US1615259A - Telephone repeater system - Google Patents

Description

A. B. CLARK TELEPHONE REPEATER SYSTEM Filed March 18, 1921 Jan. 25 1927.

Patented Jan. 25, 19,27,

y UNITED STA TESy PATENT OFFICIEL ALVA B'. CLARK, OF BROOKLYN, NEW YORK,. ASSIGNOR TO .AMERICAN TELEPHONE i .AND TELEGRAIBH COMPANY, A CORPORATION OF NEW YORK.

TELEPHONE REEATER; SYSTEM.

Application led March 18, 1921. Serial No. 453,304.

The principal object of my invention is to provide a new and improved repeater in stallation adapted for connection with telephone lines of various characteristic impedances. This and other objects of my invention will be appreciated after the dis-v closure of a specific embodiment thereof and a discussion of the principles involved in itsv operation. I now proceed wlth such d1s-` closure, leaving the invention to be defined in the appended claims'.

The accompanying drawing is a diagram showing my invention embodied in a repeater adapted to connect any two of a plurality of lines radiating therefrom.

A plurality of different lines 11, 12, 13, 14, 11, 12', 13 and 14 come to terminals at one repeater station. These` lines may be widely different in their characteristic impedances. One of them may be an open wire line, another may be a cable pair, one may be an unloaded line, another may be loaded continuously, another. may have lumpedloading. It is desirable to terminate` each of these lines when it is in use by an impedance approximating its own characteristic impedance in order that waves reaching the repeater may not suffer partial reflection which would cause transmission losses and distortion at this repeaterand would affect detrimentally the impedance balance of repeaters at other stations on the line. Instead of specially de-v signing the repeater to adapt it for the characteristic impedance of a particular line with which it is to be used, I have devised a system by which a given repeater can be immediately connected in between any two of the lines without occasioning. serious reflection effects. I

For each of the linesll, 12, etc., a respective balancing network is provided,I such asl, 16, etc. Each of these networks is an artificial line having approximately lthe same impedance characteristic as its.cor responding geographical line,

V'A hybrid-coil transformer provided. Its coils 21 have two,y pairs of terminals for connection between any one of theflines and its corresponding artificial line. Thefacility for making these various connections isl indicated'by the flexible conthe primary coil 23 ofan auto-transformer,

a minor portion of whose windings are included in the circuit leading through the` wave filter .25' to the potentiometer 26, from which the input circuit goes to the transformer 24 and thence to grid 29 and filament 28 of the audion 27. From the flament 28 and plate 30 of this audion 27 the output circuit goes l`through the conductors `31 to the coil 22', which -matches the coil 22 of the hybrid-coil transformer already described. The repeater has bilateral symmetry which is indicated by the use of the primed reference characters, and the duplicate/parts need not be described verbally.

There is one magnetic circuit for the coils 21 and 22. Linked with this magnetic circuit the coil 22 has la large number of turns and the coils 21 only a comparatively small number.

The energy of incoming currents from the left over the line 12 will mostly go through the transformer windings 21 and on into the artificial line 16. It is true that there is a bridge across the coils 21 comprising the coil 23, but this is a path of high impedance, so that comparatively little current will be divertedthrough the bridge. It is also true that the coil 22 stands in inductive relation with the series coils 21, so that some current will be induced in the circuit of the coil 22,

but since the windings of the coils 21 are com- 'energy to be diverted at the transformer 95\\ 21-22 or its bridge 23 as just pointed out, it follows that the wave from the line 12 will go through into the artificial line 16 with `very little loss, and the impedance that the repeater is substantially that of the artificial line which in turn has the correct value for terminating the line. In this respect my inventiondiiers from an arrangement that is incoming wave encounters on entering the 21-22 1s common',"where practically none of the in., 10

coming ener reaches the artificial line and the impedance encountered by the incoming wave on entering the jrepeater depends upon the structure of the repeater circuit and not on the articial line.

While it might appear as a disadvantage that the major portion of the energy of the incoming current along the line 12 goes to the artificial line 16 instead of to the input element of the repeater, this disadvantage, such as it is, is easily overcome by using asuiiicient degree of amplification. In other words, while the currents-put through the filter 25 have only a fraction of the energy of those that come in over the line 12, they have the same wave form and are readily amplified to the necessary degree by the audion 27.

The connection between the coils 21 and -22 that prevents excessive energy diversion |`sated by the amplification by the repeater element. t v

It will be 'seen that I have provided a system in which the passive impedance of the repeater approximates closely to any type of line on which the repeater may be required to Work. In other words, I have succeeded in securing the result that the passive impedance of the repeater circuit will be correct for any line that maybe connected therewith, whether it be loaded or not, and whether it be of the open wire kind or a pair of conductors in a cable. The respective network for the line considered is the electrical counterpart of the'line and termi- Y nates the line perfectly, that is, it introduces no irregularity in transmission. This as-v sumes that in the case of lines having a lumped loading, the termination of the line :sok

is with a half coil or a half section between the coils. By,connecting the repeater apparatus between each line and its balancing ynetwork so that it will have a negligible ef* feet, I secure and preserve the result that the repeater system will havel a perfect passivel impedance. Y

To illustrate the resultant gain with my system, let us consider a special case.` It is well understood that al repeater of the type illustrated in the drawing may ordinarily give a net gain of about 21 standard miles, a standard 'mile "corresponding to the formu a prior to my invention,l somethingv like 6% further loss in the wave filters, etc. Assume for the purpose of the illustration that the net total gain available with a single audion element is about 271/2 miles. This corresponds to a current amplication of about 20 times and an energy amplification of about 400 times.

The foregoing is based on the assumption that the repeater input is connected to the bridge terminals of the hybrid coil, but instead of this, I connect it through an autotransformer 23, thus causing the amount of energy taken from the line to be as small as I please. This makes the effect of the bridge upon the passive impedance of the repeater to be correspondingly small. Similarly I make the effect of the output coil 22 upon the passive impedance of the repeater to'be as small as I please. For the purpose of this illustrative example, let us assume that the bridge and the output coil are so designed as to aHect the line impedance by only 10 per cent. The passive impedance of the repeater, then, will not depart more than about 20 per cent from the ideal, indeed the departure willbe much less because the series and shunt'impedances of the repeater circuit tend to produce opposite effects. In order that the e'ect of the bridge circuit shall be limited to 10 er cent of the impedance of the network its 1mpedance must be 9 times the impedance of the network and about 316 of the energy entering the repeater will reach the bridge circuit. Similarly if the winding 21 is such that its effect upon the impedance is'limited to 1Q per cent of the network impedance only about ,15 of the possible power output of the tube can be utilized inthe line. Accordingly each repeater element will now be required to produce an energy amplication ofV approximately 10 10=10O times, to overcome the losses due to the inefficient transformer couplings. However, since each audion is capable of amplifying the energy 40() times, we get a net amplification of 40G/100:4 times, which corresponds to a gain of about 61/2 standard miles. 'llhusA it will be seen that` in spite of the special transformer designy with its Ainecient coupling, I secure a gain of 6% miles and accomplish this result without substantially aecting the passive impedance which the repeater offers to the fline.

By using more than one audion element in'tandem instead of the single one shown inthe diagram I could evidently increase the gain obtainable withI the reice Imi

-line'12 to line 18 by attaching the conductors 19 and 20 respectively to the line 13and the artificial line 17. I claim:

1. In combination, a telephone line and a repeater, said combination comprisin an artificial line connected with the telep one line, the series impedances offered by the repeater to the line bein made small and the shunt impedances ma e lar e so that the input impedance is determine by the impedance of the artificial line.

2. In combination, a plurality of tele honeA lines, respectiveartificial lines there or, a

repeater, and means to connect any two said telephone lines through said repeater, the series impedances oered to the incoming waves being made small and the shunt impedances large so that the input impedance to each telephone line is determined by the impedance of its artificial line.

3. In combination, a telephone line and a balancing artiicial network, a three-winding transformer between them with input and output connections, the series impedance of these connections being small from the line and the shunt impedance being large from the line whereby most of the 4. 'Ihe method of repeating between any two of more than two lines of various characteristic impedances with respective balancing networks which consists in passin incoming energy from one such line throng to its balancing network with diversion of only a small portion of energy for amplification, and applying the amplified energy to the outgoing line through a relatively ineiicient connection.

5. The method of repeating with am litication on a telephone line without mate ing the impedances of the repeater circuits to the line which consists in sending the major part of the incoming energy on the Ime through and diverting only a small part thereof, amplifying this small part and applying the amplified energy relatively ineiiciently to the line.

In testimony whereof, I have signed my name to this specification this 15th day of March, 1921.

ALVA CLARK.

Images