Updated 20110410

Survey of amateur ESA requirements

The subject of antennas is arcane and boring to the crowd. But to anyone that wants to talk (or type) to the world from a small plot in crowded England, they are an interesting headache. Transmitting on a low frequency demands a BIG antenna. Grappling with the antenna size problem, we find the answer is NOT on the Internet, I have to find it for myself...

The antenna system is electrically small if its maximum length is less than λ/2π. So a 10m band antenna of under 1.6m long qualifies, for instance. This is the (Harold) Wheeler definition of a small antenna, which was derived mathematically 50 years ago. The most significant limit on small antennas is the Chu limit. This defines the minimum Q factor (or maximum bandwidth) that a certain size of antenna can ever achieve, with no losses:

Where a (sometimes and more logically called ‘r’) is the radius of the smallest sphere enclosing the entire antenna system, and the free space wave number k = 2π/λ. This rule is derived from near-field charges around an antenna. Unfortunately losses cannot be avoided in reality, and though they reduce the ‘Q’ they’re otherwise very unwelcome. The Q measurement is where the reflection coefficient rises to double that at resonance. For example, assuming a 1.0:1 at resonance, and frequency of 10MHz, if the SWR is 3:1 at 9MHz and 11MHz, Q = (2x10) / (11-9) = 10.

Unfortunately the Chu limit refers to an idealised mathematical situation. Practical problems with ESAs just begin with the Chu Limit, as real antennas never get close to it. They are always 1.5 times above it and often much worse.

A small antenna has small radiation resistance and big capacitive reactance. A large mismatch to practical transmitter designs. Attempts to match the transmitter using LC networks (tuners) give even narrower bandwidth, because we’re introducing additional resonant elements. Passive components, no matter how carefully made are intrinsically lossy. Coaxial feeder has high losses when mismatched, and though open wire line is better it has practical disadvantages. Our problems are getting worse!

Turning to professional research, there are designs to approach the Chu limit and match a 50 ohm system. The lowest at Qchu = 1.5 I can find is the 4-arm folded spherical helix designed by Steven R. Best. That has a 50Ω match, at ka = 0.26. So for the 80m band at 3.7MHz, we have an antenna 3.3m in radius, 6.6m diameter. It would be a complex shape with high windage and expensive to manufacture. Even then, scaling up the results I saw for a 300MHz design give only about 90kHz of bandwidth at 3.7MHz. It is unlikely that design can be tuned over a useful range easily. So we have a monster bastard of a problem now!

Step back a bit from theory, to consider the parameters of a good HF amateur antenna:

1. 1.Small size, taken in all three dimensions
   

2. 2.Good mechanical strength for exterior location
   

3. 3.Good matching to a practical feeder cable, typically 50 ohm coaxial
   

4. 4.High efficiency >75% (low losses)
   

5. 5.Wide bandwidth (low Q)
   

6. Good directional response

Of these, point 2 is achieved more easily if the structure is physically small. Point 3 requires the system be designed for the appropriate feeder impedance. But meeting points 1, 4 and 5 together means breaking the Chu limit. If we relax the requirement and exclude one of those points, the design becomes feasible. Another loophole in the Chu limit is antennas with multiple resonances.

The 6th point is left to one side, as a directional (beam) antenna adds a whole other set of problems. I consider an omni-directional antenna better than no antenna at all !!

So we can have a small efficient antenna, but bandwidth is narrow. Provided the ESA can be retuned automatically, without introducing large losses, the bandwidth limitation is mitigated. A simple dipole has 2:1 SWR bandwidth of 9%. The amateur bands at 7MHz, 10.1MHz, 18MHz and 24MHz are much narrower than this. A dipole specially for those bands is wasting about 80% of the antenna’s capability! So accepting a narrower bandwidth is fine, provided the system can be retuned easily. It allows the size and efficiency trade-off to be met.

What we want is something that comes close to the Chu limit, and can be retuned easily or has multiple resonances. Let’s have a look...

Inductive or Capacitive Loading

These methods are commonly and traditionally used by amateurs. Capacitive loaded antennas are favoured by my “antenna hero” Les Moxon (G6XN SK), and are generally preferred over inductive loading because they have much lower losses. The exception is wide helix or spherical wound types. But reducing the size to meet the ka < 1 limit leads either to losses or the impractical spherical shape to try and claw back some bandwidth.

Designs like the Pro-Antenna give fair efficiency by using both capacitive and inductive loading. Having thin end-elements for capacity loading does make them neighbour friendly, but falls far short of capacitive “space filling” end-loading.

Pro-Antenna are basically selling an end loaded dipole that doesn’t resonate on any amateur band, but presents a reasonable impedance to a tuner. The base version has a transformer for stopping common mode currents, and perhaps to present a more reasonable VSWR to the coaxial line. There’s nothing new in any of their designs, despite receiving rave reviews in the amateur press.

The practical problem with capacity loading at HF is the high windage such structures suffer. Capacitive/inductive loading has been explored extensively, and cannot offer anything really new in optimising performance.

Linear (line) Loading

The loss from inductively loaded antennas comes from the resistance of the coil and the inter-turn capacitance. It is possible to use a transmission line for loading. This has been theorised as a more efficient way to make a small antenna. This is explored in the excellent Antenna Designer’s Notebook by KF2YN. His conclusion is the losses are higher than previously thought. Also the size reduction does not meet the ESA ka < 0.5 parameter. Keep looking..!

Dielectric Resonator Antennas

Ceramic (dielectric) antennas use a slab of high permittivity material which radiates. That is, the antenna elements are actually insulators. The low velocity factor reduces the effective wavelength, in some cases upto 10x. These antennas are used in pagers, bluetooth devices, GPS and some mobile phones.

Their big disadvantage is they are lossy much below 1GHz. Below about 1GHz it is necessary to use metal elements, which introduces losses. In some cases half the energy is lost as heat. This makes them unsuitable for transmission. I have a 433MHz dielectric antenna and find it performs close to a larger “rubber duck” helical antenna. But transmitting more than a few watts makes the antenna heat up. The antenna I tested was made by Phycomp, and their part number is CAN4313121200431B.

Lower loss materials are being researched, but they only work at UHF and high VHF frequencies. Thus dielectric antennas are not currently the answer for small HF transmitting ESAs.

Small Magnetic Loops

The “Antenna Designer’s Notebook” has a number of simulations of magnetic loops. These antennas are a closed loop, but with a high Q, they actually produce a significant E field voltage, though the near field is predominantly H. A 10m band loop of diameter <2m qualifies as an ESA. Their efficiency is good when thick copper material is used for construction. Bandwidth is small, so easy retuning is essential. The book shows designs with two capacitors, of which one could be switched and the other mechanically tuned air-spaced.

Disadvantages with loops are they require high quality tuning capacitors, and heavy gauge copper because radiation resistance is very low. Loops for bands below 20m must be large to get good efficiency. The recommended minimum is 0.1λ measured on the circumference. That means a diameter of >1.3m for 40m, and >2.6m for 80m. Such a lot of heavy copper would be expensive to make and difficult to mount outside. They are thus only really practical for the higher bands.

For the moment I am not pursuing loops, but if other things fail I will come back to them. My feeling is they don’t offer “something for nothing”. The size vs. bandwidth (Chu) limit for loops was quite recently revised by H.Thal. His highly mathematical paper suggests (I think?) that the effective Chu limit is worse for loops than dipole derivatives. So the conclusion in terms of an amateur antenna is they have a lower “goodness” in efficiency, but compensate by being less sensitive to local noise which is largely E field.

There is a distinction between small and large loops. A loop of nearly 1λ (a quad) has E and H fields in similar proportion to a dipole, whereas a much smaller loop is predominantly an H field device.

LINK TO TOP OF PAGE

EH Antenna

This design has wild claims made by inventor Ted Hart. It’s supposed to break the Chu limit by a 90 degree phase shift of the E and H near fields using an LC network. Of all the esoteric designs this one looked most likely to work. They have been “debunked” by many people, but there are also many websites claiming successful operation of them. I decided to find out for myself who is right.

I made several EH antennas for 10m, and one for 20m. 10m types use 40mm diameter plastic pipe, the 20m one uses 50mm. Copper foil is used for the elements. Two ways to improve on the cylindrical dipole construction are shown by a photo of the 20m band version:

Copper tape for the main coil (left of pic) minimises inter-turn capacitance and doesn’t wander about like round wire. Bare wire tends to make shorted turns, and holding it in place with hot-melt glue is ineffective. I find tape takes 15% more turns than wire, indicating reduced inter-turn capacitance. 3mm wide tape has the same surface area as 1.9mm diameter round wire. The bottom end of the coil below the tap carries most current, and wider tape can be used there. The tape comes with a peel-off backing, so the coil can be tuned by moving the top few turns and temporarily taping into place. When the position has been confirmed, stick the tape on permanently.

The connection from the main coil to the cylinders is 1.5mm silver plated wire. The cylinder length/diameter ratio is 5:1. Electrically, the EH antenna is an autotransformer with a capacitor connected across the ends. The capacitor consists of the cylinders plus coil inter-turn capacitance. The antenna is not a high impedance circuit as it first appears. The radiation resistance of a short dipole is very low. So optimal length ratio of the coil will be 1:1 or slightly greater, and the conductor surface area has a big effect. My engineer’s guess is the length ratio should be 1.0:1 < 1.5:1. EH antenna proponents say the phase shift increases radiation resistance to 10Ω (?).

A second improvement is to cut slots out of the pipe, to allow access to the inside, making construction easier.

There has to be an effective common mode choke on the coax cable, otherwise tuning is impossible. Without a good high impedance choke, the feeder can form a tuned circuit with the transmitter ground. I have 3 turns of RG58 through 3 large beads of type 43 ferrite 2m down the coax cable for the 20m version.

An important realisation is that resistive loss in context of stopping common mode current is actually good. The combined R + jx impedance is directly in the way of the current. If R >1000Ω and jx is also high, the overall current flowing and therefore the overall loss in a 50 Ω system is small.

A coil of cable is not a good way to make a common mode choke, because R is only large over a narrow range of frequencies near resonance of the coil. Additionally, the resonant frequency of the coil is easily disturbed by nearby objects. This does not happen with a high permeability choke.

A table showing optimum resistive choking is given by G3TXQ. This is very relevant to EH antennas. A high resistive impedance will stop common mode current sharply, and allow the outside of the cable to radiate effectively. In combination with cable having a good low-resistance outer shield, the system should work better.

The distance of the balun down the coax is difficult to determine, but the antenna picks up very little with the balun close to the antenna. That it needs to be a distance down the coax is very evident from the received signal strength.

I tried measuring the near E and H fields directly with a loop and stub antenna on a dual channel ‘scope. The position of the probes is very critical. Movement of only a few centimetres affects both magnitude and phase. The loop and stub need to be fixed relative to the antenna. However, I found changing the tap point on the coil makes no difference to the relative phase of the near fields. The conclusions from this are discussed further down the page.

With a network analyser, I calibrated out the effect of 5m of RG58 with the choke in-line. This can be done simply with the VNWA by an open/short/load calibration on the end of the cable. The EH antenna has a BNC connector, the back of the BNC connector is shown in the photo.

The next photo is the VSWR and complex impedance result directly at the base of the EH antenna. I found the bandwidth is narrower than measured from the end of the cable. The effect of the source coil is to add series inductive reactance on the HF side of the resonance curve. Thus at the 50Ω point, we perform a conjugate match and get close to a pure 50Ω resistive result. The red trace is reactance, and it can be seen to dip to the centre line (j0) where the blue resistance curve is 50Ω.

It’s a flat match, but SWR is not the real issue! Does the coax or the antenna itself radiate? Using a field strength meter, the answer is some of the RF radiates from the antenna itself. Moving the meter along the cable from the transmitter shows nothing below the choke, demonstrating its effectiveness. Above the choke there is some radiation, which drops off alongside the main coil. It increases over the bottom cylinder, but is a maximum over the upper cylinder. This does not mean the upper cylinder contributes most to far field radiation. Because the radiating section of coax is much longer, it can contribute equally or greater to far field radiation, as opposed to the local E field strength.

EH proponents say the “Poynting Vector” from a 90 degree phase shift is the most critical thing to make the antenna work. My experiment EH Phase.pdf is definitive proof the relative phase of the fields does not change when the electrical phase feeding the dipole elements is changed. Nor is there a sharp peak in either Rx or Tx performance as the coil tap is moved, which there would be if the Poynting Vector principle worked.

Compared to the OCF dipole described elsewhere on this site, signals are 6-10dB lower. It hears about as well, because noise is correspondingly lower. By the reciprocity rule, the outgoing signal must be down by 6-10dB points also.

It doesn't break the Chu limit because the coax adds to the dimensions. Also I disagree with the notion that the EH is not available commercially because the market is too small to make it viable to manufacture. The reason it is not available commercially is it does not work well, and is difficult to tune. The EH antenna does not meet the ESA spec set out at the top of this page.

My conclusion is the EH + coax combination works as a loading system for a short wide radiator. Radiation comes from both the upper element and coax tail. The coax does radiate, but as the antenna needs a feeder anyway, which can be elevated on a support pole, using it is not a disadvantage. The EH antenna is suitable for monitoring, and short range transmitter testing. I see no evidence of Poynting vector radiation, from either the phase shift or field strength measurements.

Internet searches for ‘EH antenna’ only show some websites, because many are not in English. For interest I present these links:

Original EH site from Ted Hart - snake oil warning!

EH antenna group on Yahoo (membership required)

Interesting ideas from ZL1CLG

Site of WB5CXC

A semi-scientific test by I1RFQ

Site of SM5DCO (Conny)

Lloyd Butler - his results agree with mine

Clear instructions from a French amateur

Non-English sites requiring Google Translate or Babel Fish:

UA1ACO site for flat EH designs

IS0FQK

IW0BZD

TLC Antenna

After simulation and practical experiments, I concluded the EH antenna does not work. Recognising the phase shifted signal sent to the elements makes no difference, the antenna can be simplified with a focus on reducing losses.

The positions of the coil and short elements of the cylindrical EH antenna are a hangover from the CFA (Crossed Field Antenna). Having dismissed the “Poynting Vector” principle, there’s no reason to follow that design. The flat-EH antenna favoured by Russian amateurs avoids the long feed-through wires of the dipole version.

Combining aspects from the Isotron, MicroVert and EH, I arrived at a design that will be called the top loaded coaxial (TLC) antenna. Initial performance tests indicate it will be useful for my amateur station, though it does not meet the ESA requirements set out at the top of this page. From the few tests carried out, the TLC antenna is more “live” than the isotron. I define live as picking up more signal, including background noise.

Further details of the TLC antenna cannot be revealed yet. There are three discoveries which combine nicely to give it unique abilities. But it will take some time and research to make the antenna easily reproducible.

Isotron/MicroVert

The Isotron is made by the Bilal Company. I bought a 10m Isotron and set it up on the roof here. Signals received are 12dB (two ‘S’ points approx.) down on a 1/2 λ vertical. That means efficiency of 6%. Bandwidth is 900kHz at 2:1 SWR. It performs fine for local voice contacts, and PSK-31 throughout Europe during the sporadic-E season. The receive noise level is correspondingly lower than my other antenna, so the isotron hears about as well.

The 10m isotron is a simple device consisting of just a coil and flat aluminium plate. There’s a moveable lever under the coil, as seen in the photo (left). The lever allows fine tuning by capacitive coupling to the metal plate and coil. I slung a thin nylon cord around the lever, so it could be moved from below. The antenna then can be tuned between 27.5MHz and 29.0MHz, covering CB and 10m together.

The isotron is very dependant on ground for its operation, contrary to statements in the instructions. It’s a single ended device and without sufficient counterpoise will not produce a good match. The 10m version requires a metal mast, and I also had to connect it to a metal window frame to get a low SWR. Connecting a wire >5m long to mains ground did not alter the SWR, again contrary to the instructions from Bilal.

The coax also radiated, and therefore touching it affected the SWR. This indicates where most of the losses are with the Isotron. The ground is undefined, consisting of various types of metal which have significant resistive components. The coax radiates by common-mode current like the EH antenna. A major part of the design is the coil, which has loss resistance higher than the radiation resistance of the small aluminium plate. Such a coil has large inter-turn capacitive/dielectric losses. Isotrons look like an ESA from the antenna itself, but taking into account the ground structure required, they don’t qualify as an ESA.

The Isotron’s bandwidth is undoubtedly increased by losses, but even Bilal admit it’s narrow on the 1.8MHz (160m) and 3.5MHz (80m) band versions. Moving the tuning arms can tune it over a small range, but Bilal are not developing that idea. The Isotron documentation was done on a typewriter, and from that evidence the design dates from the 1980s.

As with the EH antenna, the Isotron does “work” but with severe losses. Tuning one without a network analyser may prove difficult. I have not tested the lower frequency design which has two plates. However, it’s clear the two plates do NOT act as two halves of a “capacitive” dipole, there is no such thing. The lower plate is there to form a tuned circuit, so stabilising the frequency, and does not radiate as the lower half of a dipole does. Consider it like putting an RC tuned circuit on the end of a coaxial cable. The transmitter does not see zero impedance at resonance. It sees the losses in the cable and the tuned circuit itself as resistance. This is how the isotron works. The antenna itself radiates little, the cable screen and ground get most of the energy.

Talk of effective aperture in context of these antennas is phooey. The physical area of the plates has nothing to to with the effective aperture of the antenna at all.

Making an Isotron is easy, as they are just metal plates with a coil between. Efficiency may be improved by using copper instead of aluminium plate(s), and copper foil for the inductor. Given the ground structure does most radiating, such efforts will produce small improvements.

I removed the ground from my isotron, and connected a 1/2 λ piece of thick copper wire. The wire was laying on the roof, and the resonant frequency of the system went up to 28.5MHz. Pushing the tuning lever up brought it down to 28.3 again. The wire didn’t produce any measurable improvement, but tended to flap about in the wind.

A development of the isotron is the MicroVert. This has a copper cylinder instead of a plate at the end of the coil former, and the coax is openly regarded as part of the antenna. I have not tested one, but it should be similar in performance, and the designer admits it was derived from the isotron.

I disagree with the MicroVert designer that the coax does not radiate as much as the short vertical element. As with an EH antenna the radiating coax section is much longer than the top element, so will contribute more to far-field radiation despite the apparent result from a  field strength meter.

Slow-wave Antennas

This description covers designs like the G2AJV toroid and the Fishbone. The principle is to slow down the propagation velocity (phase velocity) within the antenna structure. Therefore effective wavelength is reduced, making resonance possible in a shorter length.

Size reductions of >30% have been reported for designs like the fishbone dipole. It is a normal dipole with many short wires (bones) fixed at 90 degrees to the main elements. Intuitively this looks like a low loss design. But the size reduction does not qualify it is an ESA.

Back in the 1990s, the toroidal antenna by G2AJV appeared. It has two toroids mounted a distance apart, and with circular plates off either end. The small size of the G2AJV “toroidal dipole” easily makes it an ESA. Designs for various bands from 2m to 80m have been described in books I have read. Again, I have not tried this antenna. It would be an interesting one to do, as in my opinion it has as much chance of success as any other design. It is definitely claimed to work by the slow-wave principle.

I find the reviews of it too good to be true. If this really worked, they would have saturated the market long ago because spiral antennas are far from new. A patent lapses after 40 years, so they have re-patented a design from 1969 by Allan Brown, that shows an electrically identical layout. Spiral antennas in general have been studied for a long time, and nothing earth-shattering discovered.

I have not tested a Tak-tenna, so have no results of my own to give. Having copper-plated aluminium wire has some strength advantage over plain copper wire, but they could have used silver as it has an even better conductivity. I would certainly expect silver at the high price they want for these!

As with all other small “miracle” designs, the loss resistance will  turn out to be much bigger than the radiation resistance, and also the feeder will radiate due to imbalance.

LINK TO TOP OF PAGE

Non-Foster Matching

So far on this page, nothing has seriously looked a good amateur ESA contender. Loops or the use of combined capacitive/inductive loading fare best, though they have been researched for years and hold out hope of only minor innovation. Looking at more esoteric concepts is necessary.

It’s possible to make active electronic circuits that produce negative reactance. They have been used in audio filters for years. If the principle could be applied to antenna matching, the need for a loading inductor or any conventional passive matching circuit is removed. Lossy passive conjugate loading is eliminated, and a method of tuning the antenna without mechanical means would be a reality.

Non-Foster matching is where a negative reactance is used to counteract a positive one. So the large capacitive reactance of a small antenna can be matched with a -C. I don’t understand the practicalities of designing such a circuit, but can see the basic principle, which  is summarised by this fragment of circuit diagram:

This seems to be a breakthrough for an efficient and broadband antenna. For sure it will be used in broadband receiving applications. But there’s a problem for transmitting, because the active devices have to handle energy in both directions.

The power going into the active matching elements must be greater than the transmitted power. It would mean having the transmit PA and front end receive physically as part of the antenna. The distinction between transceiver and antenna would then disappear. Such technology would be a greater revolution in HF communications than the arrival of SDR, and give the company selling it a major advantage over those selling “rigs”.

The non-Foster matching concept could be a paradigm shift, but implementing it and even fully understanding the theory would take a long time. Commercially available equipment based on this technology may never happen if amateurs do not do the research for themselves. But don’t look at me to do it!

Parasitic Elements for Impedance Multiplication

Putting additional elements in the near-field of a dipole can affect radiation resistance. With a Yagi, it’s usually reduced. It is also possible to increase radiation resistance. Because a short antenna has low radiation resistance, the parasitic elements must be used to increase it if we want to match a 50 ohm feed line.

Returning to the “Antenna Designer’s Notebook” there are chapters on the Boxkite Yagi and Twin-C. Both these use parasitic elements to reduce size while matching radiation resistance to 50 ohms. The twin-C described in the book is 1/6 λ or 1/3 of a dipole length, though it’s a 2 dimensional structure which also has a width of λ/6. That means a 3.5m square is enclosed on 20m, and the twin-C does not quite qualify as an ESA.

The formula for radiation resistance multiplication by additional parasitic elements is simply Ri = n2 x Rs where Ri is terminal resistance, n is number of extra elements, and Rs the the radiation resistance of a single element. So a 5.5 ohm antenna can be transformed to 50 ohms by two additional elements. This method can also neutralise the capacitance of a shortened radiating element.

The book does discuss adding more than one parasitic element, and with 2 the concept does qualify as an ESA. Bandwidth is reduced with more elements. There is more research to do with the parasitic element concept. The effect of loading the parasitic elements is unknown. It maybe possible to tune the antenna in some way, and fulfil the ESA parameters set out at the top of this page.

From experiments with the EH antenna, I concluded a single element antenna is like a 2-port electrical network with the output as EM fields. If it’s linear, nothing from the input signal can affect it’s input/output characteristics. In other words, nothing extrinsic (e.g an ATU) can improve it’s intrinsic performance. To improve the fundamental efficiency/bandwidth/size of an antenna, its necessary to change the network, which is the near field surrounding the antenna.

I do not have access to, or knowledge of advanced antenna simulation software. To poke about with parasitic element designs blindly would waste significant time, so I will not make anything outside of a published design. Over the last year, something very interesting related to parasitic element designs has come to my attention...

Metamaterial Antennas

There is much new research on metamaterials, which manipulate EM waves in unnatural ways. They have permeability (magnetic) and permittivity (electric) designed in by structure, as opposed to the natural molecular property of a bulk material. Metamaterial can be made with both negative permittivity and permeability. EM waves passing through are refracted “in reverse” of Snell’s law.

Metamaterial antennas have few commercial applications. Mobile phone type BL-40 from LG, and a wireless router from Netgear use them. Academic research from places like Imperial College London, and Arizona University looks encouraging. Metamaterial research has come from outside the field of radio antennas by heavyweight scientists. It is early days for the technology.

Even apart from antenna applications, metamaterial is a fascinating area of research. Most of it is beyond the realm of amateur experimentation or understanding. The manufacture of double negative materials has been demonstrated beyond doubt. What is in doubt is if the losses in such material are too high, and if the cost of manufacturing “slabs” in a commercial environment is too high.

A possible solution comes from metamaterial inspired antennas. They use a piece of metal shaped like a single cell of true metamaterial. This overcomes the complexity of making hundreds of individual structures and mounting them around the antenna. The metamaterial can be viewed as a form of matching network, and there is clearly common ground with other parasitic element designs. To design such antennas from scratch requires use of advanced simulation software, like Ansoft’s HFSS.

What many fail to realise is dipoles, etc are always looked at when radiating into free space. That situation can be altered.

I manufactured a prototype MM inspired antenna. It’s 46mm high, 50mm wide, and possibly the first such device made by an amateur. My results in terms of resonance and impedance are broadly in line with the Arizona University group. Following on were several prototype antennas of this type, with predictable and pleasing, but not spectacular results. Here is a picture of the first prototype, reflected in the copper ground plane used for the tests:

It will NOT provide a top-band dipole in your loft, but they are a hell of a lot more likely to provide something which works than the EH antenna!

One result I will share is that these antennas do not break the Chu limit. But they get closer to this desirable limitation than other designs I looked at.

Antennas look simple but are extremely complex, and unfortunately mathematics cannot be completely avoided. Melding theory and practical measurement is the way to go.

I have limited but useful facilities at M0RZF for practical antenna investigation. Facilities including a network analyser, roof-top antenna test fixture, and of course my ham licence to transmit in many approved bands. The major thing I’m lacking is knowledge in antenna modelling software. That needs to be addressed...

An old adage is “the antenna is 80% of your ham station”. Perhaps the percentage can be argued over, but it is definitely more important than the radio itself. Much as I like the Softrock concept, it will have to share my spare time with antenna development!!

LINK TO TOP OF PAGE