  For fixed-frequency use with digital modes like WSPR and FT8, it's easier to use an antenna analyser for a perfect match.  If you can get the 'dip' in the matching to reach your desired frequency of operation, but the SWR is too high, you need to move the feeding wires around a little.  Try moving the base wire from the balun up the side opposite the long wire feed first.  If you have an antenna analyser, you will see the 'dip' get deeper, reaching a more perfect match.

(16) For a simple, remote tuning system, have a read of this earlier blog post and look at the photographs for how the mechanical assembly fits together. 


Here are some early results on 14MHz WSPR, 200mW.