US1694248A - Telephone line - Google Patents

Description

Dec. 4, 1928. 1,694,248

A. 'BYK TELEPHONE LINE Filed Jan. 5, 1924 F .l. V J. 1:?

Fig.2.

0 4,1 a: 0,8 44- 4! a 0.7 q: ,9 40% a q 0,2 0,5 0,4 as a, 7 0,5 9 0% Invent or Alfred Byk,

His Attorney Patented Dec. 4, 1928.

UNITED STATES PATENT QFFICE.

ALFRED BYK, OF CHARLOTTENBURG, GERMANY, A-SSIGNOR [T GENERAL ELECTRIC COMPANY, A CORPORATION'OF N'EW'YORK.

TELEPHONE LINE.

, 7 Application filed January 5, 1524, Serial No. 684,845, and in Germany January 30, 1923 In telephone, especially long distance telephone engineering, it is often very important to reproduce the impedance of a line. Loaded lines can bereproduced as such by means of circuit elements connected on the other side of the inserted apparatus, such for example as amplifiers, on the artificial line side and this reproduction is facilitated as much as possible by allowing the long ,line to-begin at 10 a suitable point of the periodic interval. 1f the ohmicresistance of the real line is disregarded, which is permissible for all practical purposes, the line can be extended by means of a condenser connected in parallel with it.

Kiipfmiiller in German Patent No. 330,96d proposed a changeof the real line exceeding the mere choice-of its starting :point within the periodic interval. By connecting an oscillating circuit and a self-inductance at the end of the real line he changed its impedance into that of a. homogeneous line the real part of which is independent of frequency. Now the kind of change leading to homogeneous lines is obviously important, but there is no basis for making an exact reproduction. The object of my invention is to obtain, with the same number of circuit elements especially for higher frequencies, a much more exact reproduction than that obii tained by Kiipfmiiller, if we disregard the demand for provisions for homogeneity which is unimportant for the purpos at hand.

The novel features which I believe to be 3 characteristic. of my invention are set forth with particularity in the appended claims. My invention itself, however, will best be understood by reference to the following descript-ion takeninconnection with the accon panying drawing, in which 1 shows diagrammatically one embodiment of my tion, and Figs. 2 and 3 show characterise curves of telephone lines illustrating the improvement which may be obtained by means of my invention.

Let us start with the loaded line, which is ordinarily used, beginning with half the distance between coils, aline whose ohmic resistance we can practically neglect as a rule. According to K. W. Wagner (Archiv. f.

cy dependence of the real part can be reproldlelztr. V. II, 1919, p. 76), this has pure real impedance,

if weinsert the cut off frequency,

w 2 Vi L indicates the total self-induction, C the total capacitance of acomplete section of a line, to, 2n times the frequency. This impedance must be changed so that the frequenduccd as exactly as possible by a small number of elements at the artificial side of the line, whereby the imaginary part of the impedance must disappear as before for all frequencies; The change in the frequency dependence of the riial part is effected by means of a shunt condenser Q of suitable capacity connected to a loaded line A as shown in I 1, while the undesirable imaginary impedance part, caused by this connection, is oi'npensated byan oscillating circuit 6 L, at he side of the real loaded line. l Vhen the hunt condenser has suitable dimensions the real impedance of the changed line can be reproduced by an ohmic'resi stance B on the artificial line side, which is connected in parallel with a capacitance C, and a self-inductance L,- in series. Since this connection on the artificial side of the line produces a new imaginary impedance-component, the oscillating circuit O L, on the real side of the line will have to be calculated so that it not only compensates the imaginary component of the impedance introduced by the shunt sndei but will also reproduce at the same time in the real line the imaginary component of the impedance of the reproduction R C 'L V r The ii'npedance lti of-the loaded line supplemented by the shunt condenser Q with the impedance i s as follows:

if the real and imaginary parts are simplified. Inserting the value of W from (1) we have:

1 I f we choose y We have 1" 4;1 0, and. 62' has the form of the impedance of an oscillatin circuit which 1s. is well known b 7 1 l as follows:

L-w 1 L C 20 If we use: L=L -y the oscillating circuit L, O on the real side,

will exactly compensate the imaginary impedance bi introduced by the shunt condenser.

The real part of the impedance of the loaded line supplemented by Q is, according to Y rer) For 3 whereby the denominator of (7) equals 1,adecreases as the frequency increases and the same relation obvlously occurs when 1 from smaller values y approaches the value A real part of the impedance of this kind, which decreases with the frequency can be reproduced by a suitably calculated circuit W R C L of Fig. 1. Tests show that the best reproduction of the real part of the impedance can be produced by means of the following data:

Here

, 2H mm indicates the natural frequency of the oscillating circuit composed of C and L Due to the imaginary impedance part introduced with the circuit 11 L 0 the oscillating circuit L G as assumed previously in Equations (5) and (6) has to be altered. It is found that this oscillating circuit, in ad dition to compensating the imaginary impedance part of the loaded line supplemented by Q also performs the necessary reproduction of R L C if one uses the following equations with (8) to (11):

R 279, L 0.24 henry,

(R is the ohmic resistance of a complete periodic interval of the cable.) For the sake of comparison let us also compute the repreduetiou of the cable according to Kiipfmiillers method, mentioned above.

Taking into consideration the ohmic resistance the in'ipedance of the loaded cable beginning with half the distance between coils is (see K. V. WVagner, l. o.)

With the quantity ratios concerned, this might be writteniaccording to Kiipfmiiller as follows (see 1. 0., p. 2)

/ i Co al J c t 1 1 1 If, however for the changing of '(15) Kiipfmiiller uses the following approximation formula:

it is not permissible for the frequency range concerned for all practlcal purposes, because for 08, a value whlch is prominent as a limit for balancing the impedance, 1t 1s for instance which means that one side of Equation (16) is about three times larger than the other.

The absolute value or" tl'iequotients of the imaginary and real parts under the radical in (15) whereas is so small for all values of to which have to be taken into account that We can practically disregard its higher powers. Atxthe limits of the l'requency range concerned 7? =0.2and

% 0.9, A is 0.0366 er 0;041 1,that "is, Sllffi- O ciently small. One can easily be convinced that A assumes no larger values in the intermediate range. The extreme of A is of course expressed "by the condition:

whereby =arc tan A or I s1n 2 2 cos 2 1 aretrue. Introduced into (18) if simplifications are made for the real and ln'iagmary parts of VV'.

After switching 111 any kind of a shunt condenser Q the loaded line has the following impedance (dwhm or, if we insert:

Era-m If, as is permissible terms higher than the this gives the following:

Fig. 2, curve I shows the ath of the real part of the impedance of the hineland cable changed according to the present invention by a shunt condenser and oscillating circuit,

w as a function of big. 2, curve II shows the path of the corresponding reproduction. Fig. 2, curves III and IV illustrate, with the sign altered, the imaginary part of the impedance of the changed cable as Well as the imaginary part of the impedance of the reproduction. Fig. 3, curves I to IV give the corresponding magnitudes for the Rhineland cable reproduced accordin to Kiipfmiillers method (shunt condenser I 0.315C oscillating circuit of the change, L, =0.315L.,,

comparison with Ki'ipfmiillers, especially with higher frequencies.

TABLE.GomparaM1Je reproduction error.

Accord- According to g to this in Kiipfvention mullet w Per cent Per cent E; 0. 2 2. 8 0.23

w a 0. 4 l. 7 0. 74

In the above table the comparative error in reproduction made according to the method described here and according to that given by Kiipfmiiller is recorded. If Z is the complex impedance of the total changed line on the real side, and N that of the reproduction on the artificial side, the following equation represents the comparative reproduction error:

The superiority of this method is again evident in the values of V especially with higher frequencies.

As the calculations with the Rhineland cable show, it is usually unnecessary to consider the ohmic resistance in reproducing a line. If one wishes to do this however, in cases of especially large ohmic resistance, the imaginary part of the impedance dependent on the ohmic resistance may be compensated by a self-inductance and capacitance.

lVhat I claim as new and desire to secure by Letters Patent of the United States, is,

1. In a telephone line loaded with inductances, an arrangement for changing the impedance of the line to facilitate impedance simulation, said arrangement comprising means to build out the terminal of the line to a fraction of a section intermediate between a mid section and a full section so that the real art of the impedance of the line may be simu ated by the real part of the impedance of a network comprising resistance shunted by capacity and inductance in series, and compensating means connected to the terminal of the line, said compensating means being such that its impedance has no real component and its imaginary component compensates for the difference between the imaginary components of the impedances of the built out line and simulating network, respectively.

2. In a telephone line loaded with inductances, an arrangement for changing the impedance of the line to facilitate impedance simulation, the said arrangement comprising a capacity shunted across the terminal of the line to build out the terminal of the line to a fractional section intermediate between a mid section and a full section so that the real part of the impedance of the line may be simulated by the real part of the impedance of a network comprising resistance shunted by capacity and inductance in series, and compensating means connected to the terminal of the line, said compensating means being such that its impedance has no real component and its imaginary component compensates for the difference between the imaginary components of the impedances of the built out line and simulating network, respectively.

3. In va telephone line loaded with inductances, an arrangement for changing the impedance of the line to facilitate impedance simulation, said arrangement comprising means to build out the terminal of the line to a fraction of a section intermediate between a mid section and a full section so that the real part of the impedance of the line may be simulated by the real part of the impedance of a network comprising resistance shunted by capacity and inductance in series, and compensating means in series with the line at the terminal thereof, said compensating means comprising an oscillating circuit whose im- IOU pedance has no real component and the imaginary component of Whose impedance is such as to compensate for the difference betWeen the imaginary components of the impedances of the built out line and simulating network, respectively.

4. In a telephone line loaded with inductances, an arrangement for changing the impedance of the line to facilitate impedance simulation, said arrangement comprising a capacity shunted across the terminal of the line to build out the terminal to a fraction of a section intermediate between a mid section and a full section so that the real part of the impedance of the line may be simulated by the real part of the impedance of a network comprising a resistance shunted by capacity and inductance in series, and compensating means in series With the line at the terminal thereof, said compensating means comprising an oscillating circuit Whose impedance has no real component and the imaginary component of Whose impedance is such as to compensate for the difference between the imaginary components of the impedances of the built out line and simulating network, respectively.

In Witness whereof, I have hereunto set my hand this 12th day of December, 1923.

ALFRED BYK.

Certificate of Correction.

Patent No. 1,694,248. Granted December 4, 1928, to

ALFRED BYK.

It is hereby certified that error appears in the printed specification of the above numbered patent requiring correction as follows: Page 3, lines 13 to 15, for the formula $3 1 read Same page, lines 23 to 25, for formula and that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 15th day of January, A. 1). 199-9.

[SEAL] M. J. MOORE,

' Acting Commissioner of Patents.

read

Images