Quad Phaser Network for CFA Testing
Some of the unsuccessful operation of the small compact transmitting experimental Crossed Field Antenna (CFA) may be due in part to failure to provide the correct power ratio at the correct matching impedance, along with a failure to provide the correct quadrature (90-degree) phase relationship of the R-F energy existing between the antenna elements proper. The CFA inventors have claimed that failure to meet these requirements is the reason many operators have such poor non-repeatable results with CFAs. Let’s find out about that.
In this article I describe a method of meeting all of their identified requirements I know of by presenting an antenna phasing matching network design approach that offers simple flexible adjustments for providing 90-degree relative phase signals, impedance matching for each antenna element (plus the input co-axial cable), and varying the relative power ratio between the antenna elements. I call this unit “Allen’s QUAD PHASER Antenna Matching Network”, or “QUAD PHASER”, for short. This method was shared with the “GARDS” members many months ago.
The QUAD PHASER has met with good success. The basic version was briefly referred to as Figure 2 in the November 1999 antenneX article titled “More DDA Matching Networks” by Richard Morrow, K5CNF.
Here is an encouraging quote from an e-mail I received from GARDS’ member George Sharp, KC5MU!
“The main reason for this Email is to tell you the QUAD PHASER works as advertised! Its great!”
“The bulbs light lit up at resonance and I found that when all adjustments were made for 1/ 1 SWR, on the MFJ Analyzer, they were the brightest and the field strength meter was at maximum, too. Its nice to know the 90 degree phase difference is there and not have to guess at it.”
“Thank you, ….. Old George”
Here is another encouraging quote from an e-mail I received from antenneX Publisher Jack Stone in telling about tests performed by Richard Morrow, K5CNF using the QUAD PHASER Network with a Duo-Disk Antenna (DDA)!
“Harold, yes, Richard has made several successful eXperiments & sez the Network/DDA combination is working better than anything else tried before…”
At this time I would like to use this article as a way of sharing the QUAD PHASER Network in much greater detail with all antenneX subscribers.
This unit can do a good job providing flexibility in meeting identified CFA interface requirements. But, to say Allen’s Quad Phaser Antenna Matching Network does a good job is not to say that all the various CFA shapes, sizes, and spacing configurations of the experimental antennas it drives do a good job, too. The QUAD PHASER can meet the identified interface parameter requirements. More important, the Quad Phaser can be used to remove any doubts as to the worthiness or worthlessness of a given CFA configuration or other small compact antenna under test.
This article does not address the advantages or disadvantages of CFAs of various sizes, shapes, and element spacings, that have been discussed and curssed in other articles and in the very active antenneX “Forum Discussion Boards”. The Quad Phaser is a tool to help YOU do your own real evaluation. This is a quadrature phase generator plus an antenna matching network interface between the antenna and the co-ax feed line from the transmitter. It matches the antenna’s impedance and it matches the feed line’s impedance. It is not an antenna tuner. The only thing that “tunes” an antenna is the changing of the configuration of the antenna itself.

Figure 1 – Allen’s Quad Phaser Basic Version.
The Basic Version of Allen’s QUAD PHASER antenna tuning and matching network features:
1) A simpler, efficient, easy method of generating R-F quadrature phasing, (by tuning C1 and C2 to peak resonance),
2) Independent adjustments of source impedance to matching each antenna element (by positioning output-taps on L1 and L2),
3) Adjustment of the load’s input impedance to match the input feed line’s impedance (by positioning the input-tap on L1).

Figure 2 – Allen’s Quad Phaser Expanded Version
1) Select the antenna element to receive the most power (using S1),
2) Adjust the ratio of power fed to the antenna elements (by adjusting transformer T1’s mutual coupling),
3) Select between quadrature phase lead or lag feed to the antenna element receiving the lowest power (using S2),
4) Select isolation of return path to the primary circuit (using S3).
Quadrature Phasing
A topic covered to a much lesser degree is quadrature phasing, the 90-degree phase relationship required of R-F energy existing between some of the elements of antennas at the operating frequency. It is being specified by some to be a very critical requirement for CFAs and for some of the other experimental compact transmitting antennas, also. A major purpose of this article is to increase the awareness of this different, much simpler, and considerably easier to tune method of generating a quadrature phase relationship, compared to the inefficient and very difficult to tune standard methods used previously for CFA experiments.
By using the QUAD PHASER method of quadrature phase generation that I used and described in the article “An Answer to Narrowband Frequency Modulation Reception”, published on page 28 of the February 1948 issue of QST, it is very logical that this same approach be used to also provide a different more positive, easily controllable, practical approach to generating quadrature phase for CFAs, or any other antenna system requiring it.

This approach, much to our advantage and ease of operation, requires that each of the two resonant circuits be tuned exactly to resonance (as indicated by maximum R-F voltage near the stator of each of the two tuning capacitors C1 and C2 of Figures 1 and 2). This approach does not require the tricky, awkward, inefficient, critical “off-set tuning from resonance” that has been required by other phasing/tuning networks for CFAs.
Using the QUAD PHASER method of generating quadrature phase, the correct 90-degree phase differential between the output signals will now occur at resonance!
Theory of Operation
The Quad Phaser theory is based on the fact that the magnetic flux surrounding R-F transformer T1 ( of Figures 1 and 2 ) is in-phase with the current flowing through its primary coil winding L1. (The primary can be tuned or un-tuned, however it should be tuned to resonance to yield highest efficiency.)
The Change in the magnetically coupled flux generates the energy (a new signal) in the transformer’s secondary coil winding L2, thus the new signal generated in the secondary winding has a 90-degree phase lead relative to the phase of the original (old signal) current flowing in the primary coil winding L1. Then, the inductive reactance in the secondary winding L2 delays the phase of the induced signal (as it also does in power and audio transformers that are untuned). However, this inductive reactance XL2 can be canceled by an equal value (but opposite sign) capacitive reactance XC2 introduced by tuning capacitor C2. When XC2 is adjusted to be equal to XL2 then the secondary circuit is at resonance and the true newly generated 90-degree leading phase is restored!
It is very much to our advantage that this occurs at resonance, not at an inefficient off-set frequency position. With this method, both C1 and C2 can now be easily tuned to the highest efficiency peak resonance positions while providing quadrature phase at the same time!
This method is the basis upon which the majority of FM radio demodulators, both discriminators and FM ratio detectors, found in most home entertainment receivers use to generate the quadrature phase they need in order to demodulate FM and PM signals (most FM BC stations actually use PM).
For example, a discriminator fed a limited amplitude R-F (or I-F) frequency modulated signal, changes the original frequency modulated signal into a new phase modulated signal, then combines this new phase modulated signal with some of the old un-phase shifted original signal to create an amplitude modulated signal. At this point, it is a simple task for diode detectors to rectify an amplitude changing R-F signal into an audio signal. (The ratio detector does the same task but it does it without requiring the input R-F signal to be limited to a constant amplitude, thus it can be built for less because it does not require a “limiter” stage).
The key feature to the operation of both of these types of FM demodulation is the method of converting the original frequency-shifting signal into a new phase-shifting signal having an average leading-phase of 90-degrees.
In the first two paragraphs of this section it was pointed out that it is necessary for the capacitive reactance to cancel the inductive reactance in order to recover the new 90-degree phase leading signal that was created as a result of the changing flux in the transformer. If the frequency changes even slightly the two reactances (XC and XL) are no longer balanced and complete cancellation no longer exists. When the reactances no longer cancel, a phase lead other than 90-degrees results. If the frequency shifts lower, the phase lead becomes proportionally less than 90-degrees. If the frequency shifts higher, the phase lead becomes proportionally greater than 90-degrees.
This new shifting phase-leading signal (first swinging less than, and then swinging greater than, the 90-degrees phase-lead with modulation, while averaging a 90-degrees phase-lead) is combined with some of the original (0-degree) signal, which (except when the new signal has a phase-lead of exactly 90-degrees) first adds to it a part of the time, and then subtracts from it the other part of the time, but averages neither adding nor subtracting from it. In essence, the result is an amplitude changing R-F signal.
This method of generating quadrature phase shift has been the system most commonly used in FM receivers since I was a boy (and I am 74 years old now). So, this is the simple method of quadrature phase generation that I am suggesting be used as a way to greatly simplify phasing the signals to be fed to CFA networks, as well as to other systems that need a quadrature phase. I have used this approach to generate quadrature phase for over 50-years in some other applications. So, even though this is not a new invention, it is a very valuable new application of some important old principles to making the QUAD PHASER. I have presented these old, but frequently misunderstood, principles in order for those experimenters who intend to do serious evaluations of the CFA and other small compact transmitting antennas be thoroughly aware of them, appreciate them, and use them in an informed manner.
This method of quadrature phase generation requires each variable capacitor to be easily tuned for maximum voltage across them to obtain both maximum efficiency and 90-degree relative phase at the operating frequency. A simple peak voltage indication can now be used to indicate when optimum performance of phase and efficiency exists.
The Standard CFA Way
Compare this approach to the standard awkward, difficult to tune, inefficient standard way the CFA people keep promoting to create quadrature phase. Their standard methods of generating quadrature phase for the CFA actually de-tune one circuit feeding one element of the antenna enough higher in frequency to introduce enough additional capacitive reactance to exceed the circuit’s inductive reactance, so that, at a selected operating frequency, it will yield a 45-degree phase lead. While the other circuit, feeding the other antenna element, has to be de-tuned enough lower in frequency to provide less than enough capacitive reactance to equal this circuit’s inductive reactance at this one selected operating frequency, thus leaving this circuit inductive enough to yield a 45-degree phase lag. Thus a differential 90-degree phase relationship between the two antenna elements is achieved, but it is done in an inefficient way that is so awkward that most operators give up on it.
In some of their approaches, they use a parallel LC tank circuit inserted in series with the output feed going to one of the elements. At the resonant frequency of this parallel LC tank circuit, it is actually a high impedance “rejection trap”. This method of being near to, but critically off-resonance to achieve 90-degree phase differential signals does pose problems in tuning and efficiency. The standard CFA way, for most operators, is both awkward to describe and even more awkward to successfully tune.
General Instructions
Here is a practical factor. One does not need to know which antenna element leads or lags the other in phase at this time, because both modes should be thoroughly tried and tested to determine the selection and adjustment of parameters that yields the best results for a given compact transmitting antenna configuration at the desired operating frequency. Being consistent in testing and taking notes with the assumed phase direction you chose is all that is required to achieve optimum results. Then later on, at some convenient time, the actual phase can be verified. If your Quad Phaser is built and labeled according to Figure 2, the phase should be correct. In essence, please do not let the matter of your not being able to verify phase cause a delay in your building and testing a CFA or any other compact transmitting antenna.
Throughout all of the discussion in this article, and in any following article, it should be kept in mind that all of the optimum settings, tuning, and adjustments are made to result in the highest FSM indications and in the lowest SWR indications. After the Quad Phaser has been adjusted for optimum, with the tweaking of C2 being the last network adjustment, the transmitter’s output should always be adjusted to the constant reference power level in order to maintain accurate test results. It is important to keep good test notes about each configuration and the final test results on a given frequency. Also, make a note of the antenna’s height and of the type of soil that exists under the antenna being tested when it is being tested in various locations. The extent to which soil conductivity, moisture, and weather have, or do not have, on the test CFA antenna is not fully known, so please keep good records of all possible parameters as tests are being conducted.
(Note: An R-F voltage indicator as simple as a neon lamp can be very useful. One lead of each bulb can be attached to the rotor of the variable capacitors—C1 and C2—and the other lead serve as stray capacitive R-F voltage coupling when positioned somewhere near the stator of its variable capacitor. Adjust the spacing between the open lead and stator to provide a good tuning indication of peak resonance. Other useful tuning indicators are R-F Ampmeters in series with output terminals A and B. In the absence of antenna R-F current meters, small flashlight bulbs, having a tiny wire loop connected to their terminals, can be positioned near to—but insulated from—the output terminal leads.)
Some experimenters find it is easier, for the health of their transmitters, to do the initial adjustments while listening to a receiver and observing maximum “S” meter readings, while their receiver is tuned to a constant strength local station.
With reference to Figures 1 and 2, each of the two tank circuits (L1, C1 and L2, C2) must resonate at the operating frequencies chosen by the experimenter. Both of the variable capacitors (C1 and C2) must be able to tune completely through resonance at the selected operating frequencies while loaded or unloaded in order for the Quad Phaser to have the flexibility needed to provide all of its features.
The experimenter should select a pair of the best low loss variable capacitors he has available (split stator capacitors are the most efficient), then obtain, or make efficient, large-conductor, low-loss L1 and L2 coils to resonate over his desired operating frequencies. (We are presenting the methods and concepts to provide a good method of generating quadrature phasing, plus versatility in a matching network, but not a step-by-step construction article with a parts list for each operating frequency band. The frequency is your choice and there are plenty of good reference books and handbooks to show you the amount of inductance needed for your selected frequency.)
With the Quad Phaser method, tuning C1 and C2 to the peak resonance positions will include the load reactances introduced by the antenna elements proper, so expect all adjustments to be somewhat interactive. Though not required, it is better for the coils to be identical. Even though the variable capacitors may track over the range of frequencies used without any load connected to the output terminals (A and B), I suggest no attempt be made to mechanically gang the capacitors together. This is because there will probably be large variations between the separate unique loads presented to the two output terminals by the individual antenna elements as various CFA configurations are experimented with at different frequencies.
When testing any compact transmitting antenna, the experimenter should initially position the Quad Phaser under the antenna being tested and operate it there manually until becoming familiar with its operation, rather than initially trying to remote control it. If operated close to your body, I suggest keeping the transmitter’s output power very low; only use enough power to conduct the experiment.
There are three “output-taps”: one is located on R-F transformer T1’s primary coil L1 (for the antenna element requiring the most power), while two are located on T1’s secondary coil L2 (one being used to match the impedance of the antenna element requiring less power while it has a leading 90-degree phase, and the other being used to match the same antenna element while it has a lagging 90-degree phase). The positions of these power output-taps are selected to provide the best source impedance matching position for each of the antenna elements. The farther up the coil from the chassis ground return they are positioned, the higher will be the source impedance they provide to feed the given antenna element. They will also have a secondary effect on the relative power ratio between the elements.
One experimenter reported that his CFA’s best power ratio between the two antenna elements was 1.8:1. If this is the case, then the output-tap on L1 should be connected to the element designated to receive the most power. Please notice that switch S1 is provided on the EXPANDED QUAD PHASER version, as shown in Figure 2, to allow the antenna elements to be selected, and swapped if necessary, as different configurations and frequencies are selected. The mutual coupling would then be adjusted to 0.55 by varying the effective “coupling” between coils L1 and L2 (the ratio of 1/1.8 = 0.55 ).
Better yet, experiment with determining the best cylinder to disk power ratio to best satisfy the CFA’s requirements, by adjusting the “mutual coupling” between L1 and L2, while striving for a maximum indication on the field strength meter (FSM) for each frequency and antenna configuration variation. If an optimum power ratio cannot be gone through, then, with S1, swap the output terminal connections to terminals A and B feeding the cylinder and the disk on a CFA.
Next, re-establish optimum impedance matching, then again vary the mutual coupling between L1 and L2 for maximum field strength on the FSM. If this does not provide enough field strength, then swap the major phase of the signal delivered by L2’s output-taps from leading to lagging the phase of the signal on the output-tap of L1 by 90-degrees. Switch S2, on the Expanded Quad Phaser version, is provided to allow the major relative signal phase at the output-taps of L2 to be easily swapped from “Lead” to “Lag”, in reference to either leading to lagging the signal on the antenna element requiring the most power, by 90-degrees.
Again, re-establish optimum impedance matching. The two output-taps are provided on L2 to allow impedance matching to be made optimum on the selected output that controls the major phase changes from lead to lag.
CAUTION: Avoid making any changes of any of the switches or to any coil tap positions when power is applied. As mentioned earlier, preliminary adjustments can be made while listening and watching the “S” meter on the receiver.
If you cannot adjust transformer “coupling” through an optimum position with either a lead or lag phase position with S2, relative to the antenna element requiring the most power, and the coupling position between the coils beings physically as close as possible without arcing (approaching 1.00 mutual coupling), then it could be concluded the required power ratio between the antenna elements is 1:1. (This would be contrary to the previous reports of an optimum power ratio of 1.8:1.) If you encounter a condition like this, then compensation for it can be made by adjusting the output-tap on L1 to a slightly lower impedance (lower on L1) and thereby slightly reduce the power fed to the antenna element requiring the most power, thus compensating for the transformer’s mutual coupling not fully attaining 1.00 .
Isolation
How very disappointing, and even embarrassing it would be, if we concluded, or worse yet, announced we have a special new design for a special compact antenna only to find out later that we only had the electrical equivalent of a top-loaded short whip antenna that required an earth ground (or an equivalent to it, such as a the shield of a co-axial cable acting as a radial).
As with other types of matching networks, verification of independent operation from an actual earth ground, or even the braid/shield of the co-ax cable, should still be done with a 1:1 balun having a high power rating (without even one common direct-connection) sometime during the testing process for each antenna configuration tried. If it is verified with actual isolation tests to truly be ground independent, only then should the balun be removed from the evaluation test. If a two-terminal antenna (as a DDA most likely is) is being tested (using output terminal B and the chassis ground RETURN terminal to feed the two elements) isolation can be made by opening “Isolation Switch” S3 on the Expanded Quad Phaser version (as shown in Figure 2). (Resistor R1 serves only as an electrostatic drain. Its value can be from one to five meg Ohms.) S3 is provided to make the unit a more versatile “antenna matching network”. With S3 closed (Isolation switch: “off”), short whip antennas (such as bed springs, shower rods, coat hangers, rain gutters, random wires, etc., can probably be fed against an earth ground, or radial system, or even the co-ax feed line’s shield/braid. Adjust the loading for maximum field strength and minimum SWR. If those monitors are not present at the time, you can get “in the ball park” by adjusting for maximum R-F voltage and/or R-F current at the antenna terminal, because the antenna is a fixed load that is not changed by your network adjustments.
Even though toroids contain, or concentrate, the magnetic flux better than air cores for very low power, I hesitate to operate using a toroid or powered core device with other than very low power. When a core saturates, many strange things can happen like overheating, cracking and thus completely changing it from the design specifications, generating a new family of harmonics, and even causing the insulation on the copper windings to melt and burn off.
Not many people in the field of electronics are aware of some of the other types of “active electronic devices” used in severe environmental conditions such as in some of the military computers used in rugged battle conditions and in some guided missiles. One type of very successful “active device” in this category utilized the controlled modification of magnetic core characteristics to achieve a family of operations. One of these devices was the almost mechanically indestructible “magnetic amplifier”. My appreciation of these, gained by using them, is what makes me reluctant to encounter uncontrolled problems that may be unintentionally introduced by these magnetic cores on my antenna experiments, especially when there are already enough other variables affecting the antenna that I need to become aware of.
Conclusion
The Quad Phaser is designed to provide the necessary operational flexibility to give the CFA, or almost any compact antenna under test, the benefit of a doubt. I think I have covered all possible parameter categories and have provided sufficient design philosophy about a better way to generate quadrature phase and versatility in network matching to satisfy all of the requirements for any compact antenna to be tested (with the exception of small transmitting loops and DLAs, which I plan to discuss later). If you know of any CFA antenna matching requirements that I missed, please let me know. More information about Allen’s Quad Phaser is scheduled to be discussed next month. Keep on experimenting. Maybe you can come up with a good CFA configuration! Please share your results with us!
Quad Phaser Network - Part 2
Introduction
This article expands on this simple antenna matching network design approach that offers simple, flexible, easy adjustments for providing 90-degree relative phase signals, impedance matching for each antenna element (plus the input co-axial cable), and varying the relative power-ratio between the antenna elements of an antenna under test. This Quad Phaser Network can meet all of the interface parameter requirements that have been identified for a Crossed Field Antenna (CFA). More important, the Quad Phaser can help remove any doubts as to the Worthiness or Worthlessness of a given CFA configuration under test.
This article does not address the advantages or disadvantages of CFAs of various sizes, shapes, and element spacings, that have been Discussed and Cussed in other articles as well as in the very active antenneX “Forum Discussion Boards”. The Quad Phaser is a tool to help YOU do your own evaluation of how good the antenna works, or does not work, regardless of the antenna theory involved.
The Quad Phaser is a quadrature-phase generator with an antenna matching network interface, which can be used between the antenna and the co-ax feed line from the transmitter. It can match the antenna’s impedance while matching the feed line’s impedance. Even though it would be called an “antenna tuner” by many, it is not really an antenna tuner because the only thing that actually “tunes” an antenna is changing the configuration of the antenna itself.
Theory of Operation
The theory of operation was discussed in depth in the first article. Briefly, this simple method of generating a quadrature phase relationship between its two output terminals used to feed antenna elements is based on the fact that the magnetic flux is proportional to the current flowing in the transformer’s primary coil winding L1 while the signal generated in the secondary coil winding L2 is proportional to the Change in flux and is phase shifted by 90-degrees as generated.
This method of obtaining quadrature phasing, with its simple easy tuning to maximum r-f voltage, has the big advantage over the previously used standard method that requires an inefficient hard to tune approach to generating quadrature phase by off-setting the tuning of each variable capacitor from resonance; one is off-set to a frequency above resonance and the other off-set (hopefully) by the same amount below resonance.
The important feature is that the Quad Phaser method of generating quadrature phase, the correct 90-degree phase differential between the output signals, now occurs at peak resonance. In other words variable capacitors C1 and C2 (Figures 1 and 2) are easily tuned for peak resonance, which results in the r-f voltage across each of them being at maximum when the desired quadrature phase and maximum efficiency occur.


General Information
Overall Test Parameters…….. It is not necessary to know if the phase is leading or lagging the other antenna element’s phase, because both modes should be thoroughly tested to determine the selection and adjustment of parameters that yields the best results for a given antenna under test at the desired operating frequency and location.
To simplify tune-up, keep in mind that preliminary adjustments can be made while listening and tuning for a maximum on the “S” meter on the receiver.
Throughout all of the discussion in this article please remember that all of the optimum settings, tuning, and adjustments should be made to result in the highest Field Strength Meter (FSM) indications and in the lowest voltage standing wave ratio (VSWR) indications. After the Quad Phaser has been adjusted for optimum, with C2 being the last network adjustment, the transmitter’s output should always be adjusted to the constant reference power level in order to maintain accurate experimental test results. It is important to keep good test notes and the final test results obtained on a given operating frequency about each antenna configuration, along with weather and site information.
In the interest of taking and keeping good test data L. B. Cebik, W4RNL, has contributed the following worthwhile suggestions:
“Any antenna site factors known at the time of conducting tests should be noted if by chance they may have some later significance.”
The reference to site characteristics may have at least one of the two following significances.
1. Ground immediately beneath the antenna: The antenna disk height above ground and the ground quality may (or may not) affect tuning/phasing settings for a given antenna geometry.
2. Ground conditions in the Fresnel zone: ON4UN’s “Low Band Dxing” has had some notes on this since the 2nd Ed. This region is normally beyond control, since it is some wavelengths away. However, if it is correct to say that the CFA fields are independently produced, then the characteristics of the Fresnel zone may have some bearing on the initial reflection and mixing and hence upon the phasing/tuning settings.
Thank you LB for contributing these important suggestions.
In keeping with LB’s suggestions of being aware of your average soil conductivity corrections in your area, if you live in the US, you may find information at the FCC web site: http://www.fcc.gov
Design Factors
Even though this is not a step by step construction article, several design factors are discussed in order to offer more design parameter options for the experimenter to draw from when construction is being done.
With reference to Figure 2, in order to reduce wasting r-f power in the form of heat, use two heavy-duty low-loss double-pole double-throw (DP-DT) knife-blade switches. Use one switch for S1, to select which output terminal (A or B) is fed the “highest power”, and use the other switch for S2, to select which output terminal (A or B) is fed the “lagging phase”.
While constructing the Quad Phaser, I suggest adding S3 for conducting antenna experiments in addition to the CFA. A heavy-duty low-loss single-pole single-throw (SP-ST) knife-blade switch for “isolation” switch S3, which can assure experimenting independent from earth ground and co-ax braid/shield ground without having to use an external 1:1 isolation balun transformer.
But, when conducting CFA experiments, a high power 1:1 isolation transformer (having a heavy duty core such as a T200-2) should be used initially, and S3 should be closed. An external resistor having from 1 to 5-megohms of resistance should be added between the input co-ax braid and the QUAD PHASER’s chassis ground return to serve as an electrostatic drain to prevent an electrostatic voltage build-up.
Power-Ratio……… The amount of mutual coupling between the transformer’s primary and secondary coil windings, L1 and L2, primarily affects the ratio of r-f power available to be fed to the two output terminals, which in turn is fed to two antenna elements. To allow for worse case design, it is assumed (based on reports from others) that an un-equal amount of power is required by the two antenna elements (such as a CFA’s cylinder and disk), and it would be logical and desirable to be able to accurately control the relative ratio of power provided to each element. The mutual coupling is the result of L2 intercepting a part of the magnetic field generated by L1.
The most obvious approach for doing this is to provide linear adjustable spacing between the transformer’s primary coil L1 and its secondary coil L2 while maintaining the axis of each coil mutually parallel and in the same common plane. Another method is to have the axis of each coil in the same plane but let the axis of L2 be varied from being in parallel with the axis of L1. In this manner, only one end of L2 has to be moved while the other end of L2 can pivot about a fixed point. Still another approach is to have the two coils co-axial and continue sharing the same axis as spacing between coils is varied. A variation on this last approach is to leave the spacing fixed but increase the magnetic concentration by incrementally inserting a ferrite or powdered iron rod (core) to supplement the air core, but it may be limited to low power. All of the approaches will affect tuning to some extent and require re-tuning, but this last approach will change the tuning the most as it increases inductance.
Reference marks should be provided on the selected coupling mechanism to permit repeatable setups between different steps in the experiments. Even though the primary function of this adjustment is to change the power coupled from L1 to L2, in order to provide the optimum power-ratio between the two elements of the antenna, there will be some interaction, or secondary effects, on the impedance at all of the taps, the resonant tuning of both C1 and C2, and some on the relative phase which has to be corrected by trimming C2 to quadrature phase as indicated by maximum r-f voltage across C2.
If a larger amount of power is required by one of the elements (compared to the other element) at a given operating frequency, it may be possible that the optimum power-ratio requirement may change with changing frequency. To expand on this thought to an extreme, we should be able to accommodate any such variations if they occur, even to the extent of a reversal as to which element may require the most power. The element requiring the most power should be fed by the output-tap on primary coil L1, with the other element connected to the output-tap on secondary coil L2. If such a crossover in these requirements does occur from one operating frequency to another operating frequency, then the connections to the two elements would need to be swapped. The Expanded Quad Phaser has provided S1 to make this element swap easy.
A major design goal is for the transformer to transfer as much energy as is practical from the primary coil winding L1 to the secondary coil winding L2 when the power-ratio between the two output terminals needs to be as close as possible to a 1:1 ratio. This is another way of saying the mutual coupling should ideally be able to provide coupling as close to 100% as practical, if the need exists. “If ” is use here because one report gave a requirement of 1.8:1 (or 55%) which would be easy to achieve. However, it would be logical to have L2 intercept as much of the magnetic flux generated by the primary coil L1 as possible until more confidence can be gained in the questionable (and probably varying) power-ratio requirement.
Coil Configurations……… Coils can take many shapes and positions. Sometimes they are positioned side-by-side, sometimes end-to-end, sometimes bi-filar wound, sometimes one coil is inserted inside of the other coil, etc.,.
Considering that the r-f current flowing through a coil is constant at all positions within the coil, the secondary coil winding is sometimes placed at the “cold” part of the primary coil in order to reduce the chances of the higher r-f voltage at the “hot” end of the primary coil from arcing across to the secondary coil. Also, the probability of capacitive coupling of harmonics from the primary to the secondary is greater at the “hot” end of the primary coil. The possibility of the inductive magnetic coupling of harmonics exists at the top, bottom, and at all locations coupled by the magnetic flux.
Concerning coils, since the r-f voltage increases on a tap as the tap is moved toward the end of a coil away from the coil’s connection to chassis ground return, many have falsely concluded that the coil’s current, and the resultant magnetic flux at the position of the tap, decrease. The fact is that the series circulating r-f current flows equally through the coil at all points in an LC tank circuit, including the variable tuning capacitor.
Power Loss…….. Wiper contact ohmic resistance in rotary and adjustable coil taps, as well as in variable capacitors, can be another major reason for power loss.
Using split-stator variable capacitors is highly recommend for C1 and C2 because they have improved efficiency since they do not use wiper contacts, therefore wiper contact ohmic resistance losses are eliminated.
If power loss is suspected in any circuit, one helpful test used to isolate it is to quickly feel of the suspect areas with the back of your finger, or fingers, (and with the other hand safely behind your back) immediately AFTER removing power to detect any localized heat build-up due to any power loss problems.
(Only the back of the fingers should be used to do this test in order to avoid any reflex action by the muscles closing tightly around a conductor being touched when voltage is present. But, using the back of the fingers as a sensor has a small disadvantage, as pointed out to me by one of my former college students. He told me about knocking himself out that way by hitting his jaw when he touched a live circuit that was supposed to have been turned-off! He thanked me for having taught him this method, because he thought that his electrical accident would have otherwise electrocuted him if he had turned his fingers the other way.)
Impedance….. Since the r-f voltage increases as the output-tap is moved toward the end of a coil from the chassis ground return connection, and since impedance is determined by the r-f voltage divided by the constant circulating r-f current at a given spot on the coil, the source impedance at the output-tap position on a coil also increases as the tap is moved toward the end of a coil away from the coil’s connection to chassis ground return.
Selecting the position of the output-tap to yield the best FSM reading is a method of selecting the source impedance needed to best match the load impedance of almost any short antenna such as a whip (and even odd things like bed springs, or coat hangers). As the operating frequency is changed the output-tap position will have to be changed to match the resulting new load impedance presented by the antenna. Even though the primary function of moving an output-tap is to optimize the source impedance feeding the antenna element, there will be some interaction, or secondary effects, on the power-ratio, resonant tuning, and some on phase.
Phase…… When secondary coil L2 is tuned to resonance with C2, the phase at one end of L2 is + 90 degrees, and the phase at the other end of L2 is – 90 degrees, relative to the phase of the output-tap on L1. (The relative phase of one end of L2 is 180-degrees in reference to the phase at the other end of L2, and will be in quadrature with the phase at the output-tap on L1 only when C2 is adjusted to peak resonance. Incorrect adjustment of C2 can result in the two ends of L2 having undesired phases, such as: +80 / -100; +70 / -110; +60 / -120; or going the other way, +100 / -80; +110 / -80; +120 / -70, instead of the desired +90 / -90 degrees, as compared to the 0-degree reference at the output-tap on L1. (The phase of the incoming co-axial cable signal should NOT be used for reference; only the relative phase between the two antenna elements is important.)
The choice of the major phase required by the antenna element receiving the smallest power is available from the two output-taps on L2. Selection of major phase from either -90 or +90 degrees (same as either +270 and +90 degrees, also called either “lead” or “lag” phase) is selected by the position of switch S2 on the Expanded version of the Quad Phaser Network.
Of course the output-taps will have to be re-adjusted to provide the optimum match whenever a large change to a new operating frequency is selected. It is hoped that a swap for power-ratio selection will not prove to be necessary but provision is made for it.
Other Quadrature Antenna Systems
In addition to CFA, there are other antenna systems that require quadrature phasing. Many of the commercial TV broadcast stations feed all of their north-south radiating portions of their multi-level turnstile batwing antennas from one co-ax feed line, while all of the east-west radiators are then fed in quadrature by another co-ax feed line that is a quarter of a wavelength longer than the first co-ax cable, at the operating frequency.
Some FM broadcast stations use different lengths of co-ax to generate a quadrature phase difference, too. However, most FM broadcast stations now use only one feed line to drive their large vertical stack of horizontally mounted narrow-band small loop antennas. TV broadcast stations would like to use only one feed line, also, but the instantaneous frequency band width loop of the loop antennas is too narrow to properly radiate the 6-mHz wide-band TV signals.
Stations who depend on quadrature phase being generated by using differing lengths of transmission line to provide 90-degree delay are confined to operate over a narrow band of frequencies. Some success in expanding the operational bandwidth has been achieved by some in extending their delay line’s length using a variable “Co-axial Line Stretcher”, made by General Radio. Others have expanded their operating band width by adding a variable capacitor across the quarter wave length horseshoes to allow re-tuning of the delay section. (I observed this technique being employed at an international short wave transmitter site that has two 50,000 Watt transmitters, and twelve 100,000 Watt transmitters, that change frequencies on schedule several times a day.
Some contend that the Duo-Disk Antenna (DDA) and Single-Disk Antenna (SDA) should be fed in quadrature. This unit can provide their assumed three-terminal feed requirements, too. (However, I think it would be overkill and un-necessary to use the three-terminal features of this network to feed two-terminal antennas.)
Instantaneous Band Width…… Even though the Quad Phaser Network can allow low power amateur operation over a wide range of operating frequencies by re-tuning, its instantaneous band width is very narrow. (If by chance your unit does not have a narrow band width, then the probabilities are very high that there is a problem with a bad connection in the feed line, the antenna system, or the antenna matching network. Check for bad connections and thermal hot spots immediately after removing power. It is very important that solid connections (preferably soldered) be used.
The design of an instantaneous wide band-width quadrature phaser that does not require re-tuning is certainly achievable, but it will not be as simple nor as inexpensive as the Quad Phaser described in this and the previous article. One such very flexible quadrature phase approach that was in use at all major airports in the US prior to 1932 used two final r-f amplifiers.
Two Terminal Feed…….. The Quad Phaser can also be used to feed a two-terminal antenna (as a DDA most likely is) by using output Terminal B and the chassis ground RETURN as the other terminal. Isolation can be made by opening “Isolation Switch” S3 on the Expanded Quad Phaser version (as shown in Figure 2). S3 is provided to make this unit a more versatile “antenna matching network”. Of course output Terminal A, working against the co-ax braid, can be used if S3 is closed. But, this mode does not afford any isolation from the co-ax braid and any possible “earth ground”path. Many types of antenna experiments can be done with this versatile unit feeding different experimental antennas.
DLA
As shown in Figure 2, the presently shown version of the Quad Phaser Network circuit is not optimized for a small Dual Loop Antenna DLA and cannot effectively match the extremely low impedance (considerably less than one ohm) required to feed a small transmitting magnetic loop antenna. Logic would indicate that the small dual loop antenna (DLA) would require even smaller impedances.
A different but similar version of the Quad Phaser Network I used in 1946, while doing small loop antenna research, can provide matching and quadrature phasing to give the DLA a similar benefit of doubt. (I doubt that the DLA will perform any better than [or even as well as] the older compact transmitting magnetic loop antennas.) (To get a head start on a version of the Quad Phaser Network for DLAs please review the four-part series in Archives III of this magazine titled “LOOPS Of Olden Days”.) I plan to make available the design of this “Loop Quad Phaser Network”. If sufficient interest develops in DLAs, then this Loop Quad Phaser Network may be discussed in another article.
Construction Factors
When the term “ground” is used in this text, it refers to the Quad Phaser phasing/matching network’s chassis ground return and NOT to an actual earth ground connection, nor to the external input co-ax cable’s braid/shield. If an actual earth ground is intended, it will be identified as an “earth ground”.
This network requires that L1 be the primary and L2 be the secondary of r-f transformer T1. (See Figures 1 and 2.) Because many problems have occurred with smaller toroids saturating, I suggest initially using an air core for T1. (Later tests could be done with toroids, but I will use nothing smaller than a T200-2 for this application.)
L1 and L2 should have approximately the same inductance, and they should use large size wires, such as #12 or larger (or use parallel wires), because the circulating current flowing through them can be very high. Also, both C1 and C2 should both have the same approximate range of capacitance and must be able to tune through peak resonance at the operating frequency when the antenna elements are attached.
- Connect the ends of L1 to tuning capacitor C1, one to one side and the other to the other side.
- Connect the center conductor of the coax connector to an input-tap on L1. Position this “input-tap” initially about 1/4 of the way down the coil (L1) from the center-tap.
- Position the “output-tap” initially about 1/2 of the way up the coil (L1) from the center-tap position.
- Connect the center-tap of L1 to the outside of the co-ax connector.
- Connect the center-tap of L2 to the chassis ground return.
- Connect the ends of L2 to tuning capacitor C2, one to one side and the other to the other side. Initially position the two “output-taps” about 1/2 of the way between each end of the coil (L2) and it’s center-tap position.
- Complete all other wiring and connections in accordance with Figure 2.
Sensing R-F
Small neon bulbs can provide very simple inexpensive r-f voltage indications. They can be mounted on the frame of each tuning capacitor and positioned near, but not connected to, the stators of C1 and C2 and can be used to provide maximum r-f voltage indications for easy tuning.
If available, two old small florescent bulbs can provide an even a better “bar” tuning r-f voltage indication than the neon bulbs. Similarly, a small flashlight bulb having a tiny coupling pick-up wire loop combination can be mounted near terminal “A” on the lead that connects S1 to output terminal “A” to indicate r-f current. Another flashlight bulb with a tiny coupling loop combination can be mounted near terminal “B” on the lead that connects S1 to output terminal “B”, to provide an indication of the relative r-f current flowing to each of the antenna elements.
As an option to providing greater awareness of the signals driving the antenna elements, an oscilloscope, preferably a synchronized two channel or dual beam oscilloscope, may be used to observe these two r-f voltage amplitudes and the relative phase between them. Connect an oscilloscope high-voltage sampling-probe resistor-network 1 between terminal A and chassis ground return to feed channel A on the oscilloscope and use this signal to provide reference sweep synchronization. Similarly connect another oscilloscope high-voltage sampling-probe resistor-network 2 between terminal B and chassis ground return to feed channel B on the ‘scope.
Adjustments
C1 and C2 are both adjusted to initially provide maximum r-f voltage across each tuning capacitor, or at output terminals A and B.
These maximum r-f voltage indications should co-exist with the maximum r-f current indications. (Note: The maximum peak r-f voltage appearing across C1 and C2 will not be as great when the antenna is taking power as compared to when the antenna was not connected.) Final adjustments still require both of the capacitors to be tuned to peak voltage with a given set of other settings. C2 should always be the last capacitor tuned, thus assuring quadrature phase.
While both C1 and C2 affect efficiency, C2 primarily adjusts the relative phase angle difference between the two output terminals. (C1 will shift the phase of each output signal equally (relative to the unimportant phase of the input co-axial cable), but it will not shift the important differential phase angle between the cylinder and the disk; that’s the phase that counts. That explains why C2 should always be the last thing tuned.
If the shape, size, or spacing of any part of an antenna under test is modified, interactions are anticipated, so start the phasing and matching process all over again. The Quad Phaser Network should give any configuration of a CFA under test the benefit of any doubt in proving its worth, if it has worth.
Unfortunately, I am not aware of provable repeatable, efficient CFA’s test results of performance of -12 dB, or better, coming forth at the time of writing this article. I would very much like for some configuration of the CFA to be made that will repeatedly result in it working efficiently. In the meantime, research into other types of efficient small compact transmitting antennas should continue to be made.
Anything operating two “S” Units down is probably not worth continuing research on. That equals -12 dB. That is a power ratio of 1:16, meaning when the antenna is driven with 80-Watts it would radiate 5-Watts. Throw in some feed line losses and you are now down into QRP power levels with your expensive new 100 Watt transceiver. Of course QRP is fun, if you have the time to do it.
The coax “input-tap” on the transformer’s primary coil L1 winding determines the match to the coaxial transmission line (and the transmitter). The “input-tap” on the transformer’s primary coil can then be adjusted to provide the best VSWR to the transmitter with the antenna under test connected. Remember, everything is interactive and will probably need re-adjusting several times. So, please be patient. You can do it!
Field Strength Meter
All final adjustments should be based on obtaining maximum field strength meter’s indications, along with keeping very low VSWR indications at the transmitter, after accurately adjusting the transmitter’s output power to the same consistent reference power level, throughout the tests. It is very important to maintain a constant sensitivity on the FSM. A relative reading FSM is very important tool for antenna experimenting and it would be even more useful if it were calibrated.
It would be worthwhile to compare and calibrate your FSM measurements against the field strength radiated from some nearby AM broadcast station using a non-directional one-tower high-efficiency quarter-wavelength antenna. AM Broadcast radio station power and pattern data useful in determining your baseline may be found at: http://207.91.54.150/radiostation/ (Sadly Been Gone for a Long While).
At your transmitter site, place the FSM in a distant position where things like people and vehicles do not vary the readings. (A position in a simulated wooden “bird house” above a flower bed, that still allows its indications to just be readable with some binoculars, has works well for me.) A better, but more expensive, approach is to use a calibrated remote reading FSM.
Earth Ground Independence
If there is a question about the antenna system being earth ground independent, a 1:1 balun isolation transformer, using a T200-2 or larger core, can be inserted between the external input co-axial cable and the input co-ax connector on the Quad Phaser Network to fully isolate any possible direct or indirect ground effects. If the antenna under test is truly earth ground independent or co-ax cable braid independent, there should be no difference between isolated vs. direct connection as measured on a FSM.
An electrostatic discharge path consisting of a carbon resistor having from 1 to 5 meg ohms of resistance can be connected between the external co-ax cable’s shield and the chassis ground return and is recommended.
Later, after it is clearly determined that the CFA is, or is not, earth ground independent, then there will be time to consider removing the isolation transformer, as an option.
Obtaining Test Results
Most compact antennas will exhibit very poor results when operated horizontally polarized while near the ground. If operated two or more wavelengths above the ground the shorting effect the ground, or other conducting surface, has on shorting out horizontally polarized antennas will be greatly reduced.
However, the present major thrust of our experimental research efforts should be to find small compact transmitting antennas that can be effectively operated near ground. So, I suggest research efforts be concentrated on those antennas that can efficiently produce vertical polarization near ground. Better repeatable tests can be conducted if the not too distant receiving station uses vertical polarization, also.
A vertical monopole, or a vertical dipole, would be a better reference antenna for signal strength A/B comparisons than a horizontal one.
However, most dipoles, and inverted “V” antennas, are not vertically polarized, and they may also be far from ideal, which could make the antenna under test seem to perform a lot better than it really is.
No other antenna in the vicinity of the test area should be tuned to the operating test frequency when the antenna under test is operating, especially any reference antenna used in A/B signal strength comparison tests. The reference antenna should be de-tuned when the antenna being tested is receiving or transmitting, and then be re-tuned to the operating frequency when it is time to use the reference antenna. Otherwise, stray energy coupling between the antenna being tested and the reference antenna could result in the test antenna appearing to perform much better than it actually could by itself. A calibrated FSM can be a more accurate meaningful monitor for your transmitted signal.
While the antenna under test is transmitting or receiving, it is suggested that the reference antenna be de-tuned so it will not act like a parasitic element to the test antenna. This can be done by inserting a reactance (an inductor or capacitor) across its feed line, but not by shorting directly across it with a jumper or test clip lead. Remember, a “high gain” Yagi beam antenna only has one driven element; all the others are parasite elements. All of the Yagi’s directional apparent gain is achieved by using parasite elements. As stated before, a remote reading calibrated FSM would be even better.
Many “signal reports” given by a lot of amateurs seem to be in keeping with their trying to be friendly and maintaining good public relations. In other words, they are inclined to greatly flatter other stations for having exceptionally good signals, transmitters, and antennas (and themselves for having such sensitive “S” meters) with generous inaccurate reports. Its not uncommon to hear reports like, “You’re 5/9+30 here, but.…ah.… ah.…would you….please repeat your call and your handle one more time!” Very critical “Honest” signal reports are extremely more useful.
RF Coil Design Factors
A physically larger coil having the same inductance has greater heat and r-f radiation.
Larger diameter conductors have less ohmic loss.
Larger diameter conductors afford more surface, or “skin”, to conduct r-f better.
Larger spacing between turns reduces the amount of stray between-the-turns capacity and permit the coil to be tuned over a wider frequency range with a given size of variable capacitor.
Larger diameter coils require longer conductors to equal the same inductance because they are less reinforcing.
Coil mounting should avoid coupling of the primary coil to anything other than the secondary coil of the same transformer.
To prevent any two adjacent turns of a coil from resonating with its stray capacity at the same undesired frequency (which could be at a harmonic frequency, or could serve as a resonant circuit in the generation of parasitic oscillations, as sometimes occurs between evenly spaced turns), variable spacing of turns made progressively wider toward the “hot” end can be used.
However, if uniform spacing between coil turns is used, a non-parallel “grounded” metal bar mounted close to the windings serves to modify the stray capacitance existing between adjacent turns can be used, but with the disadvantage of increasing the total stray capacitance. The bar should be spaced farthest from the coil at the “hot” end of the coil. A variation of this uses a bar equally spaced from the coil windings but having the bar’s variable width made smallest at the “hot” end of the coil.
Faraday shields can be used to reduce capacitive coupling between coils.
Some coils are made of copper pipes and are cooled with water flowing through them to keep them from melting with high circulating current.
Spacing between turns determines the voltage breakdown rating.
Some coils use ribbon or bar shaped conductors to increase their voltage breakdown rating.
Sharp edges and burrs should be avoided since they can serve to generate corona discharges.
A simple useful tool called a “tuning wand” having a piece of copper on one end and a piece of iron on the other end can saving a lot of service tune-up time in determining if it is positioned close to an L/C tank circuit being tested. If the copper end is positioned close to the circuit (such as being inserted into a small I-F transformer housing), the circuits frequency will be increased by the “shorted-turn” effect in variable amounts (depending on proximity) by as much as removing one turn from the coil, thus decreasing inductance.
If the iron end of the tuning wand is similarly used, it will lower the frequency of the circuit by increasing the inductance due to increasing the concentration of the magnetic flux, and it also increases loading on the circuit due to hysterissis losses and Eddy currents.
When making tap connections it should be noted that the routing of the leads can either add to or subtract from the total inductance, but by amounts less than the effect one full coil turn would have.
These same factors should be considered when a coil is mounted too close to a metal cabinet. The larger a coil the more it can be affected by nearby metal. Avoid Eddy currents and hysteressis losses resulting from energy being magnetically coupled into and absorbed by iron and steel cabinets. Similarly, avoid “shorted-turn” effects when positioned too close to non-ferrous metals.
Close spaced metal cabinets can have unintentional coupling to an inductor. If mechanical vibration causes variation in the spacing between the close cabinet and a coil, the inductance of the coil can vary as a function of the spacing. This is a form of the “variable reluctance effect”.
Alternate Configuration
A simpler variable coupling between L1 and L2 has been suggested by Heikki Antman, OH2BGC from Finland. He e-mailed the following suggestion: “One possibility to change the coupling between the 2 coils is to use coupling coils and an adjustable capacitor for adjusting the coupling. This might be easier to make than to turn the whole coil.”
Heikki’s approach certainly provides a means of varying the coupling, but unfortunately it would also introduce variable phase shifts with changes in coupling which could negate the unit’s major feature of simply maintaining quadrature phase whenever C2 is adjusted to peak resonance. The approach of using two small link coupling coils, but without a capacitor being in series with them, may well provide easier mechanical flexibility if only one of the coils were to be movable, considering that the link coupling coils do not have to have any taps on them. Thank you Heikki for your comments.
George Sharp, KC5MU, from Roswell, New Mexico, USA, e-mailed asking, “What is the possibility of using a center-tapped coil for the T1 transformer, with other taps being added for the input and outputs?”
If after determining the power-ratio and preferred major phase, along with the ideal matching impedances as required by a given CFA configuration that provides good results on the preferred band of operation (as determined by first using a two-coil QUAD PHASER Network), then a single winding coil auto-transformer could be custom configured.
However, it cannot be link coupled with a link coupling coil fed by the input co-ax cable and still retain the QUAD PHASER Network’s main feature of being able to be tuned by the variable capacitors being set to maximum peak voltage at resonance and still provide quadrature phasing.
If this single coil auto-transformer approach is attempted, then the circuit shown in Figure 1 could possibly be used. The coil’s center-tap should go to the chassis ground return. Then consider the top-half of the coil to be L1 and the bottom-half of the coil to be L2. The variable capacitor across the top-half of the coil will be C1, and the variable capacitor across the bottom-half of the coil will be C2. The co-ax input-tap will be positioned on the top-half of the coil.
If the center-tap is made adjustable, problems can arise because moving the center-tap will result in varying the power-ratio, both output impedances, input impedance, and some phase shift, as well as requiring both variable capacitors to be re-tuned. In other words, primary control, vs. secondary control, of most individual parameters can not be isolated from one another in an easy manner. This auto-transformer approach is somewhat similar to some previous attempts made to match CFAs, which rarely met with customer satisfaction.
Also, this approach does not lend itself to reversing L2’s output signal’s major phase from lead to lag, if it became necessary to do so. To come anywhere near achieving optimum settings all at the same time while using the adjustable center-tap approach would be possible but highly unlikely. In other words, the phasing and matching flexibility provided by the original two-coil approach may be severely limited by the center-tap approach at this time. After confidence is gained with the present two-coil method with an antenna configuration that is found to provide good efficiency, then the center-tap method may possibly be refined.
Thank you George for your comments!
Feed Line Radiation and Impedance
Radiation from the outside of a co-axial cable feed line can frequently distort the radiation pattern of an antenna being tested to the extent that it can make a major undesired contribution when a mismatch and a high standing wave exists. When the polarization of the radiation pattern changes as a result of re-positioning the feed to an antenna under test, then any test data obtained about the test antenna is of questionable value. The current flowing on the outside of the outside conductor can also result from part of the energy radiated by the antenna being intercepted by the co-ax and be re-radiated and contribute to the total radiation pattern in an uncontrolled manner. Toroid chokes mounted on the outside of the co-axial cable can be beneficial in controlling this problem. A mismatch between the antenna and the feed line can present an even bigger problem.
As has been said by many, an “antenna tuner” does NOT tune the antenna, it transforms the load impedance presented to it by the antenna into an impedance that the co-axial cable and transmitter can accept. This is usually 51.5 ohms, and its based on very good reasoning but it is not too well known.
Co-axial cable’s impedance is determined by the ratio of the outer diameter of the inner conductor compared to the inner diameter of the outer conductor. A large center conductor can handle a larger amount of current, but the reduced spacing will break down easier with less voltage. A smaller center conductor can offer increased spacing resulting in a higher breakdown voltage rating, but it can not provide good conduction for a higher current. 51.5 ohms is chosen because it is the characteristic impedance of co-axial cables that permits the transfer of the most r-f power through a given limited size of outer conductor diameter.
Or stated another way, 51.5 ohms is the impedance of co-axial cable that transfers the most r-f power for the money. (However, using means other than co-axial cables, such as open transmission lines, can transfer even more power with less power loss for a given amount of money.)
Most amateur transceivers and transmitters have been designed to operate most efficiency when delivering power at the source impedance designed to match the popular optimum co-ax impedance of 51.5 ohms. This does provide the best power match to the 51.5-ohm co-axial cable commonly used.
It has been said that a perfect 51.5-ohm resistive (non-reactive) load on the far end of a perfect 51.5-ohm impedance co-ax cable will absorb all of the power and there will be no returning power to combine with the outgoing power to form any standing waves. If this ideal condition were to exist, then a mismatch between the transmitter and the co-ax cable would not cause standing waves on the co-ax cable but instead would only result in some power loss due to less efficient matching by the source.
In theory this is correct, but since the load on the far end of the co-ax cable usually has some degree of mismatch most of the time, then in practice standing waves result first from the mismatched load (antenna) end of the cable, and secondly from the mismatched source (transmitter) end of the cable not absorbing the returned power.
Variations from the designed characteristic impedance of the co-ax cable can be introduced by departures from the conductor ratios, such as can result from kinks, poor splices, strains, etc., can serve to create discontinuity reflections in varying amounts.
To summarize, two or more separate standing waves patterns can be found on a co-ax cable with varying voltage amplitudes to contribute to the highest voltage existing on the co-ax cable. Each type of co-ax cable has a breakdown voltage rating based on a perfect cable being used under ideal conditions. If non-perfect conditions and/or a non-perfect cable exist then the rating has to be de-rated . This amounts to the power handling ability of the co-ax cable having to be reduced when the voltage standing wave ratio (VSWR) increases as a result of combined standing waves.
When the center conductor of a co-ax cable is no longer centered (probably due to being stepped-on, poorly spliced, or kinked) the impedance at the damaged spot is usually altered, but even more important, its voltage breakdown point will be reduced. A high percentage of co-ax cable failures are due to voltage breakdowns.
A breakdown of a co-ax line at one spot frequently introduces failures at periodic spacing (such as arc welded type spots burned through the outside copper conductor of a co-ax line spaced every ½ wavelength apart) due to even more severe standing waves being introduced by the first failure point.
Commercial fixed frequency transmitter operations usually locate the antenna matching network at the antenna (such as the base of the AM Broadcast station’s tower) in order to match the antenna system’s impedance to the co-axial cable’s to maximize power transfer. This reduces the chances of having arcing and breakdown failure within the co-ax cable itself.
By contrast, amateur operators usually compromise and only try to protect their transmitters by merely trying to present a good impedance match to just their transmitters. By having their antenna matching network located in the transmitter’s operating room, they usually disregard the standing waves on their co-ax cable for the sake of being free to operate on more than one operating frequency and treat their co-ax as a part of the antenna system. But, by disregarding the standing waves on the co-ax cable, the operator does run a much greater risk of voltage breakdown within the co-ax cable itself. By ignoring the standing waves on the co-ax cable the operator does avoid the cost and the inconvenience of having to operate the antenna matching network via remote control as could be required when the antenna matching network is ideally located at the antenna.
Conclusion
This method of generating a quadrature phase relationship between antenna elements is based on the fact that the magnetic flux is proportional to the current flowing in the transformer’s primary coil winding L1, while the signal generated in the secondary coil winding L2, is proportional to the CHANGE in flux.
This method of obtaining quadrature phasing, with its simple tuning to maximum r-f voltage, has the big advantage over the previously used standard method that requires an inefficient hard to tune approach to generating quadrature phase by off-setting the tuning of each variable capacitor from resonance; one is off-set to a frequency above resonance and the other off-set (hopefully) by the same amount below resonance.
By way of review, the power-ratio between the antenna elements is determined by the mutual coupling BETWEEN L1 AND L2, the position of the taps on the coils determine the impedance for matching the loads, while the tuning of the secondary variable tuning capacitor, C2, will primarily determine the phase differential between the elements (without any regard for the phase on the incoming co-ax cable). As is to be expected, interaction by the antenna element (loads) placed on these circuits will result in having to re-tune the variable capacitors to peak resonance, as can be easily indicated by a peak in the r-f voltage across each capacitor. Final adjustment should always be made with C2 to restore the proper 90-degree phase said to be required by CFAs and by some other experimental compact antenna candidates.
Consider trying the DDA too; it’s looking better each day. Or, you may wish to verify the report that claims the Single Disk Antenna (SDA), a single disk and cylinder combination, does better than the DDA. Then, another antenna test has been suggested to experimentally try feeding two disks without a cylinder, by letting one of the disks replace the cylinder and be fed by the terminal that previously fed the cylinder.
LUse your imagination and test your own antenna ideas using this flexible Quad Phaser Network “work horse”! Who knows, there may be a good small compact antenna configuration out there somewhere! If you find it, please share it with your fellow hams!
The anticipated load impedance requirements of various experimental compact antennas vary greatly. The Quad Phaser Network provides a method of giving the benefit of a doubt to most of the experimental antennas (with the exception of the small loop and Double Loop Antenna which has an extremely low impedance of much less than one ohm).
If there are other parameters that have not been addressed for CFA’s by Allen’s QUAD PHASER Network, please let me know. Your comments and suggestions have been appreciated in 1999 and will continue to be appreciated even more in Y2K! May the New Year allow you time to greatly enjoy your experimenting!
Originally posted on the AntennaX Online Magazine by Harold Allen, W4MMC
Last Updated : 14th May 2024