# Патент USA US3068339

код для вставкиDec. 1l, 1962 ` 3,068,329 R. W» DE MONTE ETAL NEGATIVE-IMPEDANCE REPEATER Filed April 28, 1959> 3 Sheets-Sheet 2 FIGS NEG/„VVE /MPEDANCE CON VERTE/P /B V »VEGA r/vE- f /MPEDANCE CONVERTE/P /4 b5 /0 BR/DGED-T NETWORK TERM/NAL /MPEDANCE-Za/ F0f? SER/E5 0R BRIDG/NG BRANCH FÍ G. Í0 5,5% i TERM/NAL /MPEoA/vcE-z„2 Foi? ssn/Es on @maal/v6 @RANCH /A/VE/VTORS ??. W DE MONTE w J KOPP BWÍMM ,4T TORNÈV United êtates Èatent QQ@ 3,068,329 Patented Dec. 11, 1962 l 2 must be seen looking in either direction from these ter 3,068,329 minals. For this condition, lalEGATlWV-MPEDANCE REi’EATER Robert W. De Monte, Berkeiey Heights, NJ., and William J. Kopp, Richmond Hiil, NX., assignors to Beil Tele phone Laboratories, incorporated, New York, NX., a ZI-i-‘Zg t’dlnh Ü where 0 is the transfer constant of the repeater 1, and will corporation of New York be negative in sign, and Z1 is its image impedance. By Filed Apr. 28, 1959, Ser. No. 809,421 6 Ciaims. (Ci. 179-170) image impedance is meant one of the two equal iin pedances which will simultaneously terminate the two This invention relates to wave transmission and more pairs of terminals 2_3 and 4_5 in such a way that, at particularly to a two-way, negative~impedance repeater each of these pairs of terminals, the impedances in both adapted to operate between unequal impedances. eliminate reflection at one end of a two-way repeater directions are equal. Equation 1 is essentially the same as Equation 48 on page 137 of the book by K. S. John son, entitled “Transmission Circuits for Telephonie Com operating between unequal terminal impedances, An munication,” published by Van Nostrand Co., New York, other object is to widen the band over which an im 1925. However, we have set the driving-point impedance Z equal to the terminal impedance Z1 and -use the symbol The principal object of the invention is to reduce or pedance match is obtainable. Z2 for the other terminal impedance ZR. As indicated by the arrow in FIG. l, the driving-point impedance Z of Two-way repeaters are often required in wave trans mission systems such as loaded voice-frequency transmis sion lines. These may be located at the end of the line 20 the repeater 1 is the impedance seen at the terminals 2_3 when the impedance Z2 is connected betweenthe terminals 4_5 and the impedance Z1 is removed. ñection effects and singing in the system, it is important or at an intermediate point: In order to reduce >re »From Equation 1, that the repeater should present a good match to the terminal impedance, at least at one end, over a broad band of frequencies. This is sometimes difficult when 25 the repeater operates between unequal terminal im~ pedances, especially if one or both of the impedances are complex. and ` Zïmiiïa tana @ii/[2 man vii +2125] (3) In accordance with the present invention, a good im pedance match over a band of frequencies is obtained at 30 It is seen from Equation 3 that there are two choices of i one end of the repeater by a special choice of its image impedance. The required image impedance has one of to be made in evaluating Z1. Therefore, Z1 may have four values which are determined by the terminal im~ any one of four different values. These `are determined by choosing both i signs as -|-, both as _, the first as -|-- pedances and the gain of the repeater. The repeater may be built as a lattice, a bridged-T, or any other equiva 35 and the second as ---, or the first as -- and the second lent structure, generally requiring two or more negative impedance converters. The nature of the invention and its various objects, features, and advantages will appear more fully in the following detailed description of the typical embodiments 4 illustrated in the accompanying drawing, of which: FIG. 1 is a block diagram showing a negative-im pedance repeater in accordance with the invention operat ing between unequal impedances; FIG. 2 is a set of graphs showing, the frequency char acteristics of the resistance and reactance of terminal im pedances assumed as an example; FIG. 3 shows therresistance RIA and reactance X12 of one image impedance ZIA suitable for the repeater; FIG. 4 shows a symmetrical lattice network and FIG. 5 a balanced bridged-T network suitable for the repeater of FIG. l; FIGS. 6 and 7 show impedances suitable, respectively, for the impedances _Z2 and -Zb of FIG. 4 to realize as -|-. However, in order to facilitate the synthesis of the network, the signs are preferably so chosen that the real part of Z1 is positive in the frequency range of in terest. An example in which the terminal impedances Z1 and Z2 are both complex will not be presented. The broken-` line curves of FIG. 2 show the resistance R1 and the reactance X1 of the impedance Z1. The solid-line curves show the resistance R2 and the reactance X2 of the other impedance Z2. These characteristics are plotted over a frequency range of 100 to 10,000 cycles per second, on a logarithmic scale. The impedance Z1 is typical of that encountered at the oiiice end of a telephone cable. The resistance R1 is constant at 900 ohms and the re~ actance X1 is that of a capacitor having >a value of two microfarads. The impedance Z2 represents that of a long, loaded, 22-gauge cable with a building-out network at the near end and an image-impedance termination at the other end. It is seen that there is some irregularity in R2 and X2 in the neighborhood of 3,500 cycles, the cut-olf fre quency of the cable. FIG. 8 presents a graphic comparison of the im It will be assumed that the repeater 1 has a uniform pedance of the terminated repeater and the terminating gain of 0.7_ nepers (about six decibels) and negligible impedance to be matched; phase shift over the band of interest. Therefore, the FIG. 9 shows the resistance RIB and the reactance XIB of another image impedance ZIB suitable for the repeater; 60 transfer constant is and FIGS. 10 and l1 show impedances suitable, respec One possible image impedance ZIA is now found from tively, for the impedances --Z„7 and -Zb of FIG. 4 to Equation 3 by substituting _0.7 for 0 and choosing both realize ZIB. i signs as -|-. The curves of FIG. 3 show the required FIG. l shows a two-way, negative-impedance repeater resistance R111, which is positive, and reactance X111, which 1 with an impedance Z1 connected to the terminals 2_3 is negative. and an impedance Z2 connected to the terminals 4_5. The next step is to synthesize the repeater network. The impedances Z1 and Z2 are unequal and either or both FIG. 4 shows one suitable configuration in the form of a may be complex. symmetrical lattice structure with two equal series It will be assumed that reñection is to be minimized 70 branches each of impedance Za and two equal diagonal at the terminals 2_3. Therefore, the same impedance branches each of impedance Zh. To simplify the draw ZIA; 3,068,329 n à ing, only one series and one diagonal branch are shown in detail. The other branches are indicated by broken ¿i in series, as shown in FIG. l1. Of course, the approxi mation may be made more exact by adding elements to lines connecting the appropriate terminals. These imped ances are found from the image impedance and the trans _Zag Or _2112. In the two embodiments described, the required values fer constant by using the relationships: of the resistors in ohms and capacitors in microfarads are Z..=ZIA caring as follows: (5) R1 ___ 273 R2 ____ 0 z_b;Z,rA 09th '2" (6) 725 R3 2485 10 R., Since 0 is negative, both Z,L and Z1, will have negative real parts. Impedances with negative real parts are easily obtained by means of four negative-impedance con verters, each having an impedance conversion ratio ap proximately equal to _1. Each series branch of the lat tice includes such a converter 8 terminated in an imped ance _2a. Each diagonal branch comprises a converter 9 terminated in an impedance _2b. FIVG. 5 shows a balanced bridgedfi" network, equivalent 20 to the lattice of FIG. 4,- which may be used for the re peater 1. The bridging branch comprises a winding 11 closely coupled to each of the two series windings 12 and 13 in the two sides of the line, and a negative-impedance _____ 1232 R, ______________________________________ __ R6 __ ___ C1 ___ _ C2 ___ 300 2660 2.33 ___ ___ 0.416 C3 _________ __ _ C4 ___ 0.432 ___ 0.664 ments are only illustrative of the application of the prin ciples of the invention. Numerous other arrangements may be devised by those skilled in the art without depart ing from the spirit and scope of the invention. What is claimed is: 1. An active transducer adapted to operate between a converter 14 terminatedy in an impedance _2a. The. 25 terminal impedance Z1 at one end and a terminal imped shunt branch, connected between- the midpoints of the ance Z2 of different Value at the other end> substantially series windings 12 and 13, comprises a second negative without reflection at the one end, the image impedance impedance converter 15 terminated in an impedance of the transducer being approximately equal to -Za z Gc/aod simulation of the impedance _Z,v over the band 30 of interest may be provided by the impedance branch _2,11 shown in FIG. 6, comprising a resistor R1 in series with the parallel combination of a second resistor R2 and a capacitor C1. The impedance _Zb or _Zh/2 may be simulated satisíactorily by the branch _21,1 shown in 35 FIG. 7, which comprises the series combination of a re sistor R3 and a capacitor C2 in series with the parallel where 0 is the transfer constant of the transducer. 2. A transducer in accordance with claim 1 in which one of the terminal impedances is complex. 3. A transducer in accordance with claim 1 in which 0 has a negative real part. 4. An active transducer adapted to operate between a combination o_f a resistor R4 and a capacitor C3. One or terminal impedance Z1 at one end and a terminal imped more of the component elements may be made adjustable, ance Z2 ojf different value at the other end substantially as indicated by the arrows, to permit an adjustment of 40 without reflection at the one end, the image impedance the repeater gain or to allow for changes in the imped of the transducer being approximately equal to ancesZ1 and Z2. The curves of FIG. 8 show how well the driving-point impedance Z of the terminated repeater 1 matches the terminal impedance Z1. The solid-line curves R and X ì/íZ tanh 0] +Z1Z2ïi ‘ Z2-Z1 2 where 0 is the transfer constant of- the transducer. are the resistance and the reactance, respectively. The 5 ._A transducer in accordance withclairn 4 in which one resistance R1 and reactance X1 of the impedance Z1 are of the terminal impedances is complex. plotted in broken-line curves for comparison. It is seen 6. A transducer in accordance with claim 4 in which that the resistive match and the reactive match are both á' has a negativereal part. excellent over the voice band, and are close enough out 50 side ofthe band- to prevent singing. Of course, the match References Cited in the iile of this patent can> be made closer by adding elements to the impedances UNlTED STATES PATEN’IS »Z311 and _2131. FIG. 9 shows another possible image impedance Z113, 2,582,498 Merrill ____________,_____,___ 1an. l5, 1952 foundv from Equation 3 by choosing both i signs as _. , Here, also, the resistance R113 is positive and the reactance X113 is negative. This image impedance may be closely approximated in the lattice network of FIG. 1 if the impedance _Z1l is simulated by the impedance branch 2,685,066 2,694,184 Barney ___.,_____ ______ __ July 27, 1954 Rounds _______________ __ Nov. 9, 1954 2,742,616 2,788,496 2,844,669 Merrill ____ _____,____,_,___ Apr. 17, 19,56 Linvill _____ ____ __~______ Apr. 9, 1957 Arndt ________________ __ July 22, 1958 _Z112 shown in FIG. 10, which is simply a resistor R5, 60 2,878,325 Merrill _ ________ -______ Mar. 17, 1959 2,904,641 Radcliíïe _ ____ ____ ____ __ Sept. 15, 1959 and the impedance _Zrb is simulated by the impedance branch _Zbz comprising a resistor R6 and a capacitor C1

1/--страниц