US2510162A - Aerial array - Google Patents

Description

June 6, 1950 E. 0Q WILLOUGHBY 2,510,152

AERIAL ARRAY Filed Oct; 1, 1945 2 she t -sheet 1 In uentor Ema 0J80PN wiLzaasHr' By Alto y June 1950 E. o. WILLOUGHBY 2,510,162

AERIAL ARRAY Filed Oct. 1, 1945 2 Sheets-Sheet 2 F/G6. F/GZ 2 Attor ey Patented June 6, 1950 AERIAL ARRAY Eric Osborne Willoughby, London, England, as-

signor, by mesne assignments. to International Standard Electric Corporation,

New York,

N. Y., a corporation of Delaware Application October 1, 1945, Serial No. 619,653 In Great Britain October 31, 1944 1 Claim. 1

The, present invention relates to electric aerial arrays, and particularly to those intended for operation over a wide band of radio frequencies. It is often of great advantage to be able to radiate a wide band of. frequencies from a single dlrectiye aerial array in order that the wavelength may be changed as may be necessary to secure the best transmission throughout the day.

It is the principal object of this invention to provide an aerial system which can be used over a range of frequencies in which the ratio of the maximum to the minimum frequency is of the order of 3:2 with relatively small impedance mismatch between the individual aerial units and the feeder lines over the range.

Another object is to provide means for shifting the operating wave band.

The invention accordingly provides an aerial array comprising a plurality of dipoles arranged one above the other and connected effectively in parallel to a feeder line by two or more tandemconnected transmission line links, the characteristicimpeda-nce of any link being higher than that of the link or line which feeds it.

The invention will :be described with reference to the accompanying drawing in which:

Fig. 1 shows a diagrammatic plan view of a folded dipole which may be used in carrying out the invention.

Fig. 2 shows a perspective "view of a preferred form of the dipole according to Fig. 1.

Fig. 3 shows a vertical array of dipoles connected by transmission line links according to the invention.

Fig. 4 shows another vertical array according to the "invention.

Fig. 5 shows a circuit diagram of one form of -.conn e ct-ing link which maybe used Fig. 4.

Fig.6 shows a diagrammatic view of two dipoles connected by a link and used as corresponding dipoles in a main and reflector array according to the invention.

:F g. 7 shows a diagram used to explain how the dipoles of Fig. 5 maybe mutuallyadiusted.

' Figs. .8 and 9 show dia rammatic details .of ano her tyne of main aerial array accompaniedby vention.

Fig.- 1 shows diagrammatically a fol ed dip l element used in carry ng ut the invention. ,It comprises two similar loops l and .2. hav ng rurin s and 4 to which. a transmission l ne (not shown) may be connected for feeding the dipole. A switch 5 is provided for connecting together the free ends of the two loops. The total length of each loop should be substantially equalto one wavelength at the mid-band frequency of one range, so that the overall length of each loop will be about half the said wavelength. When the switch 5 is open, the dipole functions as a full wave dipole over one band of frequencies, and when it is closed, it operates as a half wave dipole over another frequency band in which the frequencies are half those of the first band.

In order to obtain a pass band of the desired width for the aerial, it is necessary to proportion the dipole elements so that the impedance measured at the feed terminals has a substantially flat frequency characteristic over the operating range. This is done by suitably proportioning the characteristic impedance Z? of each loop regarded as a balanced pair transmission line, and the characteristic impedance ZA of each loop treated as an unbalanced transmission line consisting of the two limbs of the loop in parallel to earth. It is desirable that the impedance of the aerial should be low, so each loop may be made of wire cage construction as shown in Fig. 2, the wires being well spaced apart. The necessary impedance adjustments can be effected by adiusting the mean width w and depth d of the cage. Adjustments of w will principally affect ZP without having much effect on ZA so long as d is relatively large, so that there is some degree of independence of the two adjustments, which are preferably made by trial, and it is found that best dimensions for the loops are not very critical since the characteristics of the aerial change quite slowly in the neighbourhood of the best adjustments. It has been shown both by analysis and experiment that it is possible to obtain a good performance for both full wave and half wave operation over respective frequency bands in which the ratio of maximum to minimum frequency is 3:2 the phase angle of the impedance measured at terminals 3 and 4 not being more than 25, and the magnitude of the impedance not varying from the mean by more than 5:15 per cent. The terminal impedance of the dipole when adjusted in the manner described is substan tially the same in both frequency ranges, the

switch 5 being appropriately set in each case.

In the case of a vertical array of ordinary dipoles arranged one above the other half a wavelength apart, it is the usual practice to feed such dipoles by connecting them in tandem by means of half-wave transmission line links all of the of Z12, Z23, Z34 as indicated, and

-poles are matched in stages. duces substantially the same result as the known errors at frequencies near the ends of the bandi 7 become excessive. It can be shown that the feeding process according to the invention, which is about to be described with reference to Fig. 3, is far superior when the array is to be used over a wide frequency band.

Fig. 3 shows diagrammatically an array of four folded dipoles which may be similar to those described with reference to Figs. 1 and 2. These are supposed to have impedances of Z1, Z2, Z3 and Z4 as measured at the feed terminals. These dipoles are spaced apart by half the wavelength at the mean frequency of the operating band.

The transmission line connecting links have characteristic impedances at the mean frequency the main feeder line has a characteristic impedance Z0.

Then according to the invention the transmission lines are designed so that approximately:

In other words, the characteristic impedance of each link or line is chosen to be equal to the actual impedance which it feeds, so that the diprocess at the mean frequency of the band, it gives much better results at other frequencies in the band. It will be evident that the same principle is applicable to an array having any number of dipoles, and the characteristic impedance of the links should decrease progressively as the main feeder line is approached. In the case of dipoles which have been dimensioned as ex- -plained with reference to Fig. 2, so that their frequency response characteristics are substan- 4. In this figure, four dipoles 6, 1, 8, 9 are spaced Vertically above one another by a distance equal .to a half wavelength at the upper mean frequency 2f (Or a quarter wavelength at the lower -mean frequency 1). 8 are connected by a link transmission line The alternate dipoles 6 and of characteristic impedance Z13 equal to the terminal impedance Z1 of the dipole 6 at the mean .frequency. This link should be a whole wave- I length long at the frequency f (or two whole wvavelengths long at the frequency 2 ..Wise, the alternate dipoles 1 and 9 are connected by a link transmission line H of the same like length as It, and of characteristic impedance Z24 equal to the terminal impedance Z2 of the dipole 1 at the mean frequency. The dipole 8 is connected to the feeder transmission line I2 by a link transmission line l3 having a characteristic impedance Z01 equal to Zl3/2. and the dipole 9 is connected to the line l2 by a link transmission line l4 having a characteristic impedance Z02 equal to 224/2. The lengths of these While this prolinks l3 and I 4 should be such that the dipoles 8 and 9 are fed in the same phase: for example l3 could be two wavelengths long at frequency f, and i4 could be one wavelength long at frequency f: or, is could be of zero length and I3 only one wavelength long. Thecharacteristic impedance Z0 of the feeder transmission line should be approximately equal to if the input impedances of all the dipoles are equal to Z, then evidentl Z0 should be equal to Z4.

The spacings chosen for the dipoles in Figs. 3 and 4 are such that no vertical radiation components are produced at either of the frequencies f or 2).

Since the lengths of the links Ill and II are approximately twice the'distance between the dipoles which they connect they may be bent or strained outwards as indicated in Fig. 4. Al,- ternatively these link lines may be shortened by including series loading inductance coils with appropriate shunt capacities for halving the effective velocity of propagation, in the manner indicated by the circuit shown in Fig. 5. The coils I5 may conveniently be wound on porcelain formers which can act as strainers, and the condensers l6 can be conveniently provided by spacing insulators which can readily be designed to introduce the proper amount of localised capacity. The characteristic impedance of the low pass filter so formed should be designed to have the value according to the foregoing-explanation, and the filter should have a sufiicient number of sections to ensure that the cut-off frequency is well above the higher operating band.

When using an array of four dipoles arranged as described with reference to Fig. 4, it is possible to obtain satisfactory operation over two wave bands of the type 0.8f to 1.2f and 1.61 to 2.4! according to the position of the switches 5.

When an attempt is made to employ an array of broad bend dipoles with a reflector array either fed 01' unfed, it is found that with the spacing necessary to give the best results, the mutual impedance between corresponding dipoles of the main and reflector arrays is sufficient considerably to modify the operation, and satisfactory performance over a wide frequency band requires a mutual adjustment of the characteristics of the main and reflector arrays. This can in general only be obtained at the cost of appreciable reduction of the operating band width.

Fig. 6 shows one of the dipoles I I of a vertical I array connected to the corresponding dipole I8 of a reflector array by a quarter wave link I9 at the mean operating frequency. The switch in each of the dipoles will be either always open or always closed; in other words, the system can only be employed in one of the two bands of frequencies. because phasing cannot be arranged to give efficient operation of reflector at both frequencies. The dipole I1 is connected to the other dipoles of the main array (not shown) by links such as that indicated at 20. and the connections are made in the manner which has been described with reference to Fig. 3 or 4.

Each pair of dipoles such as IT and I8 is adjusted to obtain the desired band pass characteristics, and the variables available for this adjustment are ZP and ZA of each dipole as previously described. and in addition, the characteristic impedance of the quarter wave link l9. Control is also possible to a "small extent by adjusting the overall length of the dipoles.

While it is theoretically possible to determine by calculation the necessary adjustments which will take into account the mutual inductance between the two dipoles, it proves to be much simpler to find them by experiment. This may be done for example with a scale model in the manner shown in Fig. 7, which shows a model consisting of two half dipoles HA and I8A arranged as an unbalanced system with respect to earth and connected by an unbalanced quarter wave link HA. The model is mounted over a standing wave detector 2|, and is excited in a suitable manner and adjusted until the desired performance is obtained, after which the actual dipoles can be immediately designed by enlarging the scale in known manner.

Fig. 8 shows a modification of Fig. 6 in which the dipoles l1 and I8 respectively forming part of the main and reflector arrays are not directly connected together, the arrays being separately fed from the base in the manner indicated in Fig. 9. In Fig. 8, 22 and 23 represent the link lines connecting the dipoles respectively to'the other dipoles (not shown) in each array in the manner previously explained. The two dipoles l1 and 18 should be mutually adjusted (preferably experimentally) so as to obtain the desired band pass characteristic. In this case, of course, there is no direct link line so the only variable available for making the adjustment are ZP and ZA. In Fig. 9, 24 is the main feed transmission line, and 25 and 26 are the links respectively feeding the main array and the reflector array. The links are connected by a quarter wave line 21 in order that the reflector array may be fed at 90 with respect to the main array. An adjustable short circuited branch line 28 about a quarter wave long is provided for obtaining a satisfactory adjustment of the feed to the reflector array.

Fig. shows a modification of Fig. 6 in which the reflector array is unfed. The dipole I8 is spaced a quarter of the mean operating wavelength from the dipole l1 and is terminated by a load resistance 29 equivalent to the feeding load on the dipole H. The dipoles l1 and I8 should be mutually adjusted in the manner explained to obtain the desired band pass characteristic.

It will be understood that the dipoles shown in Figs. 6 to 10 could alternatively be operated on the other wave band with the switches 5 open, in which case the adjustments to take account of the mutual impedance between corresponding dipoles in the two arrays would be different.

What is claimed is:

An aerial array comprising a plurality of horizontal dipoles disposed one above another, each dipole having each of its arms composed of a plurality of wire loops which extend in horizontal planes, the uppermost and the lowermost plane being separated by a dimension d, and the loop portions which extend linearly from each dipole center being separated by a dimension w, a common terminal for corresponding ends of each loop in a given dipole arm, the other ends of each loop being interconnected, and switching means for at times interconnecting said other ends of the loops in the two arms of each single dipole, said dimension d being relatively larger than w and the dimension to being suitably chosen to control the characteristic impedance of each loop regarded as a balanced pair transmission line, whereby a substantially flat frequency characteristic is obtained over either of the two operating ranges that are provided by opening and closing said switching means.

ERIC OSBORNE WILLOUGHBY.

REFERENCES CITED The following references are of record in the flle of this patent:

UNITED STATES PATENTS Number Name Date 1,643,323 Stone Sept. 27, 1927 2,183,784 Carter Dec. 19, 1989 2,226,728 Lalande et al. Dec. 31, 1940 2,234,744 Thomas Mar. 11, 1941 2,267,889 Aubert Dec. 30, 1941 2,278,660 Lehmann Apr. 7, 1942 2,283,914 Carter May 26, 1942 2,283,938 McKesson May 26, 1942 2,338,564 Aram Jan. 4, 1944 2,350,916 Morrison June 6, 1944 FOREIGN PATENTS Number Country Date 378,652 Italy Feb. 20, 1940

Images