Tag Archives: how to build a magnetic loop antenna

Guest Post: SM0VPO’s 3D Printed 10KV Tuning Capacitor

The completed antenna.

Many thanks to SWLing Post contributor,Harry Lythall (SM0VPO), who shares the following update to his excellent homebrew 20 meter magnetic loop antenna post:

3D Printed 10KV Tuning Capacitor

by Harry Lythall (SM0VPO)


You may have seen my 20m (14MHz) loop, or frame, antenna, and the ease of construction with just a bit of wire and a bit of plastic tube. The tuning arrangement is a little primitive, using just a “gimmick capacitor”, comprising two bits of wire twisted together.

The original “Gimmick” capacitor that burns.

This arrangement works very well for QRP, where the average RF power is about 5 Watts or less. If you exceed this power level, then the twisted-wire capacitor tends to warm up and the tuning drifts a little. But if you use more than about 10 Watts of continuous RF power, then things start to burn. Cheap insulated wire also smokes. This is because the impedance at the ends of the coil is so high that you can get many 100s of volts and the insulation, normally intended for house wiring, breaks down.

In this page I will show you how to build a super-cheap tuning capacitor that will tolerate up to 10,000 Volts of RF and allow you to use up to about 100 Watts of RF into my 14MHz (20m) antenna. The capacitor is also tunable so that you can adjust it by hand (when the RF is removed, of course :-). The tuning range is about 8pf to well over 30pf when really compressed. The normal range for the antenna is about 12pf to 15pf.


My prototype does not look very pretty, and it is not supported on anything other than the connection wires from the antenna. Very few components are used:

  • two plastic foam pan scrubbers or one thick plastic bath sponge
  • two pieces of metal 4cm x 6cm. Copper-clad board works fine
  • one plastic nut and bolt – see text
  • one heavy-duty 3-pole block connector with centre-pole removed

Components for the 10KV tuning capacitor

The plastic nut and bolt need to be about 6cm long and totally non-conductive. If you want to make my bolt, then do NOT use black plastic because some black plastics use carbon as a colouring agent. You can use a nylon bolt, as used to secure IKEA toilet seats, but you will also need a washer to spread the stress, otherwise the copper-clad board tends to bend under the stress with time. I chose copper-clad board because it is easy to solder – no need for drilling or connection bolts. I made my plastic compression tuning bolt using my 3D printer. I have included the project files for you to download.

Ikea toilet-seat bolt

Note that the connector for the wires needs to be well spaced between the metal inserts, in order to tolerate up to 10KV, so I used a 3-pole connector and took out the centre pole. At 100-Watts continuous there is a very slight warming after a few minutes, but no sign of smoke, sparks or corona. 🙂

The connector with the centre-pole removed.

I tried a few different types of plastic dielectric and all worked well, providing they are 100% dry. The best ones those I stole from the kitchen cupboard (when Maj-Lis was not watching). I think it costs about $1.50 for a packet of 10 pieces. Perhaps I should have used a nicer colour? A pretty pink? Heart-shaped? No! maybe that would be going a bit too to far ;-). Here is my finished capacitor using my 3D printed tuning screw.

The assembled 10KV capacitor.

3D Files

Once more, this project is ridiculously easy to make with a 3D printer. The hardest part was to get the pitch of the threads right, then clone/connect the pieces to get a longer thread. They were printed with the bolt vertical, so the slope under each thread is a steeper angle than the upper slope. This makes the printout a lot easier to print. If the angle is too steep, then it may extrude PLA into mid-air. My nut and bolt heads are about 3cm Diameter, and the 10mm thread for the nut was cut using boolean subtraction. I then enlarged the nut by about 3% so it still fits but there is a little slack so that it does not bind. I may have overdone it a little, but not much.

Project rendered in 3D Studio Max

Here are the files:
3D studio MAX file – 3d-cap-01.max
ASCII STL file – 3d-cap-01.stl
ASCII OBJ file – 3d-cap-01.obj
the GCODE file for my Wanhao (Prusa) Duplicator i3 – 3d-cap-01.gcode

Just right-click on the file and select “save as”. Some web browsers try to open ASCII files instead of saving them. I usually export STL files, but on this occasion, I tried comparing OBJ files since CURA slicer will accept both. Although I use a WANHAO replica of the PRUSA Duplicator i3 printer, the GCODE files are rather generic and will probably work on most printers.

My printer settings in this GCODE are:

  • Nozzle temperature = 200°C
  • Bed temperature = 60°C
  • Support structure = brim
  • Layer height = 0.1mm
  • Print speed = 60mm/s
  • Fill density = 40%
  • Shell thickness = 1.2mm

The printer temperature is 200°C, which is 5°C hotter than recommended for PLA filament. I find that I get a better print at 200°C as it sticks to the bed a lot easier. When printing this screw thread, it may be advisable to start at 200°C and then turn down the nozzle temperature to 195°C after a few layers have been printed.

I hope that you find this project interesting. It is a bit small, but it gets me active on 14MHz from my car with this portable antenna. I have a new car and I don’t want to start throwing long aluminium tubes in it that scratch the interior to bits.

Don’t forget to visit my messageboard if you have any questions about this or any other project. I always look forward to receiving feedback, positive or negative ?

Very best regards from Harry Lythall
SM0VPO (QRA = JO89WO), Märsta, Sweden.
EA/SM0VPO (QRA = IM86BS), Nerja, Spain.

Thank you so much for sharing this, Harry! I love both the frugality and ingenuity in this unique capacitor design!

Post Readers: be sure to check out Harry’s website which is loaded with radio projects of all stripes. You’ll easily spend a few hours digging through his tutorials and downloads. Harry also maintains an alternate mirror server located here.

How to build SM0VPO’s 20 meter magnetic loop antenna

The completed 20 meter loop antenna.

A few days ago, SWLing Post contributor, Robert Gulley (AK3Q), pointed me to an excellent website by Harry Lythall (SM0VPO) which is chock-full of various homebrew radio projects. In particular, we both were impressed with Harry’s 20 Meter Loop Antenna–it’s such a simple project and requires no special order components. In fact, all of the components (save, perhaps, the antenna connector) can be purchased at a DIY store.

I reached out to Harry and he has kindly allowed me to republish this project as a guest post:

20m Loop Antenna

by Harry Lythall – SM0VPO


I recently saw that my 80m (3.5MHz) loop (or frame) antenna has been really popular, and that there are loads of other radio amateurs who have taken my design and “ran with it” to produce variations that all have some great improvement. There have been many in-depth tests and simulations, all with exceedingly good results and reports. This is exactly what I am aiming for with my homepages – free information for all and my designs being improved upon. That way we all win 🙂

One small point all variations have in common is the need for an expensive tuning capacitor and a very restricted RF power level. Of course, you can throw money at the problem, but for me this hurts. I got to thinking that there must be a way of adjusting the design a little and finding another technique to tune the antenna, and to make the best use of the little radio transceiver I have in Sweden, given the limited space.

My limited space apartment.

As you can see there is not much opportunity for grand antennas. And to add to this, the equipment I have in Sweden is also limited to a single 5-Watt unit.

My limited equipment – only 5-Watts.

The Design Thoughts

Today I have no area of land to use for antennas. I have a glassed-in balcony on the 4th floor of an apartment block. I really like the 20m (14MHz) band so I will concentrate on that. I am not really interested in the CW end of the band, except perhaps 14.070MHz for the digimodes. So my requirements are:

  • As efficient as possible (useable)
  • Small size, also portable so I can use it for field use
  • No expensive components, everything available locally
  • No TVI, QRM or interference to stereos or computer sound
  • Total price less than $2

The antenna I have created is based on my original 3.5MHz loop (or frame) antenna. This time I built it out of scrap components. I cadged (tiggade) some plastic conduit tubes from an electrical contractor at work. The same guy also gave me the remnants of a roll of 2.5mm C.S.A. multi-strand mains cable. That was all I needed.In my junkbox I found no tuning capacitors, but WAIT!! Why do I need to tune the antenna? Once it is tuned I should not need to tune it again, just set the centre-frequency to 14.175MHz. If I can get the Q-factor to around 100 then my useable 3dB bandwidth should be more than 150kHz. That will give me 14.10MHz to 14.25MHz.

Ok, I need a 1-off, preset tuning capacitor. Why not use a Gimmik Capacitor? Just twist two bits of wire together and cut it short to get the resonant frequency I want. So I need to get the coil wound so that there is sufficient cable length and self capacitance to give a resonance of about 14.5MHz without any extra capacitance what-so-ever. That means I need just a few pf. That sounds like a good plan.


The 15mm Diameter plastic tube I “aquired” were 80cm long. After much trial and error I found that exactly 3 turns, with 2.5cm spacing, gives about 14.9MHz self resonance. The wire support holes are exactly 4cm spaced, beginning 1cm from the end of each tube. The two tubes are fixed into a and X using zip-straps (tie-wraps, buntband). The feed loop is 1/2 turn.

Note the size and position of the feed loop. Also the Gimmik capacitor.

One problem I had with the original loop antenna was that of RF coming back down the cable braid. Using on old FT-101ZD it was possible to feel the RF on the microphone with your lips. The cure for this is to use a balanced feed and at least 5m of RF cable.

I robbed the ferrite ring for the balun from an old ATX computer PSU and made a triflar wound torroidal transformer. That is to say, twist together three lengths of 1mm x 7-strand insulated hookup wire together. Use this to make a 7-turn coil and connect the three coils in series, with four connections. Feed connections (numbered in the picture below) 1 and 3 are connected to to the antenna feed loop. Connect the coaxial cable braid to connection 2, and the coax centre to connection 4. My balun is self-supported on the connection leads.

The 1:1 Balun I used.

The coaxial feed cable was found to affect the resonance slightly, so I fed that through an extra bit of tube to make it stay in one place. It works fine.

Feeder cable secured in the support tube.


Testing is very easy. I used my GDO-2 to check the middle-turn of the loop for a dip. Twist the two tails together to form the Gimmik capacitor and adjust the length of the twist until the centre-frequency is 14.175MHz. With the GDO you can get it within about 100kHz to 200kHz, but then you can check the VSWR using your HF radio. You can also sweep the band for maximum noise and get a very close approximation.

The Gimmick capacitor.

The centre frequency of my 20m Loop antenna is 14.175MHz, and the VSWR is better than 1.05:1 (I can hardly see any movement on my meter). The Q-factor is somewhere approaching 100. The useable bandwidth is just a little narrower than I would have wished, but the antenna certainly works well and meets all the other criteria. But the slightly less useable bandwidth criterion is at the expense of better performance, and it still allows me to use 14.070MHz, although it is a little quieter down there.

The completed antenna.


No-matter how you play with the figures, the best indoor antenna cannot replace a full-size dipole antenna. But the indoor antenna can give some extra features, such as just reaching out your arm and trimming a little, which you cannot do with a long-wire antenna up a tree out in the garden, especially when it is raining.

This antenna gets me on the air on 14MHz, and it has a useable frequency range. The VSWR is almost perfect at the centre-frequency, and this time I don’t burn my lips on the microphone (not that I am likely to do so with just 5-Watts of power). The design uses no expensive components, in fact the only item I bought was the block-connector for the balun. That cost me US$1.50 for a pair of 12-contact screw-terminals. The construction is ridiculously simple and easy to build.

On the air I can hear traffic on 14.070 digimodes, and from 14.130 to 14.220MHz I have a near-perfect VSWR aqnd good clear reception of SSB. I can also rotate the antenna to cut out rubbish, and most of all, using the Gimmik capacitor I don’t need to re-tune it: it seems temperatore-stable. The weight is less than 500g and when I poke it out of the balcony window the reception improves, the VSWR does not change, and I can make myself heard among the big boys.

I hope that you have some fun building and using this antenna. If you have any ideas for further improving it then please use my forum.

Don’t forget to visit my messageboard if you have any questions about this or any other project. I always look forward to receiving feedback, positive or negative.

Very best regards from Harry Lythall
SM0VPO (QRA = JO89WO), Märsta, Sweden.
EA/SM0VPO (QRA = IM86BS), Nerja, Spain.

Many thanks, Harry, for sharing this excellent project on the SWLing Post!

Post Readers: be sure to check out Harry’s website which is loaded with radio projects of all stripes. You’ll easily spend a few hours digging through his tutorials and downloads. Harry also maintains an alternate mirror server located here.

Side note: I’m impressed with the fact that the main SM0VPO website is actually hosted on a bedside Raspberry Pi computer (running the Linux-based Lighttp server).  Very cool!

Check out other homebrew mag loop antenna projects on the SWLing Post by clicking here.