Below the Noise Floor — Episode 22: "Height, and What Is Underneath It"

Last episode was the noise you make yourself. This one is the question that follows it: not what antenna to build, but how high you can get the one you already have.

In this episode:

• Why height is measured in wavelengths and never in feet, and how the same thirty-five foot tower is a good low-angle antenna on ten meters and a near-vertical radiator on eighty.
• What ground reflection does to a horizontal antenna at a quarter wavelength, a half, a full wavelength, and two, and the resulting takeoff angles.
• Why a wire dipole at sixty feet beats a beam at twenty, and why gain aimed at the wrong angle is not gain you get to use.
• The Fresnel zone: why the ground that sets your low-angle performance is out in the next field, why a downhill slope is worth real decibels, and why you cannot fix low-angle takeoff by improving the soil at the base of the mast.
• Why horizontal antennas are forgiving of poor ground and verticals are not, and what saltwater is actually doing for coastal stations.
• The two separate jobs of the ground under a vertical: near-field current return, which radials fix, and the far-field mirror, which they do not touch. What Brown, Lewis and Epstein found in 1937, where the curve flattens, and why four elevated radials can match a large buried system.

Next: lightning protection and station entry panels, done properly.

Transcript

Copy

SPEAKER_00 0:00

Last time we spent an hour on the noise you make yourself, the switching supplies, the common mode current, riding on the outside of the Cokes, the three different things people mean when they say grounding. And at the end of it, I said that once you have quieted your own station down, the next question is not what antenna you should build, it is how high you can get whatever antenna you already have, that is tonight, height and what is underneath it. This is one of those subjects where the physics is genuinely simple and the folklore around it is genuinely dense, and I think the folklore has cost more people more contacts than any deficiency in their equipment ever has. Let me start with the single sentence that carries most of the weight. An antenna does not radiate in a vacuum. It radiates over ground and the ground reflects. What you hear at the far end is the sum of the energy that left your antenna going up and outward, and the energy that left going down bounced and came back up. Those two waves arrive at any given distant point, having traveled slightly different distances. Sometimes they arrive in step and add, sometimes they arrive out of step and cancel. The angle at which they add and the angle at which they cancel is set almost entirely by one thing, and that thing is how high the antenna is above the ground in wavelengths, not in feet, in wavelengths. This is the first place people go wrong, and it is a very forgivable place to go wrong, because we buy towers and feet and we push antennas up in feet, and the number on the tape measure feels like the real number. It is not the real number. 35 feet is a respectable height. On 10 meters, 35 feet is just over a wavelength, and you have a genuinely good low-angle antenna. On 80 meters, 35 feet is about an eighth of a wavelength, and you have something that puts nearly all of its energy nearly straight up. Same tower, same tape measure, completely different antenna. So whenever you are thinking about height, do the division first. Height in feet divided by the wavelength in feet. 984 divided by the frequency in megahertz gives you the wavelength in feet. Near enough. Do that division and hold the answer in your head because everything that follows is about that number. Here is what that number does. Take a horizontal dipole. If you put it a quarter wavelength above ground, the reflected wave comes back and reinforces most strongly straight overhead. Your pattern is a broad dome. Almost everything goes up at a steep angle, 60 degrees and above, and comes down again a few hundred miles away. Practically nothing goes out at the low angles that make long-distance contacts. Now take that same dipole to a half wavelength. The reinforcement pattern shifts. You get a main lobe at roughly 30 degrees above the horizon, and the energy straight overhead has begun to fall off. This is the height at which most people first notice their antenna has become a different antenna. Take it to a full wavelength. Now the main lobe is down around 14 degrees, which is squarely in the range that supports multi-hop propagation, and a second lobe has appeared higher up. Two wavelengths, and the main lobe comes down to about 7 degrees. Those numbers are approximations over average ground, and you should treat them that way, but the trend is not an approximation. It is a consequence of geometry. It holds for every horizontal antenna, and it does not care at all how expensive the antenna is, which leads to the conclusion that I think is the single most useful thing in this episode. A simple wire dipole at 60 feet will outperform a beam at 20 feet for long-distance work on the bands where 60 feet is a meaningful fraction of a wavelength, not sometimes, reliably. The beam has gain, certainly, and it has front to back ratio, and it is a better antenna in every respect except the one that dominates, which is that at 20 feet on 20 meters it is aiming its gain at the sky. Gain pointed at the wrong angle is not gain you get to use. So the ordering of priorities, if you are spending money or time, goes height first, then a decent feed line, then antenna design. People do it in exactly the reverse order because the antenna is the part with a model number. Now let me spend some time on the ground itself, because this is where it gets more interesting and I think more misunderstood. The reflection we have been talking about does not happen at the base of your tower. This is worth pausing on. For a signal leaving at a low angle, the reflection point is out in the field, hundreds of feet away, and for very low angles, it can be many wavelengths away. The region that matters is called the Fresnel zone, and for the low angles that carry long-distance signals, it can extend out 10, 20, 100 wavelengths from your antenna. That has two practical consequences that come out in opposite directions. The first is that the ground you can do something about, the 20 feet around the base of the tower where you might be tempted to lay wire or dig, is largely not the ground that is setting your low angle performance. You cannot improve your low angle takeoff for a horizontal antenna by improving the soil near the mast. The reflection is happening in your neighbor's field. The second is that the terrain out there matters a great deal, and it is not soil chemistry, it is shape. A clear downhill slope in the direction you want to work is worth real decibels because it effectively lowers the ground under the reflection point and raises your antenna in wavelengths relative to it. A hill in that direction costs you. This is why some stations in ordinary suburban lots with modest antennas do remarkably well in one direction and poorly in another, and the operator concludes something about their antenna when the answer is the shape of the land. Ground quality itself, meaning conductivity and dielectric constant, does matter, but it matters unevenly depending on the kind of antenna. Horizontal antennas are relatively forgiving. The reflection off poor soil is a little less efficient than off good soil, and you lose a fraction of a decibel to a decibel or two at low angles, which is real but rarely decisive. Vertical antennas are a different story entirely, and here poor ground is expensive. A quarter wave vertical over genuinely poor soil can be several decibels down at low angles compared to the same antenna over good soil, and over saltwater it can be several decibels better again. This is why coastal stations with unremarkable verticals sometimes sound impossible on the air. They have the best ground reflector on the planet, starting a few hundred feet from the feedpoint, and there is nothing you can buy inland that substitutes for it. Which brings me to radials and to a distinction that I want to make carefully because it is the source of a lot of circular argument. The ground under a vertical antenna is doing two entirely separate jobs and they are not related to each other. Job 1 is being the return path for the antenna current in the near field. A quarter wave vertical is half of a dipole. The missing half has to be supplied by something, and that something is the ground system directly underneath, within a quarter wavelength or so of the base. Current returning through lossy soil dissipates as heat. That is a straightforward, unglamorous efficiency loss, and it is what radials fix. Radials are not a lightning measure and they are not a safety ground. They are a low resistance path so that your transmitter power ends up in the air rather than warming the dirt. Job 2 is being the mirror that produces the reflected wave out in the Fresnel zone, hundreds of feet away, where you have no radials and never will. Radials help enormously with job one and do essentially nothing for job two. That is the whole distinction, and once you have it, most of the arguments about radials resolve themselves. On how many, the classic work here is the Brown, Lewis, and Epstein study from 1937, which is where the very long radial recommendations in broadcast practice come from. They tested up to 113 radials. For amateur purposes, the useful shape of the curve is this: going from 4 radials to 16 buys you a great deal. Going from 16 to 32 buys you something worth having. Going from 32 to 60 buys you a fraction of a decibel. Going past 60 is for broadcast engineers with a license condition to satisfy. 16 to 32 quarter wave radials on the ground is a genuinely good ground system, and almost nobody needs more. And a note on elevated radials, because this comes up constantly, if you can get the radials up off the ground, even a few feet, the picture changes because the soil is no longer part of the return path. Four elevated radials, properly resonant and reasonably symmetrical, perform comparably to a large buried system. That is not a trick, and it is not controversial anymore. It is the reason ground plane antennas on a mast work as well as they do. So let me draw this together into what I would actually do. If you are choosing where to put an antenna, choose height and measure it in wavelengths on the band you care about most. If you can put it where the land falls away toward the horizon you want, do that and value it as highly as you would value real gain because it is real gain. If you have a horizontal antenna and it is below about a half wavelength, understand what you have. It is a regional antenna. It is very good at that job, and on 80 and 40 meters, a low depole is genuinely the right answer for talking to people 200 300 miles away, better than a high one. Do not think of it as a failed long-distance antenna. Think of it as a different instrument, and if you want the other job done, that is a second antenna and not a modification of this one. If you have a vertical, put in radials, 16 if you can, 32 if it is easy, and then stop and spend the remaining effort somewhere else, because you have collected nearly all of what is available, and accept that the ground out where the reflection happens is not yours to improve. It is a property of where you live. And underneath all of it, the thing to hold on to is that height and wavelengths is the variable that dominates. Everything else is a correction term. Next time, lightning protection and station entry panels done properly, which is to say a single point where everything enters the building, bonded together with the reasoning behind it rather than just the picture, because it is the one area of station construction where the folklore is not merely inefficient, but occasionally dangerous. Until then, keep the noise down and get it higher.