Critique of CFA Experiments and Papers
By Ralph Holland, VK1BRH

Introduction
he CFA remains a controversial antenna for there are more counter-claims than there are claims of success, despite seemingly successful deployment in Egypt.

I am not an E-M theorist, but I attempt to apply reasoning and known or learned knowledge in my analysis of this controversial antenna. This article is a critique of the data I have on hand, and the data that I have derived, during my recent research when preparing the feature article CFA Field-Strength Measurements in this month’s antenneX issue for July 2000.

Basis for Controversy
There is a lot of controversy surrounding this antenna and it has so far not been dispelled. On one hand we have actual physical deployments of the CFA in the form of commercial broadcast stations in various parts of the world and on the other we have numerous experiments that have produced null-results. Granted there have been some experiments with positive results, but professional measurements performed with calibrated equipment, that would provide more precise field-strength reports, are absent.

This controversy surrounding the CFA is evidenced by:

1. The inventors touting the "Heavyside theory" which attributes radiation to the direct synthesis of the Poynting vector in an Interaction Zone (IZ) which they state surrounds the antenna. This theory has not been accepted.
2. The antenna is physically very small and its height is reported as being largely independent of its radiation efficiency. Smaller CFAs are claimed to be harder to feed but almost as efficient as the larger versions.
3. The antenna field-strength measurements, so far released into the public domain, show that the antenna has outstanding gain over a 1/4 wave monopole. An as yet unprecedented event, seemingly counter-intuitive to antenna theory.
4. We only have field-strength measurements for the Egyptian CFAs, which appear to support the inventor's claims. One set of independent measurements were restricted to the Tanta CFA site, where it was reported that the transmitter power was calculated from the inspection of the plate voltage and current. To my knowlede, there has not been any direct measurement of the input power by means of a directonal coupler and a power meter.
5. Independent studies, through both numerical and physical modelling, have resulted in the publication of null-results, thus, a general failure to reproduce the inventors’ claims.
6. Counter-claims to the failures state that the failures were due to inappropriate feed arrangements or incorrect modelling tools.
7. Some other organizations that have studied the antenna are not prepared to release their findings; this coupled with the rather scant data makes it difficult to that lay credence to the inventor's claims.

What we need is cold hard data, and perhaps the explanation and theory will follow. The inventors have been reluctant to cooperate sufficiently in the presentation of the much-needed verified data, so we are left to speculate without conclusion.

Chronology
The first round of studies were produced by the inventors, I believe, whilst Inventor Kabbary was a PhD student circa 1988. The sequences of events are not definitive because I may have missed sources and I haven’t been privy to all the occurrences, but you can take them as an indication of the activities.

Several articles appeared in Electronics World and Wireless World, March 1989 [1,2] and others followed [3,4]. Several articles by the co-inventors appeared in antenneX [5, 6, 7, 8, 27, 29] from September through February 1999. To the best of my knowledge, no field-strength readings were released to the public until circa 1995.

The first published field-strength findings, of which I am aware, are those attributed to Dr Kabbary and the EUTR at Cairo available in [7] and reproduced in [15]. These are believed to be from measurements performed circa 1991 and released to the public in 1995.

In June 1999 a team of Brazilian Engineers went to Egypt to specifically study the CFA and collect field-strength measurements from the CFAs that have been deployed there for some years. These measurements were limited to the Tanta station. Their findings indicate an apparent gain over a 1/4 wave monopole, and in excess of Dr Kabbary's earlier claims. These gains may be studied in reference 15. Please note that the input power to the Tanta CFA has not been verified by direct measurement.

I recently received my copy of reference [16] which shows the CFA D-plates fed by a voltage transformer and this article contains relative field-strength readings, which were compared against a 1/2 wave dipole. The recent Sydney CFA For Sale article reference [14], featured in antenneX June 2000 contained two absolute and also several relative field strength measurements.

This brings us to the present date:

Critical Analysis
Poynting Vector Synthesis
References [1,2,3,4] and [7,8] discuss Maxwell's equations and the Poynting vector. Reference [2] in particular attempts to define the zone of interaction around a 1/2 wave dipole stating that this occurs in a small annular region where . The inventors state that this is where Poynting vector direct synthesis of the radiation occurs for a dipole, which I can assure you is utter rubbish. The radiation around a dipole is generated as a result of the vector potential established from the conduction currents, not by direct Poynting Vector synthesis.

The Poynting vector analysis of energy flow, or power flux is a convenient mathematical analysis. The fact the a Poynting vector exists does not necessarily mean that there is radiation. Traditional wire antennas have been modeled with extreme veracity by numerical software and physical models. To shed light on the questions I have had recent correspondences with Dr G. J Burke at Laurence Livermore National Laboratory (LLNL) where I have raised this question of Poynting Vector synthesis which can now be debunked.

Models have been made, which were specifically designed to arrange the near E and H fields so they are in the ratio of E / H = 377 and in phase quadrature in an attempt to model the inventors’ claims, i.e., E and H were placed in the correct phase and correct amplitude to determine if any significant improvement in radiation may be had and the results showed no significant increase in radiation. Further, they found that the suggested interaction zone (IZ) is very small and that either side of the IZ that the magnitudes of E and H altered considerably (by orders) in magnitude and phase.

Of course the inventors also claim that the IZ around the CFA is more efficient than the IZ around a traditional antenna.

Dr. David Jefferies [9] also points out this danger of considering E x H as constituting the generation of power. The "Heavyside theory" with direct Poynting vector synthesis touted by the inventors is flawed and controversial. Follow the links on Dr Jefferies’ site if you want to find more out about this controversy.

Let’s not explain this antenna in terms of some unsupportable theory, rather, let’s make precise measurements of field-strengths, currents and radiation; present facts not fiction.

The D-plate capacitor radiation
The "Heavyside theory" is also based around the premise that the D-plates are good radiators of R.F. energy (not heat). Many traditional experiments have shown that a capacitor is a relatively weak generator of R.F. fields. Thus this coupled with direct Poynting vector synthesis appears to be nonsense.

The capacitor experiment quoted in reference [6] is also flawed for the following reasons:

1. The capacitor was a cut-out in a plate, which would produce fairly intense E-fields between the edge of the capacitor and the plate, which certainly does not model the CFA D-plate.
2. The probe is shown as being parallel to the plates, in which case it would not be expected to couple into the H-field generated by the displacement current. (It may have been drawn incorrectly though.)
3. The current profile present in the graph is uncalibrated and we have no idea of the magnitude in relation to any other field. The current may indeed be very weak, indicating a very weak H-field.
4. According to a well informed source, the current profile indicates that the probe was probably measuring radial current in the capacitor and the E-M field which is known to extend outside a capacitor i.e., it does not appear to be measuring the field due to the displacement current.
5. The feed situation for the CFA D-plates requires that a single feeder be connected between the D-plate and the Ground-plane (GP). This feeder contributes its own field due to the conduction current which would oppose the field due to the displacement current. Whereas the capacitor experiment employed a voltage feed referenced to the sheet that the capacitor is cut-out from. This is not the way the D-plate is fed in practice.

It is possible that people in the academic community are put off by the nonsense in the inventors’ explanations and by the lack of veracity in their experiments. If the CFA is to recover from such form of ridicule, then some proper science better start surrounding the CFA investigations. We need truthful reporting and proper experimentation in support or counter-support of the CFA. Null-results as well as positive results help to improve our understanding, but we must take care in any supporting evidence, whether pro or con.

Numerical Modelling
There has been some doubt cast on the veracity of NEC to model electrically small antennas [20,21]. I questioned whether, firstly that NEC could model electrical small antennas and I forwarded these questions. Dr Jerry Burke's responses were:

1. NEC2 can model elements down to 0.001 wavelengths in length.
2. NEC-2 double precision should work down to about 1.E-8 wavelengths
3. NEC-4 should do antennas without loops to where the radiation resistance underflows (~1.E-32).

Jerry also said that NEC has difficulty modeling small loops because the basis functions may result in large currents - which may result in numerical errors. He recommended not employing loops in a wire-grid model when modeling the electrically small CFA. Although, he did say the loops may be okay in the model, provided large currents are not present in the simulation. If large current are observed in the small loops the model should be discarded.

I also questioned whether NEC can model a capacitor in its own right and the response was yes even including the fringing fields.

We discussed the effect of large reactive power in some models and Jerry did had some concerns over the observation of large reactive powers circulating in and out of the model's ports, this too may lead to numerical errors when modeling via NEC.

Physical Modelling
The inventors have claimed that the best signal strength arises when the E-plate voltage and D-plate voltage are in the correct ratio and fed in phase quadrature. Other experimentalists have NOT found that the best signal-strength occurs at the 90-degree phase point. Indeed, Dr John (Jack) Belrose [19] states that he obtained better performance when they were fed in phase. Further, Joel Hungerford [20] reports that the field-strength actually increases and decreases with the voltage on the E-plate and that there was no peak when they were fed in quadrature. Counterclaims by Dr Stewart, state that the ratio of voltage between E and D plates must be correct before the proper conditions exist for maximum radiation at the 90 degrees phasing point. This correct ratio argument is further supported by the reports in the recently received paper reference [16], by Mr. Hately, who diagrams a voltage transformer feeding the D-plate on the smaller CFA models. (This step-up transformer may be necessary, given that others have found that a capacitor is a weak radiator of H-fields.)

Dr. Belrose has also described how he experienced difficulty controlling the feed on his model, and has observed that large reactive powers were reflected back to the transmitter. His simulations have also found large reactive powers circulating in and out of the antenna ports, which is undesirable (from a numerical accuracy view as well as practical view). Perhaps a different feed arrangement must be sought by Jack as others have not reported this difficulty.

More difficulties are associated with physical modelling where there may be coupling between the antenna and its surroundings. Experimenters must be careful to ensure that only the antenna components are radiating. This requires meticulous insertion of isolation (current) baluns and careful feed arrangements to ensure that nothing untoward is happening. By some reports we have heard that touching the GP on the CFA has caused its signal to quieten down. This is possibly an indication that the GP and feeder may be undesireably taking part in the radiation. We also need to assure ourselves that the CFA is not coupling into something in its near environment such as wiring or other hidden or unsuspected building structures.

Different physical dimensions have been used in each model. This is quite interesting as data was quoted as being provided from the same or similar sources, but quite understandable, as each author makes assumptions to obtain a principle dimension. I have gone through the references and measured the proportions in the CFA photos and arrived at my own set. These data are presented in the next figure, where I have reported all sizes in milli-Lambda (Lambda is one-wavelength). It is interesting to note the apparent changes in dimensions working from Prototype I, Prototype II, Barnis and Tanta, which I believe may be the order of development. It is of some note, that the size of the D-plate hole in Jack Belrose's model appears to be deficient compared against the other models. By the same reasoning, Gary and Jay's 160 m CFA [12, 13] seems to be deficient in the E-plate and top-hat dimensions. (See reference [26] for the spreadsheet data.)

Field-Strength Measurements In regard to the Sydney CFA field-strength measurements:

1. Only two absolute measurements were performed, one in the front yard of an attendant and the other some further distance down the street in an urban environment. The first measurement, by reports of a third-party, was the result of taking a peak reading! These measurements can be seen in reference [15] where it is shown that the first measurement is higher than expected. Two measurements are insufficient to fit data to any known theory or conclude any definitive result. From both values all that can be concluded is that the CFA appears to be producing more field-strength than a reference dipole in free space or less than the maximum field expected due to gain from the ground. This does not quite match the outstanding performance shown by the Brazilian measurements for the Tanta CFA.
2. Some relative field-strengths were performed by comparing the Sydney CFA with, I assume conventional antennas, from other stations operating on similar frequencies. These figures were adjusted for the differences between the other stations' outputs and the output from the CFA station. However, these readings were taken at large numerical distances (see reference [15]) and over terrain for which I have received no description nor profile. It is quite possible that the signals are affected by Fresnel-zone diffraction. We need some field-strengths that are closer and over predictable terrain.

John said that the CFA measurements are the property of the Sydney station, so he is unable to release them. There may be more than two measurements.

The Field-Strength measurements provided in reference [6, 10] and reproduced in reference [15] indicate that the CFA has apparent gain over the 1/4 wave monopole, and in the later case significant gain. The Brazilian measurements are the best independent measurements that we have on hand, but these show substantial improvement over the original measurements provided by Dr Kabbary. These Brazilian measurements were restricted to the Tanta CFA installation only and the input power to the Tanta CFA has not been verified by direct measurement.

The comparative measurements provided by the results conducted by the School of Signals, Blandford camp, during March 1990 present some difficulty too because

1. They are quoted in dBm. In order to equate these back to absolute values I have to assume a certain effective gain (and hence effective height) for their receive antenna.
2. The comparison antenna was quoted as a half-wave dipole mounted at the same height, which suggests that they employed a horizontal dipole. The field from such does not exhibit ground gain, unlike a vertically polarized counterpart (one could say this is a comparison between apples and cheese!).

I cannot equate the received input power readings back to field strength (without assumptions) on their 2-kilometer range because I do not know the parameters of their receive antenna, nor enough about the comparison antenna's orientation.

The Isle of Man Deployment
This proposed deployment is under license with the CFA being the only accepted antenna due to environmental and practical reasons, i.e., it has been mandated because it will be smaller than the equivalent long-wave quarter-wave tower [830+ foot (245m) tower versus the less than 90-foot (27m) CFA]. You can read about the Isle of Man CFA in references [23, 24, 25]. This acceptance does not mean that it works more effectively than a monopole, it just means it was accepted for other reasons.

It also seems that the Marconi School of Wireless has raised an objection to the use of this antenna on the grounds that it is new and has not undergone environmental impact studies to determine the effects of radiation on neither man nor beast. This may be the delay behind its commissioning, which is an interesting aside. However, this appears to be more of a question about effects of RF than about CFA technology. A critical meeting is scheduled for September 2000 to determine and try to resolve these environmental issues.

Conclusions
The CFA, at the very least, is certainly a controversial antenna. The theory surrounding its explanation appears flawed and is one of the detractors as to why the academic community may seem distance and hesitant when asked to provide input. We can alleviate this situation by ignoring the proposed theory for the time being, and by concentrating on obtaining scientific measurements.

The following points are very relevant:

1. We need actual field-strength measurements stemming from the near-zone to the far-zone, i.e., from short to large numerical distances, and in a progressive fashion. With such data we will be able to fit and compare the output against the monopole antenna. These need to be actual and absolute field-strength measurements, not relative measurements. Relative measurements are not calibrated, and as can be seen from the diversity in several articles, they are not easy, if at all, to relate back to absolute field-strength values, as too many assumptions must be made about the receiving equipment.
2. The taking of measurements over terrain that does not have a flat profile (unlike Egypt) complicates the relationship of said measurements back to the isotropic (theoretical) source. If possible, experimenters should perform measurements on flat ranges and paths where there is little or no obscuration.
3. We need data that has been independently verified. So far we have only one source of independent data from the Egyptian CFA, however this data shows substantial apparent gain over the original data attributed to Dr. Kabbary and the input power has not been verified by direct measurement. As a result, two CFA sites are under construction in Brazil and we have been promised the freedom of access to take measurements at these sites and promised that any measurements will be publically available.
4. We need current distributions of the E-plate and D-plate elements.
5. We need more details of the feed conditions, such as voltage ratios and phases.

If we can produce findings, which are performed scientifically, and referenced to standards, and these findings show that the CFA is effective, we ought to be able to obtain interest from the more academic circles. But until then, the CFA will remain an enigma. We should be able to put this wide debate to rest one way or the other! -30-

References

1. "Maxwell's equations and the cross-field antenna", F. M. Kabbary, M. C. Hately, and B. G Stewart, Electronics and Wireless World, March 1989.
2. The Cross-field Antenna in practice, by C. Bryan Wells, Electronics World and Wireless World, November 1989.
3. "CFA: Working Assumption", by M. C Hately, F. M. Kabbary and B. G. Stewart, Electronics and Wireless World, December 1990
4. "CFA - RIP?", by Colin Davis, Electronics World and Wireless World, 1993
5. "The Cross-Field-Antenna Part I", Maurice C. Hately GM3HAT and Ted Hart W5QJR, antenneX, September 1998, www.antennex.com/archive2/Sep98/Sep1/cfa.htm.
6. "The Cross-Field-Antenna Part II", Maurice Hately GM3HAT and Ted Hart W5QJR, antenneX, October 1998, www.antennex.com/archive2/Oct98/Oct1/cfa2.htm
7. "The Cross-Field-Antena Part III", Maurice Hately, GM3HAT, Dr Kabbary, Dr Stewart, MM1DVD and Ted Hart, W5QJR, antenneX, November 1998, www.antennex.com/archive2/Nov98/Nov1/cfa3.htm
8. "Cross-Field-Antenna Part IV", Maurice Hately, GM3HAT, Dr Kabbary, Dr Stewart, MM1DVD, and Ted Hart, W5QJR, antenneX, September 1998, www.antennex.com/archive2/Dec98/Dec1/cfa-4.htm
9. "The Poynting Vector, Power Transmission and the CFA", Dr D. Jefferies, reprint antenneX, November 1999, www.antennex.com/library/shack/Nov99/pv_cfa.htm
10. Brazilian Broadcast Engineering Report on Tanta CFA's visit June 1999, reprinted by permission at www.antennex.com/compact/images/braz_rep.gif
11. "CFA for 160 Meters Part I", Richard Morrow, antenneX April 2000, www.antennex.com/archive4/Apr00/Apr6/160cfa.htm
12. "160-Meter CFA - Part II", Gary Nixon, WA6HZT and Jay Lemmons, N6YIP, antenneX, May 2000,
     www.antennex.com/archive4/May00/May3/160cfa2.htm        
13. “The 160-Meter CFA - Part III”, Gary Nixon, WA6HZT and Jay Lemmons, N6YIP, antenneX, June 2000, www.antennex.com/archive4/Jun00/Jun5/160cfa3.htm
14. "Sydney CFA for Sale, by Jack L. Stone, antenneX, June 2000.
15. "Analysis of CFA field-strengths using Norton's Plane Earth Attenuation", by Ralph B. Holland, VK1BRH, antenneX, July 2000.
16. "The Cross Field Antenna, Its implications for Military Radio", by Maurice C. Hately, MSc, FIEE, Journal of the Royal Signals Institution, Vol XX, No 1, ISSN 0374-3519, Summer 1991.
17. Dr Jefferies Poynting vector article and site references www.ee.surrey.ac.uk/Personal/D.Jefferies
18. "CFA phasing result", private correspondence from Joel Hungerford, June 2000.
19. "CFA Experiments ", by Dr Jack Belrose, antenneX, June 2000, www.antennex.com/shack/Jun00/jb_cfa.htm
20. Private communications between John Knot and Ralph Holland, June 2000.
21. Private communications between Dr Jerry Burke and Ralph Holland, June 2000.
22. "MF HF Groundwave model", by Ralph Holland, antenneX, June 2000.
23. Longwave broadcast station on Isle of Man, http://www.g4loe.freeserve.co.uk/music.htm
24. Tynwald Parlimentary sessions papers from the Isle of Man http://www.tynwald.isle-of-man.org.im/papers/hansard/191099T-41.htm
25. Tynwald Parlimentary sessions papers from the Isle of Man http://www.tynwald.isle-of-man.org.im/cgi-bin/search/fcp.pl?words=Cross-field+Antenna+CFA&wt=be&bl=or&d=/papers/hansard/210499T-04.htm
26. Spreadsheet CFAdimensions.xls
27. "How to Build a CFA for 75/80 Meters", by Maurice Hately, GM3HAT, Dr Kabbary, Dr Stewart, MM1DVD, and Ted Hart, W5QJR, antenneX, January 1999, http://www.antennex.com/archive3/Jan99/Jan1/cfa_5.htm
28. "Care & Feeding of a Crossed-Field Antenna", by Maurice Hately, GM3HAT, Dr Kabbary, Dr Stewart, MM1DVD, and Ted Hart, W5QJR, antenneX, February 1999 http://www.antennex.com/archive3/Feb99/Feb6/cfa-nw.htm
29. Private communications, John Innes, June 2000.

Send mail to webmaster@antennex.com with questions or comments.
Copyright © 1988-2010 All rights reserved - antenneX©
Last modified: December 31, 2009