Plates in the Sands - More Thoughts on the CFA at Tanta

sandshdr.jpg (30379 bytes)

More thoughts on the CFA at Tanta
David Jefferies
E-Mail

t was suggested to antenneX publisher, Jack Stone by Ing. Serge Stroobandt (MSc MBA, ON4BAA) of Oostende, West Flanders WV Belgium, that small HF antennas on buildings in the desert transmit rather well, if they are isolated from the ground. Consequently, Jack asked me to write about this in the context of the CFA structures at Tanta in Egypt, after we had done some research into the pertinent facts known about the Tanta CFA site, including the size and construction of the buildings.

I was not convinced that the insulating properties of desert sand were necessary for the performance of such a structure. After all, a length of wire connected between transmitter and ground radiates rather well if it is not short compared to a wavelength, and if the ground is a good one.

I decided to apply the analysis of "Useful Radiation from Compact Antennas - Plates" (article #84 antenneX Archives IV - February 2001) to the Tanta CFA to see what the expected efficiency and bandwidth might be.

EARLY CONCLUSION
To cut to the end of the story, this plate mode of radiation does seem to explain the reported performance of the particular compact antenna structure at Tanta. Other CFAs, which do not appear to work so well, do not share the same local environment that Tanta has, and so we have now a consistent view of why the Egyptian antenna functions when the consensus view from professional antenna engineers is that the CFA mechanism is, at best, a very inefficient radiator.

sandtant.jpg (12071 bytes) First let us take a look at a photo of the Tanta CFA. The first thing that strikes us is that there are two antenna structures, side by side, and in close proximity to each other. These will presumably couple currents capacitatively to each other, and, we are told, are built to operate on different frequencies. The tuning circuits on these structures will probably act as good traps for each other's frequencies, as otherwise there would be strong intermodulation products generated by driving one transmitter with the output of the other, and the installation would generate spurious signals.

INFLUENCING FACTORS
The building size is reported by Alan Boswell to be 15 metres by 8 metres by 7 metres high. The wavelength appropriate to the operating frequency of 1160 kHz is 258 metres. If we consider a current path from a point on the roof, underneath the CFA structure, to the ground then it will be at least (7 + 8/2) = 11 metres, or about 1/24th of a wavelength.

When we consider that there is also a current path through the adjacent CFA structure which will be longer and also contributes to the radiation, and we also add in the height of the CFA structure itself we arrive at an estimate of antenna size that puts the Q factor at around 100-200 (say 150) and the bandwidth at around 6-12 kHz. For reasons stated in the article on radiation from plates the efficiency will be high, because the current density is low and the cross section carrying current (skin depth times width of current path) will be much larger than in a wire antenna.

Now Alan Boswell remarks that the channel spacing in the broadcast band in Europe and the Middle East is 9kHz, giving two 4.5 kHz sidebands. Thus, the Tanta CFA plus building structure is quite capable of radiating the reported signal strength and bandwidth, at high efficiency, without invoking Poynting Vector Synthesis or any other novel CFA radiation mechanism.

Now, the current supplied from the feed to the plates of the CFA spreads out over the plates and couples (via displacement current if you like) to the same total current in the roof of the building which forms the ground plane. The question that arises is, what path does this current take in getting back to the transmitter? (we recall that currents always flow in closed loops; the transmitter pushes current into the hot lead of the feed and pulls the same current from the shield or cold lead).

It is likely that the inductive and resistive impedance of the return path through the building metal is considerably less than that straight back up the short return side of the feed, as the area of conductor via the roof and sides of the building is so much more. Therefore we expect the current to take a circuitous route back to the transmitter, and radiating as it goes. Without knowing all the details of the earthing arrangement at the site, it is difficult to be any more specific and quantitative than this.

A CASE IN POINT
To illustrate this ability of current to return through building frames, I offer the following report of a consultancy I did some years ago for the Electricity Supply Board at Reading, UK.

The problem presented to us is that on a small light industrial estate, one of the manufacturing units had computers in a back room several metres from an electricity supply sub-station for the site. The writing on the screens was snaking from side to side through a distance of a few millimetres, making use of them by the secretaries tiring and almost impossible.

The hypothesis was that the refresh rate of the screens, at 50Hz, was beating with the nominal 50Hz mains supply frequency. The computer clock was derived from an internal quartz standard for each machine and was not phase-locked to the mains.

The further hypothesis was that there was a magnetic field at 50Hz moving the electron beam back and forth on its way from the electron gun to the screens.

We wound a large 100-turn search coil on a cardboard box, and plotted a cylindrical magnetic field, appropriate to a current of 90 amps running parallel to the back of the building about one metre below ground level.

The electricity board consulted their maps, and found that this was where the three phase supply cable ran. Normally the current in the three phases is balanced and there is no current flowing in the adjunct neutral wire. However, different industrial units were on different phases, and there was a varying load on each phase which resulted in a 90-amp out of balance current on the cable. In normal circumstances, this would have returned along the neutral in the same cable channel. For some reason, the net current flowing in all four conductors was not zero.

The site had a "protective multiple earth" arrangement whereby the neutral was bonded to the metal building frames and earth at each building unit on the site. The current was returning "the long way home" by flowing through the earthed building frames and back to the substation.

The Electricity Board engineer went into the substation, tightened the "rusty bolt" on the neutral return where it was bonded to the supply, and immediately the magnetic field and screen wobble vanished!

djj-intro.jpg (9196 bytes) CAREFUL MEASUREMENTS ESSENTIAL
So we learn from this anecdote, both the importance of careful magnetic field measurements, and also of ensuring the integrity of current return paths in power installations.

I would advise antenneX engineers to pay particular attention to trying to establish the magnetic near-field distribution and current return paths on any CFA structures they investigate on-site. -30-

Dr. David J. Jefferies
School of Electronic Engineering, Information Technology and Mathematics
University of Surrey
Guildford GU2 7XH
Surrey
England
Click Here for the Authors' Biography

~ antenneX ~ September 2001 Online Issue #53 ~