This is a reasonably compact, linearly polarised and well behaved antenna, with a gain , when well made, of around 1dBD    ( 1dB more than a dipole). It is smaller than the Big wheel and the Skew Planar wheel, and I have used it very successfully on 868MHz on our SurVoyeur UAV.

It is derived from a 1/2wave ring, located some distance above a large conducting ground plane. The ring is at a point on its circumference shorted to the ground plane by means of a conducting copper strap. The coaxial cable feed line has the center conductor attached to the ring a short distance away from the shorting strap, and its braid shield connected to a corresponding point on the ground plane. directly below the center conductor connection. The Antenna radiates from the gap between the ring and the ground plane - a Slot radiator antenna. Since the ground plane is inconvenient, we can 'roll' it up into a cylinder, the same diameter as the ring, located coaxially directly below the ring. The antenna performance remains intact. However, the long cylinder is just as inconvenient, so we shorten it till its length is the same as the thickness of the wire forming the ring, and the end result is the picture above. Two concentric rings, spaced apart by plastic insulators, with a 4mm wide copper shorting strap, and the feed coax in place. When the dimensions are proportioned correctly, the antenna gives 1dB gain over a dipole, and when laid horizontally, has a vertical radiation pattern identical to the dipole, with a slight null in the direction of the shorting strap.It is linearly polarized.

For 433MHz, the antenna dimensions shown will give an SWR of 1.1:1 at 433MHz, and 1.25:1 at 429MHz and 437MHz


Seen from the side :

The rings are 100mm diameter, 2mm clean copper wire, no plastic insulation.

The plastic insulators can be Servo push rod tubes, or any plastic tube or solid rod, about 6 to 10mm diameter, and about 20 to 25mm long, with 2mm holes drilled in the sides, 15.5mm apart, to take the wire rings. 

To make the rings, take the copper wire, place one end tip in a vice, and with a pair of pliers clamping the other end of a 1meter long piece, pull hard and gently to stretch-straighten the wire. Find a former ( coffee tin, etc) exactly 100mm diameter, and wrap the wire around the former overlapping the wires. With sharp sidecutters cut through both overlapping wires, to make a butt joint in the ring.

Remove the wire from the former, slip over the plastic spacers ( at least 4 prefer 6) and then, using  a 10mm x 5mm piece of thin copper foil, wrap it around the but joint and solder the wire ends together this way. 

Make two such rings, feeding the second ring gently through the second set of holes in each spacer before butt jointing the ends.

Then position the two butt joints over each other and connect a shorting strap of copper foil, 18mm x 4mm wide, from the top ring to the bottom ring. solder the bottom connection, and then the top, ensuring the strap is straight and taut.

Now connect the inner conductor of the 50ohm coax to the upper ring, 9.5mm away from the center of the shorting strap position. then connect the braid of the coax to the lower ring, at the same place.

The shorting strap end of the antenna is low impedance and low voltage. the direct opposite end of the ring is a very high impedance, high voltage point, sensitive to the proximity of auy other objects, servos, wires, etc, as are the end tips of a dipole, for example. Objects close to that region will affect the tuning...Do not place any support plastic pillars  past the two pillars farthest from the coax feedpoint, as depicted in the last image at the bottom of this blog.

                                             Here you can see the shorting strap location.


No spacers past the two on the right of this image...


It is convenient to use tubes for the spacers, as you can fill them with hot melt glue to fix the spacers and rings in place, forming a rigid structure.

The antenna works equally well right side up, or upside down....

Tuning the antenna :

This is a sensitive process, and you will need and RF source at the right frequency, and a decent SWR meter.

Connect the rf source to the SWR meter, and the meter to the antenna via the intended cables.

There re two adjustment to be made - feedpoint impedance, and frequency of resonance.

Feedpoint impedance is adjusted by moving the coax connections to the ring ( the inner conductor and the braid) either closer to or further away from the shorting strap. Why does this work? Well, the shorting strap is a short, so impedance at the strap is 0 ohms.  The opposite side of the rings circumference is a high impedance point, many thousands of ohms, from an RF standpoint. So, somewhere in-between 0 and 'thousands', will lie 50ohms - that is the point you seek, and it is very close to the short.

Frequency adjustment is done by squeezing the end of the two rings opposite to the shorting strap together to lower frequency, or apart to raise it. In other words, the spacing of 15.5mm is reduced or increased appropriately.  If you find you need to increase this spacing by more than 2mm or so at that quadrangle, then the rings are to long in circumference. Cut 2mm out at that point, and join with the wrapped foil trick again, SOLDERING WELL...and tune again.

When you achieve lowest SWR by squeezing, then adjust the coax spacing to the short, try closer or further, for lowest SWR. Work back and forth between these two adjustments, till no more improvement occurs, and you should strive for an SWR better than  1.4:1. 1.2:1 is easily achieved.

This antenna works very well. I have achieved nearly 30km with 300milliwatts on 868MHz, with one of these on the aircraft at 100meters AGL, and a sleeve dipole on the ground at 2meters AGL.

Here are some pics of the 868MHz implementation , the rings made from FR4 PCB material.


This is the datalink and video tx module used on our SurVoyeur aircraft. The ring of copper is the 1/2wave ring radiator ( the other ring lies below it)

The Skew planar wheel is for 2.4GHz Video TX.

To the left of these antenna are the datalink modem and the video TX module.

                                       Here are views of the shorting strap and coax connection

                                                                    to the rings



E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • Morning Will,

    Forgive the omission - the polarisation is linear, and vertically polarised - it is the same as a dipole vertically orientated when the ring lies horizontally. Most linear antenna have a dual in a slot version.

     Imagine a large sheet of conducting material, say 1mm thick, 300mm X 600mm, standing upright with the 600mm length left to right. Cut a horizontal slot out of the material, in its center. Make the slot approx 10mm tall by 340mm wide, and place the coax feed line across the center of the slot, coax center to one side of the slot, braid the the other. You now have a slot dipole on 433MHz, the slot lies horizontally, BUT the radiation is linear and vertically polarized. the SAME as the vertically orientated wire dipole. 

    I did not know that 868MHz is not permitted in the UK - I thought it was permitted in all of Europe. A pity that, it just makes the antenna so much more manageable. Although a 1/2 wave ring at 433MHz is still considerably smaller than a turnstile at 868MHz, ignoring the other differences between the two types ( polarisation,etc)

    I use Ansoft HFSS, but there are some really good , inexpensive software that can take you a long way in understanding antennae better, and have fun along the way - such as   EZNEC and NEC-Win-basic


  • "when laid horizontally, has a vertical radiation pattern identical to the dipole, with a slight null in the direction of the shorting strap.It is linearly polarized."

    you don't mention the actual polarisation tho, as the conductor is perpendicular to the conductor in a dipole then the polarisation is presumably perpendicular to a normal vertical dipole. So do you then have a fancy (not a "standard" vertical) antenna on the ground or do you have to carefully align the ground antenna as you say that you are using a dipole on the ground? or am I confused/mistaken? antennas are not my strong suit.

    your 0.86GHz antenna looks great in its own right but super cool with the FPV antenna combo, its a shame the 0.433GHz model is so big as license laws restrict uk users to the longer wave length...

    also do you build and optimise from theory and analysis or are you using microstrips or HFSS or the like?

  • Moderator

    Ah I know those problems, we had issues with a rhino yesterday thankfully resolved. Had some very interesting collar data from a jackal as well that spent days looking or making a hole in a game reserve fence and has now legged it. Flying tomorrow AM to find it. I'm not sure how I would create a lion chasing solution, generally I think I would be going the other way.

  • Ok, I see now ..

    Sorry, this blog was really a result of discussion with Big Plane Builder Vladimir, and aimed at the 433MHz single frequency version - so Iost the plot...I failed to realise your reference was toward the 868MHz/2.4GHz version.  My bad..

    My problem is that I am supposed to be 'retired' but cannot stop doing this stuff - to young to retire or something..

    And so I am fearful of getting into the product supply rut again!  Maybe 3DR would like to take my designs and manufacture them??? Imaging all those dollars rolling in!  

    Right now I am working on how to get a small UAV to go and irritate and chase a Lion that insists on habituating a local village for its Donkeys - it is soon going to get shot! It has a GPS collar so we know where it is....The problems of Africa!


    ( sorry for off-topic..)

  • Moderator

    You are operating two frequencies in less space than most people run one antenna. That would have to be a very useful product if you felt like producing it for those (like me) less able.

  • Huh??...


  • Moderator

    Wow stacked multiband goodness, that's very impressive. Can we form an orderly queue for the end product?

This reply was deleted.