US1091193A - Telephone-exchange system. - Google Patents

Description

1,091.193. Patented Mar. 24, 191* 10 SHEE'I'HHEET 1.

E. E. CLEMENT.

TELEPHONE EXCHANGE SYSTEM.

APPLICATION rum) JAN. 19, 1906. nznnwnn D20. 31, 1910.

Patented Mar. 24, 1914.

10 SHEETS-SHEET 2.

B. B. CLEMENT.

TELEPHONE EXCHANGE SYSTEM.

APPLIOATIOI rILnn .ux. 10, 1006. nnnnwnn 111:0. a1, 1910.

1,091,193. Patented Mar. 24, 1914 10 SHEETS-SHEET 3.

We twzooeo/ E. E. CLEMENT.

TELEPHONE EXCHANGE SYSTEM.

APPLICATION FILED JAN. 19, 1906. RENEWED DBO. 31,1910.

1,091,1 93. Patented Mar. 24, 191:1

10 SHEETS-SHEET 4.

IPQ

E. E. CLEMENT.

TELEPHONE EXCHANGE SYSTEM. nPLmA'rIolnLnn June, 1006, RENEWED DEC. 31, 1910.

1,091,193. Patented Mar.24, 1914.

wanna-$11 12": a.

W; human E. HULEMENT. TELEPHONE EXCHANGE SYSTEM.

APPLICATION FILED JAN.19,1906 1,091,193.

RENEWED DECK'31, 1910v Patented Mar. 24, 1914.

10 SHEETSSHEET 6.

Suveufoz We moose:

' E. E. CLEMENT.

TELEPHONE EXCHANGE SYSTEM.

APPLICATION FILED JAN. 10, 1906. RENEWED DBO. 31,1910.

1,091,1 93. Patented Mar. 24, 1914 10 SHEETS-SHEET 7.

lvil'ncoom 72* a [1 E. E. CLEMENT.

TELEPHONE EXCHANGE SYSTEM.

APPLICATION nun JAN. 1a, 1006. RENEWED no. 31, 1910.

1,091,1 93. Patented Mar. 24, 1914 10 SHEETS-SHEET B.

,F 'yf J Wz'zasu'ea E. E. CLEMENT.

TELEPHONE EXCHANGE SYSTEM.

APPLICATION FILED JAN. 19, 1906. RENEWED DEC. 31, 1910.

1,091,193. Patented Mar. 24, 1911 10 SHEETS-SHEET 10.

Conn ec tors mar elec 2'0 r;

)fiazn Fkame In termediafe fllstruiba t1,

Sa6.scnz'6ers Lines.

1 gamma; gamma} 1 $111M Fatima .-l((((((, ,[(([[L ltt&, I 1([u1 N INVENTOR M UNITED STATES PATENT OFFICE.

EDWARD E. CLEMENT, OF WASHINGTON, DISTRICT OF COLUIVIBIA, ASSIGNOR, BY MESNE ASSIGNMENTS, TO FREDERICK C. STEVENS, OF ATTICA, NEW YORK.

TELEPHONE-EXCHANGE SYSTEM.

Specification of Letters Patent.

Application filed January 19, 1906, Serial No. 296,853. Renewed December 31, 1910. Serial No. 600,276.

To all whom it may concern:

Be it known that I, EDWARD E. CLEMENT, a citizen of the United States, residing at \Vashington, in the District of Columbia, have invented a certain new and useful Improvement in Telephone-Exchange Systems, of which the following is a specification, reference being had therein to the accompanying drawings.

My invention relates to telephone exchange systems, and has for its object the rovision of means whereby incoming calls from subscribers stations or other Stations may be automatically apportioned among the idle portions of the connective apparatus.

This invention is applicable to, andI shall describe it in connection with exchanges equipped both with manual and with automatic switching devices for interconnecting lines. It is capable of application in a number of ways, and may be embodied in various forms, some of which I shall indicate, but I have selected for specific description herein what I consider to be the best form devised up to this time.

The theory upon which my invention is based is this: In all telephone exchanges a certain proportion of the apparatus is idle most of the time, and all of the a paratus is idle some of the time. In a manuaFexchange, wherein double cords terminating in pairs of plugs are employed for purposes of interconnecting the lines, the number of pairs of such cords and plugs required in an average city exchange is theoretically equal to about ten per cent. of the total number of lines to be served. In practice, this percentage is somewhat increased, because some of the cords are almost always out of service, and short periods are apt to occur in which the number of calls rises abnormally. In some exchanges the percentage of connecting cords is as high as 17, but this is because the load is apt to jump unequally on different parts of the board. That is to say, that one operator will receive a great many calls from her particular group of subscribers, when others are comparatively idle.

One purpose of my present invention is to correct this unequal distribution of the calls, by causing the first line calling to be connected automatically to the first idle cord or other connective circuit, the second line calling to the second idle connective circircuit, and so on.

number of others, through which it is finally connected to the line wanted. Such an arrangement, however, is expensive, for there must perforce be a number of the first selector switches as they are called, which is equal to the number of lines, and from ten to thirty per cent. of other or auxiliary switches according to the size of the exchange. In an exchange of 1,000 lines, the number of switches is thus 110 per cent. of the number of lines; in a 10,000 line exchange it is 120 per cent. Since most city exchanges have between 1 and 10,000 subsciibers, it is customary to figure on 1.2 switches per line, and as compared with the manual switchboards, the inherent economies in the automatic system, which are due to their low operation and maintenance charges, do not begin to make themselves apparent until after the 4 or 5,000 line size is reached. Manual multiple-boards ac commodating that many lines begin to get very expensive by comparison, especially in operation.

My present invention enables me to ecouo mize switches in an automatic exchange by providing only a limited percentage of first selector switches, and distributing the incoming calls among these, the first line to call being automatically connected to the first idle switch, the second to the next idle switch, 620.

Since I used the same character of linecircuit in my automatic systems and manual systems that is to say the same plain metallic circuit, the a )paratus which I employ to efi'ect the distribution in both manual and automatic systems is the same. It consists merely of relays. It has heretofore been proposed to use moving switches, which have been called starters and finders by some persons, and which I have called primary selectors, for the purpose of picking out a calling line in a group, and connecting it to a switch whereby further connections might be made. Such devices, however, have the inherent uncertainty and liability to derangement which are inseparable from all moving pieces of apparatus, especially of an electrical nature, where they are governed and moved by electromagnets. According to my present invention there can be no uncertainty in operation, because I use simple positive acting relays, which work Without margin on current or no current in their circuits; and there is very little liability of derangement of these relays, because they are mounted and housed in a manner now Well understood in the art, that is to say on iron racks and in dust-proof cases.

Briefly stated, the invention consists in dividing the lines into groups, in which they may be distributed through an intermediate frame for answering purposes, as is now customary with manual, multiple switchboards, assigning to each group a fixed percentage of trunks o1- cord-circuits, as the case may be, and providing a set of connecting re ays for each trunk whereby it may be connected if the proper combina' tion of relays be energized to any line in the group to which it is assigned.

Each connective circuit has a bridged battery and a pair of bridged controlling relays, and local circuits are provided such that when any line calls and is connected,

it will control the trunk circuits. Only one set of relays will be energized, that one which corresponds to the first idle connective circuit, because I arrange the controlling rela s of the connective circuits so that as each ecomes busy it will disable all of its own selective relays except the one through which it is connected, at the sametime connecting and rendering active the relays-of the next trunk or connective circuit in order.

Where the invention is used with a manual system, the connective circuits to which I refer may be ordinary single cord-circuits terminating in plugs or jacks, and each having an associated signal, whereby an operator will be apprised of the call when it comes. Where the invention is employed with automatic switches, the connective circuits are trunks, if I may so call them, leading to first selector switches, the number of the latter being thus reduced to a number less than the number of lines.

The best embodiment of the invention now known to me requires two relays per line, or two hundred relays for each one hundred lines. One'half of these relays are what I shall term multiple contact relays, and the other half what I shall term single-contact relays. The multiple-contact relays consist each of an electromagnet having a movable armature and either ten or twelve double pairs of contact springs with platinum points. Such a relay takes the place or ten single relays, and receives ten lines. Each line is connected to the same contact in ten of these relays, in multiple. The single relays, on the other hand, are of the ordinary type commonly used in telephone exchanges, each comprising a single electromagnet with one of more double pairs of contact springs.

The multiple-contact relays I shall refer to also as the tens relays, and the singlecontact relays as the units relays. These relays are mounted in banks, ten of each kind to a bank and one bank to each trunk or connective circuit. The circuits of the tens and units relays are so arranged that when aflgiven line calls it closes the circuit of all the tens relays corresponding to the tens digit of its line, and all the units relays corresponding to the units digit of its line. Only those relays are active which lie in the ban appertaining to the first idle trunk or connective circuit. Assuming No. 34 to be the number of the calling line in the group, the third tens relay and the 4th units relay will pull up simultaneously, and their local. circuits being combined by this joint action connect the line With the trunk orcordcircuit to which the bank belongs. All other relays in the bank are instantly disabled, the operative relays are themselves locked as long as the calling line remains in use, and the relays of the next idle connective circuit become active.

It will be observed that in this system the calling line is selectively connected to an idle trunk through the mere closure of the calling subscribers line circuit. I am aware that systems have been proposed in which single relays are actuated by means of grounded circuits controlled by keys at the subscribers substations, Such systems however, are ineliicient and commercially impracticable.

The selected form of my lnvention to which I have referred is illustrated in the accompanying drawings, in which Figures 1, 2, 3, 4, 5 and 6 form parts of a complete diagram from calling subscribers stations through the connective apparatus at the central ofiice. Of these, Fig. 1 shows three subscribers stations and the line-circuits leading to central. Figs. 2 and 3 show a terminal set of two tens relays for each of three trunks. Fig. i shows the units relays for three trunks. Fig. 5 shows the bridging relays for the said trunks. the battery connections thereof, and the solenoid or controlling relays. Fig. 6 shows a selector switch and connector switch which are worked from the subscribers stations through the intermediation of the relays in Fig. 5. Fig. 7 is a separate diagram show ing the relation between the tens relays and the units relays of a single trunk. Figs, 8 and 9 are a. side view with parts in section,

and a top plan view, respectively, of a ten point relay such as shown in Figs. 2 and 3. Fig. 10 is a face view and shows a type of automatic switch which may be employed in the circuits shown in Fig. 6. Fig. 11 is a section on the line a:- a; of Fig. 10. Fig. 12 is a schematic diagram showing the method of primary distribution.

Before proceeding to the circuits, wherein the inventicn in the present case resides, it will be advisable to obtain a clear understanding of the pieces of apparatus which I employ, and which are shown in Figs. 8 to 11, inclusive. I do not give any special illustration of a single point or ordinary relay, for the reason that every one skilled in the art is so familiar with these devices as to render it unnecessary. That portion of my system which is illustrated in Figs. '2. 3 and 4 constitutes what I have termed the primary selector or answering portion of the system. In this, as alrea y set forth, ap pear a number of ten point relays and a number of single point or ordinary relays. These latter may be of any suitable type. The only special feature required is that of double winding, which. as I shall point out hereinafter, can be dispensed with it sonieother means he provided for locking the relays when energized. The apparatus at the subscribers stations also forms no part in itself of my present invention. except as it coacts with the apparatus and circuits at the central oflice. I contemplate using an ordinary common battery telephone outfit the same as those employed in standard manual switchboard systems, and with these subscribers senders or signal transmitters shown in one instance in Fig. 1 as a key, and in the otherinstance as a handwheel with projectiingpins passing over and under spring contacts. Referrin .then to Figs. 8 and 9, I will briefly descri c the ten point relay, N. This comprises an iron back yokc', n, a pair of magnets, 11,, mounted thereon, an armature, n, extending across the front end thereof, and a set of contact :springs, 71. a", n n, overlying the upper bell-crank extension of the armature, so as to be worked thereby. There are eleven sets of these springs, all secured at the rear end of the relay upon the return member, if, of the back yoke. The front ends or pole pieces of the cores of the olectroanagnets,n, receive the front yoke. of, of non-magnetic material, such as brass, this also having an upper return portion, a, overlying the magnet. windings, and alined with the return portion, 11,, of the back yoke, 01. Upon this return portion. n", of the .front yoke the upper bell-crank arm, n", of the armature, n-' lies flat under normal conditions, Cut from end to end of this upper bell'crank portion, It", are two grooves in passingthrough a hole in thespring, a.

,while the uppermost stud of each pair rests upon the spring, 15 su 'iports the spring. n, upon its shoulder, and carries the spring. M. on its head, its stcm passing through a hole in the spring, a. The actuating studs ai t'ect only the springs, 11%, a". of each set, these constituting the movable springs, while the others. n, n, are tixcd. Each actuating stud carries the spring. -n. on its shoulder, and the spring, 11., on its head, its stem passing through holcs in the springs. 11?, n. The entire relay is secured to aninsulating' back block. 71', by means of the bolts. n. constituting extensions of the cores and threaded to receive lock-nuts. The rear ends of the contact springs cxtend out through openings made in this back block. and an inclosing shell or casing. n, is fitted to the block, with which it forms a tight joint all around the ed c. This shell is secured in place by mcans of the stud, a. threaded to receive the nuts. a, As the armature, it, is attracted, the springs, n, n. are lifted by the actuating, stud, until their outer points niakc contact with the corresponding points of the springs, 11?, 91 in each set. Ten of thcsc sets are line-contacts.and the eleventh is for locking purposes. as will sufliciently appear from the description of the circuits.

Referring now to Figs. 10 and 11, l have therein shown a piece of apparatus which I do not present as a part. of my invention. nor as an example of linished design in an automatic switch. but which is typical of the class of switches commonly known Strowgei or automatic electric switches, As I have previously pointed out, my present invention may .he practised with various connecting means, but in order to render this description complete and coherent, I include this form of switch. and its circuits.

As illustrated in Figs. 10 and 11. the switch is desiguatet'l in toto by the letter and represents a selector switch, that. is one oi the intermediate switch units to transfer a connection from the calling line to an idle trunk and switch in the. desired thousands. or hundreds group. lt has a frame. 5'. pr vidcd with two postx, extending downwardly to support the contact-hanks. x .c". which :I. shown in Fig. it are arranged in a c\'lindrical figure around the central spindles. There are lcn r ws incur-licontactwhich rest the insulating studs of the 1 bank, with ten pairs of cnlitih'il in each ltl 'v l lit) The ma net. has an armature. v drawn of the bank, 8 and ten single contacts 111 each row of the bank, 8. The contacts are radial strips of brass or German silver clamped solidly between upper and lower plates, 5", andseparated by suitable insulation. Upon the spindle are mounted the wipers, a, 5 a the latter being a connected pair adapted to rub against both faces of the single contacts in the banks, 8 and the former being separate wipers adapted to rub upon opposite faces of the pairs of contacts in the bank, 5. Connection is established with these wiper springs through a flexible cord or similar means, which I have not deemed it necessary to show.

The spindle, s, is a long, straight piece of steel, journaled in the upper and lower members of the frame, 8, so that it can both turn andslide therein. Its own weight normally keeps it down in the position shown in Fig. 10, with the wipers out of alinement with any of the contacts. The parts are not drawn to scale, but it is supposed that the first upward step of the spindle produced in the operation of the vertical magnet will bring the wipers opposite the respective first rows of contacts in the banks, and each successive step upward will shift the wipers opposite another row. After being stepped up until the wipers are opposite the desired rows, the spindle is then supposed to be turned one step at a. time by the rotary magnet until the wipers rest upon the contacts of an idle trunk. For the purpose of effecting these vertical and rotary movements, the spindle carries two ratchets, s, .9 the former having a number of annular teeth, usually eleven or twelve, and the latter having a circular series of vertical straight teeth, which'irsually do not extend entirely around the ratchet surface, but. only far enough to cover the are through which the wipers must sweep when passing over all the contacts in a row. The vertical movement of the spindle is effected by the magnet, v. and the rotary movement by the magnet, '1", both being mounted on the frame.

down by a spring, 1' and carrying a pivoted pawl, 0 provided with a tail spring, P, which tends to throw the tip of the pawl normally prevented from doing so by the stop pin, 0 set in the back plate. 8, which is supposed to cover the entire back part of the frame, 8. When the magnet, r, is energized'it draws up its armature, and the point of the pawl, r. is immediately thrown forward to engage the upper tooth of the ratchet, s", the continued upward motion of the armature then raising the spindle and its wipers one step.

The rotary magnet, 1', has an armature. 1",

appearing edgewise to the observer. in Fig. 10, and pivoted at its nearest edge in a ver- 1 tical line, so that the edge away from the observer swings to and from the magnet oles, moving the extended arm, 7 back and orth, and by means of the terminal pawl, *1, turning the ratchet, s. The pawl normally remains out of engagement with the ratchet, the armature being retracted by the spring, 1*.

In order to retain the spindle in any position to which it may be raised and turned by the magnets, 41, and 1', I provide a pair of swinging pawls, t, 15 whose ends engage the respective ratchets, s, and s", and whose bell-crank body, 13 is pivoted between posts, 11*, at the edge farthest from the observer. Upon the front edge of the body is a projecting pin, t adapted to be engaged by the link, 6 on the upper end of the armature lever, 23 of the release magnet t. The link and the lever are normally drawn over by the spring, i, into the position shown in Fig. 10. The pawls, 6, F, are normally held in engagement with their respective ratchets by a suitable spring, not shown. When the magnet, t, is energized it draws over the lower end of the armature lever, t, to the right, throws the link, t, to the left until it passes over and engages the pin, t; and

'when the magnet is again deenergized the spring, t, again draws the link to the right and swings the bell-crank body, t upon its pivots so as to raise the pawls, t, t from their ratchets and permit the spindle to be rotated so as to disengage the wipers, s s 8 from the rowsof contacts and then to drop'i'n'to the normal position of disuse. At the end of its travel the link, 23, passes off the pin, t, and the pawls are restored to engagement with theratchets.

These are the operating members of the switch, in addition, the switch frame carries an escapoment or controlling magnet for the side switch and for certainv contacts which will be specified in the description of the circuits. The side switch is shown in front view in Fig. 10, at it. It consists of a plurality of switch-arms mounted one over the other with interposed insulation, the entire set being carried u on a lever, u, pivoted at 11 and normally ield in the position shown, against the tension of the spring. a, by the engagement of the upper end. u, with the first one of a number of teeth, it", carried upon the armature lever 11 and constituting an escape'ment. \Vhem ever the magnet, U, is energized the lever, u, is thrown over to the left, the tip, u", passing a half step to the right, when the armature is again retracted the tip passes a fullstep to the right, and so on, two steps placing the arms, u, successively in contact with the first. second and third sets of fixed contact points, it, on an insulating block carried by the frame. When the switch spindle is all the way down in its normal P iti II t e lower edge of the ratchet wheel, 8 engages the short upper arm of the bellcrank lever, u whose lower arm engages the lever, u, to push the same back into initial position as shown in the figure.

The magnet, U, at certain times has its armature mechanically held up, and for this purpose a projecting finger r, is attached to the armature, r, of the rotary magnet r. Upon the upper part of this finger a button of insulating material is placed to engage and normally keep closed a pair of contacts, 1', in the. circuit of the magnet, 1". These contacts and the projecting finger, 1, are used only in the selector switch, in which the r0 tary movement is accomplished by constant and automatic vibrations of the armature, r, until an idle trunk is found by the wipers.

The magnet, ll, controls a set of contact springs, u a", the former normally closed, and the latter normally open. When the armature, 11", is attracted the conditions of these springs are reversed.

In order to restore the spindle to its nor mal position in a rotary direction, a coiled spring is provided housed in a cap, 8 at the top of the spindle, one end of the spring being attached to the same, and the other end to a lever arm, 8", loosely journaled at one end on the spindle between the cap and the collar, 8", and loosely sliding at the other end on a post, 8". The collar, 8, carrics the stop arm, a, which abuts against a turned down leaf on the arm, a", when the parts are normal. The travel of the arm. 8, up and down the fixed post, 8 and the vertical extension of the ratchet, 8 enables the spindle to be turned and restored through its full sweep in any vertical p0- s tlon.

Having thus described the mechanical parts a knowledge of which is essential to the comprehension of my circuits, 1 turn to the description of the latter, beginning with F l. Therein I have shown three subscribers stations, A, A',A each equipped with a telephone transmitter. a. receiver, a \\ltt'llll0()l-I. a ringer, and condenser, all of usual or any suitable type for use with common battery circuits. When the hook is up. that is when it is relieved of the weight of the rc 'civer, the circuit is through the transmitter and receiver; and when the hook is down. the linc-circuit is completed for alternating current only, through the ringer and condenser. At each station I have shown a sender; or signal transmitter for the use of the subscriber. At station A this is a key. and at each of the other stations it is a rotary wheel with. pins. the designating letter in all three cases being K. Since the circuits are the same for all the lines, I will refer to the line from station A only. This consists of conductors 1 and 2, extruding from the substation to the central oflice. At the substation the talking circuit is completed through the wires 3-4 when the hook is up, and the ringing circuit through the wires 5-6 when thehook is down. These figures are useful for the purpose of tracin the circuits only.

Turning now to Ilg. 2, the linewires 1 and 2 appear at the left-hand side of the figure, terminating in contacts of the cut-off relays, M, which are normally deenergized, so that in each case the line-wire 2 is grounded at 7 and the line-wire 1 extended through wire 8 to the linerelay M, and so to main battery B and ground. The armature, m, of each line-relay M is grounded at 9, and when attracted is adapted to close upon terminal contacts of Wires leading to the tons and units relays of the primary selector sets. I preferably divide all the incoming lines in an exchange into groups of say one hundred each, and to each group I apportion a fixed number of trunks leading to connective apparatus. In a. small or 001m paratively inactive exchange the percentage of these trunks may be as low as five, and in busy exchanges it may be raised to ten or even fifteen. I assume, however, that ten per cent. of connective apparatus will be Sufiicient to care for the present system, hence I provide ten trunks for each one hundred lines, and a complete set of tens and units relays in a primary Selector group for each trunk. The circuits herein illustrated show six lines, which for convenience I will call Nos. 500, 501, 502, (Fig. 2) and 510, 511, 512, (Fig 3). The subscribers circuits of Fig. 1 will fit any of these, and in order to fully comprehend the connections of these lines and the arrangement of the tens and units relays, it is necessary to consider Figs. 2, 3 and 4 together. In Figs. 2 and 3 tie subscribers extension lines are shown coming in to their initial terminals on the tens relays, which are of the type shown in Figs. 8 and 9. Each of these relays carries the terminals of ten lines, and for each trunk there are ten of these relays, so that for 100 lines. provided with ten trunks, there will be 100 ten point relays, containing 1,000 line-terminals; from which it appears that each line can be multiplied ten tunes, which it should be in order to obtain access on occasion to every one of the ten trunks. The contacts on the ten point relays are arranged in duplicate pairs. That is, the incoming lines terminate each on one pair of springs, opposite which another pair of springs appears to which the line is connected when the relay is energized, and from which a local circuit. passes to the front c ntact springs of a particular units relay or single point relay, For each group of ten tens relays there are ten units relays. or twenty relays in all for each trunk. A group of this kind constituting one primary selector or trunk terminal for thirty lines, is shown in Fig. 7, apart from the other figures, so as to free the lines from their multiple con- I nections. In order to make clear the coni nection of a line coming into the exchange, I will first refer to this figure and then trace i the circuits through the others. in Fig. 7. N. N. N are three relays of the type shown in Figs. 8 and 9, having ten pairs of contacts each for the line-circuits. These relays correspond to the relays shown in and 3, correspondingly lettered. .\t the top of the figure I have shown the y terminal ends of lines 500 to 529, inclusive, 5 connected to a pair of contacts, n 71 such as shown in the mechanical Figs. 8 and 9. i For lines 500 to 509 these terminals are in the relay N. For lines 510 to 519 they are l in the relay N. For lines 520 to 529 they I are in the relay N Thus each relay contains line-terminals, n n for ten lines. When any one of the lines appropriated to l a particular relay is calling, the electromag net at, of that relay becomes energized. i This will be explained in connection with the circuits of Figs. 1 and 2 farther on. For 1 example, if any line from 500 to 509 originates a call. the magnet, n, of the relay N y will become energized, and simultaneously connect all of the lines from 500 to 509 onto the extensions 10 to 19, inclusive. If, however, the calling line is in the group from 510 to 519, the magnet, n, of the relay, N, i will become energized, connecting all that group onto the extensions 10 to 19, inclusive. Similarly, if the calling line he in the third 1 group, from 520 to 529, the magnet, n, of l the relay N will be energized, connecting that entire group onto the extensions 10 to y 19, inclusive. hus extensions 10 to 19 are I seen to be common to all of the tens relays. l and in fact each extension pair is multiplied l to corresponding contacts in all of the tens relays. No. 10' extension goes to the first 1 pair of contacts, n, n, in each of the relays. N, N, N Extension No, 11 is multiplied to the second pair of contacts, at, 12", in each i of the relays. and so on with all the extenl sions up to No. 19. I At the lower edge of Fig. 7, I have shown j ten units relays, marked P to P", inclusive, These are the same as the relays in Fig. 4, similarly marked. Each is a simple, stand ard relay, of the type in common use, and preferably of the same general type as that shown in Figs. 8 and 9, hereof, viz., that; type in which the electromagnet is mounted horizontally on a frame or rack, carrying the contact springs upon it, and the whole is incloscd within a dust-proof shell. Each relay has an armature, 1 adapted to be actuated by an electromagnet, p, to close together the contacts, 32 p, and p p. In the present instance, that is in the trunk group of each primary selector, all of the "branch going to the appropriate tens rel and the other branch contacts, 72 of the entire set of relays, P to P", are connected together and all of the contacts, 7), are connected together, forming a set of double parallel branches marked 20 to 29, inclusive, which are really so many bridges across the trunk circuit, 3031.

- This trunk-circuit, as indicated at the letter,

w, in Fig. 7, may pass to any preferred form of connective apparatus; for the automatic system herein fully illustrated the trunk terminals, m, pass directly to the bridging relay circuits shown in Fig. 5; for a manual system, they terminate in a cord-circuit of the same general type as the circuits of Fig. 5, but without the solenoid relays, these being unnecessary except where automatic switches are to be controlled.

The outfit shown in Fig. 7 is suflicient for 30 lines, and by its use any one of the lines, 500 to 529, may connect itself. to the trunk 30-31, by merely closing its line-circuit, and energizing the appropriate tens and units relays The energizing circuits are not shown in Fig. 7, but will be described in connection with Figs. 2, 3 and 4. They comprise parallel branches closed by the line-relay of each line when calling, one ay, going to the appropriate units relay, which thus become energized simultaneously and act in conjunction to connect that particular line and no other to the trunk. For example, suppose line No. 512 is calling the subscriber having taken down his receiver from the hook, so asto close his line-circuit as shown in Fig. 1, and energize his line-relay, M, shown in Fig. 2. This line-relay closes two branches, one leading to the tens relay, N, in Fig. 7, and the other leading to the units relay, 1", which is the third in the set of ten. The actuation of the tens relay, N, connects all of the lines 510 to 519 onto the extensions 10 to 19, inclusive; but it will he observed that these extensions are open everywhere else, hence no effect is produced by this action alone. The third units relay, I, however, connects the third extension 2 to the trunk 30-3l, through the brim-1 2'2, and by following the circuits in Fig 7 it will be found that this connects the lin 51% onto the trunk direct, whereupon the trunk relays become energized, and furtherproceedings may be had of a kind suitable to the system employed. If the trunk 3031 terminates in a cord-circuit, then the action is the same as if a plug had been inserted in the subscribers line-jack, and by using suitable means the subscriber will find himself connected directly onto the operators telephone. If, on the other hand the automatic circuits be employed, then the subscriber will be able to proceed at once to manipulate the automatic switches to which the trunk 30-451 leads. Only one line is thus connected to the trunk at a time, assuming of course that only one tens relay and one units relay of the 'same set be energized simultaneously. It is true that the tens relay connects all of its lines onto the extensions; but all of these extensions are open except the. particular one which is closed onto the trunk by the particular units relay encrgized. It is lso true that the extension thus connected to the trunk is multipled to the corresponding pair of contacts in every one of the tens rela 's, but this produces no effect, because all but one of the tens relays remain deenergized.

Before passing from Fig. 7 it is neces- Sary to call attention to the fact that only three tens relays are there shown. accommodating thirty lines. This limitation is due to the necessary limitations of a Patent Ofl'ice drawing, and it is to be understood that the same principles apply to any desired group up to 100 or more. It is most convenient in a scheme like this to make use of the decimal system, using ten relays per trunk, ten lines per relay, and ten trunk multiples per line. For 100 lines there would be ten of the tens relays, N, 1*, N etc., but still only ten of the units relays. audeaeh of the extensions 10 to 19, inclusive, instead of being multipled three times, would be. multipled ten times, into all the tens relays.

Turning now to Figs. 2, 3 and r we have here the same relays described in connection with Fig. 7 but somewhat differently arranged. Only two tens relays and three units relays of each set or of each trunk are shown. I have assumed the numbers 500, 501, 502 for the lines entering Fig. 2. and 510; 511; 512 for the lines entering Fig 3. The first three lines being within the same tens digit of each other go onto the same tens relay in each set. For the sake of convenience I have given the designating numerals of the relays coeflicients, and have marked the three trunks in Figs. 4 and 5 so as to be readily followed. Thus in Fig. 2 all three of the lines come into the relay N, which is adapted to put them onto the extensions, 10, 11, 12, etc., leading to the units relays, P, P, P etc., in Fig, 4. (I should here remark that the tens relays in Figs. 2 and 3 are not shown with their full complements of springs, for the sake of simplicity. Each of them is supposed to have ten pairs of contacts, in accordance With the showing of Figs. 7, S, and 0.) Each of the lines 500. 501, 502 not only terminates in contacts of the relay, N, but is mnltipled to similar contacts in the relays 1N. 2N, etc. There would be one of these added for each trunk, that is appropriated to the group of lines. In the present case I only illustrate three trunks. so that each line need he mnltiplcd only three times.

In Fig. 3, the lines 510; 511; 512 arc con- 'nected to the tens relays, N, 1N and 2N in the same manner as the lines in Fig. 2, each line being multipled three times so that each may hare access to any one of the three trunks. It will be understood of course that the relays, N, N, of Figs. 2 and 3, and l, l P of Fig. 4 constitute one set, appropriated to the trunk 30-31, and have precisely the same relation among themselves and the same connections, as the set shown in Fig. T. Similarly the relays 1N, 1N, 11, ll, 1P all belong to one set, appropriated tothe trunk line 130-131, Similarly again, the relays 2N, 2N, 2P, 2P and 21, belong to one set, appropriated. to the trunk line 230-231. The extensions from the tens relays to the units are marked 10, 11, 12 for the first set, 110, 111 and 112 for the second set, and 210,. 211 and 212 for the third set. As these of course bear a logical relation throughout, and follow the same arrangement as that in Fig. 7, it is thought no difiiculty will be experienced in comprehending the meaning of Figs. 2, 3 and 4. If Fig. 7 is understood, the description of the other figures may be put into one sentence. They show several groups like Fig. 7, with the line -wires multipled to each group instead of terminating as in Fig. 7 in single pairs of contacts.

The actual connections controlled by the relays N and P being thus understood, it is most important to explain the means by which I control these relays. Referring to Fig. 2, the line- wires 1 and 2 terminate as I have stated, in the contacts of the cutofi' rclay M, through which they are normally connected to the line-relay M and ground. \Vhen the subscriber at station A takes the telephone from the hook, he closes his linecircuit, the line-relay M is energized, and the two branch circuits 50 and 51 are immediately closed. The branch 50 passes in multiple to one winding of each of the three tens relays, N, 1N and 2N, these relays being double-wound for the purpose of making them self-locking when attracted. Any' one of these relays might thus be energized by any one of the three lines shown, but it is my purpose that only one of thcin shall be so cner 'ized at once, and that no two lines shall be a le to energize the same relay at the same time. Hence, on the other side of these relays 1 take away the circuitwires 54. and 56 to three separate controlling means presently to be designated. Each of these relays has its locking coil open on one side, but adapted to be grounded by the relay it--elf in pulling up: and on the other side of this second or locking winding it is connected through a wire .37, 58, or 59 to con trolling means similar to those I have just mentioned, and which will also be specified later.

The second branch from the line-relay is marked 51 for line 502; 52 for line 50].; and 53 for line 500. in other words, each of these lines has a difi'erent units wire, which is proper, as will be apparent at once if Fig. 7 has been understood. The three wires, 51, 52 and 53 pass along the bottom of the Figs. 2 and 3 and into Fig. 4, where they termi-- nate in the first or operating windings of the three relays, I, P and P also by multiples in the corresponding windings of the relays, 1P, 1P, 1t and 2P, 2P and 21. The three relays, P, P, P since they belong to the same set, have a common return, 60, which connected to the common return, 54 from the tens relays, N, N, of the same group or trunk. This wire 54 passes into Fig. 5, and certain contacts on one of the relays, by which it is controlled. This will be fully explained a little farther on. Each of the units relays in Fig. 4 has a locking winding separated from the actuating winding, and each locking winding terminates on one side in a contact normally open but adapted to be grounded when there ay is energized. All the locking windings of the first set or group have a common return, 61, passing to the wire 57, which is the control wire of the locking windings on the tens relays, N, N. This wire 57 passes into Fig. 5 and is there carried to one of the relays, by which it is controlled.

The actuating windings of the units relays 1P, 1P, 11, have a common return, 62, passing to the wire which goes to a relay contact belonging to trunk l3O -l31. The locking windings of these relays. 1P, 11", ll have a common return, 63, which goes to the wire 58, passing into Fig. 5 and there controlled by one of the relays. in the same manner as in the first set. Similarly, the actuating and locking windings, respectively. of the units relays. 2P, 2P and 2P have common return wires til--65, connected respectively to the wires 56-59. which pass into Fig. 5 and are there controlled by the relay of the trunk No. 230-23L his thought the scheme will now'be plain. Suppose that one of the line-relays, as that of line No. 502, pulls up. It grounds the local circuits 50 and 51 and energizes whichever one of the tens relays, N, IN or 2N happens to be idle, together with the appropriate units relay. This would be one of the three relays, P 1P or 2]? in Fig. 4. The tens relay and units relay thus selected having pulled up, their locking windings are grounded, and hence they are maintained energized through the locking common return as long as a subscriber keeps his receiver off the hook, being finally deprived of current when'the subscriber hangs up, so that the relays are then restored.

I have thus shown, first. a single trunk group of primary selector relays, and second three such groups, or at least portions of three groups, belonging to three several trunks, and havin the subscribers lines connected to them in multiple. I will now point out how a. calling line is permitted to select an idle trunk only, by working idle relays only. For this purpose it is necessary that I should refer briefly to Fig. 5. This figure shows portions of three trunk-circuits, ear-h of which commences at the letter 1' on the left of the figure, which is supposed to be the terminal point of attachment of the selective set shown in the preceding figures and in Fig. 7. Since the three trunks are identical, I will describe the upper one only, with its apparatus. The trunk-wires are numbered 303l, and the control wires which accompany the trunk are marked 54 57. The wire 54 is the common return from the actuating circuits of all the relays, both tons and units, belonging to the trunk 30 31, and the wire 57 is the common return wire from the locking windings of the same relays. It will be perceived that each subscribers line is shown in Figs. 2 and 3 to be equipped with a cut-off relay, whereby the line-relay may be disconnected, and the line put through onto its extensions, the line itself being in fact normally disconnected from the contacts on the tens relays. Since the tens and units relays which act to put any calling line onto a free trunk must be maintained energized after the line-relay has been cut olt', locking windings are provided on all the relays; and the common return 57 from these locking windings is taken through thetrunk solenoid relays, so that at the expiration of a communication all may be restored, the continued actuation of any set of relays being thus controlled through their own trunk. On the other hand, since only an idle trunk must be accessible to calling lines, I carry the control wire 54 from the actuating windings of the relays of one trunk to the point Where it can be. controlled b the condition of the next preceding trunl c in becoming busy. Gen erally stated, the method adopted in this last-named arrangement is this: Assuming all the trunks to be idle, the first trunk has its relays connected to battery. As soon as the first trunk becomes busy its tens and units relays which have been actuated to make it busy are looked through the wire 59 controlled by its own solenoids; but the common return 56 from the actuating windings of all the other relays of the same set is disconnected from the battery, so that no other line can come in on the same trunk. Simultaneously with this action. battery is supplied to the actuating windings of the next trunk in order, and consequently the next call that comes in will go on to that trunk and lock the relays, thereby actuated, at the same time disabling all the other relays of the same group, and in turn putting battery onto the third trunk. If now one of the intermediate trunks becomes idle, while those on both sides of it are busy, it will immediately become available for furthercalls, because the preceding trunk being busy supplies it with current, and the succeeding trunk being busy its relays are locked and not disturbed by the freedom of the intermediate trunk. If now the succeeding trunk becomes idle, while the intermediate trunk remains so, the succeeding trunk will not thereby become available, because as long as the intermediate trunk is idle, everything that follows it in order is cut off. I wish to say at this point that while the description must proceed in a lame and halting manner, as if the operation were slow, in reality the actuation of this system is so quick as to be correctly called instantaneous. None of the relays herein shown require more than a small fraction of a second for the operation, and hence the difie-rent changes and shiftings of the lines and trunks take place without any perceptible interval of time whatever. This would be dangerous, in that it might lead to confusion were it not for the careful pro vision made whereby all the relays are positive in their action. There is no marginal Working; I do not cross-connect any relays to work in opposition to each other, nor do I depend upon time intervals between relays of the same type. The only way that two lines can interfere by coming on the same trunk at the same time, is for them to close their line-circuits, and perform all of the resulting operations within an interval of less than say one-tenth of a second. The figure representing the probability of such a concatenation is so small as to be absolutely negligible.

Referring now to Fig. 5, I have therein shown what is really the connecting link between the primary selector portions of the circuit and the automatic connector portions of Fig. 6. All the apparatus in Figs. 2, 3 and 4 is merely for the purpose of answering calls and connecting the calling lines onto the circuits of Fig. 5, which I have called trunks for want of a better term. In other Words, the relays of Figs. 2, 3 and 4 take the place of operators in plugging in with an. answering plug, when a subscriber calls. In Fig. 5, the three trunks are shown as 3031, 130131, and 230-231. These come in at the left-hand side of the sheet, and terminate at condensers, C, C. On the other side of the con densers, and passing out at the right-hand side of the sheet are the extension trunk Wires -7l, 170-171 and 270271. As all three of these trunks are the same, I will describe the upper one only. This comprises the two main or talking circuit wires 30 or 31, with their condensers, C, C, and

their extensions, 7071, which pass into Fig. 6 and there connect with the automatic selector switch, S. Bridged across the wires 3031, in the branches 72-73., and with the main battery B between them, are the trunk relays, D, D. The relay, D, controls the battery supply in each case to the wire 74 and thence through the armature, d, of the relay, D, to the contacts of that relay. The relay D becomes energized as soon as a line is connected to the trunk 308l, a path for battery current being then immediately formed out through the line and back, and also to ground through the cutoff relay M. This relay is therefore held up as long as the line is on the trunk, by cur rent through the cut-01f relay. The relay D on the other hand responds to the move ments of the subscribers hook and is employed for the purpose of repeating the impulses from the calling line into the switch circuits of Fig. 6. The manner of doing this will sufficiently appear from the statement of operation hereinafter. It suffices at this moment to say that when the armature, d, is back, the solenoid relay Q is drawn up, and when the armature, d, is up the solenoid relay Q is drawn up. These solenoid relays are preferably made with their spools set vertically, the windings surrounding inner tubes within which the cores slide up and down after the fashion of plungers in dash-pots. The upper ends of the tubes are fitted with small check valves opening outwardly, so that each core may be drawn up quickly, but will retract or fall down very slowly. The solenoid, Q, performs several functions. To understand these perfectly it is necessary to have in mind the type of automatic selector and connector switches which I employ. These may be of any wellknown or suitable type, such as those which I have myself designed, but they are herein shown for convenience as of the Strowger or automatic electric type. In order to operate this type of switch, the actuating impulses are sent over one side of the line, and the controlling or shifting impulses are sent singly over the other side. In finally re leasing and restoring the switches, an impulse is sent over both sides of the line simultaneously. In my circuits, the actuating impulses are transmitted over the trunkwire 70, in groups corresponding to the digits of the number wanted. The controlling impulses are sent over the wire 71, singly, and in bet-ween groups of actuating impulses. For example, suppose the number wanted to be 468, four impulses go over the wire 70, then a singleimpulse over the wire 71, then six impulses over the wireJO, then another single impulse over the wire 71, and finally eight impulses over wire 70, followed by another single impulse over wire 71. The purpose of my solenoids, Q,

Q", is to produce these impulses in the'two wires and 71 in 'their proper order and arrangement, in response to makes and breaks only in the metallic circuit 30-31; for it will be observed that at the subscribers station in Fig. 1 Ishow no grounds, all of my present designs dispensing with grounds outside the central office. The solenoid, Q, pulls up at the first energization of the relays, D, D, and stays up until the subscribers finally clear out. In coming up it opens the contacts 92 and 93 and closes the contacts 90 and 91, these latterbeing together with the contacts 8586 of solenoid Q, the controlling contacts of the relay locking circuit 57. At the first break produced by the sender or key, K, at the subscribers station, after the line has become connected, the armature, d, falls back for an instant, and the solenoid, Q, pulls up. The impulses into wire 70 are repeated direct by the armature, d, by way of the wire 80, contacts 87-88 and wire 89. The contacts 87-88 are closed together when the solenoid pulls up. The cont-acts 84 are momentarily closed as the solenoid pulls up, without effect, because they derive current from the wire 81 attached to the front con tact of the relay D, which is open when the solenoid pulls up and closed when it falls down. The latter movement takes place between the' groups of impulses, and a single pulse of current is then sent by way of wire 83 to the wire 71. The solenoid, Q, is thus seen to be really theclearing out solenoid, while the solenoid, Q, is really the controlling or actuating solenoid.

'1 he wire, 54, it will be recalled, is the common return for the actuating circuits of all the primary selector relays appertaining to the trunk 3031. This actuating circuit derives its current from a normally closed back contact of the armature, d, of the relay, D; and this armature, d, derives its current in turn through the wire 98 from we wire 74 controlled by the relay, 1D of the next trunk 130131. This battery supply circuit may be traced as follows: bu tery B, wire 75, armature d of the relay 2D, wire 74, 77, 99, armature d of the re lay 1D, wire 98, armature dof the relay D, to wire 54. The common return 55 from the actuating coils of the relays of this trunk 130-131 takes current from a back contact of the armature, d, of the relay, 1D, which in turn derives its current through the wi 99 from the wire 74-of the preceding trunk 230231. The common return 56 from the actuating windings of the relays appertaining to this last trunk, derives its current from a back contact on the armature, d of the relay 2D, this deriving its current from the main battery. The effect of this ar rangement is to make the trunk 230231 normally active, since its relays, 2N, 2N;

2P, 2P, 2P are normally on the main battery; but all the other trunks are normally inactive. As soon as a calling line becomes connected With the trunk 230231, the relays 2D, 2D, pull up, the common return wire 56 is disconnected from battery, the pair of relays that have already pulled up remaining energized by their locking coils through the wire -59, but all the other relays of that set, tens and units alike, being disabled by the cutting ofl of the common return. At the same time, battery is put on the wire 77 leading to the next trunk 130-131, and through the armature, d, of the relay, 1D, thereof, to the wire 55, which is the common return of the relays, IN, IN; 1P, 1P, 11, etc.-, of that trunk. The next call will therefore find the trunk 130131 active, and as soon as that trunk has become connected, it will in turn cut off its own actuating coils, except those we have already operated, and will shift the battery to the next trunk 3031 through the wire 98. When that trunk operates, it will cut ofi its own actuating common return54 in the same way, and throw battery onto the wire 77 to render the next succeeding trunk active, and so on.

Referring now to Fig. 6, which completes the chain of circuits, two switches are therein shown, as types of an organization suflicient to serve 1,000 lines. These switches are, in ordinary technical parlance, a second selector, S, and a connector S. I'have described the mechanical construction of these switches as shown in Figs. 10 and 11, so it will merely be necessary to refer briefly to the circuit connections before passing to the operation of my system as a whole. The Wires 70, 71 terminate in switch-arms, 2, 3 of the side switch of the second selector, S. It should be explained that the present equipment being for 1,000 lines or "less, it is only necessary to hate these two sets of switches, the second selector, S, serving to pick out the desired hundreds group wanted; and the connector switch, S, of that group then serving to pick out the desired tens and units. The switch-arms, 2 and 3, normally rest upon terminals of wires 100 and 101, connecting the yertical and rotary rclays, V and R to the wires 70 and 71, and maintaining this connection until the third step of the switch. The wires 118 and 119 leading to the wipers s, 8", are normally open, and the test relay T has its circuit 142 also normally open at the switch-arm, 4. The connector switch has the same parts as the selector switch, with some additional parts. The vertical magnet, v, is the same, but the rotary magnet, r, does not have the vibrator contacts, 1, responding only to single impulses. The vertical and rotary relays, V, R, control the magnets, together with the private or controlling relay, U.

Images