Practical Antenna Theory

(Part 9)
by AD5XJ, Ken Standard
ARRL Technical Specialist

This series of articles deal with the more technical aspects of antennas and the practical theory behind their use. You will find this guide to be a pragmatic approach to antenna theory. The views expressed here are the author’s and are not meant to be the definitive text on the subject. You should also be armed with the knowledge that there may be changes from time to time where technology advances and the author deems it approprate to update the text.

You should do your won research and fill in where there are “learning gaps”.

Much of the material in this series is derived from publications by the ARRL. Namely,

The ARRL Handbook ®


The ARRL Antenna Book ®.

You should consider those publications reference material in this cursory study of a very complex subject. The most recent editions are available from the ARRL Store online and can be seen at many hamfests, but may vary in quality from used to mint conditon, over many past versions. Any should be considered a valuable addition to your reference library on antennas.

I hope you enjoy this. Please email me if you have questions or would like to see a particular subject enhanced.

Multi–band Antennas

In our discussion so far, we have been concerned with single band or single element antenna types. Quite often, it is convenient and desirable to operate multiple bands with a single antenna and coax. In order to do this successfully, a number of factors are to be considered in design and construction or purchase of a multi–band antenna.

The non-resonant Sloper Antenna

diagram courtesy Alpha Delta Communications, Inc.

While horizontal dipoles required two large support masts, this antenna type only needs one large mast. It is therefore widely used by radio amateurs with limited space. In particular for low frequencies this antenna form is interesting. The angle of the slope is usually between 45°–60° and the lower end of the wire is at least 1/6 wavelength above the electrical ground. Due to the low–angle radiation pattern this antenna has, it performs well for long distance QSOs.

The dipole is typically fed with a coaxial cable in the center. At least ¼ of the wavelength of feedline must be at 90° angle to the antenna. It is also possible to feed the antenna asymmetrically. Asymmetrical meaning only one side of the dipole is in the air, and the other side of the dipole is provided by another structure (like the pole or tower it is mounted from). In such a configuration, the coax must not follow the angle of the wire. It must be the same as the alternate element (usually the tower) and move away from the wire element at at least 90° separation.

The reason for this is simple. The coax may very well become part of the antenna acting as part of the support structure (and by definition, the second antenna element). Care, then, should be taken to reduce the amount of common–mode current allowed on the coax shield and possibly entering the shack. The use of so–called "bead" chokes is highly recommended at least at the bottom of the tower or support pole and surely before entering the shack.

This requirement is somewhat reduced if the sloper is a full dipole with both elements being fed at the center and the down–lead positioned at a 90° angle from the slope.

However, as we discussed earlier, there may still be the need for a balun at the feedpoint to avoid unwanted common–mode currents due to unequal loads on each side of the dipole. The sloper type antenna is much more complex that it would appear on the surface since everything around it affects feedpoint impedance. Your particular installation should be evaluated carefully before making gross conclusions. If in doubt, get an Elmer involved. — adapted from Wikipedia

The Multi-Band Vertical

Let’s take the case of the 6BTV antenna from Hustler. This is a very good example of a well designed multi–band monopole vertical antenna. Whew! What a mouthful. It is also a good illustration of how multi–band antennas in general theory are constructed, whether vertical or duplicated for horizontal orientation as a dipole.

The overall length of the antenna is 25 feet. With just a little calculation, we can observe that it is a shortened vertical for the bands below 15 meters. It is a 1/4 to 5/8 wavelength vertical for 15 meters and above. So, how is this possible?

This type of antenna uses resonating elements along its length to either load a portion of the element or electrically eliminate part of the length. In the case of the lower bands (80 thru 20), the resonating components provide loading that causes the shortened vertical to appear to be electrically much longer. This technique was discussed in an earlier chapter. For the bands 15 meters and above, the resonating components act to “cut off” part of the antenna so that it becomes electrically the right length. These components are often called “traps”. This name originates from the characteristic nature of parallel resonating circuits. The flow of current in a parallel circuit is mainly within the parallel components and not in the series connection. So RF current flowing along the antenna element will become “trapped” in the parallel resonators and not allowed to flow further down the length of the element to the end. The reality is that some of the RF does flow the full length, but at a dramatically lower current. The lower frequencies are allowed to flow the full length of the element but meet considerable loading at key points in order to maximize current flow in as much of the antenna length as practicable. Since we are talking about 5 different frequency bands, this becomes a very complex “juggling act”.

The Multi-Band Dipole and Yagi

Now, consider duplicating the one element and positioning it horizontally. Now we have the multi–band horizontal dipole. Commercial versions of this type antenna are available from several manufacturers including MFJ, Mosley, Cushcraft, and M2. The phenomenon was first observed before WWII by two Japanese scientists H. Yagi and S. Uda. Hence the name Yagi Uda style directional antenna.

Positioning other resonate dipoles near this horizontal radiator, at specific distances, we can direct our signal and improve overall gain over the dipole alone. Voia La! We have the multi–band beam antenna.


In beam antennae, directivity and gain are the two most important factors of design. It is quite often the case that one is slightly sacrificed for the other in order to meet other design factors such as cost or size. That is why there is no less than 30 different 3 element Yagi designs available for the amateur. Each major antenna manufacturer has one or more of these designs for various frequency ranges – single band and multi–band and with differing gain and directivity characteristics.

A unique variation of the multi–band Yagi is the log periodic dipole array (or LPDA). This design attempts to provide a smooth impedance curve over a very wide frequency range. It is not unusual to see log periodic designs that span several bands. There are no “traps” or other resonating elements in this design. Instead, many self–resonate elements of varying lengths are combined. The variation in length and spacing follows a logarithmic progression (hince the name “log periodic”). In this design directivity and gain are compromised (vs. the Yagi design) to provide a very good match on a very wide frequency range. Commercial designs for marine and ham bands span frequencies from 4.5 mHz to 30 mHz with very good SWR. The chart below is an actual graph of SWR for a commercially available log–periodic.

Courtesy M2 / Array Solutions 2006

It is easy to see why this type of antenna is so appealing. The SWR curve only goes above 1.8 at the very high frequencies. An antenna of this type makes it possible to operate comfortably on all but the 80 and 160 meter frequencies. The prohibitive factors for most hams is #1 Cost and #2 Size. This guy is BIG! It has a turning radius of 65 feet! And you should know that it takes the budget of a small country to buy it. Homebrew construction techniques could render a similar design cost effectively if the size could be accommodated. The SWR graph is of one particular commercial antenna spans a 7:1 frequency range. Fewer elements and a less severe frequency span would make construction simpler and probably cheaper. Similar performance should be expected from any log–periodic design. Software design programs available online, are available to aid the homebrew constructor. It should be noted that fewer elements usually result in a more dramatic SWR curve over the same frequency span of the design. Adaptation to supported wires in a frame could make a log–periodic design very affordable for 40 – 10 meters with very good results given adequate real estate to put it up. The down side is it is stationary – so you would have to construct it in the direction most used or desired.

Vertical and Horizontal loops and quagis

In an earlier chapter we discussed the traditional Yagi as an array of horizontal dipole radiators in close proximity. The proximity / gain / directivity displayed in the Yagi is also displayed with other antenna types, such as the vertical loop. An array of vertical loops in an array similar to the Yagi exhibits similar characteristics where proximity effect of the elements, increase in gain, and directivity are concerned. A comparison of these characteristics will reveal some interesting features of this antenna type. Below is a side by side comparison of the reactance displayed by a 4 element Yagi and a 2 element Quad beam.

One of the first things to notice is the dramatic difference in the range of the real values between the two antenna types. The Yagi beam ranges from about 22 ohms to 56 ohms over the 20 meter band. The Quad beam on the other hand ranges from 50 ohms to 150 ohms over the same band segment. In terms of transmission line matching, the obvious conclusion to be drawn is that the Yagi can be directly connected to 50 ohm coax with some variations in SWR while the Quad beam requires an impedance transformation to 50 ohms in order to lower the overall SWR to an acceptable level. Assuming these parameters, the next illustration compares the SWR over the band segment used for these two antenna types.

Observation of this comparison reveals one obvious feature of Yagi antennas that make them very attractive. The SWR [for the Quagi left] over the entire band is 2.0 or lower even to the band edges. This wide band performance can be one of the compromises designed into a Yagi. The Quad Beam [right] on the other hand has a distinctive “notch” in the SWR curve at the resonance point with a much narrower bandwidth below the 2.0 SWR value. This can be a good thing where adjacent band interference is common for your location. You should also note that the ultimate SWR value at resonance is almost a perfect match (1.06:1). If your common operating frequencies are in the frequency range that yields a 1.5:1 SWR or lower, This type of antenna tuned and optimized for the center of your normal operating segment can be extremely effective.

There are some myths associated with the Quad beam that we must discuss. One such myth is that the Yagi will outperform the Quad beam in all applications. One only has to look at the origins of the Quad antenna to completely break this myth wide open. The quad was invented because a Yagi used at high attitudes generates a considerable amount of corona discharge due to high voltage being present at the high impedance points along the antenna. The inventor Clarence Moore (W9LZX) correctly reasoned that a loop antenna would not develop nearly as much high voltage and consequently less corona than the dipole based Yagi (see the ARRL Antenna Book pp 12–1 20th edition). In the thin air of the Andes mountains his new antenna indeed did lower the corona discharge while transmitting.

To understand this reasoning one must recall the current distribution of a dipole antenna. Current is very high at the feed point and low at the end points. Applying Ohm’s law we can conclude that the real part of reactance is low at the feed point and high at the end. Ohm’s law would also prove that the voltage present at the feed point would be much lower than at the end point. If we apply power as a variable, Ohm’s law for power circuits again proves that it is possible to generate very high voltages at the end of a dipole while only modest increases in power at the feed point. The Quad loop antenna has no such endpoint. The dipole is effectively folded over upon itself with the ends connected together. However, the Quad loop is usually one wavelength long rather than the traditional quarter wave or half wave of the dipole.

Since its development and popularity, the controversy has raged as to which is the better performer – Quad or Yagi. The only objective way to evaluate this myth is to compare calculated and measured performance using modern methodology. Computer modeling reveals that performance related values are of negligible difference where gain and front to back ratios are concerned where bandwidth and boom length are roughly the same. The Quad has a slight (1 db) gain advantage while the Yagi always wins the bandwidth battle where both are optimized and tuned for the same band segment. The obvious differences become apparent when viewing the characteristic radiation patterns of these two antenna types. The next illustration compares the calculated radiation patterns of each over normal surroundings and no obstructions.

The obvious difference between the two are the apparent wide main lobe of the quad and the relative narrow main lobe of the Yagi. Another apparent difference is the broad rear lobes of the quad and the much smaller and narrower rear lobes of the Yagi. This difference will become less apparent when the same number of elements and boom length are maintained for each antenna type. A less obvious feature would be the elevation of the main lobes of the two antennas. The Yagi has only one main lobe at approximately 22 degrees. The quad on the other hand has two main lobes – one at 14 degrees and another at about 45 degrees. If we apply a small amount of knowledge of propagation for DX to these radiation patterns, we would conclude that the Yagi is a great performer where a signal arrives at a low angle. But when high angle signals are prevalent, the quad would out perform the Yagi. This broad DX performance characteristic is what makes the Quad Beam so popular, notwithstanding the significantly lower cost of construction. The Quad Beam has suffered some durability weakness in the past. New construction materials and techniques have made modern commercially available, and home built Quad beams quite durable in warmer climates. They are, however, still very vulnerable to heavy ice and snow damage.

The next myth to dispel is that the Quad beam performs better than a Yagi at low heights. A comparison of the two antenna types with the same number of elements and similar boom lengths would reveal that each would have roughly equal performance at the same heights. Antennas of all types suffer the effects of ground proximity.

One note is worthy of comment at this point. The Quad beam radiating element can be electrically polarized vertically or horizontally without changing the position of the array. This accomplished by feeding an adjacent side or corner of the radiator. Quad Beam arrays are constructed for one polarization or the other and rarely accomplish both with equal performance. Depending on the band, matching techniques are very similar. The feed point impedances are very similar and a balun is used to convert the nominal 100 ohms to 400 ohms or 50 ohms for the feedline if coax is used. Quad beams using “ladderline” and no balun are very common although more difficult to construct.

The Quad Beam does enjoy one advantage over the Yagi. That is that the quad element configuration provides some noise immunity from terrestrial man–made noise sources. The key word here is SOME. The horizontal orientation of the conventional Yagi antenna lends itself to receiving noise from horizontally polarized sources like electrical power distribution lines. This type of noise is marginally lower with a quad element antenna.

The horizontal loop is well known to be a very quiet antenna where terrestrial man–made noise sources are concerned. This type of antenna is of simple construction but requires considerable real estate where the lower bands are desired. It may be fed from the center of any side or from a corner, depending on the impedance desired and the radiation pattern selected. It can be of long wire or current loop design (both feedline connections to the antenna).

For instance a horizontal loop antenna for the 80 meter band would be 259.2 feet long, or a square of approximately 62.5 feet to a side. The horizontal loop is very sensitive to ground proximity so it must be located at a height that produces the desired radiation pattern and feedpoint impedance. The illustration below is for a horizontal loop optimized for 20 meters at the optimal height and 2 meters higher

Either case presents a manageable impedance to the feedline using a 2:1 balun for 50 ohm coax. Other heights may be chosen for other feed line impedance match situations. However, impedance matching is only one compromise we must make for this type of antenna. The next illustration shows the difference between the radiation pattern of the same antenna at the two meter difference in height mentioned.

The radiation pattern varies significantly as does the feed point impedances. Note: the oblique angle of the radiation pattern is due to the position of the feed point – a corner. You must make your height selection and feed point selection based on your operational practice. Obviously, an antenna designed for use in DX contesting will be of a different height than one for general communication. The point to remember from this section is that no matter what configuration the antenna may take, the ground exhibits a significant influence on the characteristics of the antenna impedances and radiation pattern.


Long Wavelength And Traveling Wave Antennas

Often amateurs are faced with limitations placed on the space available and the length of wire or number of wires that may be possible in a operational antenna at the shack. There is a particular class of homeowner that does not want to see a web of wires suspended between trees or supporting structures in neighboring property, or an unsightly tower adorned with multiple beam antennas in the air over a neighboring home. If neighborhood civility is to be maintained, compromise is often undertaken to restore peace. Astute amateurs can make compromises and yet maintain surprisingly good performance from limited antenna systems.

One such compromise could be the Long–wire antenna versus a multi–band, multi–element dipole or Yagi type antenna. Long–wire antennas are not known for their gain characteristics. But long–wire antennas have distinctive advantages over other types that compensate for this deficiency. Long–wire antennas come in a number of versions as do dipole type antennas. Among these are the random long–wire, the inverted ‘L’, the inverted ‘U’, the ‘V’ beam, the so–called “Zep” end–fed long–wire, the fishbone, the traveling–wave and beverage antenna. The last two are what is known as a terminated receiving antenna. There are differing variations of each of these long–wire antenna types, but all exhibit some similar characteristics. They are generally large ( i.e. Electrically much longer than the lowest frequency it is used for), non–resonate, single feedpoint antennae. Most are not bi–polar as the dipole would be. In addition, it will accept power and radiate well over any frequency for which its overall length is not less than about a half–wavelength. All are relatively simple to construct and cheaper to build than array type antennas.

A particular advantage of long–wire antennas over the traditional dipole antenna is the power gain over the length of the antenna (assuming a electrical length over one wavelength at the lowest frequency used). The reason is that radiated fields do not combine at a distance from the antenna in the same way as with half–wave dipole based antenna arrays. “There is no point in space, for example, where the distant fields from all points along the wire are exactly in phase (as they are in the optimum direction, in the case of two or more collinear or broadside dipoles when fed with in–phase currents).” – ARRL Antenna Book 20th edition pp. 13-1.

Also, the longer the antenna, the greater the number of significant radiation lobes and the sharper the angle of incidence (after several wavelengths). The illustration shown is a comparison between the traditional half–wave dipole and a 3–wavelength long–wire antenna. Of significant value is the number of major lobes in several directions. This multi–lobe characteristic is particularly valuable where directionality is not necessarily desired. The dotted line indicates the radiation pattern from a conventional half–wave dipole for the same band. One should note that the power gain of the four broadest major lobes indicates better performance than the dipole. This type of antenna has characteristics that are similar to transmission lines that are unterminated – that is; the wire contains standing waves. Unterminated long–wire antennas are often called standing wave antennas. As with dipole antennas, the height above ground is the main determining factor in the radiating pattern along with frequencies at which it operates.

A long–wire antenna is normally fed by a single wire at the end, Sometimes at a current loop, but since the current loop will change with frequency, this is only the case for very narrow band operation. The end fed antenna is the only true long–wire antenna. Almost all long wire antennas are non–resonate.

The most often imitated of long–wire types is the Zeppelin or “Zep” as it is referred to. The “Zep” is a long wire fed from the end by one conductor of an open wire feed line. It is so named due to the popularity of this antenna during the days when Zeppelin airships were flying over Europe. The ships’ radio used this antenna, as evidenced by the long antenna wire trailing the belly of the ship.

On traditional “Zep” antennas, the other conductor of the ladder line is left unconnected at the antenna, purposely. This arrangement takes advantage of the transmission line effects of a long–wire antenna to provide very broadband performance with only one antenna and very low losses. The disadvantage is that the ”Zep” end–fed long–wire must have carefully routed lead–in and a balanced open–wire tuner in the shack in order to make the impedance matches and provide a method to tune the open wire line for optimal performance.

There has been an attempt to mis–name a dipole fed at the center with ladder line as a “Double Zep”. This is a misnomer because the “Zep”, by tradition, is not a dipole antenna and conversely a dipole cannot be a “Zep”. The so–called “Double Zep” is simply a high impedance non–resonate version of the conventional dipole fed directly with ladder line of high impedance (typically 400 to 600 ohms). Variations of the “Double Zep” include the G5RV, which attempts to use the ladder line as a balun and impedance transformer for coax attached to the input side of the ladder line. The performance of the G5RV has proven to be highly variable in differing installations and HF applications. This peculiarity should be factored when considering the G5RV for situations that may involve widely varying application frequencies or locations. The distinct advantage of any high impedance dipole over the conventional low impedance version is there is an increase of 3 db or more in the main lobes over the conventional dipole with much wider bandwidth where SWR is at an acceptable level. Optimized “Double Zep” designs can easily be constructed to have considerably better performance and a bandwidth SWR under 2:1 for the entire 80 meter band for instance (see the ARRL Antenna Book 20th edition pp 6–5,6–6,6–7).

Even the “Double Zep” has permutations that have been well documented. One such variation is the “JF array”, so named for the linear loading on each radiating wire that resembles the “J–pole” antenna on its side. As with other wide band designs, the bandwidth SWR below 2:1 spans most, if not all of several bands. Details of this type antenna are in the ARRL’s Wire Antenna Classics 1st edition, in chapter 10. It is a reprint from an April 1983 QST technical article. Computer simulations confirm that this can be a very effective multi–band antenna.

One interesting variation on the long–wire antenna is the terminated long–wire receiving antenna. Most of our discussion to this point has been about resonate wires. Resonance is not a necessary characteristic of a good receiving antenna. In some cases, non–resonate wires may be used to create a very stable and quiet receiving antenna from very long wires. However, this approach is not without compatibility problems. One problem to resolve is reactance at the feed point connection. This problem can be addressed if the opposite end of the long wire is terminated to ground with a resistive load. The value of the terminating resistance depends on the impedance characteristics of the long wire. We have seen from earlier chapters that a number of factors including height, length, and proximity to objects, play a factor in determining feed point impedance. If measuring instruments are available, determining the antenna impedance is much easier. A detailed explanation of the reasons for termination and other technical and theoretical explanations are beyond the scope of this book and are in the ARRL Antenna Book.

From a practical standpoint, a number of hams have reported exceptional weak signal results during contests using antennas constructed in this manner. Your particular results depends on the variable factors discussed previously. In particular, the available route the long wire will follow determines whether the directional characteristics of the terminated long–wire will be beneficial to your installation and usage.

Perhaps the best known terminated long–wire antenna is call the Beverage antenna. The received–only Beverage antenna variation is characterized by a terminated long–wire of more than 1 wavelength at a very low height, usually only a few feet above ground. The termination resistance has its ground connection at the focal point of a radial ground system. As with the conventional long–wire, this antenna has some directional characteristics that follow the length of the wire axis. Due to the very low height, the signal to noise characteristics of this antenna are exceptional, especially in the lower HF frequencies. It is a favorite of 160 meter contesters for its exceptional weak signal performance. The significant drawback to this antenna design is the considerable real estate necessarily for low frequency applications. Even the minimal size Beverage antenna for 160 meters is well over 500 feet in length and 10 feet above ground. The signal sensitivity pattern for this type antenna resembles a multi–element Yagi array in the direction of the wire path toward the termination resistance. A two–wavelength Beverage can have a maximum gain elevation angle of only 15 degrees on the maximum strength lobe. This low angle of maximum gain and exceptional signal to noise characteristic accounts for the popularity with DX contesters. Improved directivity and receive system gain can be achieved by longer wire lengths and careful optimizations.

“Even though Beverage antennas have excellent directive patterns if terminated property, gain never exceeds about -3dBi in most practical installations. However, the directivity that the Beverage provides results in a much higher signal–to–noise ratio for signals in the desired direction than almost any other real–world antenna used at low frequencies.” — ARRL Antenna Book 20th edition pp 13-21

It is not unusual to find multiple Beverage antennae aligned in oblique paths and switched from the shack, in order to provide the maximum directional sensitivity and performance from multiple directions.


Test Equipment and Analysers

There is a very real myth that persists in ham radio that the average ham cannot design, build and make meaningful measurements of antennas. This is completely erroneous. This discussion is all about how to burst that myth wide open without going bankrupt.

We have discussed several measurable aspects of antennae and feedline. We will take each and demonstrate affordable ways to measure each aspect.

The first is SWR. This is the most commonly referenced aspect of antenna/feedline performance measurements. SWR is measured by an instrument appropriately named an SWR meter. The pictorial below is of a widely used commercial SWR meter. The Bird 43 thru-line watt meter has been around for more than 50 years and proven to be extremely durable, reliable, and accurate.

The detecting elements are selectable and capable of detecting RF current in the forward (transmitter to ant) and reverse direction (ant to transmitter). The detecting element can be seen as a small disc shaped object just below the meter indicator.

Professional instruments of this type can run from several hundred dollars new to a couple hundred dollars used in mint tested condition. Great values can be obtained at local hamfests and online.

mfj wattmeter
MFJ-822 Meter

A more affordable alternative is shown in the next pictorial. It is commonly available from many ham equipment suppliers for well under $100. The obvious differences (other than appearance) are that there are no selectable detector elements and both forward and reflected power are indicated by opposing needles that cross in the middle of the meter face. This “cross needle” design is quite popular and appears in many SWR meters on the market in all price ranges. The detector circuit in this type of meter is designed to operate over a wide range of frequencies. The accuracy is not linear over the entire range. Some time should be spent verifying the accuracy with a known good SWR wattmeter.

When selecting a SWR meter, be sure you understand the frequency limitations of the model selected. The unit shown is accurate for 1.8 to 300 mHz. Many do not have that wide an operable accuracy range.

bird wattmeter
MFJ-269 Antenner Analyser

The illustration below shows the percent of power as a function of SWR. You must recall our discussions of SWR, power loss and coax loss factors for this chart to be meaningful.

This chart is included to demonstrate how much power can be lost with SWR that is uncontrolled.

However, we have seen over the last several articles that SWR is only part of the picture. We need a means of measuring the antenna impedance. This is a much more complex aspect to measure reliably and accurately. To do this an instrument known as an antenna analyzer may be used. Articles have been published in 2006 QST issues comparing commonly available antenna analyzers for accuracy. Some did not make the grade for the cost given. Those that did, are in the hundreds of dollars.

The pictorial below is of one such analyzer, the MFJ–269. The retail price on this instrument is several hundred dollars and still has some shortcomings that cannot be ignored. If your resources allow the purchase, it is quite capable and handy when doing serious antenna work.

This instrument contains an RF source for a range from HF to UHF and impedance detection circuits. It displays SWR and general impedance reactive values on two analog meters and a digital display (it indicates the magnitude only - not the + or – value of the reactance). Quite often the “average” ham does not have the disposable income to fund purchases of this type. Stay with me here, all is not lost.

mfj watt wattmeter
MFJ-269 Antenner Analyser

Another valuable tool that can be used for antenna design and test, is far less expensive and moderately accurate. It is called a noise bridge (sometimes called a noise phase bridge). Ordinarily, the ardent ham avoids noise of most types. But in this case we want noise.

The theory behind the noise bridge is similar to other types of bridge networks. One of the most familiar would be the Wheatstone Bridge circuit for DC measurements. In this circuit, four resistances are placed in a configuration that allows a meter to display when the current in all parts of the bridge are equal by indicating a zero- center reading. The diagram below is of the common Wheatstone Bridge for DC meters.

wheatstone bridge
The classic Wheatstone Bridge

The noise bridge operates in a functionally similar fashion. Instead of using a meter in the bridge, the zero balance indicator is your receiver.

While the Wheatstone Bridge measures resistances, the noise bridge measures complex impedances. This means that the relatively simple resistive components of the Wheatstone Bridge will become reactive components in a noise bridge. We have learned in past lessons that resonance occurs where reactive values become a zero sum and only resistance remains. Therefore the noise bridge can indicate not only the resistance (the R of our impedance expression R+jX) of an antenna, but the reactive components as well (the +jX part of the impedance).

The next pictorial displays a commonly available noise bridge. It has a retail price well under $100. Quite often you can come across an older one at the local hamfest for $10 to $20. It should be obvious from the labels on the adjustments of the front panel, that both resistance and reactance can be measured and read directly from the indicator knobs of the front panel (older models may have different but similar labeling).

mfj 202b
MFJ-202 Noise Bridge

What is not shown in this pictorial is the connections on the back panel. One is for your receiver and one is for the unknown (antenna, load, or coax). When the bridge is connected to your receiver and antenna in this fashion, it is possible to measure the antenna impedance directly (or coax impedance if terminated at the far end with the appropriate impedance).

To make this measurement, the noise bridge and receiver are powered on with the receiver attached to the appropriate connection and an antenna or load (antenna, terminated coax, or even a tuned circuit).

NOTE: It is important to lock out the transmit function of a transceiver. Do not under any circumstances transmit while the noise bridge is connected or you risk permanent damage to the bridge.

The first paradoxical thing to notice is that noise level indications on the receiver are very high. This is why it is called a NOISE bridge. Noise will be high for every frequency on the dial EXCEPT the resonant frequency of the load.

NOTE: When using a purely resistive load, the resonant frequency is anywhere the receiver is tuned, as pure resistance is resonant by definition. The reactance is by definition zero. Resistive loads are used to calibrate an instrument of this sort or provide a non-reactive load to coax. In which case the noise will null at the resonant frequency that is the fractional wavelength of the coax.

For reactive loads like an antenna or coax, near the resonance point the noise becomes very low. The receiver frequency should be tuned over a very wide frequency range in the same band to find the lowest noise indication using the speaker and/or S–meter indication. On some bands and with some antennas, this noise “null” could be very narrow, especially on 80 and 40 meters. It is helpful to know the design frequency of the test object so finding the null is not as difficult. You could easily hunt for the null point for a very long time before giving up unnecessarily.

Start by setting the front panel controls of the noise bridge to the expected resistance and zero reactance. Next set your receiver to the frequency band that the antenna or load is expected to handle. We will assume an antenna at 40 meters for the sake of our discussion. While listening to the speaker noise on the receiver, tune across the 40 meter band very slowly (actually begin below the band limits and tune well above to make sure you don’t miss the resonant point), listening for (and/or observing the S–meter on the receiver) a dip in the noise level. The noise may not go all the way to zero, but there should be a pronounced drop near the measurable resonance.

When this point is found, rotate the resistance knob in each direction for the same kind of null. Then the reactance knob for an even lower noise level. When no more noise can be eliminated from the receiver by tuning the frequency on the receiver, the resistance control, and the reactance control of the bridge, you can assume that the reading on the front of the noise bridge indicates the antenna impedance. The limitation using this method is the range of the noise bridge controls and the accuracy of the markings for each control. The nominal frequency range of the noise bridge is 2 to 28 mHz for most consumer grade products commonly available to hams. It is not appropriate for VHF and above nor 160 meters. The resistance control may measure from 10 to 500 ohms and the reactance control may measure +/- 150 ohms. Some newer models such as the one shown extend this range, but results become less accurate in the extended range. Your antenna could very well exceed the measurement ability of this bridge. You will, however, get a very good indication as to the possible solutions available. From the material presented so far, you should be able to take the impedance readings and null point frequency to interpret whether the antenna load should be shortened or lengthened (or values changed in the case of a tuned circuit).

If your measurements are close enough, the indications could show what the possible working bandwidth will be. The noise bridge does not however, indicate SWR for antenna loads directly. Our discussions previously indicated SWR is a result of antenna to feedline mismatch. SWR then, can be inferred by calculating the measured impedance and dividing by the feedline impedance or measured directly with the SWR meter mentioned earlier after disconnecting the bridge and substituting the SWR meter in–line. As an example, let’s say we measured the impedance of our antenna under test at the shack to be 55 –j102 on 7.245 mHz. Our feedline is RG–8U coax at 50 ohms, cut to a ½ wavelength for the center of the 40 meter band. Using some simple math we can infer a SWR some where near 1.1 to 1. This is not a precise measurement because there is a reactive component that may influence SWR to some extent (more influence the farther away from resonance and less closer to the resonant frequency), and the length of the coax may not be exactly ½ wavelength for our test frequency. The instruction sheet that comes with the noise bridge has formulas that will render a more precise impedance, taking into account the reactive values and feedline length. A few minutes with a calculator and you have an accurate, fully compensated, impedance value for either the transmitter end or antenna end. This method is for the most particular and mathematically capable of users only.

Current editions as well as older editions of the ARRL Antenna Book contain homebrew projects for a well–capable noise bridge as well as other antenna measuring instruments that can be easily built for reasonable costs. Even vintage issues of the Antenna Book and ARRL Handbook have valuable projects for the antenna experimenter that can be constructed for a minimum of cost with accurate results.

One such project is the homebrew antenna or coax impedance meter found in the ARRL Handbook 1986 (Chap. 25 p. 33 R.F. Impedance Bridge for Coax Lines). This is a direct reading impedance meter using a single frequency RF source – not wideband noise as discussed before. However, operation of this type test instrument is very similar in operation yielding similar but more accurate results when compared to a currently available noise bridge (albeit at a possibly higher cost).

The serious VHF and UHF antenna experimenter may also wish to construct a standard antenna for measurement and test of antenna gain. This antenna is built from commonly available copper or brass components from your local hardware or home improvement supplier. The standard antenna for VHF is easy to construct and is very useful when making gain and pattern measurements of VHF or UHF designs (ARRL Antenna Book 20th edition p. 25–51). The dimensions given are in wavelengths, but may be converted to metric or English measurements using the appropriate formulas for frequency to wavelength conversion. Pay particular attention to construction tolerances. Accuracy in construction is necessary to insure proper operation. Wavelength at VHF and UHF is quite short compared to HF frequencies. An error of a half inch (12.5 mm) in VHF frequencies and one sixteenth (1.5625 mm) in UHF frequencies could be fatal to the intended design. The higher the frequencies, the closer the construction tolerances must be. This is true for operating antenna measurements as well.

In this discussion we have demonstrated that with inexpensive instruments we can learn a lot about the characteristics of antenna and feedline systems and obtain acceptable results. With an inexpensive noise bridge and SWR meter or homebuilt impedance bridge, the experienced antenna builder can obtain accurate and reliable results. Antenna efficiency (i.e. The amount of signal actually radiated versus being lost to coax attenuation or ground loss absorption) can be measured or calculated, and increased through such measurements with little effort.

For the beginner, much can be learned about antennas, feedlines, impedance, and SWR without breaking your pocketbook.

73 AD5XJ





Rev. 1.00 2018-01-09 AD5XJ