Tournai of Research of the National Bureau of Standards Vol. 46, No. 2, February 1951 Research Paper 2182 Effect of a Metal Mast and Guy Wires on the Performance of the 600-0hm Multiple-Wire Delta Antenna Harold N. Cones This report describes the resul ts of measurements made to determine the effect of a metal mast and guy wires on t he radiation pattern a nd the effect of guy wires on t ile ter­ minal impedance a nd the radiation e ffi ciency of the 600-ohm mult iple-wire delta antenna. Model techniques were used to obtain radiation patterns of identical antennas s upported in vario us ways. Curves are presented showing t he ter mi nal impedance over t he frequency range 1 to 25 megacycles of full -scale a ntennas wit h a nd wi t hout g uy wires. A graph of t he quasi-radiatio n effici ency of t he antenna as a fun ction of frequency is s hown . I. Introduction at 25 :Nfc. Pattern measurements were made over a grou nd mat of metallic mesh. When the model The 600-ohm multiple-wire delLa antenna 1 (fig. 1) was supported by a metal masL, Lhjs mast was used in ionosphere studies requires only a single electri cally con nected to the ground mat. large wooden mast. This mast, which is lc s than The model an Lenna was used as a receiving an­ 80 ft. long, can ordinaril y be procured and erected te nna. A baLLery-operated high-frequency trans­ in most locations. However, in some locations it mi tter mod ulaLed at an audio frcq uency was mounted may be desirable to u.se a sectionalized metal mast. at Lhe apex of an A-frame and moved abou t the As the mas I, lies in a neu Lral plane of Lhe antenna, model in a semicircular arc of 13 ft. in radiu s. The it wo uld appear that if the antenna was electrically angular displacement of the Lransmitter wiLh res pect and mechanically symmeL rical and was fed by a to grou nd was Lr ansferred to Lhe t urnLable of a balanced transmitter there would be no radiation recorder by means of a sclsyn system. The signal from a metal supporting mast. It is quiLe im­ received by the model was reclin ed and th e res ul tlng probable however that such a compleLely balanced audio-frequen cy volLage passed through a seri es of condition will prevail. Any unbalance in Lhe sys­ selective amplifiers and se rvomechanism Lo cause tem will cause a residual current flow in Lhe mast, a radial displ acemen t of a pen on the recorder which at resonance could become quite large caus­ tu rntable. Thus as Lhe transmitter was moved ing appreciable r adiation. I t would also seem through an arc of ]80 deg. , Lhe verLical radiation desirable to usc guy wires to help support eiLher a pattern of the anLenn a was automatically ploLted. wooden or m etal mast. Because the effecLs of a F igure 3 is a diagram of the model range showi.ng metal mast and guy wires on the radiation pattern the paths of th e target transmi tter with reference to and impedance of th e delta antenna are d iffi cult the antenna up.der test. The model antenna was to predi ct analytically , experimental measurements located in the X -Z plane. Radiation patterns were were made to determine these eff ects. This report measured in. th e X-Z and Y-Z planes. In measuring describes the results of these measurements. patterns in the X-Z plane the polarization of t he transmitting antenna was always such that the II. Instrumentation electric vector was tangent to the m eridian arc XZX'. In measuring patterns in. th e Y-Z plane, the Radiation pattern measurements were made by electric vector was always parallel to the X axis. using model techniques. That is, all dimensions The patterns obtained show the relative power of the antenna were scaled down by a certain factor th at would be radiated by the antenna at different in Lhis case 30, and th e frequency was scaled up by vertical angles in a given plane for a given frequency. the same factor. The model antennas were sup­ The patterns were normalized, that is, the gain of ported by nonconducting or metal masts of a diam­ the pattern plotter was adjusted for each frequency eter corresponding to 1 ft. in diameter on a full-scale to give a full-scale deflrction at the point of maximum antenna. Since ordinarily a single mast supports gain. Pattern m easurements were made only for the one transmitting and one receiving delta antenna frequencies equivalent to 14 through 25 Mc because, mounted at right angles to each other, guy wires with the scale factor of 30 which was used, the model when used were spaced as shown in figure 2. In range was too small for measurements at lower all cases the guy wi.res were insulated from the pole frequencies. and from ground. Wnen continuous guy wires Impedance measurements of full-scale antenna were used . these wer e the only insulators. The were made with a balanced recording impedance sectionalized guy wires had insulators at a spacing meter that automatically plots the absolute magni­ equivalent to every 776 ft., or about 0.2 wavelength tude of input impedance of an an.tenna over the frequ ency range 1 to 25 Mc. A block diagram and I II. N . Oones, H . v. Oottony, a nd J . M. Wa tts, A 600·ohm multiple·wire delta antenna for io nos phere studies, J . Research N BS 44, 475 (1950) RP2094. brief description of this instrument have been 11 3 published.2 Basically, it consists of a sweep fre­ tinuous guy wires. At this frequency the long guy quency generator with' a constant-current output, wires are 2 wavelengths long and the short guy the voltage across the output terminals being, wires are 1 wavelength long. Also at this frequency, therefore, directly proportional to the impedance the metal mast is approximately 1% wavelengths impressed across the output terminals. The instru­ long. From 22 through 25 Mc the patterns are very ment is calibrated by substituting noninductive nearly alike, although the use of the sectionalized resistors for the unlmown impedance. The absolute guy wires results in a slight change in the radiation magnitude of the unknown impedance at any pattern at 24 and 25 Mc. frequency can be determined by interpolating The difference in the radiation patterns observed between the calibrating lines. at 15 Mc between antennas employing nonconduct­ This instrument can also be used to obtain an ing and metal masts (both without guy wires) is at­ indication of the radiation efficiency of a nonresonant tributed to the resonance of the mast close to this antenna. The curve plotted by the impedance frequency. This effect would not be observed ex­ meter when an antenna is connected to its output cept for the existence of an exciting voltage that terminals shows in arbitrary units the voltage across would have to arise from an unbalance incidental to the antenna imput terminals. The curve also shows the model antenna system under test. Therefore, the impedance of the antenna, as the voltage is as the unbalance is likely to vary in different installa­ directly proportional to impedance. Since the input tions it may be expected that the resonance effects voltage and the input impedance of the antenna are discussed above will likewise vary in magnitude. lmown, the input power (in arbitrary units) can be Figure 5 shows comparative radiation patterns computed for frequencies at which the antenna is in the Y-Z plane for all four methods of support. purely resistive. These are, for all practical pur­ At anyone frequency the patterns are substantially poses, the frequencies at which the impedance is alike. either a maximum or a minimum. If measurements It is seen that tho use of continuous guy wires is are made of the voltage across the terminating very effective in suppressing side lobes in the X -Z resistor, the power dissipated in the resistor (in the plane at 14 and 15 Me and to some extent effective same arbitrary units) can be determined. Knowing in reducing low-angle radiation at other frequencies. the power input to the antenna and the power They do not appear to affect the pattern in the Y-Z dissipated in the terminating resistor, a quasi-radia­ plane. tion efficiency of the antenna (taken as the quotient of the power input less the power dissipated in the IV. Results of Impedance Measurements terminatin.g refistor and the power input) can be evaluated. This quasi-radiation efficiency always Impedance measurements were made of full-scale exceeds the true radiation efficency, since it in­ antennas. As no metal mast was available, tests cludes the copper, ground, and dielectric losses as were limited to determining the effect of continuous power radiated. guy wires on the terminal impedance of the antenna. The following impedance measurements were III. Results of Radiation Pattern made: 1. Measurements of an antenna without guy Measurements wires; 2. measurements of an antenna with long continuous guy wires, that is, guy wires fastened to Radiation pattem m easurements were made of the top of the mast; 3. measurement of an antenna four identical antennas supported by a noncon­ with long and short continuous guy wires, that is, ducting mast, by a metal mast, by a metal mast and guy wires fastened to the top of the mast and guy continuous guy wires, and by a metal mast and guy wires fastened to the mast 35 ft above the ground.