Vol.1 No.2 March 2004
A Really Cool RX Antenna and
Some Really Dirty Talk About It
It is now mid-March at Fish Lake and one of the magical features of Michigan weather is starting to unfold - the winter thaw. After 3+ months of a countryside completely covered by snow and ice, just like in a Norman Rockwell painting, the delicate balance between the wobbly-top dynamics of our cold planet and the fiery conflagrations powering our sun undergo a shift which probably appears on paper only as a change at about the fifth or tenth significant figure. To us earthlings, however, this manifests itself as a temperature increase of 30 or 40 degrees which brings about the inevitable state change from solid (ice) to liquid (water). This is accompanied by an annual, microscopic revolution in the plant and animal organisms that share the planet with us in overwhelming numbers. The plants start turning their incident fields (i.e. sunlight) into lifeforms via the amazing chlorophyll molecule instead of just scattering them back to wherefrom they cometh. The animals begin again to metabolize with a frenzy and soon the frozen, barren countryside teems with life. And alas, my new 160M. RX antenna heads south – literally!
The big project this winter at the lake was my first serious foray into the wonderful world of 160 meters. The OM has never been QRV on this band before, since he was aware of, and completely lacking in, the three main requirements for success in the land of the long waves. They are, in increasing order of importance, long antennas, high antennas, and co-operative grounds. Of course, there is yet a higher requirement , and that is the great need for water. In particular, very salty water in quantities which are capable of filling the gaps between continents..
One of our first rude awakenings was our almost immediate confrontation with noise. I mean with REAL noise, the kind that actually destroyed three sets of cheap headphones and renewed our interest in AGC. It brought back, in big bold type, the old fashioned, nearly forgotten adage that “you can’t work ‘em if you can’t hear ‘em”.
Our little haven here on the lake is not exactly a DX’ers dream property. In fact, at first glance the real estate situation is positively grim. But despite the limitations, we have been quite pleasantly surprised by our ability to make peace with Mr. Maxwell, rather than fighting his tyranny. Our meager antennas have shown amazing ability to launch waves with a vengeance – at least enough vengeance to make it consistently across some of the oceans. But on a band where a signal that is just 10db above the noise level is likely to elicit a 589 report , RX antennas reign supreme. The big guys have really big antennas, but they also have fields full of those clotheslines called ‘beverages’. So we took a little time to study the situation, using our primary tool for quick education and review of subjects like this: a trip to the University archives for a browse through the gospel – The Proceedings Of The IRE. This journal chronicles in exacting detail virtually every major development in the history of electronics, and RF electronics in particular, since the early 1900’s.
We arrived at a few very general facts during the first hour of library time. First, the basic theory of antenna design was formulated and put into practice starting in the early 1930’s, a time at which even the word computer didn’t fully exist yet. But those pioneers struggled their way through ridiculously long calculations using slide rules and, where necessary, hand arithmetic, to lay down for us the paradigm by which successful antennas could actually be designed and built.
Second, we found that the presence of the earth has a profound effect on antennas, especially as the antenna gets closer to the ground. Everyone knows this, but I don’t think they know just HOW profound it is (I will be elaborating on this shortly). What I found particularly disturbing is how much this really complicates ‘the problem’, the problem being how to calculate precisely how an antenna radiates when fed a voltage at some pair of terminals, or how it responds to an incident wave field by generating a voltage that our RX can detect and decode.
There are a number of alleged ‘facts’ about the beverage antenna which have been circulating since its inception in the 30’s, by, amongst others, Mr.Beverage. He was one of the leading radio/RF pioneers at Bell Labs who was mandated to find a way to establish transatlantic telephone communications. A number of these facts have always aroused suspicion in me and one of my missions is to validate or refute them. But my first mission, for this year, was to find a way to hear DX on 160.
The QTH here on the lake is not especially noisy, and I am far enough from the big city that the normal levels of industrial noise have been attenuated by the time they get here. But typical forms of atmospherically generated noise and QRN are plentiful, to the extent that for over75% of the time the noise level on my TX antenna is too high. Having heard so much praise for the beverage antenna I felt compelled to try it, but the location is not really suited for it.
But one thing that I do have that is unusual, compared to most, is an 80 acre fresh water lake in my back yard!. Although slightly unusual in shape, it does feature one dimension that allows a continuous 1500+ feet straight run from my dock on a direct European heading (about 40 degrees NE from here). Would it be possible to somehow use this?
After wading through the basic theory of wave propagation over lossy media (the official description of radio waves over the earth) a few facts became evident, and I would like to briefly comment on them. A key concept in all of the theory (and practice) of this type of wave propagation is the concept of conductivity, for which a short review is in order.
A QUICK TREATISE ON ELECTRICAL CONDUCTIVITY
To the pure physicist the relevant quantity is called ohmic conductivity and it is a basic characteristic of all material objects and liquids. It is simply the ability of the substance to conduct an electric current when a voltage is applied. The generic model for studying conductivity is a pair of parallel plates, between which some material is placed. When voltage is applied to the plates, an electric current (which can be simply measured in terms of the number of electrons per second which cross the plates surface) flows. This current is motivated by the electric field (the so-called electromotive force, or EMF)
A little bit of basic physics shows that under normal circumstances the current that flows will be proportional to the ratio of the area of the plates to their separation. In the metric units, this ratio will be a quantity with the dimensions of meters(squared) divided by /meters, or simply meters. Now, the current itself will also be proportional to the applied voltage, which we can express as so many amps per volt. It can now all be put together as follows. The unit of 1 amp-per-volt is called a mho because it is the reciprocal of an ohm (1 volt-per-amp). We agree by definition that if the voltage is 1 volt and a current of 1 amp results for one meter of conductor material (actually, one meter(area)/one meter(thickness) the conductivity of the material is 1mho-per-meter.
What I have just done here is explain the basic derivation of the concept of conductivity, which is probably the most important basic quantity in electronics. It is the natural way to look at real materials, in which an applied electric field (i.e. a voltage) causes the flow of electrons in that material (i.e. current). Generally speaking, most materials in nature are NOT good conductors. In fact, most are good insulators, and it takes a special set of conditions to cause all of the atoms of a material to enter the state that we call "conductive”.
Every material possesses a specific value of conductivity which is nearly constant over a wide range of temperature, pressure, etc.. The unit of mhos-per-meter is the MKS standard for this and most antenna theory published after 1950 uses this unit. It is universally denoted by the Greek lower case sigma (i.e. s). However, I am finding that the lower case greek letters are always getting QRM’d by these computer word processors, so I will continue to spell it out in my discussions of ground conductivity.
For the record, many places in the world are more comfortable with an alternate description of this quantity, which is resistance. It is simply the reciprocal of conductance so that it measures volts-per-amp across a conductor. It is no more or less accurate, only an alternative viewpoint.
I would like to make one more brief diversion before returning to the main subject at hand. You may have noticed an amazing similarity in the above model of conductivity to that of a parallel plate capacitor, and you are absolutely correct. Let me now remove the sigma-material (a conductor) between those plates and replace it with a dielectric like air or plastic. If the applied voltage is now an AC sine wave instead of a DC voltage, there will also be a current flow. This current flow is a result of what is commonly referred to as reactance, but whose true mechanism is too deep to go into here. Just let it suffice to say that everything that holds for a conductor now holds for the dielectric, and the current that flows is commonly called the ‘displacement’ current. Its major distinction from conductive current flow is that it is 90 degrees out of phase (leading) with the voltage, whereas the conductive current is perfectly in phase. This distinction has profound effects.
For materials which produce current flow in this way (displacement current) the ratio between voltage and current is epsilon (e) but again I want to avoid any typographical problems It is the constant of proportionality between current and voltage applied to a dielectric in the same way as sigma does for conductors, and all materials posses a typical value which is nearly constant below the microwave region.(Many of these material properties begin to change seriously at infrared frequencies and above, where the RF wavelength starts to become comparable to molecular dimensions).
It is common to use an alternate unit dielectric constant , which is ratio of the permittivity of the material to the permittivity of vacuum, known as epsilon-sub-zero (e0.=8.87 pf/meter)
Typical RF materials have dielectric constants in the range of 2-10.
This was a brief review of the nature of current flows in materials as a result of applied field, and most materials that we deal with are primarily one or the other. But there are many cases where actual materials exhibit both types of current flow. In fact, most materials really do both at particular frequencies. For example, PVC pipe at DC (i.e. frequency=0) is an excellent insulator (sigma=0) and a perfect dielectric with k=5. But at a frequency of 100MHZ the PVC will become quite “lossy”, which I will now clarify the meaning of. It simply means that upon applying an AC field(i.e. 100MHZ) there will appear a component of the current flow which is IN PHASE with the voltage, and hence, will result in direct loss. It is valid to say that the material has now taken on a ‘conductive’ component of current, which is identical to what would happen if it were a weakly conductive metal. At 100MHZ this conductive component is about 10% of the reactive component, resulting in about a 10% loss in any RF current flow.
In other words, we have here the appearance of both a resistive and reactive component of current, and in formal engineering terms this is expressed by saying that the permittivity is now a complex number. I do not want or need to go into detail about that now except to say that it is directly analagous to an electric circuit having both R and X. All of the usual analogies about phase angles, power factors, etc. all hold true. Some RF designers use these analogies to reduce RF field problems to circuit theory, which to many seem more basic to analyze.
Now I am ready for the main point of this subject. The earth itself is a ‘complex dielectric’ in exactly the sense that I have illustrated. It is a mixture of sigma and epsilon, with each quantity covering quite a wide range. The values of these parameters are known in a very general way, and I will cite some actual numbers shortly.
Before that, however, I need to talk about the role they play in antenna theory. That is what brought us here and that is how we will leave. For the remainder of this discussion we will be concentrating on a special class of antennas – antennas close to the ground. By close we mean less than a wavelength, which on 160 is about 250 feet. So this includes pretty much EVERYBODY’s antennas on 160 and 80 meters.
One property of dielectrics and conductors alike is that they both produce significant reflection of waves impinging on plane layers of them . In particular we are going to be concerned with the case of air and earth, with the surface of the earth defining the infinite plane boundary between them. The entire history of the theory behind grounded antennas is based upon this simple model, and we will follow it here.
In addition to reflection they also produce refraction, which is similar to reflection except that the refracted wave enters into the region beyond the boundary layer, or the earth in our case. The whole picture of antenna radiation from grounded antennas is determined by how these various processes interact with the antenna and affect the field that manages to get radiated into the far field.
If you look at the pattern of a generic antenna (e.g. a resonant dipole) in free space you will see a nicely directional pattern with major lobes broadside to the wire with about 3-4 dbi of gain. In addition, the radiation is totally non-directional in the elevation plane. That is, it radiates equally at all elevation angles including 0 and 180 degrees.
If you now look at it over a perfectly conducting ground (sigma=infinity) the following behavior appears. As the antenna-to-ground spacing decreases, a peculiar spatial resonance pattern emerges due to the reflection of the field of the antenna from the ground. I will not dwell on the details of this other than to note that as the height approaches zero, the radiation at low angles is extremely attenuated. At zero spacing the radiation is mainly upwards – the familiar ‘cloudwarmer’. There is no radiation at 0 degrees and series of elevational lobes that appear and disappear as the angle increases.
The vertical radiator shows a somewhat different behavior. When it is placed at ground level, the usual scenario, a perfect ground plane greatly enhances the radiation field, preferentially at very low angles all the way down to 0degrees. In the process it produces an overall 3db gain because all of the radiation that went off in angles below the horizon in the free-space model now are added by reflection to the primary field. If we could actually build antennas like this they would blow your socks off! But Mr. Maxwell always prevails, in this case abetted by a few others who will go unnamed.
We now consider what happens as sigma drops from infinity (for a perfect conductor) to the values found in MY backyard and yours. At this point I am going to bring forth some of the numbers I promised earlier about real world soil. In other words, get ready to hear the real dirt now.
The two most basic references we have for discussing ground properties are pure water and silicon sand. The first is a very pure dielectric with k=80, and the latter is a simple inorganic molecule that is nearly a perfect dielectric with k=8. The value of sigma is essentially 0 for both, a fact which I have verified to my own satisfaction numerous times. Real earth, on the other hand, is not nearly as well behaved. First, it contains additional types of solid molecules with different values of K, mostly higher. Second, it contains water, but unfortunately this water is not just pure water with k=80 and sigma=0. The dirt contains significant quantities of ionic salts which dissolve in the water to form conductive liquids that greatly enhance its sigma value.
I have personally measured numerous dirt samples and found that the value of sigma can be radically altered with relatively small change in water content, with the value of k also varying but not as radically. The commonly quoted figures for wet soil are in the range of .01-.1 while ‘dry’ soil is quoted at .001-.01 mhos/meter. By comparison, sea water is 10. Measurements in my area indicate a range of .001-.005 . Values of greater than .1 require the soil to be wet to the point of ‘muddiness’.
This whole area is quite fuzzy to me at this point, as these measurements show values which generally lie in the expected range, but show so much variation that you cannot confidently make accurate calculations with them. My summary of the situation is that for the midwest soils that I have tested, sigma probably lies in the range of .001-.05.
Based upon this we now make another calculation, that of the skin depth of a lossy conductor with sigma=.001 and mu=1. We find that at 1.8MHZ the skin depth turns out to be 30Meters (yes, I did the arithmetic correctly and double checked it!) The implications of this are profound and I am going to touch on them only briefly for now. What it shows is that for the situation of a wire antenna over a lossy ground, that ground is a LONG way from being a nice reflecting copper plate. In fact, it is more like a big block of plastic with a k in the range of 10. It reflects moderately well because of its dielectric properties, but it also moderately well absorbs and dissipates all of the radiation that is refracted into the earth.
In order to understand the intricacies of the beverage itself, you must carefully analyze the process of wave propagation over this type of earth. I have done some of this and will state some of my findings relevant to this topic. For the study of this wave propagation the most important parameter that enters the calculations is the ratio [sigma/epsilon*2PI*f], also known as the index of refraction. Its value, denoted by N, appears in every single wave propagation factor and amplitude term, and this is how (mathematically) the physical properties of the dirt get translated into their real effect on waves.
For values of N=10 or less, the ground conductivity contributes very little reflection. It does not begin to start reflecting like a copper plate until it gets up into the range of 100, which is true for sea water.
The beverage antenna ideally responds only to the horizontal component of electric field incident upon it. However, a horizontally polarized wavefront is highly attenuated by the conductivity of the earth which essentially shorts out the electric field. But a vertical field is a different story. As the vertical E-field strikes the earth, some quite complex changes in the wave field occur. They are a result of phase shifts that are induced by its complex permittivity (i.e. its combination of epsilon and sigma). The details of these changes are very difficult to analyze, requiring the full use of formal EM field theory, and in another installment of these chronicles soon to be appearing, I will attempt to explain this process with good clarity without heavy doses of vector analysis. For now, just let me summarize the main point to be dealt with -–the one that makes the beverage antenna possible.
It is generally called ‘wave tilt’, which was its historical title. It is now known that this is not quite correct. The wave does not actually tilt forward, but takes on a small amount of a peculiar type of elliptical polarization. This net result of this is an equivalent horizontal component of the electric field that lies right along the direction of propagation, so if you place a wire oriented in this direction, that horizontal field will induce a current in the wire. In other words, the wire responds to a component of the field that was NOT actually transmitted in the (vertically polarized) incoming wave. It was produced by this exotic phase-shift and reflection effect of the earth.
This effect, and the resulting antenna action that occurs from placing a wire oriented parallel to this field, cannot be properly analyzed by the popular antenna ‘modelling’ programs. I will deal with this subject also in the near future, to show why this is true. It is necessary to go back to basics, as I have been doing, to study the effects of the lossy ground on wave propagation and how they subtly modify the wave field in our favor.
For now I would like to simply summarize this issue by stating that all of the phenomena involved here are highly dependent on ground characteristics (our friends epsilon and sigma), but one thing can be stated as a general rule. That is, for the range of sigma and epsilon of interest to us, the earth is a very fuzzy interface. What I mean by that is profound: the concept of ‘ground’ as we like to use it in antenna theory, is not nearly as concrete as everyone believes. The prevailing impression is that ‘ground’ is a copper-like conducting plane which almost perfectly reflects waves that strike it. In using modelling programs we casually connect feedlines, terminating resistors, etc to ‘ground” as though they were perfectly conducting, equipotential layers of basic electrical theory.
One fact, which you may not have at your fingertips, is that whenever a boundary surface between two regions and/or media is subject to an incident wave, there will be some reflection of that wave , as well as an electrical current near the surface of that boundary. That current is an inherent part of the reflection process – you cannot have one without the other. For a typical metallic boundary (i.e. a copper ground plane) the current never penetrates further than the skin depth into the absorbing region. But as I showed above, the skin depth of our earth is in the range of 10-100 meters!. Hence, there is no ‘surface current’ to clearly identify a sharp boundary, because it (the induced boundary current) is now distributed gradually in the vertical direction, as you go down, over a range of many meters.
My point about this boundary business (I apologize for having to drag you through all of this, but it is the only way to the truth) is that those effects that happen near this particular boundary cannot be very sharply localized because the boundary itself is not! In particular, the ‘tilted’ wave field itself does not vary rapidly with height near the surface of the earth. That is, it varies only slightly as the field point goes from 0 (at the surface) to several meters, and possibly many meters, above the surface. Therefor, this antenna should perform approximately the same whether on the surface or several meters above it (or below it!), with regard to its ability to receive the ‘tilt’ field. This was the basis for the Fish Lake Beverage. I am going to now leave the theoretical discussion of the beverage, realizing that there are numerous questions to be answered. As promised earlier, I will be revisiting that subject soon with some new and specific information which will enhance our understanding greatly.
The Fish Lake Beverage was erected (I use that somewhat loosely) in December just after the lake was frozen enough to support the weight of myself and my sidekick/assistant engineer Eddie. Starting at my dock, we unreeled a continuous 1500 feet length of #14THHN on a heading of about 40 degrees. This puts it in line with our bearing to central Europe, and hopefully all of the European DX that we hoped to be hearing during the winter.
Our initial feed systems proved minimally effective and we ended up with simply connecting the coax to the wire and grounding it to our dock. The latter is a fairly ‘high-rent’ ground as such things go. It is an aluminum framework supported by 12 three-inch aluminum pipes buried several feet into the lake bottom. Compared to a 6 foot copper rod pounded into the ground, it is a giant step forward. But again I keep hearing that nagging question “what is ground?”.
By the end of December the FL Beverage had assumed it final home (for this year), embedded about 1-2 inches below the surface of the ice covering the lake. In other words, embedded in a block of dielectric with a k of around 50, which floats on the surface of the lake (k=??), which has an average depth of about 10 feet, with a few areas where the depth goes to 35 feet. The bottom is pure mud with very high conductivity (sigma>1).
If I had some nice equation to plug all of those numbers into and get a magical result in return, I would have done it in a flash! But that possibility is some way off, so I just wired it up and started listening. The following is what I heard.
My only viable choice for a reference antenna is my TX antenna, which is a highly unorthodox vertical that will be the subject of another of these chronicles. It is a very well-performing antenna within the category of basic, single-element verticals. The output signal from the FLB is about 20-25db below this antenna, so I immediately built up a pre-amplifier with 20db gain and excellent selectivity and IM specs. (This pre-amp will be sold publicly by my friend Joe, N8EA to other interested Lfers.) With a little switching circuitry I can instantly switch between the amplified beverage and the TX antenna, which I did relentlessly for the first month of operation.
The beverage signal is extremely quiet, not just because of its lower gain. Most of that has been equalized with the pre-amp, as evidenced by the fact that I have copied many European signals that were basically equal strength on both antennas. There certainly are differences, especially for more off-axis signals, but generally speaking the beverage+pre-amp produces comparable output to the TX antenna. Also, the signal is ALWAYS louder with the TX antenna, signifying that I probably could use another 10 db of pre-amp gain. That is, I have never heard a signal on the FLB that I could not hear on the vertical.
But the SNR difference is enormous. It is so large that when I first started using the FLB, I would switch it in and out excessively just to verify that it was actually working! I have compared it to several local stations, particularly N8EA, who has a good assortment of traditional beverages. It is actually noticeably quieter than every other one I have tested, yet we seem to have comparable RX gain on the basis of relative station performance. During the ARRL CW DX contest I was able to work more DX in 24 hours on 160 than I have in my entire lifetime.
The directivity is not nearly as great as many of the claims I hear for beverage antennas. I find it to be moderately directive, consistent with a cosine directivity factor that I would expect on theoretical grounds. During the one VK QSO in the contest I was only able to hear him on the vertical, which may be explainable by his bearing from me being about 60-80 degrees off-axis. This is one of the additional areas that I am interested in analyzing along with the general study of beverage and other grounded antennas.
The feedline is about 150 feet of RG213, connected as described earlier. The measured impedance at the shack is fairly low, in the range of 30-40 ohms and not highly reactive. I state this for no particular reason, since I have not yet arrived at a point of knowing what is really meant by the impedance of this antenna. The problem is, as usual, the ground connection. The whole ground ‘system’, and the manner in which it passes current from the antenna endpoint into a slightly conductive dielectric, is as much a part of the antenna as the wire itself. So trying to attribute some value of radiation resistance to the tiny ‘tilted’ component of the electric field is a daunting task.
Well, that’s about the end of the story for now. The Fish Lake Beverage has now gone due South to the bottom of the lake. It still works but I have lost a very audible number of db of overall gain. With sections of it in 35 feet of water now, I am not going to pursue it much more for this season. It still copies fairly well and is still a savior for local QSOing when the QRNancy is bad. Next winter I may try a more advanced design to achieve some directivity. But before doing that, I still have to figure out what “ground” means, especially if getting directivity requires grounding one wire. When I have achieved that level of understanding, you will be the first to hear about it.
73 and please say Hi if your hear me on the air.
Updated March 26, 2008
 |
  |
| magic supplies |