Strowger System

Almon B. Strowger in his lateryears, fram "Telephony" tor Oc­tober 15. 1949.

the central office for each subscriber line.The switch consisted essentially of a hollowcylinder with a shaft capable of both verti­cal and rotary motjon. This shaft carried awiper arm capable of wiping over all thehorizontal rows of contacts on the inside ofthe cylinder. All of the subscriber lines weremultipled to the rows of terminals aroundthe cylinder of every switch, while one ofthe subscriber lines was connected to thewiper of each switch. Strowger did not spe­cify in his patent the maximum number oflilies which could 'be terminated in hisswitch. but in a circular issued in April,1891, he claimed that it would "connect tele­phones in aIsystem from two to ten thou­sand." In describing its method of operation,however, most writers have assumed tenhorizontal rows of 100 contacts each. (Eventhis number would have required a switchprohibitive in size and cost.) The verticaland rotary motions of the shaft were con­trolled by electromagnets actoated by im­pulses sent from the subscriber station. Fourpush buttons on the subscriber telephonebox each controlled a wire to the centraloffice, while the fifth wire was for talking.

To call the number Bl5, Strowger statesin his patent, the subscriber pressed his"hundreds" button (d) three times, thuslifting the shaft of his switch three notchesand bringing the wiper opposite the third ,row of terminals. He then pressed his "tens"

95

R. B. HILLGeneral Staff

of the

The Early Years

Some of the early attempts to devise anto­matic telephone systems were describedbriefly inc a previons article", and it wasstated that, while some of the early inven­tions embodied ideas that were later devel- 'oped into important featores9f the modernautomatic art, none of the systems them­selves went into commercial use. It remainedfor Almon B. Strowger, an undertak~r ofKansas City, Mo., to devise the first auto­matic telephone system to be used commer­cially. The important featore of this inven­tion was that of imparting to the switch m;m£lrst a vertical movement, and then' a rotarymotion in a horizontal plane. Strowger is saidto have constructed the first crude model ofhis switch from a cardhoard collar box and apaper of pins. He stock the pins from theoutside of the box toward the center untilhe had ten horizontal rows of ten pins each,one row above the other, to represent tbeline terminals. The vertical shaft for carry­ing the switch arm was represented by alead pencil through the center of the box.

Strowger filed his patent application onMarch 12, 1889, and it was issued on March10, 1891 as patent No. 447,918. The mecha­nism of Strowger's patent is shown in Fig­ure 1. While this particular arrangementwas never used commercially, it illustratesthe inventor's original ideas. As may beseen, an individual switch was required at

RECORD, January, 1955, page 22.

March, 1953loratories Record

.panies are re­stomers regfird,-feature, some

hods of piovid,ns and the spe­areas post cards,hen the serviceLOges occur. In.s are ,adequate.ate what,opel'a­'ormation.nst "listen to allitY of the eight­'ach subscriber,e teti:party sys­riber hears £lve,'ide a 'ten-party,bscriber wouldrequire an addi­lis was found to

hipstre designed toch in the funda­Since they were{Ships nave beentnt $4,000 to the.hip and $1,500

"""ducted creativelading centers of"pient being freeit which he willJps are awarded)ommittee of sci­staff of Bell Lab­are the research

1e importance ofd the likelihoodist. The ,awardshave recently re­l' who are about

Courtesy www.historyofphonephreaking.org

button (H') once, which gave a single rota­tion to a 10-toothed ratchet wheel (E') at­tached to the shaft and moved the wiperarm to 310. Pressing the "units" button (I')!lve times operated a 100-toothed ratchetwheel (E) attached to the shaft and movedthe wiper arm five more points to 315.

Ringing was done with a magneto or bat­tery. When through talking, the calling sub­scriber pressed the (p') button, which ener­gized the release magnets and brought theshaft and wiper arm back to their normalpositions. The use of five wires and groundwas, of course, an impractical feature, butStrowger planned to later reduce the num­ber of wires. A heavy battery was to be em­ployed at each sub-station for operating thecentral office switch. There was nO provisionagainst the calling subscriber being con­nected to a busy liue.

The first company was incorporated onOctober 30, 1891, under the name of theStrowger Automatic Telephone Exchange.Early in 1892, A. E. Keith, who had pre­viously been employed by' the Brush Elec­tric Company, of Baltimore, entered theservice of the Strowger company and wasfor many years thereafter one of its leadingtechnicians and inventors.

On November 3, 1892, the first Strowgerexchange was opened for public service atLa Porte, Ind., with about seventy-five sub­scribers. This was the first automatic tele­phone exchange to be installed anywhere,and a considerable amount of ceremony wasattached to the affair, with a special trainrun from Chicago and a brass band on handto greet the guests. Since this system wasdesigned for less than 100 lines, there wasno need for a switch with a two-way move­ment, and a flat rubber disc type of switchwas employed, with but one (rotary) move­ment for the wiper arm and one circular rowof contacts. There were still five line wires,as in Strowger's patent, aud each telephoue.was equipped with four push buttons, al­though the hundreds button was not used.The battery for operating the switches wasnow located at the central office. There wasstill no provision against a subscriber beingconnected .to a busy line. The systemworked with reasonable accuracy wh'Jn thesubscribers operated their push buttons cor­rectly and remembered to press the release

96

button after a conversation was finished.In 1893, a small Strowger equipment was

exhibited at the Chicago World's Fair.About the beginning of the year 1894,

Frank A. Lundquist and the Erickson broth­ers, John and Charles J., entered the serviceof the Strowger company, and, with A. E.Keith, started work on an improved systememploying the so-called "piano wire" ter­minal banks with a Hat-faced multiple anda Switch arm capable of both a rotary and alongitudinal movement. Switches of this

.type, which were covered by patent No.540,168, were installed at La Porte, Ind., inthe fall of 1894 to replace the original equip­ment, and at Michigan City, Ind., in the sameyear. These new installations employed asimple form of busy signal. When the call­ing subscriber rang, both bells respondedif the called line was free. If it was busy,the calling subscriber bell remained silent.They also utilized the !lrst automatic re­lease arrangement, invented by A. E. Keithin 1893 and covered by patent No. 573,884,embodying a mercury switch connected withthe subscriber switehhook, which restoredthe central office switch to its normal posi­tion upon the hanging up of the receiver.

The "piano wire" form of switch did notprove successful, and in 1895 a return wasmade to the principle of Strowger's originalpatent. Patent No. 638,249, issued to A. E.Keith and the Erickson brothers, covered aswitch which more nearly resembled themodem step-by-step switch. Twowews ofit in the patent are shown in Figure 2. Itemployed semi-circular banks and the famil­iar "up-and-around" motion of the selectorrod and wipers. One-hundred-line switchesof this type replaced the "piano wire"switches at La Porte in June, 1895. In Aug­ust, 1895,' a similar system, with 200-line,"witches, was installed at Michigan City.

At this time, the Strowger system em­,ployed., betw'een a subscriber's premises andthe central office, two hard drawn copperline wires, and a soft copper battery wirewhich branched into all the stations to sup­ply signaling current. This arrangement con­tinued in use until about the year 1900. .

In the summer of 1896, an important for­ward step was made with the introductionof the fuiger-wheel dial to replace the pushbuttons. It was covered by patent No. 597,-

Bell Laboratories RecordFig. l-:Strowger',

( March, 1953

Courtesy www.historyofphonephreaking.org

was finished.:quipment was_old's Fair.he year 1894,:rickson broth­red the serviceld, with A. E.proved systemlilO wire" ter­1 multiple anda rotary and aitches of thisby patent No.Porte, Ind.,ift

original equip"ld.) in the sam~IS employed aWhen the calI­ells responded[f it was busy,~mained silent.automatic re­

by A. E. Keithat No. 573,884,connected withwhich restoredts normal posi­the receiver.switch did not5 a retnrn wasIwger's originalissued to A. E.hers~ covered, aresembled theTwo views of

in Fignre' 2. Its and the famil­of the selector

·d-line switchescCpiano w:fre~~

" 1895. In Aug­, with 200-lineHchigan City.;er system ern­IS premises anddrawn· copper

or battery wirestations to sup­rangement con­le year 1900.I important for­ile introductionlplace the push,.tent No. 597,-

"atories Record

Fig. I-Strowger'. firet automatic telephoM syetem, De dleclo.ed In hie patent No. 447,918.

March, 1953 97

Courtesy www.historyofphonephreaking.org

Fig. 2-Front and side "iews of the first Strowger switch to bear a general resembUzlUJe $0 themodern step.by.step switch., '

98 Bell Laboratories Record

062, issued t.o Jbrothers. As "this first formholds mountedble circular piaup a spring whwas withdrawnits normal positlimited to a m,ment ~echanisthe required nutook place to ccentral office aJ

In the fall ofStrowger Comphis health. He tdevelopment 01name, and he (ida, in May, 19

It was alsoA. E. Keith ,started work oning the trtmkiJorder to remOVEan automatic mcessity of multlines to each sw:of course, but nhad produced ,commercial USE

switches whichlines would eacllines, and enou!switches wouldvarious hundre<pr~ceded by aThe Selector wo

digitthe call to a canthe latter woul,units digits andTwo switches Vi

but they might hth"n a single la

The Keith.EIsysh,m, coveredof this characteStrowger exchanin March, 1897,this installation 1of terminals at (which it could

o RECORD, Janool

March, 1953

Courtesy www.historyofphonephreaking.org

Fig. 3-1';ir8t form of Strowger finger-wheel sub­station dial, which replaced the push buttonspreviously used.

rotating the selector were provided by thecalling subscriber, through the device of in­serting a dummy "0" in the directory num­ber following the hundreds digit. The se­lector was desigoed to rotate in response tothe pulses of the dummy digit until an idletrunk was encountered, after which the re­maining pulses were disregarded. This planwas soon superseded by one in which thepulses were furnished by a continuously op­erating interrupter in the central office. Theplan could be extended to exchanges of10,000 lines by inserting another stage ofselectors in the train.

group of selectors had,.only one trunk toeach group of connectors. Because of the in­efficiency of these one-trunk groups, thesystem although operable was not commer­cially practicable.

In the summer of 1897, however, theStrowger engineers started work on theproblem of a 1000.line truriking system em·ploying automatic trunk selection. Underthis plan the selector would have severalsets of contacts in each row, leading to theseveral connectors serving a group of 100lines, and would be arranged to connectwith the first idle trunk it found in rotatingover these contacts. In the first arrangementdesigoed for this purpose, the pulses for

062, issued to A. E. Keith aud the Ericksoubrothers. As will be seen from Figure 3,this first fopu employec;! numbered fingerholds mounted on the periphery of a rotata_

C

ble circular plate. Dialing a number woundup a spriog whose tension, when the fingerwas withdrawn, caused the dial to return toits normal position. The return rotation waslimited to a moderate speed by an escape"ment mechanism, and, during the return,the required number of circuit interruptionstook place to control the movement of thecentral office apparat,!s.

In the fall of 1896, A. B. Strowger left theStrowger Company and went to Florida forbis health. He took no part thereafter in thedevelopment of the system which bore hisname, and he died in St. Petersburg, Flor­ida, in May; 1902. .

It was also during the year 1896 that'A. E. Keith and the Erickson brothersstarted work on a l,OOO-line system employ­ing the trunking or transfer principle inbrder to remove the limitation on the size ofan automatic exchange imposed by the ne­cessity of multipling all of the subscriberlines to each switch. The idea was not new' ,of course, but none of the previous workershad produced an arrangement suitable forcommercial use. Under such a plan theswitches which make connection to calledlines would each be limited to reaching 100lines, and enough groups of such connectorswitches would be provided to serve thevarious hundreds of lines. They would bepreceded by a stage of selector switches.The selector would respond to the initial orhundreds digit of the nUIflber and' extendthe call to a connector in the proper group;the latter would respond to the tens andunits digits and conuect to the desired line.Two switches would be used on each call,but they might be made smaller and cheaperthan a single large switch. .

The Keith-Erickson I,OOO~line trunkingsystem, covered by patent No. 672,942 wasof this character. It was employed in theStrowger exchange installed at Augusta, Ga.,'in March, 1897, to serve some 400 lines. Inthis installation the selector had only one setof terminals at each of the ten positions towhich it could be dialed, and thus each

• REcORD, January, 1953, page 22..

March, 1953oratories Record

~semblance to the

Courtesy www.historyofphonephreaking.org

Incidentally, after protracted Patent Of·fice interference proceedings, the patent(No. 776,524) covering the antomatic se­lection of an idle trunk was awarded toFrank A. Lundquist, who had left theStrowger Company in 1896.

The Strowger exchange at New Bedford,Mass. - the first automatic exchange to em­playa practicable form of trunking system ­was opened for service in November, 1900.It employed first and secqnd selectors, andwas equipped for some 3,600 lines.

Other improvements were also made intbis installation. The common battery sup­ply wire was discarded, only the two linewires being carried back to the central office.However, the telepbone had a third wireconnected to ground, which served as acommon return, the call being made bygrounding one line or the other the required

Fig. 4-Strowger wall set with finger-wheel dialabout the year 1902. .

lOO

number of times. For this reason, this ar­rangement, which was standard until 1907,was usually referred to as a three-wire sys­tem. Local batteries still furnished the talk­i11g currentl and the subscriber rang byqperating a push button which connecteda central office ringing generator into thecircuit.

In September, 1901, a Strowger exchangevery similar to the New Bedford installationwas opened for service at Fall River, Mass.,with about 400 subscribers.

During the year 1901, the Automatic Elec­tric Company was organized and acquiredthe United States rights to manufacture andsell Strowger equipment. The Strowger Au­tomatic Exchange remained only a patentholding company until 1908, when its stockwas acquired by Automatic Electric.

The New Bedford and Fall River installa­tions may be said to mark the beginning ofthe work which eventually led to the devel­opment of a Strowger system suitableJoruse in large single-office cities. There werestill many improvements to be made, themost important of which were: bridg;ngtheswitch-controlling relays across the line, in­stead of placing them in series with theline; commOn battery talking; automaticringing; the provision of practical forms ofmeasured rate and selective party-line serv­ice; means for eliminating the need of anexpensive first selector for each subscriberline; and the replacement of the three-wire(two-wire and ground) system with astraight two-wire system. Until these im­provements were introduced, there was noprospect of the Strowger system competingon anything like even terms with the manualcommon battery system.

What some of the iipparatus of the Auto­matic Electric Company looked like at thisperiod (about the year 1902) is shown inaccompanying illustrations. Figure 4 showsa wall set equipp!"d with a finger-wheel dial.Figure 5 shows fropt arid side views of theselector SWitch; of Which there was one foreach subscriber line. In the upper half werethe electromagnets and their associatedequipment. The lower.half consisted ofthree banks of terminals, With a wiper foreach bank. The upper, or "busy," bank wasused to show whether a trunk was bury, andthe two lower, or Une," banks provided the

Bell Laboratories Record

talking conne(played first SE

four-digit sysisecond select<'t.er were veryselectors. Latmade it possil

'terminals intosuIted in a. sh,

After the Fexchanges we:trlct of Chie<.about 1,500 suJuly 1903, wiGrand Rapids5,300 subscribThe Chicago i

March, 1953

Courtesy www.historyofphonephreaking.org

Fig. 5-Front and Me 'views of Strowger selector switch a,pout the year 1902,

ger exchange to furnish any type of mes­sage rate service, the first selectors being

. equipped with a commercial form of meter(Veeder meters). These did not register abusy call, but did register all completedcalls, whether answered or not. The Daytonexchartge was the Brst in which the control­ling magnets were bridged across the line,instead of being in series with it, thus effect­ing a considerable improvement in trans­mission characteristics.

The first of several Strowger branch of­flees, designed to connect with and supple­ment the Main (independent) manual· ex­change at Los Angeles, was completed inJuly, 1904. Several years later, the Mainexchange was cut over to dial operation.

101

talking connections, 1'wee-digitsystemsem~ployed first selectors artd connectors, whilefour-digit systems employed first selectors,second' selectors and conrtectors. The lat­ter were very similar in appearance to theselectors. Later~ mechanical improvementsmade it possible to incorporate the 200 lineterminals into a single line bank, and re­sulted in a shorter two-bank switch.

After the Fall River exchange, Strowgerexchanges were installed in the "Loop" dis­trict of Chicago in February, 1903, withabout 1,500 subscribers; at Dayton, Ohio, inJuly 1903, with 1,300 subscribers;' and atGrand Rapids, Mich., in January 1904, with5,300 subscribers, replacing a manual board.The Chicago installation was the Brst Strow-

March, 1953

) reaso:h~ this ar~

~dard until 1907,a three-wire sys­lrnished the talk­)~criber rang bywhich connected;merator into the

:rowger excb;mge,dford installationFall River, Mass.,s.e Automatic Elec­'ed and acquiredmanufacture andfhe Strowger Au­.ed only a patent)8, when its stocklc Electric.call River installa-: the beginning ofI led to the (level­'stem suitable for,ities. There wereto be made, the

vere: bridging.theICroSS the line, in­II series with the,lking; automaticpractical forins ofve party-line serv­19 the need' qf anlr each subscribert of the three-wireI system with aI. Until these im­.ced, there was nosystem competing~s with the manual

tratus of the Auto­looked like at this1902) is shown inlS. Figure 4 shows1 finger-whe"l dial.I side views of thethere was one fori

be upper half )Vere,:I their associatedhalf consisted of

:, with a wiper forr ''busy," bank wasrunk was busy, andbanks prOVided the

moratorie. Record

Courtesy www.historyofphonephreaking.org

Fig. 6-Early form of Keith line switch.

The first Strowger exchange with commonhattery talking was installed at South Bend,Ind., in May, 1905, to operate jointly with amanual common battery board in the samebuilding. It served about 700 subscribers.With the advent of commOn battery opera­tion, it became possible to employ automatic(machine) ringing, whereby the. called sub­scriber bell was rung automatically as soonas the connection was made to his line. Com­mon battery operation and machine ringingwere, of course, old in manual switchhoardpractice at this time. The Hayes repeatingcoil system of common battery supply hailbeen developed by the Bell System in the1890's and was widely used in their ex­changes, while the Hibbard machine ringingsystem had been standard in Bell exchangesfor many years. The early Strowger com­mon battery exchanges used a bridged im­pedance type of circuit.

An arrangement aimed at convenienceand economy in the grouping of the appara­tus was covered by A. E. Keith's patent No.831,876, filed on March 9, 1905. It called forten rows of ~~n frames each, every framecontaining the switches for serving 100 sub­scribers, and consisting of 100 first selectors,

102

10 second selectors aJ;ld 10 connectors. Thisarrangl"ment provided f9r 10,000 lines, eachrow serving a particular 1,000 lines. Thispatent.. also covered the important featureknown as "bank, slip," wherein the trunkmultiples are wired or "slipped" in such away that if ten selectors start at the sametime to test the same trunk group, they willall at their first step test a different trunk.This arrangement shortened the time neces­sary for a selector to find an idle trunk, re­ducecj,'tl'le possibility of two selectors seizingthe same trunk, and equalized the wear onthe switches.

About the year 1905, the Automatic Elec­tric Company dev~loped a four-party selec­tive signaling system for use with their com­mon battery equipment. This feature, ofcourse, was also 014 in Bell System manualexchanges, the Hibbard, system, employingoppositely biased polari>;ed bells~ havingbeen in use for many years. The Strowgerexchanges used harmonic ringing,

The earliest Strowger exchanges had allbeen eqUipped with a first selector for eachsubscriber line, and this practice had con­tinued when the larger exchanges, employ­ing trunking, had come into use. It involved

Bell Laboratories Record

a g~'eat wastete}' per centbe in use atalimitation onsystem was r€No. 1,304,324plunger type,issued until 11

rangement, ,around the yc90 per centeach subserilswitch, muchfirst selector, 'his,Iine autdrnas soon as hehook. The firinitial dialingnero This limconsisted essecarrying on iiwhen actuatemagnet, pres!that connectelten trunks Iehundred linechanically bywhich kept tIswitches direc

Early in 19(twocwire systceliminated, wawas accomplisOr siow-releasgized by a cmfor a short pelits winding h,Wire system, tIing' the line a

THE AUTHOR,. d~gree from Hay. tered the Engin'

can Telephone III

of that year. FOiengaged. principastudies. 'When tand Research wato it; and since fhad heen largel)economic phasesHe had been a rphone Lahoratori,Plant DevelopmeStaff ·Deparbnent

March, 1953

Courtesy www.historyofphonephreaking.org

a great waste of equipment, since only about·' stead of grounding one side of it. The vern­ten per cent of the. first selectors wo~ld cal and rotary relays of the previous three­be in use at any: one time. This very serious wire system were replaced, respectively, bylimitation on the economy of the Strowger a qUick-acting relay and a sluggish relay.system was removed by A. E. Keith's patent When a series of pulses (one dialing opera­No. 1,304,324, covering a line switch of the tion) had been received, the pause before aplunger type, filed on April 29, 1905, but not second series could be sent in allowed theissued until many years later. With this ar- sluggish relay, which was held up duringrangement, which' went into general use the dialing, to falloff, giving the signal foraround the year 1907, and eliminated about the switch arm to rotate and choose an idle90 per cent of the first selector switches, trunk. The two-wire system sinlplified theeach subscriber line terminated in a line central office wiring, elinlinated, during dial­switch, much smaller and cheaper than a ing, the disturbing influences due to differ­first selector, whose function was to connect ences in earth potential, and permitted ahis line automatically to an idle first selector considerable simplification of the substationas soon as he removed his receiver from the dial itself.hook. The first selector then received the With the two-wire common battery sys­initial dialing impulses in the usual man- tern, provision was made for measured serv­ner. This line switch, shown in Figure 6, ice by employing a differential or polarizedconsisted essentially of a magnet and a lever service meter associated with the line switchcarrying on its end a plunger. The laUer, and operated by the reversal of the currentwhen actuated through the lever by the through the line of the calling subscribermagnet, pressed together certain springs at the time the called subscriber answeredthat connected the subscriber line to one of his telephone.ten trunks leading to first selectors. One The advent of the two-wire system, withhundred line switches were controlled me- its basic simplification of the method of sig­chanieally by a cornman "master switch," naling from the station to the central office,which kept the plungets of the idle line marks the beginning of the modern phase ofswitches directed toward idle trunks. the Strowger (or "step-by-step") system.

Early in 1908, the first Strbwger straight The engineers of the Automatic Electrictwo'wire system, in which the ground was Company continued to make improvementseliminated, was installed at Pontiac, Ill. This in the step-by-step system, while the Bellwas accomplished by the use of a sluggish, engineers developed the panel type of rna­or slow-release, relay which, when ener- chine switching system. These parallel de­gized by a current, held its contacts closed v.clopments and the far-reaching effects offor a short period after the circuit through World War I led to the adoption by the Bellits winding had been broken. In the two- System after the war of step-by-step equip­wire system, the pulses were sent by open- ment for small and medium size cities anding the line a certain number of times, in- panel in the largest ones.

103

THE AUTHOR: ROGER B. HILL received a B.S.degree from Harvard University in 1911 and en­tered the EngineeTing Department of the Ameri­can Telephone and Telegraph Company in Augustof that year. For several years thereafter he wasengaged principally in appr,aisal and depreciationstudies. When the Departt;hent of Dev~Iopment

and Research was formed in 1919, he transferredto it, and since then, until his retirement in 1951,had been largely concerned -with studies of theeconomic phases of .development and operation.He had been a member of the staff of Bell Tele­phone Laboratories since 1934. first in the OutsidePlant Development Department and later in theStaff Deparpnent.

March, 1953)ratories Record

i\.utomatic Elec­:our-party selec­,with their com­Lhis feature, ofSystem manual

item, employingd bells, havings. The Strowgerringing.changes had allselector for eachractice had con­hanges,employ­I use. It involved

ounectors. This1,000 lines, each000 lines. Thisportant feature,rein the trunk,ped" in such ant at the samesroup, they willdifferent trunk.the time neces­

1 idle trunk, re­selectors seizing~ed the wear on

Courtesy www.historyofphonephreaking.org