Simple Antennas and Good Grounds
In small spaces, vertical antennas are very attractive. They can generate low takeoff angles for radiation, which means long skip distances. These low radiation angles are just what are need for communicating with DX such as "Our Neighboring States."
However, there are problems with installing verticals. Some commercial verticals address these problems and at least allege that you can use them without radials. Some like ground plane antennas have a built in ground plane radials, often drooping down to adjust the feedpoint impedance a bit. The drooping of the ground plane improves the match to 50 ohm coax, which is why ground plane radials on VHF antennas seldom stick straight out. Configured like that, the feedpoint would be closer to 36 ohms.
With the exception of the vertical dipole, a dipole on its end, vertical antennas are only "half there". The other half is a "reflection" in the ground. They rely on return ground currents and use the ground as part of the antenna system. As a result, if the ground sucks, the antenna will also. It may load up nicely, but so does a dummy load. This applies to all end fed verticals, whether they are 1/4 wave, 3/8 wave, 5/8 wave, 3/4 wave or 1/2 wave tall, or even a random height. They all need a good ground underneath them to function well.
Good grounds are hard to find. Much soil has lousy conductivity. Salt water swamps are good, if you don't sink. Ground rods, copper-plated 8 foot steel stakes driven in the ground, might work in such a place if you use an array of them. Unplated ground rods rust almost immediately and become useless for ground contact.
Ground rods alone are usually no ground at all at RF, but they are very handy for safety and bleed off wind-generated static charge nicely. They may help moderate damage in case of a lightning strike, if you are lucky. A lightning strike can easily carry 30,000 amps of RF current at about 2.4 Mhz. That is a bit more load than most antennas are designed for.
So that leaves making your own "ground". This has been extensively researched since the earliest days of radio, and the bad news is clear:
From Dean KH7B
I installed KAHU [1060 Khz. AM] in Hilo [Big Island] over 15 years ago. Its ground system was installed by hand, pick and shovel. Five miles of #10 solid copper wire installed as 120 radials, covering 5 acres.
The result was one of the more efficient BC stations in Hawaii, especially on this island. The FCC wanted the station to lower the power from the 1 kW specified in the construction permit to 938 watts. The measured 190 mV/meter at one mile exceeded the expected 175 mV/m.
When I reminded the FCC about the poor ground conductivity in Hawaii, they said "OK" to the 1 kW. My measurements showed that the average ground conductivity 0-20 miles from Hilo to be less than 1 mmhos/m (not including sea water paths).
Fortunately, one can cheat on this and suffer only minor loses, up to a point. Backing off from 120 to 60 radials is not a big deal. Trying to get by with only four ground radials is certainly going to have a big negative impact on your antenna performance.
There are indications that four ground laid radials detune an antenna greatly and make for six DB of power loss. Four elevated radials will work well if properly installed. 16 ground laid radials clear up many of the problems seen with four ground laid raidals based on modeling.
Here are rules that seem to be reasonable for ground radials (not to be confused with elevated resonant ground plane systems):
Radials can be rather small diameter wire since so many of them exist to share the return currents and they are in parallel with the ground currents in the earth as well. Each radial is going to carry very little RF current.
Ground radials need not be resonant. This is a misconception based on elevated or ground plane type elements. True ground laid radials designed to supplement ground return currents in the earth need not be resonant. They are different from the elevated ground plane radials in this regard since ground radials supplement ground currents and do not try to replace them entirely. Elevated ground plane radials, especially if few in number, need to be bit longer than 1/4 wave at the operating frequency.
Ground radials seldom need to be longer than .2 (two tenths) wavelength regardless of the height of the antenna, even a halfwave vertical radiator. A maximum of .28 wavelength seems to certainly be an upper limit for ground radial length. Due to detuning of the ground, insulated wires laid on the ground tend to be electrically 1/4 wavelength when the physical length is close to .28 wavelength.
Ground radials do not need to be much longer than the antenna is tall. A shortened antenna with loading coils will have a more compact "near field" where the majority of the antenna field is. The ground needs only reach out as far as the near field extends. Field intensity drops off with the square of the distance from the base of the antenna.
Minimum number of ground radials is probably 8, closer to 16, well you would do better with 32... You get the idea, the more the merrier. Four ground radials is going to be a horrible system. More than 32 radials gets you into the area of diminishing returns.
The ground around a vertical monopole type antenna can be viewed as strings of series connected resistors fanning out from the base. The purpose of the radials can be viewed as attempting to short circuit as many of these resistors near the base as possible. This is especially critical very close to the base where RF field density is highest, and its importance drops off quickly beyond 1/8th wavelength from the base of any vertical antenna, where the RF field density per unit area goes down sharply.
It is important not to confuse this application with elevated ground planes. We are talking about radials that supplement the return of ground currents to the base of the antenna, especially in the near field. They work "in parallel" with the existing earth ground surface to supplement it. Elevated radials are a resonant element and serve a decoupling function and establish a completely artificial ground. They should be resonant, quarter wave wires, but still in fair numbers, probably more than the four usually seen, for best results.
Another approach to grounding is an actual metal cold water pipe system. This is not that hard to find in local parks and works well for portable operation with the always popular end fed random wire and tuner setup. You can find water pipes used for watering grass, irrigation pipes that are large diameter metal pipe systems, backflow preventers on large metal water mains, and of course fire hydrants. These masses of metal act as a giant counterpoise as well as a metalic path for ground currents to supplement soil conductivity. Plus since they are buried in the ground. they have good ground contact over a very large surface area for many wavelenths.
Fire hydrants near trees are an ideal combination and have proven to work very, very well. Halfwave wires are easy to end load with the typical tuner, but they have a very bad reputation for RF on the rig. It is deserved. They put a lot of demand on the ground system. Still, tuners for end fed wires can be quite small at low power levels and such antennas are superb for portable operation if a good ground can be found to work them against. A typical fire hydrant is attached to 6 inch or larger underground pipes and is a lot of beef when it comes to ground connections.
Often you can find nice 3-inch metal pipes used for park sprinkler systems poking to the surface here and there, and small water faucets for hoses are often found hidden right next to trees, out of the range of the big commercial lawn mowers used, but still attached to the same water pipe system. Any of these have great promise as they are probably part of extensive underground metal piping.
Large gluing clamps make ideal connectors, the kind that are like giant spring clothes pins. Take the plastic tips off and bend the tips so they are flat with each other. Drill both jaws of the clamp and put in heavy shorting wires jumpering both jaws together and a heavy, short ground lead for the rig/tuner. Be sure to remember that the ground lead must be SHORT, not just physically, but at the operating frequency. You can have lots of slack at 80 meters, but even at 20 meters length of the ground wire starts to become a consideration.
The cold water pipe ground works less well in many houses, where recent practice has been to use PVC or other plastic pipes for cold water, though you may find some metal pipe outside, like on an external faucet you can attach to.
Making certain you have an actual metal cold water pipe is critical, since plastic just is not going to cut it, and tap water is usually quite a high resistance. Keep in mind that high power transmitting tubes are made with directly water cooled plates! These power tubes flow water through the plates themselves, right in contact with the thousands of watts of RF power and thousands of volts of the plate supply, counting on the resistance of pure water and plastic pipes to insulate the water flow.
A viable approach to ground rods in horrible soil conductivity areas has been to use chemicals in conjunction with the driven or buried ground rods. Books have describe putting an eight-foot ground rod in the bottom of a pit and then filling the pit with sand and chemicals like copper sulphate. Of course you will kill everything near the ground rod and poison the soil, but it works as long as the chemical lasts. A commercial version using a perforated copper pipe and more environment friendly chemicals is available called Valrods.
Do not over look the simple approach found at your local lawn and garden center. Fertilizer will help the yard near your vertical or ground system and significantly help lousy conduction like 0.2 siemen/meter. You can get two kinds. Pellets are broadcast [as in cast over a broad area, people have been doing it with seeds for millenia]. Solubles can be sprayed with water, often using special inexpensive dispensers that attach to garden hoses. Score points with the XYL by doing "lawn work" just before the next 160 contest.
When running ground wire connections, keep in mind this is an RF ground we are talking about, so the length of the wire from the tuner/rig to the ground is important. If it starts to become any part of a wavelength at the operating frequency, it quits being a ground and starts being a reactance, even if what it is attached to is an excellent RF ground.
This is not too much of a problem at 80 meters or 40 meters, but at 15 meters just six feet of wire is already near an 1/8th of a wavelength. So for the higher frequencies the ground lead wire needs to be kept short. I try to get as close to the fire hydrant as possible and wrap any excess ground wire around it after I apply the clamp.
The worst possible case would be 1/4 wave of wire attached to a good ground. It would be a very high resistance at the rig/tuner end of the wire! That innocent looking 12-foot ground lead could be a real problem on 15 meters. Surprisingly, if the lead is an electrical quarter wave, it would work MUCH better if it were not connected at all at the far end, and used as a counterpoise. If you have to be further away, you want a 1/2 wavelength of wire to the ground.
A simple test of ground effectiveness is to tune up the antenna and then try to put your hand on top of the tuner while watching the SWR meter. If it changes, you have RF on the rig/tuner etc. If it ignores your hand, you are either well grounded or lucky and happen to have a current loop at that point in the system.
This assumes QRP power levels. Be very careful about an RF hot rig/tuner at 50-150 watts or more. As a novice I had a tuner on a long wire and a rather lousy ground. The tuner had a loading knob with a metal set screw in a plastic knob. I tried to tune up on 15 meters with about 40 watts of CW. The ground lead was just the wrong length and RF backed up badly.
The shaft of the capacitor got a very very high RF voltage on it from all the reactance. My thumb was pressed against the knob right over the setscrew as I adjusted, trying to get the mess to load. Afterwards, there was an 1/8th inch white dot in the center of my thumb. A few minutes later there was a large amount of pain Eventually, the plug of dead flesh fell out leaving a hole almost to the bone! A painful lesson in the dangers of RF backing up on rig/tuner in the shack and inadequate grounds with end fed wires.
If you are stuck on the second floor of a wood frame house, you can often use a tuned ground wire. A half wavelength of wire will replicate the impedance at the other end. At 15 meters that is about 23 feet. Also there are ground wire tuners available that series tune a random long ground wire to series resonance at any operating frequency. These can be extremely useful where you have to go down the side of a building to an outside groundstake or water faucet two or three floors below. The tuner will allow you to make any length of wire look like the "magic" series tuned length and get rid of any reactance due to the wire length.
But, keep in mind that such a wire will radiate and that radiation will be added or subtracted from the radiation from the antenna, changing its pattern, maybe for the better, probably for the worse.
In fact you can use a series of quarter wavelength wires, one for each band, attached to your tuner to fake a ground. These should be insulated wires since they will be hot with RF. They will be REALLY hot at the far end! I use teflon tape over the end and slip the ends in a section of insulating tubing, leaving the tubing going several inches past the end of the wire. These counterpoise wires can be laid about the floor of a condo apartment many floors up, or wherever you need to fake a ground.
One long one can be used with a ground wire series tuner like made by MFJ Enterprises and usually called an "artificial ground". Keep in mind the wires must be left floating at the far end in this particular application. In this case the artificial ground will be tuning them as an odd quarter wave multiple, not a half wave multiple as is the case if the end is grounded.
The artificial ground tuners can also be used with grounded wires that have to go a long way to find a decent RF ground, such as dropping out a second floor window to a cold water faucet outside at ground level. In this case they will tune the wire to an electrical 1/2 wavelength and effectively "eliminate" the wire length from the circuit. Though there still can be radiation from this long ground lead to alter antenna patterns, etc.
All these mechanisms replace the missing part of antennas like the end fed random length wire and the quarter wave vertical or the end fed half wave wire. If possible a symmetrical antenna like a center fed dipole, either horizontal or vertical, or a center fed symmetrical random wire using good twin lead or ladder line, will decrease the demands on the station ground and any radial system.