|High RF Voltage||The Cure|
|High Above the Ground||Solution|
|The Shocking Truth||Impossible Situations|
|Ground System Test||Counterpoise|
|Artificial Grounds||Testing a Counterpoise|
|The Ground Loop Solution||Line Isolators|
This article contains excerpts from a book that I am writing, entitled, "Antennas....from the Ground Up!©" ".... From the Ground Up!©" is a series of books I am preparing on various antenna topics.
From the telephone calls we receive, many of you are having problems with RF ground systems. RF ground? Yes, most of us have ground systems that provide adequate DC grounding. Unfortunately, a good DC ground system may not be a good RF ground system. In fact, you may have an 'UN-GROUND.'
UN-GROUND? Absolutely. There are situations where your ground system may actually un-ground your station. The reason lies in the fundamental difference between DC and RF circuits.
Any wire will have inductance and therefore, inductive reactance. The longer the wire, the higher the inductive reactance and the higher the opposition is to the flow of RF current. The fatter or larger the wire, the lower the opposition to the flow of RF current. The effect is similar to the DC resistance of a wire. The longer the wire, the higher the DC resistance will be. The fatter the wire the lower the DC resistance for the same length wire. There is an important 'however,' that we must consider. First, when RF is applied to the conductor, the RF travels only on the surface of the conductor. Surface area is the important factor here, not the diameter or thickness of the conductor. Another major difference between DC and RF current has to do with wavelength. The wavelength of DC current is infinitely long. This is not the case for RF. When the XL (inductive reactance) is measured along the length of a wire, the magnitude of XL (the opposition to RF current flow) varies from very low to very high values. It continues to alternate between low and high values in cycles that have a direct relationship between the length of wire and the frequency. DC resistance, on the other hand, has no cycle. It simply increases linearly with the length of the wire.
When measuring XL, its value is very high when the length of the wire is around one-quarter wavelength long. Increasing the length wire to one-half wavelength, returns XL to a low value.
The length of the wire does not have to be very long for this effect to be observed. For example, at 28 MHz an 8' ground wire (or any wire for that matter) is approximately one-quarter wavelength long. If this 8 foot long ground wire connects your 10 meter rig to your ground system, the ground wire may actually prevent RF from traveling to ground. This is an UN-GROUND!
Why? As illustrated above, the inductive reactance of wire that is one-quarter wavelength long is very high and impedes RF current flow (thus the term - impedance).
On other bands, where the length of the wire is not an odd multiple of a quarter wavelength long, the inductive reactance (XL) is at some intermediate value.
Figure 2 shows a grounding diagram of a typical ham station. There is a heavy ground strap running along the back of the equipment. The ground strap eventually reaches the earth ground system, a ground rod, through a heavy gauge copper wire 11 feet in length. The ground connection for each piece of equipment goes directly to the heavy ground strap that runs behind the station equipment. The antenna is a ladder-line fed, 80 meter dipole used on all bands. The ladder line is brought directly into the operating position where it connects to the balanced output of the transmatch. The ladder line is about 60 feet long and goes directly to the antenna, but passes very close to a metal rain gutter. Such a station should be effective and trouble free. Unfortunately, this station is experiencing problems on the higher HF bands. There is RF feedback distorting the transmitted signal and there are some TVI and RFI problems. What could be wrong?
If we tune up on 20 meters, the 80 meter dipole becomes a 20 meter, center-fed, two full wave antenna. The feedpoint impedance is around 4500 ohms. The length of the ladder line feeding the antenna is about one wavelength long. It is a characteristic of transmission line that it will duplicate its load impedance every half-wave along its length. That being the case, the very high antenna feedpoint impedance appears right at the transmatch's output terminals. However, before reaching the transmatch, the ladder line runs very close to a metal rain gutter. Feedline balance is upset, and it begins to radiate at that point.
The transmatch uses a voltage-type balun to create a balanced output. Baluns do not work well in high impedance circuits, and voltage-type baluns are especially bad in this application. With a high impedance load, the voltage balun's core will saturate even at moderate power levels. Output balance is poor. Voltage-type baluns provide the best balance when feeding matched loads. This contributes to additional radiation from the balanced line.
In this illustration, we have several problems, each compounding the other. First, all of the ground system and ground loop problems still exist, but we now have a transmatch balun that is saturating and generating high level harmonics. Signal distortion may be noticeable because the balun is no longer operating in its linear region. The ladder-line is not balanced so it radiates and the equipment at the operating position becomes part of the antenna system. Here's a real shocker! There is RF all over the equipment. The Microphone is biting your lips. Your computer crashed. The packet TNC will not talk to you anymore, but none of this matters because the station power supply shut itself down and you are off the air. Sound impossible? Unfortunately, it's not, this is a true story and this isn't the end.
The ground wire is about 11 feet long. On 15 meters, this length is almost exactly 1/4 wavelength. As we discussed on the first page of this article, a length of wire or coax that is 1/4 wavelength long is an impedance inverter. One end is at low impedance, so the other end presents a high impedance to the circuit connected to it. In other words, the ground wire is near zero impedance at the ground end, but due to the impedance inverting characteristic, the station equipment 'sees' a very high impedance at the equipment end of the ground wire. In effect, the equipment is UNGROUNDED at high RF frequencies. Refer to figure 1.
On 20 meters, the 11 foot ground wire is .15 wavelengths long. Referring to figure 1 and interpolating between zero and 1/4 wavelength, the inductive reactance of the ground wire is still quite high. To our station equipment, the ground wire simulates an inductive reactance in series with the resistance of the ground wire. This is illustrated by the coil LS in figure 2. We'll disregard the wire resistance.
Without getting into great detail, let's just agree that is would be better if the station had a direct, low impedance path to ground. This is not the case in this illustration. The path to ground is a high impedance on the higher frequency bands. In fact there are alternative grounds available to the station equipment. They may present a lower impedance path to earth or may act as a counterpoise. Unfortunately, one of those ground systems is through the electrical power lines at the operating position. RF, seeking a ground path, may have to pass by or through several electronic appliances that would perform better if they were isolated from your transmitting equipment.
Due to the inductive reactance of the ground system, none of the equipment in this station is effectively grounded on the higher HF bands. If an RF potential exists on the station ground system, the entire station may 'float' up to that RF potential. Thus, the ground reference is actually several volts above ground. All sorts of RFI problems can be the result including RF feedback into to station microphone, computers, TNCs, power supplies, etc.
Solid state equipment is especially sensitive to ground problems.
Solid state equipment is especially sensitive to ground problems. Each piece of equipment in figure 1 is interconnected by two ground paths, a ground strap and the coaxial cable that interconnects the equipment. The two paths form a ground loop, as shown in figure 1. Since there is high system gain involved from the millivolts of the transceiver's input circuits to the kilovolts of the linear's output circuit, ground loops can be a serious problem. It's even worse if the ground system is ineffective and the entire station is 'floating' above ground. Breaking the ground loops can lead to the solution to long unsolved RFI problems.
Have you ever calculated what the voltage across a 4500 ohm reactive load is at 1.5 KW? It is more than a few volts. Actually, it's a few thousand volts. It's unbalanced, and it's looking for somewhere to go. As we predicted in previous paragraphs, the antenna feedpoint impedance and corresponding high RF voltage is transferred directly across the output terminals of the transmatch. Several thousand volts of RF is only a few feet away and at RF, the station is poorly grounded.
I'm not going to bore you with a lot of math, but let's simplify this situation to a simple series circuit. In figure 3, the antenna, transmatch, and ground system are represented by a simple voltage divider. This will allow me to illustrate what is happening to the ground bus in the ham shack.
First, let's assume the voltage at point 'A' on the transmatch is 500 volts. It is actually much higher. The impedance at the output terminals of the transmatch is 4500 ohms and the reactance of the ground system is 500 ohms. I did not calculate the value for the ground system, the 500 ohms is for illustration.
Reducing the problem to its simplest terms, we have a 4500 ohm resistor in series with a 500 ohm resistor. The ground system is the tap between the two. In this example, if there is 500 volts at the transmatch, the station ground system will 'float' above earth ground. The potential is about 50 volts. Your ground system and all your equipment, in effect, has 50 volts of RF applied to the equipment grounds. This is just like having a 50 volt input signal if the input circuits were at ground potential.
Another way to look at this problem is to visualize the antenna and ground system as a big coil that represent the inductive reactances of the ground system and transmatch. The antenna is at one end of the coil and the ground is at the other. We are tapped several turns up the coil. The only way to keep RF off the station equipment and station ground is to move the point were the rig is tapped into the coil closer to ground. The higher the impedance of the ground system, the higher up the coil the 'tap' is located.
Of course, it's never this simple. My numbers are only representative, but they do serve as an illustration. The RF voltage on the station ground system does reach very high levels under some circumstances. I have had hams tell me of severe RF burns and visible 'arcing' from microphones, equipment chassis, ground busses. Obviously, at these levels there will be terrible problems, but what happens when the RF voltage on the ground system is only a few volts? You may not know that it's there, but with solid state equipment, there can be problems.
There are some symptoms that may indicate the existence of station grounding problems. A list must include such obvious things as 'mic bite,' or a tingly feeling when touching metal while transmitting. A less obvious symptom is transmitted signal distortion due to RF feedback. RFI and TVI can often be traced to grounding problems. Here are a few other observations that were the result of the UNGROUND.
(1) Two SWR meters, one in your transceiver and the second in the transmatch that are in wide disagreement. This assumes that both meters are accurate.
(2) A change in indicated SWR when the station ground system is temporally disconnected from all equipment.
(3) A change in indicated SWR reading after adding a 1/4 wavelength counterpoise in parallel with the station ground system. Information on making a counterpoise is covered later in this chapter.
(4) Different SWR readings or differences in received noise levels when using two different transceivers.
(5) Adding a Line Isolator at the output of your transceiver changes the drive to your linear, alters meter readings, necessitates changes in transmatch settings or results in a different SWR reading on either the transceiver's or linear's watt-meter.
If any of these tests indicate there is a ground or ground loop problem, there are several things you can do. Eliminating a ground system problem, may clear up both existing and potential RFI problems.
Fortunately, under most circumstances we do not have severe problems with our ground systems, but there may be symptoms that go unnoticed.
Tracking down grounding problems is most often a tedious process. Adopting a step-by-step approach will produce the best results.
Here is one procedure you may want to try if you suspect you are a victim of an UN-GROUND:
Using this procedure for hunting an UN-GROUND or solving a RF feedback problem requires several (4 to 8) MFJ-701 'break-apart' toroids. They are available from the RADIO WORKS.
After completing each step, reconnect power to the rig, go on the air and see if the symptoms persist. Unless the symptom is eliminated, continue with each succeeding step.
[ ]Temporarily disconnect the ground wires from all equipment. Make sure that a shock hazard does not exist when doing this.
[ ]Disconnect all leads to ancillary equipment.
[ ]Ground only the antenna tuner or transmatch
[ ]Wrap the coax that connects the transceiver's output to the linear amplifier or tuner around a MFJ-701 core following the instructions supplied.
[ ] Install a MFJ-701 on the output coaxial cable of the linear amplifier.
[ ] If the antenna in use is coax fed, wrap this coax around a MFJ-701 toroid.
If the antenna in use is fed with a single wire, ladder-line or open wire, it may be impossible to keep RF out of the shack. Do not wrap around a MFJ-701 core.
[ ] Wrap the power cords to all equipment around MFJ-701 cores.
Determine the effect of the following. Remember to evaluate any improvement in the RFI problem after each step.
[ ] Reconnect all ancillary equipment
[ ] Reconnect the microphone
[ ] Reconnect all control cables.
| The Cure,
If the problem worsens, use a MFJ-701 core and wrap the appropriate cable around the problem cable.
[ ] Reconnect the ground system, grounding each piece of equipment independently to a single, central ground point. If a transmatch is used, it should be the central ground point.
Having completed these steps, there should be a noticeable improvement in the symptoms previously observed. If not, the problem is so severe that you will need to follow the suggestions in one of the RFI handbooks available from the RADIO WORKS' library.
[ ]Remove the MFJ-701 cores, one-by-one, making sure the problem does not return. This procedure will confirm the specific source of the problem.
If a change in symptoms is observed when connecting or disconnecting the ground system, follow the suggestions for installing an effective RF ground system that follow.
If placing MFJ cores on one or more of the coaxial cables that interconnect the transceiver, linear, and transmatch is effective, INSTALL LINE ISOLATORS in place of the MFJ-701 cores. 4K-LI and T-4 LINE ISOLATORS are much more effective than the MFJ-701.
If placing MFJ cores on one or more control, interconnect cables or power cables proves to be effective, permanently install the MFJ-701 toroids on those cables.
In most installations, it is a good idea to install the LINE ISOLATORS even if grounding problems are not evident. 4K-LI and T-4 LINE ISOLATORS are very effective in RFI prevention.
The station ground must provide both effective DC and RF grounding. Creating a good DC ground is not a problem, but an effective RF ground must be carefully planned.
The ground system should generally follow these suggestions:
(1) The ground wire should be as short as possible, preferably much shorter than a quarter-wavelength long on the highest frequency band operated.
(2) The ground wire should be very large. I sometimes use the braid removed from a piece of RG-213. Better yet, use one or more lengths of 1/2" or 1" tinned braided strap.
(3) Clamp this short, heavy ground wire to your ground rod(s) or radial system.
(4) Use several different lengths of ground wires in parallel, each connected to a separate ground rod. This provides multiple, parallel ground paths.
SIMPLE - a single ground rod driven into the earth just outside the ham shack.
INTERMEDIATE - Several ground rods, connected in parallel with very heavy wire or braided strap.
ELABORATE, and very effective -
25 short (6-12") ground rods spaced approximately 4' apart and interconnected in series by a 100' length of heavy braided wire or wide copper strapping. This system is very efficient. The original design used stainless-steel pegs for ground rods and stainless steel wire to prevent efficiency reducing corrosion. Copper will loose it effectiveness after a few months months or years due to the oxidation of the copper surface. Regular ground system maintenance is necessary.
What can we do? A lot, but all the explanations and details deserve an entire chapter or a good lecture at your ham club. Her are a few quick suggestions:
(1) Lower the ground system impedance.
a. Use multiple ground paths
b .Install a radial system
c. Use heavier ground cable or braid
d. Shorten the ground wire
e. Install a counterpoise system
f. Be sure that the ground system is not resonant on any band.
g. Use a MFJ-931 artificial ground.
h. Eliminate ground loops with Line Isolators
(2) Lower the level of RF voltage on the ground system:
a. Improve the installation of the balanced line to keep it balanced.
b .Change the length of the feedline. Don't use feedlines that are near 1/4 wavelength long.
(3) Change antenna systems
a. Closed loops - their impedance values stay much lower than open antennas and loops operated on multiple bands.
b .Use trap antennas for multiband use.
c. The CAROLINA WINDOM, CAROLINA BEAM, and SuperLoop are high performance, multiband antennas that keep the impedance excursions under control and the feedline SWR low.
Of course this isn't the end of the problem. The antenna was changed to a CAROLINA WINDOM, the ground system improved and 99% for the problems were gone. However, a few potential problems remain. You may not know you have any RFI problems until you install accessories, like a packet TNC and a computer .
There are circumstances where a RF ground is simply impossible using conventional techniques. Driving a ground rod into the ground and running a 25 or 30 foot hunk of ground wire, no matter how heavy gauge the wire, it is just not going to work. The length of wire is much too long. There are alternatives.
If you cannot get close enough to earth to run a very short ground wire and install a good, quality ground system, try a counterpoise. A counterpoise is the way vertical antennas, mounted high in the air, achieve an efficient ground system.
In its simplest form, a counterpoise can be a single wire, one-quarter wavelength long or just slightly longer. For best results, use a separate wire is required for each band. If you really want to get elaborate, use two or more wires routed in different directions to make up your counterpoise. The wires can be close together, insulated and routed in a convenient way around a room.
160 meters 123 - 136 feet
80 meters 65 - 70 feet
40 meters 34.5 feet
30 meters 24.3 feet
20 meters 17.3 feet
17 meters 13.5 feet
15 meters 11.6 feet
12 meters 9.8 feet
10 meters 8.6 feet
As you can see, from this table, the length of a counterpoise can be quite long on the lower bands. Where do you put 66 feet of wire? Before I answer that, lets look first at a suggestion for making the counterpoise for multiple bands.
A multiband counterpoise consists of several separate wires, each cut to the proper length for a single band. You may be able to eliminate some counterpoise wires for bands that are harmonically related in odd multiples. 15 and 40 meters or 80 and 30 meters are examples.
So now that you have the counterpoise made, what do you do with it? If you are installing your counterpoise, you may want to hide or camouflage it. It can be routed under carpets, along baseboards, out a window and down the side of the building. I have heard of some industrious types who took up the floor molding, laid the flat cable along the floor and then reinstalled the molding. I prefer the under-the-molding or under-the-carpet installations as the final location after the counterpoise is working properly and is ready to be permanently installed. Before permanent installation, we have to make sure the counterpoise is working or even needed.
Tune up your rig, but leave the counterpoise disconnected. The station should experience what ever problem brought you to the point of building a counterpoise in the first place. What ever the problem, RF in shack, mic bite, flashing panel lights on the equipment, whatever, you will still have the problem. Note the severity of the problem in some quantitative way so you can tell if the counterpoise makes a difference. Note the SWR readings, plate or output transistor collector current on the rigs meter. Note the ALC reading.
Connect the counterpoise and note changes. If luck is with you, there will be an improvement. Note that the counterpoise was cut slightly long. If there is an improvement, try shortening the wire cut for the band you are using by rolling it up for a short distance. If there was further improvement in the problem, continue lengthening and shortening until the ideal length is found. Repeat for other bands.
When tuning the counterpoise, it is very important that the counterpoise is very close to in final, installed location. If you are going to run it along a baseboard, that is where it should be located during the test. If it will be installed under a carpet, do the testing with the counterpoise on top of the carpet. Not only will the location of the counterpoise affect its tuning, you will have the opportunity to see if a particular location makes the problem worse. In that case, you will want to run the counterpoise in some other direction.
There are other ways to tune a counterpoise. It may not be necessary to bother. If you are putting in the counterpoise system as a preventive measure, cutting the wires to 1/4 wavelength is a good place to start.
The best way to set up the counterpoise is with a MFJ-931. Buy one or borrow one if you can. The MFJ-931 is a series tuned circuit that resonates nearly any length of counterpoise or ground wire. This makes the ground appear to be a very low impedance at the rig though the length is not ideal. With the '931 you may be able to get by with just one or two lengths of wire for your counterpoise. This saves a lot of work and makes the counterpoise easier to hide.
as a preventative measure
Most cliff dwellers (people who live in high building with dense populations) want to avoid even a hint of any TVI or RFI problems. Some of them will install the counterpoise system, use good low pass filters, Line Isolators, and every other RFI reduction trick they can think of. I guess, it's like the old saying, "an ounce of prevention......."
A ground system test
OK, so you have a good ground system set up, all the ground wires are short, all the ground loops have been minimized. Can the system be improved? Probably, and here is a quick and simple way to find out. This test is especially good for checking out radial systems for verticals and ground systems on boats.
First, purchase two, inexpensive, 75' rolls of aluminum foil. Unroll about 8 feet of foil from each roll and lay in on the ground, the rolls forming a 90 degree angle to one another. Twist the first foot of foil into a thick, aluminum wire. Then twist the two twisted ends of the foil together. Use a short clip lead or other attachment method and connect the foil to the ground system. Even better, connect a length of ground braid and route it directly to the point on the transmatch in the shack. Measure your antenna's SWR with and without the foil ground hooked up. It shouldn't make any difference. If it does, then you need a better ground system. Repeat this procedure for each band, unrolling the length of the foil so it is 1/4 wavelength long for the band being tested. If your ground system is working well, there will be no difference in SWR readings.
There are other ways to make this measurement beside looking for changes in SWR. If you have an RF ammeter, (the MFJ-931 has one built-in) connect it in series with the station ground at the transmitter. Use the aluminum foil procedure outlined above. Any change in RF ground current indicates an inadequate ground system. The MFJ RF 'Current Probe' should work as well and does not have to be placed in series with the ground systems.
The foil technique works well with vertical antennas. Roll out a coupe of lengths of foil for a particular band, connect to the base of the vertical in parallel with the radial system. If the SWR changes, the radial system could use some improvement.
Note: With an improved ground system, you may see the SWR rise slightly. Remember, a properly installed 1/4 wave vertical with low I2R losses will have a feedpoint impedance around 30 - 35 ohms. The lowest SWR you will see if everything is working perfectly is 1.5:1. Trap verticals are certainly no better than a good, full-size, 1/4 wave vertical. So, don't be surprised to see the SWR increase as the ground losses are reduced.
A tuner for a counterpoise or other un-grounded grounds.
This neat little box was designed specifically for use with a counterpoise (i.e., a wire or wires substituted for a ground system). The MFJ-931 tunes the counterpoise to series resonance which in turn, translates into improved ground efficiency and improved antenna performance.
The usefulness of the MFJ-931 is not limited to resonating a counterpoise. It can be a great help to those of you with radio shacks on the second or higher floors of a building. For you, and even many ground floor installations, a very short ground wire is not possible. The MFJ-931 helps solve the problem of long ground runs by tuning your ground system to series-resonance. The impedance of a series resonant tuned circuit drops to a very low value when resonance is achieved. In effect, all the reactance is 'tuned-out' of your ground system.
The MFJ-931 is adjusted by monitoring the RF current flowing in the ground system as measured by a special RF ammeter located on the front panel of the -931.
This little box, while a very useful device, is not a substitute for a well designed and properly installed ground system. For those not able to establish a proper RF ground system, the -931 is a very useful tool for solving problems resulting from an UN-GROUND.
Multiple ground loops around various pieces of equipment can cause all sorts of problems. Rather than getting deeper into a technical discussion, let's just try to avoid problems before they cause trouble. Solving the ground loop problem may be as simple as adding 'Line Isolators' in series with the interconnecting coaxial cables between station equipment.
First, eliminate the heavy copper strap running along the back of the station equipment. Use the transmatch as a common ground point, 'Ground Central.' The heavy gauge wire from your outdoor ground system will connect directly to the 'common ground point' on the back of the transmatch. Each piece of equipment will then be connected directly to the 'common ground point'. Actually, each piece of equipment is already connected, in a round about way, to the transmatch through the various pieces of coax that interconnect station equipment. Of course, it is this "round about way" that contributes to the ground loop problems. We can't eliminate the ground braid on the coax, but we can break up the external ground loops with LINE ISOLATORS.
The Line Isolator setup in figure 4 works well for most stations. Customers report that Line Isolators inserted in series with the cables interconnecting the transceiver, linear and transmatch have eliminated stubborn RFI problems that resisted being solved by other means.
Placing a Line Isolator at the output of the transceiver or linear amplifier, prevents RF from traveling along the outer surface of the coax's shield. Any RF current flowing on the coax braid that can be radiated or coupled to other equipment is forced to ground by the very high impedance of the Line Isolator. RF current takes the path of least resistance. Of course, the Line Isolator has no effect on the signal traveling inside the coaxial cable.
The Line Isolator installed in series with the transceiver and linear amplifier helps the transceiver's output filters perform effectively by breaking a secondary (leakage) path. As in the example above, the ground loop path to the linear is eliminated.
It's an idea worth a try.
The 4K-LI or T-4 is not a substitute for good Low-pass filters. Both low-pass filters and 'Line Isolators' should be used together for maximum effectiveness.
This article first appeared in the RADIO WORKS' Reference Catalog, page 78. Copyright 1992. Updated, February 5, 1997.