DOC 001 Stanley 1937 2
Recognised text
Text below was produced by optical character recognition and may contain errors. Check it against the original.
Page @ THE TELECOMMUNICATION JOURNAL OF AUSTRALIA June, 1988 signal due to the ground wave increases, but the indirect wave from the ionosphere increases, and the signal more readily appears at distances remote from the transmitter. There is a certain critical frequency above which the ionosphere does not return the wave to earth. This indirect wave causes fading at the receiver and may cause interference to other services if propagated to large distances; , (b) An increase in frequency causes a decrease in received field intensity due to the ground wave, but the indirect wave decreases, causing less severe fading, and less possibility of inter- ference to other services. It is thus necessary to make a compromise in the selection of frequency and in this case the frequency was chosen from a series of curves +20 cs} 3 3 & same |—w Oni NEE -80. Field Intensity db. relative to Imv/n ° 100 200 Distance - Miles. Fig. 2. such as shown in Figs. 2 and 3. Fig. 2 shows the field intensity expected at various frequencies for a radiated power of 1 kilowatt, the trans- mitter being 1700 feet above sea level and the receiver at sea level. Fig. 8 shows the gain obtained due to raising the receiver above sea level. The gain is given in decibels as a function of ha’, where.h is the height above sea level in metres and , is the wavelength in metres. From Figs. 2 and 3 for a height of trans- mitter of 1700 feet above sea-level and of re- ceiver 460 feet above sea level, distance 170 miles, the received field intensities at frequencies 30, 87.5 and 50 megacycles per sec, are respec- tively 71, 75.5 and 81.5 decibels below 1 millivolt per metre. "A frequency of approximately 40 megacycles per sec. was chosen as it was considered that the indirect wave at this frequency would be negli- gible, and the received signal would be appre- ‘ciably higher than that for 50 tmhegacycles per sec. Aerial Arrays: For normal broadcast frequen- cies 550-1500 kilocycles per second (wavelength 545-200 metres) the use of directional aerial arrays is usually not practicable because of the 40 ama pone] 2 co SOmC sete lam 30 - is “2 —|50mC| : La | % 0 20 A 3 = Po a 0 wo, 6 40 50 60. hrs Fig. 3. large spacing needed between elements of: the array. As the spacing needed is- usually of the order of one half wavelength, the required space would be large. In the ultra high frequency range, however, the use of directional arrays becomes practicable and desirable, because the overall dimensions of an array of several wavelengths extent are re- latively small. ; . At the time of this experiment, no experience was available as to the relative advantages of horizontal and vertical aerials for conditions simi- Big. 4. lar to the case under consideration, and it was decided to erect both horizontal and vertical arrays. The general principles of an array may be seen by reference to Fig. 4. “In Fig. 4, a, b, ¢, d, etc., are radiators separ- ated by a definite fraction of a wavelength, the currents in the various radiators bearing some simple phase relationship.. In the arrays used June, 1938 THE TELECOMMUNICATION JOURNAL OF AUSTRALIA Page 7 the currents in all elements were in phase and “the radiators spaced one half-wavelength apart between centres. The function of the array is to augment the radiation in a certain, desired direction at the expense of radiation in other directions. In Fig. 4 if N = the number of vadiators and ni is the spacing between adj acent radiators, the currents in all radiators being in phase, the field intensity in the horizontal plane at.an angle ¢ to the normal to the array is given by sin (N7n sin ¢) Es = EF ———_______ sin (wn sin ¢) where E = field intensity. due to one radiator. In Fig. 5 is shown a polar diagram in the horizontal plane of (a) a single half wave verti- eal aerial and (b) eight half wavelength aerials fed in phase and spaced one half wavelength (b) @) . “Radiators 1 i \ fo \ i 4 Fig. 5. apart in the horizontal plane. The gain.of the array as.shown in the maximum direction is 10 db over a single half wave aerial. By placing a similar set of radiators one quarter wavelength behind the array, radiation in a backward dirée- tiom is cancelled and the frontward radiation is increased by 3 db, giving a total gain over a single half wave radiator of 18 db. Fig. 6 shows a layout of the aerials at the transmitting site, both horizontal and. vertical arrays being shown. The shape of the radiation diagram for the horizontal array is essentially similar to that of the vertical array shown in Fig. 5. The aerial arrays at receiving and trans- mitting sites are similar, and were connected to the huts by open wire transmission lines. Transmitter: The transmitter consisted of an oscillator using 808 type tubes, the frequency being controlled by resonant lines arranged in a spiral form to reduce the required space. The transmitter was modulated by a unit using two 800 type tubes in the final stage, and either continuous tone or speech could be used. Power supply was obtained from a generator driven by a single cylinder petrol engine; the generator having windings for 3000 volts and 16 volts gave both the filament and plate supplies. A plate voltage of 1500 volts was used for the oscillator, the output under the operating condi- Horizontal Array | nn Vertical Array Fig. 6. tions being approximately 200 watts. Due to the gain of the array. in the. desired direction, the radiation was equivalent to 13 db above 200 watts (4 kilowatts) from a single half wave radiator. . - Receiver: Two receivers were used :— : (1) A super-regenerative receiver (commercial model) ; . (2) A super-heterodyne receiver (built in. the Research Laboratories). : Quantitative measurements of received signal were made by (a) measuring the change in plate current of the superheterodyne receiver with change of signal; and (b) ‘measuring-the reduc- smneey THE Mae. Or aes aegge er Wee —- Ww Fig. 7. tion in noise level of the super-regenerative re- ceiver with increase of signal. Fig. 8 shows the relationship between signal input and noise level in a super-regenerative receiver. For communication purposes a telephone line was erected between the receiver hut and the Stanley Post Office. The wires were of 200 Ib. copper and were used to transmit A.C. power to operate the receivers. Fig. 7 shows the arrange- ment for supplying A.C. power over the lines; 8 THE TELECOMMUNICATION JOURNAL OF AUSTRALIA June, 1988 alephone: circuit worked earth return and ower hum was found to be negligible in lephone when suitably balanced condensers ised. Results of Tests: eral: Receiving tests were made for a of 11 days, from 29/9/87-9/10/37 using \orizontal and vertical aerials, mn transmitting on the horizontal aerial, TNS eens amen ser ee 80! @ (0 20 50 40 30 60 90 5 ~go jue Input in db refatwe to' arbitrary level s. §.—Super-regenerative Recelver Noise Level plotted against signal input. eived level was approximately 30 db better horizontal than the vertical aerial. How- vhen transmitting on the vertical aerial, ‘ived signal was only 10 db better on the | than the horizontal. Occasionally when the vertical aerial for transmitting, the d signal was almost as strong on the tal as on the vertical aerial. This would 2 that a change of the plane of polariza-. as taking place—propagation favouring tal polarization. us: Horizontal polarization (horizontal ' proved to be consistently better than . polarization, the difference at all times least 20 db in favour of the horizontal. iminary theoretical investigations it was t that the horizontal aerial would radiate proportion of the energy at high angles horizontal (the calculated maximum en- as at 18° to the horizontal). The high f energy at an angle to the horizontal is by wave interference between directly ited radiation and radiation reflected from und, and the above figure of 18° was ob- assuming a flat earth and perfect re- from the earth. At Tanybryn, however, sund sloped downwards away from the nd dropped suddenly about 100 feet out, the flat earth theory would not apply aly, and an increased signal in the hori- at the expense of high angle radiation ie expected. _Fading: The received signal varied in inten- sity throughout the day but the variations were in all cases of slow period and no distortion was noticed even when the received field was rela- tively low. At no period during the tests was there a complete absence of signal when trans- mitting and receiving on the horizontal aerials. Fig. 9 shows a curve.of received field inten- sity during typical listening periods, and it is of interest to note that the maximum variation expected due to transmitter fluctuations was approximately 1 db. Effect of weather, time of day, ete.: During the period of the tests the weather had no noticeable effect, but it is possible that seasonal changes may be noticed. This will be investi- gated by long period observations on the channel. The signal intensity at night appeared to-be the same as during the daytime, the fading noticeable on broadcast and high frequency pro- pagation being absent. Summary: The tests indicated that communi- cation is possible over distances greater than optical when using a frequency of approximately 40 megacycles per sec. but the long period re- liability of such a circuit has yet to be estab- lished. The present intention is to instal at Stanley, Tasmania, a transmitter of 200-300 watts power and to take long period recordings toes EWAN Tees tiay, |s-so-p7, |Horiontall Reng ls Signal miensity Lb above 1muerevolt. £24, AX “CROW aa V\ M +32 Wed. jO"03; Hore ertaf| Aeriays: Fig. 9.—Variation of recelved signal intensity ' over typical Listening periods. of the field strength received in Victoria. The transmitter will be crystal controlled to prevent any appreciable frequency change, and the re- ceiver will be a superheterodyne type in which the automatic volume control voltage will be used to control the current through a triode tube operating a recording milliammeter. Horizontal aerials will be used at both the, transmitter and receiver. Recently the transmitter and receiver were completed to provide an emergency service, but the tests will be continued. telephone cables. June, 1938 THE TELECOMMUNICATION JQURNAL OF AUSTRALIA ‘Page 9 A SUBMARINE CABLE TRACER It is almost a platitude to state that submarine cable repairs are expensive. It is therefore im- portant that every effort should be made to facilitate repair work and to make best use of the time for which the repair ship is chartered. ‘As bad weather is the principal cause of delay, any means of using this time to assist the repair js well. worth while. This can often be done with the apparatus about to be deseribed. When a submarine cable is laid, its position ig carefully charted and splice sheets are pre- pared to show the position of splices and the percentage of slack laid over various portions of the route. Owing to tides, currents and sub- marine disturbances, it is possible for a cable to be moved and so change its course or from similar causes it may be buried. Naturally, either of these effects would increase the time of grappling unless some means is available for detecting the true course of the cable from the repair ship.. This facility is provided by the Submarine Cable Tracer which will also show the actual position of an earth fault, thus acting as a confirmation of shore tests. It is particular- ly useful when land sights are not available for navigation. Such equipment has been developed by the Research Section of the British Post Office and has been applied to our own requirements for work in Bass Strait, both on the telegraph and The essentials: are a low fre- quency generator and a pair of search electrodes Fig. 1—Cable Tracer Equipment. to which are connected a tuned amplifier and galvanometer, these being shown in the photo- graph herewith. It has been shown that the lower the frequency, the greater the ‘sensitivity of the method. A 16.5 p.p.s. generator with an output of 250 m.A. is connected to the cable and earth at the cable hut so. that current paths follow the cable route between the hut and the earth fault (or earth connected to far end of cable) in a fashion similar to lines of force be- G.N. Smith, B.Se. tween two unlike magnetic poles. The -current paths may be considered as parallel to the cable except near the fault and there are equipotential surfaces perpendicular to these paths, that is at right angles to the cable. . Any two electrodes in the current paths will have a potential difference between them pro- vided they are not both in an equipotential sur face. The measurement of potential drop. is consequently made along the line of these current paths, the maximum deflection being obtained when over the cable. It is not necessary, how- ever, in practice for the electrodes to be towed exactly along the line of the cable, any crossing up to an angle of 40 degrees being generally satisfactory. A correction is then applied (read- ing multiplied by cosine of angle of deflection) wen aw’ ZEAE METER SOUECH DO ONLY DURING READINGS Fig. 2. to reduce the observed reading to a direction along the lie of the cable.. As the ship pro- gresses along the cable towards the earth fault the deflections increase rapidly as the fault is approached and: cease abruptly beyond it. The electrodes may be towed “in line” about 30 fathoms behind the ship or, as is now preferred by the British Post Office, “abreast” of the ship by booms’ over the side so that they are in a line at right angles to the ship. In the latter case it is necessary, of course, for the ship to cross and re-crogs the cable approximately: at right angles. : These electrodes consist of two brass log rotators shorn of their fins, with a separation of 10 fathoms between electrodes. Electrical contact is maintained by means of cab-tyre flex. In the case of the “in line’ method the tow line is connected to a draw-bar spring and: the elec- trodes are weighted, thus preventing sudden ~ movements and consequent spurious deflections on, the amplifier galvanometer. The B.P.O. is now experimenting with electrodes which dive deeper as the towing’ speed increases thus re- ducing the tendency for the electrodes to break surface. It has been shown that the sensitivity of the method is independent of the depth of the