‘L WHIWIHI t l T | l ||H|Hl I l ____———- ————— "flT EY‘A n : .,. a!“ “E" d. I A..-.L4' l..- z?‘f‘-.’EST?G.fi..T{ON 0T CONTACTS ANS CONTACT LNBURANCE E-‘ALFL Cg; ANDRES ~.- A r) : ‘ \'}. "I I, 1. ‘ ,' Q—A ‘ 'oM V'fi’ T 53,! I.|l(*:|Y, T IHEIIS mflmK‘I Ls'trez‘t 1 m fl.‘ . .ulk I. AH IHVESTIGATIOH O? COHTACTS ASD “TATALL EX "?AHCE ;L.g".‘{,q¥ A Thesis Submitted to The Fae ulty of KICHIGAH AGRIC"LTITRZL C‘LLBGL BY 9: UL . AT3223 Candidate fbr the Degree of “lectrical fingineer JUHE 1924. THESIS Connections are estthlished in automatic telephony by means of the contacts on strings actuated by the nrmatures of relays and magnets. the contacts are either riveted or welded to sprints of various m teriels and gauges end are then aligned in prOpcr position on the spring assembly. Iheir position de- pends on the circuit duty, causing certain sprixés with their associated contacts to mane or break when the relay is ener- eized. She duty to wnich any contact is sutjected derends on its position in the circuit, that is, the oreretion of opening or closing the circuit may occur once in the establishment of a throuéh connection as in the case of Lie ring out off relay, or a number of times as in the case of the pulsing relay which responds to the impulses sent by the ceiling device at the sub- scriber's station. Again, the number of operations may be in- definite ns given in the case of the rotary line switch which Operates until it finds an idle trunk. In order to insure definite electrical connection with the back or break contact, the armature sprinhe are given a certain tension when in the normal or non-orerete position of the relay. In every case the power of the relay when energized is more than sufficient to cause preper electrical contact of the armature spring contact with the front or make spring con- ,\ tact. These tensions ere deternined by the Opercting conditions togehter with the design and winding characteristics of the relay 93779 -1- Contact reteriels; ExPeriments over e large number of yetrs here gradually reduced the kinds of materiel suitable for contacts in automatic telephony to three alloys of approximately the fOIIOWIHé cor- . positions: Metal A. Platinum with less then 2; Iridium Hetal 3 Silver 90%, Gold 10$ metal 0 Gold 70$, Silver 253, filetinmm 5; These metals will be referred to later as platinum, silver and gold, recrectivcly. The physical and chemical characteristics such as density, melting point, heat conductivity, hardness, mechanical weer, chemical corrosion and resistance to spark erosion large- ly determine the durability of a contact. Some of these cher- ecteristics in the case of the pure metals entering into contact materials are given below in Table 1. Table l. Haterial Density Eelting Point Resistance Beet Con- ductivity. Platinum 21.37 1755 10.96 Iicrohms 0.173 per cm3 9016 19.1 1063 2.22 A ‘ 0.705 Silver 10.6 961 1.#7 1.006 Tungsten 19.3 3230 6.2 0.376 Hickel 6.9 ‘ 14cc 6.93 + ' 0.1n2 Conger 8.89 1033 1.77 0.91' It may we noted that platinum hes the highest melting point, the greatest specific resistance and e relatively low‘ heat conductivity. Silver on the other hand hes a low melting -2- , U . L l 1 I . , . . .- . - . . .. \ . . _. .- ‘ . ‘v . ‘ . \ . c k . op.u « h ‘ r. | . l _, ~ 1 I l a v ‘ o .. . point, a low specific resistance and a very high heat conduc- tivity. Gold has characteristics between those of platinum and silver. gycbenicsl Construction of Contacts: Contact material is usually obtained in the form of wire of #114 and $18 3 a: S gauge. Figs. 1 end 2 illustrate the size of this material formed into rivets end the corresponding welded contacts. diveted contacts are placed in a perforation of the contact spring and consequently insure a rigid mechani— cal structure. Contacts welded by automatic machines insure an equully good mechanical structure, but because of the reduced area of contact surface between the contact and the Spring it would appear that are Llcctrical and heat conductivity of the contact might be reduced from that obtained with riveted con- tactS. Generally, large or £14 gangs contacts are attached to Springs of hesvy gauge and consequently are more readily riveted then welded. Fig. 2 shows such a large contact rivet- ed ic a heavy gauge coring. Electrical ficsistance of Junction between Coutsct end Sprinp; Measurements made by the drop of potential lethod give resistances on the order of M5 to 150 microhms for the Junction resistsnce between 18 gauge silver welded contacts and standard pulsing relay armature springs. These measure; ments were made according to the citline given in Fig.3. The resistance of the standerd pllsiug relay armature Spring from. terminal to contact is on the order of 5000 nicrohm8.' Because the resistance between the contact and its -3- _ r . . n r . . o v o A II .I e 4 . 4 Q .. . o o . , . a/ 4 g - .e . . A . . . . . _ . . . \ . J t . , . A old i . o . J 0 s . r .r, I. . n . . o x o L . . J A i c a . . r n I ha » o ,A o I. . o . , t A L o 6 . b . f no _ u n I A a. . V . a t A r u v . . . J g N .V r Jo . s , . a . l .a . O _ a ‘1 . . . A u‘ .. o I ,2 . c I‘u - u v u i p r v .‘o I. \ . O o . i d . . . . u i . ‘ . A . . c . . \1 . . A u I v _ . . n V .. .n . .A A . , «H . n o w . VI J . o . . A . l . . . . U I. . ‘ . R O. r l . I § 1 p . _ . . . . _ . V. . . . I . . 1.! -spring is so extremely snail, in fact approximately on the same order as the spring material, this is not a point of increased resistance in the circuit; however, there may be a gain in ther- mal and electrical conductivity by riveting ctmpared to welding, but this is difficult to determine. Electrical Resistance between Contacts: The electrical resistance between he point of contact where two different contact metals were used Was measured by an accurate wheastone Bridge. The contacts were assembled with their Springs in the relay structure and the pressure was grad- ually'increased and tendings tales at various pressures. The results are shuwn graphically in Fig. h. After arcing had occur- red at the contacts by Opereti;n in the standard circuit the resistance was again determined in the case of’the $1h gauge platinum contact and was found to be slightly lower for the greater pressures than when the contacts were new, although the difference is not very great. After these contacts were clean- ed with a sand blasted steel contact file the resistances as shown in Fig. 5 increased materially and the results were some- what erratic, caused probably by the irregular surfaces in contact. The PrOblem Stated: A number of years ago it WnB noted that certain relay contacts. particularly those of the pulsing or line relay showed signs of firming a crater on the negative or'back contact with a corresponding building up on the positive or armature con- tact; resulting in sticking ccntacts as well as decreased life of OperatIJU (Fig.6). An investigation was started to deter- mine the nature of this phenomenon and the results obtained to date are given below. “'1"- I Possibility of Reverse Current: Since in a carbon arc a.crater forms on the positive electrode, and a correSponding point on the negative, it was supposed that a reversal of current, as a condenser discharge, might be responsible for this action. To test this theory an osciIIOgram'sas made of the current flow through the contacts, and through the condenser circuit of a standard connector as shown in Pig. 7. The oscillogrsm 0-51 shows that the condenser charges and discharges a number of times at the beginning of the pulse, and once at the and, but in no case is the current re-. versed through the contacts. From study of the circuit, as will be shown later, it is apparent that the condenser can charge only through the relay and can discharge only through the back and armature contacts in a.dircction the same as the relay supply current. it no time does 3 tendency exist for any cur- rent to flow contrary to the direction of’the relay supply cur- rent. I Egtal Arc different from Carbon.src: To determine the formation of the crater in a net- allic are instead of carbon, two capper electrodes l/u" diam: were used. With a current of 2 amperes the crater formed on the negative. The positive electrode Operated at a higher tem- ‘ perature than the negative, resulting in a small globule of molten capper and cepper oxide being formed. Because of sur- face tension, the globulc assumed a spherical form causing the negative electrode to assume a crater since the are always played between this spherical globule and the shortest point to the negative electrode. 1 -5- Jim o |1VHfV I... I . a . . J . .| 4. .A i. V. 4 - . I _ n . v e .B e v . , a ' . .. . ) . ,. . n _ . . . y. . . . . . . . r. .. .n. s . . A \ . . 5. a i . v: \ c r . ... . ; L . . , . .4. , F w r: . .. , v: u .2 n s it . ,. .evouaeoao no use hence on oeemus- redone Bonn menu as m pagan noun eeouueoao sun» msaeoan nope: caoe uo sconce u up Hooe anon ocowwooae ebwwwuom .noaoeuauomwb nu unoa unease as ououaeoao .mod no need pnmao: «a an ee>eea mama .um cannon consumes“ .9“ Sony doggone pen» ecfiwo nuance one uemmoo no mecca can; occupooao spawneom .paeno: swam puoa cbapemmm! .ocoupeeao ahead -eon can» ones “on up . concouena net epduemen pen» cones new» sound .nnoueuhau nepeadaneo dd 0 es saw no on ee .9: no 0.” e n celhon page can» none-o one upon. an ceshcu heueho .Huo condone use ocean open cashew ceded-en Mo HancoC do»a¢: «once: an e>awmeom .uemmoe menu .mewmmm.wwssmsoo embeds l‘l uMWuem commence ebwwmmon meomq .ucfiqoo.fldnu unmvnw emeuveoHo nopmmu .ummmee menu Manama» cues colsenoe accuse .uuao» oaa mo owcaaob aqua nun: coach on o» ow noun coupe» one nuance eucuao> .ecequasoo 0mm no aneurysms.» look a an has some nu ocean noou eouo.HH< 11 weak .Nouan‘ endow mom m.m gmmHn.D acau .Houemd I xmewmv serene unconnw. n.m . nmmH.D 030 e p. 0 menu aimopeuu» egg umweknl. mm .h m gm bonaHme . nopsum iuebeubn }.di«m.. 2:61 Elites be .naa on ..n.a sewn“ Lumcmsuu .Hdom seam» enmoo eouenom sou-no ‘tfiofimmbfl “neon 1 hOQ&Mb konwmp honnnpw nomc@p season unseen noncpbw honaep Alsopmmp seamen gunmen” Hemmmb. ueflmop . qumoDI; sewn-p nonheo Monaco uuuesem and Ho .emsd necemwoeam flouaehsq: H .uqm «awn ue.¢fleo no.6neo e M 33.4. Obuuulom Naoldnq Hodudfldnloo ebuueueh eachaeedfl ‘ n... Uhen the electrodes were kept at room temperature except for the actual point of the era, by running cold water in at one end of the electrode and gut near the point of arc, the two electrode surfaces remained parallel except for slight irregularities. I This indicated that the electrodes could burn without forming a noticeable crater when the electrodes were kept cool or if the thermal conductivity was relatively large. Tests made on smaller electrodes showed that the formeti n of the negative crater increased rapidly as the diameter of the elec- trodes was decreased. Bhe significance of these tests and their verification eith contacts are again referred to lstcr. In every case metal was transferred from the negative to the positiVe electrode. Bonds of cepper and capper oxide formed and drOpped from the positive electrode from time to time during Operation. Table 2 gives the results ofinateriel trans- ference When metallic electrodes are used, tagether with the physical forms of the electrodes after an arc has been maintained. These tests on metallic arcs gave indication that thero mal conductivity and size of the electiodes may affect the form- ation of the crater. Eglsing through Resistances nsApond: From the tests on metallic arcs it was but a step to equip a number of pulsing relays with #18 gauge platinum contacts and than let these contacts intermittently pass a current throegh a non-inductive resistance of low value to hasten the contact de— terioration.- After impulsing through 8.133 for 15000 times the positive contacts develOped a long point and the negative con- -6- e ‘ I -~-— g . ' x .I i ‘ _’ J . W ~' ~ . ' e . ‘ \ A . . , M _, , .., ~ . . . , ' L . . ' V . ., . . . a ‘ , .. 1,". - ' 4Q ‘ . , h 4 I L . v . ,, I n. . . o n I I ‘ I - . fl .1 . “ t .,n' V . A . ‘ ‘ , ‘ . -\ . . . . . 1 . i i . I ‘ . A . . ‘ ' -Q ' - , l .‘ ' I I ‘ . I 1 ‘ ‘ o ‘ v - x ‘I I. ‘ . m ’ t , ~ ' \ tscts a very deep creter (Springs 15.10.18.19). These contacts ere almost identical with those obtained in practice which show a negative crater (Springs 25,26,3i,35). The substitution of larger back contacts, thet is file gauge platinum yielded s much less pronounced crater on the negative and a complete absence of black deposit cround both contacts. Spark Tuegch ing: The usual form of spark quencher consists of a condenser in series with s.non-indnctive resistance. This combination may be connected either across the contacts as shown in Pig. 8, or across the coil as shown in Fig. 9. On Opening the circuit, the condenser absorbs the energy of the magnetic field sheet the coil Which would menifest itself as e apart at the contacts‘withuit s condenser. The resistance in series nith the condenser limits the momentary rush of discharge current of the condenser at the moment of msi3.‘ ‘ In either case the condenser charges in suds a direction that the discharge across the contacts, ir’these be Opened and closed with but a smell intervening time interval, is in the same direction as the battery current. Prom Fig. 10, which gives the theoretical solutiOn of the discharge of‘e condenser, it is seen that the current at the moment of make may be varied over a con- sidereble range. Ihe ideal csse would be the one where the re- niatence in series with the condenser is very large at the moment of make, in order to let it discharge very slowly end at a low Enrrent value so as to prevent arcing 53' possibly welding at the contacts. 0n the :thcr tend the Esme condenser should have no resistance in series at the moment's: break in order that the cendenser may set as an energy absorber as efficiently as possible. -7- In practice, a resistance is uswally chosen so as to offer s estisiactory mean between these two extremes. From the curve, Fig. 10, it will be noted thst a ten ohn resistance will allow the discharge current to reach a maximum value of S amperes, but this current persists for only an infinitesimslly snail time; in fact after less than 4/10000 of a second this current has been reduced practically to zero. In the case of a 100 ohm resistance with a 0.5 mid. condenser the discharge current reaches a maximum value of 0.5 amperes end then reduces gradually to zero after 2/1000 of a second. Eben a 250 ohm resistance is used with a 2. 0 mid. condenser the discharge current reaches a veins of 0.2 amperes which reduces to zero very slowly in comparison to the previiuc cases as may be seen by reference to the figure. A ten ohm resistance, therefore, because of the shirt duratizn of the discharge current ought not to cause undue erosion of the contacts. I This point will be mentioned agein later where it will be shown thet the contact trouble creviously referred to may be eliminated without changing the value of resistance which unless noted other- wise remained at ten ohms during all succeeding tests. The action of the condenser in series with the resistance (Fig.6) stem placed across the contacts is somewhat different than when the combination is connected across the coil. In the first case the condenser, While in a discharged condition, acts as a shunt circuit to the contacts at the moment of break.1t that instant the resistance of the circuit is only that offered by the series resistance. As the condenser charges tie counts: electro motive force increases, and tr.st to all intents and purposes is the same as adding resistance. If the contacts are completel; -8- D..4 a. p s. I. y .lo. 0' . q . b .4 ." ! ’— “ $1.4 . .2 L . . . A - .. > . I. u \ a. .a. J 3. .1. . . r A . , . ' 1| 4 . _ . _ . . \ I» .I -. s . u w J o . separated before this apparent resistance of the GJUdCESGI reaches an appreciable value, the cgudenser prevents arcing at the centects. For a definite rate of break such as that of a relay, 8 minimum value cf capacitance exists wnich will offer such a low appereut resiuteuce during the separatiau of the contacts. A larger condenser may be used, but 14 that case the series resistance must be increased in order that the discharge current on make will be held within reasonable bounds. A 0.5 mid. cindeneci was used as a spark quencher in all subsequent tests, since this Value of CJMdGUBCr has apparently endugh cep— ecity to act as a sufficeutly lee resistance shunt across the contacts until these are completely separated to prevent arcing. The action of the condenser ucrose the coil, Pig. 9, may be analyzed in'e different meaner'eigce in this case the capac- itetive reecteuce BhJUId be equal and Opposite to the inductive reactence of the coil at that particular speed of current inter- rupticii. if the crudenecr breaks down; the better; is shirt cir- cuited deriug Operation and therefore this edcptetiou of the con- denser and series IGBlStLUCO has not fauud epplicetiJn and will nut be further causidered. - 3.91;, y Cheat tori 1.18;: From the oecixlogrwn 0-51 and eieiiur ones 0-5& and 0-53, it we: noted that the rainy ermeture chettered when de- energized, that is, the armetare did net cine to rest gradually on return, but caused the reley'cautaots to Open and close several time. An oscillogmm was made of the currents ee in the pre- vicue 3’51, but in addition a mirror was mounted on the residiel screw Jf the armeture of the relay. the results obtuiued L8 éiwmn on 0-55 indicute that the armature rebounds Lt least two times -9- between tLe armature step and the armature spritg, causing the current to be interrneted end realitiug in sgerking at the con- tects. The movement of the relay armature may also be checked visually by the aid of successive Sparks from an indentitn coil. files the sptrks occur separated by a time inteIan of slightl) more than the interval of a cycle of Operetien of the relay, the Operation of the relay appears to pass through its cycle of cp- erstinn at e srcatly reduced speed. Since the rebound of the armature occurs between the arm- sture etc; and the armature spring oscilloerens were made to de- tcrniue the effect of ircruesing :r diminishing this gap. In 0-80 canperieon is made between armature chattering with the arm- ehzre stop in tte n reel position, and wien entirely removed. There is no doubt that the srneture sth is net the only contribut— ing cease to the rebound. Subsequent investigetiane led to the be- lief that the residual magnetism of fine Isle; core acted as a re- straining force and rebound is established by mesns of this and the periodicity of the shrine. Apparently the clearance of the armature stop does fljt effect the IYBJLmfi. A check tent st en in 0-81 where the clearances on the s'mc relay were 0.003" 0.010" and entirely removed verifies tiese canclisions. Eliminati>n cf Relay Chetteriug: ‘i'rom the above tests it became apparent that cite Inc-ens for holding the ermeture ereinst the armature stop on releare would eliminate the chattering and the resulting sperkiug'et the con- tacts. Accordingly, teats were made with e "C" type spring ect— ing from a point under the rcsidnel screw and terminating under a screw on the heel piece. Jith the prover sdjuutment of the -10- n‘, ermntnre spring which new was inflaenced by the tension in the ”C" spring, the arcing Wan materially reduced at the contacts. An oscillOgrsm n-L2 shiws the effect in the vertical magnet cir- cuit with.and withoit the "C" or damper Spring. numerous denser springs were made as for example one wiich lies against the armature Spring es stews in Fig. 11, end othars 1"“I where the force is applied to the armature arm as shown in figs. .12 and 13. One of the latter where the Spring acts age net the erm.neur the bushing has been found to be very satisfiictory with regard to Bpsrk clininstign. The adjustment for any such camper spring in order to beeffective may vary between 10 and 2a grnms back pressure, measured on the armature Just back of the armature step and_Just as the srnstire leaves the stop. Shen two 18 gauge welded platinum contacts are Operated on the armature and beck springs of the pulsing relay of’s stsndard connector the result as mentisned above consists of’s deep crater on the negstive or back spring and a correSPJnding building up of material on the positive or armature Spring. A check test made at this point on a standard connector reelited in the verificatien of the above and is shawn by the syring cistects 31 and 32. These contacts Operatel for 1,515,140 impulses after which time the switch'becsme ineperetive due to the centscts sticking. The same connector on the identical impelsing circuit was then equipped with a similar pair of contacts, but the damper spring lest menti ned Wes added to prevent chattering of the re- lay. Spsrkilg at the contacts was prscticully eliminated and the switch Operated 765,520 impulses with the cantacts bright and clean .WItb a slight indicatiin of’n crater in the pesitive. The wcsr ou the contacts was insignificant (Springs 27 and 88; see also Fig. -11.... dé I“ “W w W W11“ new won seine mites auteur-u eqftaune wit u. “use: urge-Ma: new 33,1311; em ,gomqa “0' «81.6 {tene- ream «I: a he)” as: nude 38-0 MtgeIIIoao u ' f. a". 1! ' I': j . V ' ammo «qua w’ “o" as: “one” has am: aIno 5333! he Him” 161 e. 6686: 6:6! estrr'ee request escrow?! gtéc Ids II tam at arose as 13:1th exam ed: attests” e611 '6 T mi in male as m Cutest: ed: 93 houses pi mo) suit exert! «if may at». gauge 6:13, am 1963:in 9.913 19 0110 .CI ems SI 6‘: Kim 119303111!“ 2197 u e: we: Inner and guides on: seen In ' flw In»; an 1:61 :meu m .noiunMIIl an;- at huge: (. \ ‘ :1»? . fie . f M 68 met m m. if on meat I .~ . 9 .M. be: warm: III" no . .. u {Leger (£13 rein-am ‘; I Seed fxe 071333361! .1: anI .— Mx’i! 6%, -.. m uj I -,aawf12wv on: ..I mi? 3? one. Q 6 23:". "3 _ Mauro mullet: .‘ of: him: ”a: _ 6 was on; u m or be: mu I.» mm on. am so: mm nun» QQQQQQ -’ dim f ‘mam 3m; 6: you an mum- wan-I when I53 5” 3 1.1e titanium e6! Imam. 6d»: if 33ch an 1314:," . , . ME; 6 new“ 6m 4w- aq-ugu mega: human “#1" ' ‘ «rim 6‘! .1 issue a to nonlethal c3“. In tItIl! .611 ”In 008 538 his 7st nachos) mam-gnu 66w mu3¢66 6&3 «Lt- 14). Because of the appeerence of the indicati:n of a slight positive crater a check teet.wee made on the same connector, the cautects of which after 786,960 impulses yielded similar results (Springs 29 and 31)). These tests togeth r with similar once prove concluee ively thct the pitting of the negetive p etiuum welded contact and the building up on the corresyonding positive contact can be: eliminated by the use of tiie damper Spring resulting in a longer life for the contacts and freedom from contact tre'bles; Effect of Larger Contacts - Thermal Conductivitzg From the experiments on the thermal conductivity of the metal erce mentioned above and the fact that reduced Sparking resulted iron 8 non-chattering relay, it appeared that the 18 gauge platinum contacts were perhaps too small to carry the heat away quickly_between impulses or grodye of impulses. Springs were made with large contacts of platinum, gold and silver and assembled in the standard connector. The results of theee'teete ere given in Table 3. .Of the veriaue combitetione tried, 518 gauge welded platinum as a positive on the armature spring when mated with flu gauge riveted silver contact as a negative on the back spring formed a combination which after impulsing over three million timee showed a pair of contacts perfectly flat with but little wear; bright and entirely free from any black deposit around the contecte (Springs u} end.nn). These tests which were made withJut damper epringe indicate that even though a relay OhhttGIB and cuueee edditional area, this arcing does not cause the centecte to wear away, nor the pitting action to Occur provided the negative contact be made eufficieutly lerge. These results have been borne out hi -12- 3MI a 19 $13008!“ 0113 to oeusueoqqa 611.1 it: cueoef. .{431 e138 133mg? «be «I: he the: as! 3393 mat» a 'xenne winner; ' Q A _ .A' 40?. fine QB meshes) «am. "We! ”no ”trite (idle 13636303 b31393 cued? 35W? “3&3; {mutton 053 19 3513311: 0le 3313 vbvi ~. .. I. ’ 137331an authenticate» on“ no cw gamut! or“ has «ism: (ii mum-r mute tome 31:13 to an M13 ‘6 bonnie“. ’ “£6,013 {chime an: 331nm hue SMRQO ed: 101 at.“ .. "’ - ' P £36;ng - “games 391314: to gm i, 333' to vmsmo W333 053 m “Wm 31“» am at” WW teem-x M3 doe: um bee evade bemummn £1333. " ’ a "5%: iii 8‘3 56150qu 3!: .we: @31933‘31114-03. a m “31:.“ m gnu 93 fine e03 mung new 3W3 when. 93988 mm" " ma: to scam I ‘ \4 .. .fi- 6’} mm“ 0‘33“!“ has“? "1331‘ w .011 r w 3363313 3“ m “if e «in new: an "133'“: am a grab aw. we mom: . , 03‘ 333331“ I“ ICH£6 302! «Oh W gm: m ’ ' ":12“: W flit-hi “3’ “here We at 6023‘” aamrq' e83 f . 3’33 edit: ml and um” unite .«w. gamma. (ii-‘- practice over a number of years. Tests were made with a relay equipped with such cantacts and a damper Spring in additiin, and after Operation for several days with no naticeable Sparking, the contacts‘were examined and they showed absolutely_no deterioration nor burning at the points. All the above tests were made on the circuit as given in Fig. 7, and with a Spark quencher OOuSiStlug of a 0.5 mfd. condenser with 10m non-inductive resistance in series. It, therefore. appears that a pulsing relay when equipped with a #18 gauge platinum con- tact on the armature spring and a $14 gauge silver contact on the back spring and having in addition a prOperly tensioued damper Spring will perform with only occasional Sparking and with a. resulting long life. Check Tests. Results with various Contact fietals: ,AUeing the impuleing device Fig. ”15 and various contact metals and ommbisations on a standard connector Fig. 7 results were attained'which agree with the previJus results obtained in the general way of contact performance. The back contact was large and riveted in every case while the armature contact was a #18 gauge platinun'welded contact. 30 damper Spring was used. The contacts were pulsed between two and a half and three and a half million times in sash case. Tabalated data is given in Table 3. . TABLE 3. Syring 30. Contact Pulses Condition of Contact 39_fi_e Pt. 18 ga. w ‘ 2582500 1 Very slightly pitted (t0 - As. 1e ea. B 2582500 . ilat, Edges rise slight- y (41 + Pt.18 ga. 3 2636600 Flat :42 - Ag.ie ga. R 2636600 Flat -13- --.. 7‘- loo—o. vb \r‘ . ‘4 .- . . d . I I . , I ; . . ‘ A ‘v , v o . .-...— . -u o . J ‘ o - \ ' V y '! .15 L . A ‘ I . ,y.. C \ O . - ._ g“, A u; ‘ r- , o :o f .a o - . O . . l . . -- ‘ o n o! . “ J u . --.- . - ,. I >4 . I. - c' . .. (‘ . I.. \.' . I’ ' ‘ . .1. 5 1 0' n1 7‘ , , u .' J. v . - l n \ a, x, o 1 \ L k - .5; .. .— ' r ‘ J l '1 L- — , ~‘ai. . ..: ‘6 ,na .0. A, (43* Pt. 18 as. x. 3:283L0 Slet,Slightly irreg- ( ulsr E i - Ag. 14 gs. R . 39285n0 ilzi, Slightly irreg- (45+ Pt. 18 gs. J. 30000b0 Flat :46 - Ag. 1% ga. 3‘ 3000060 Flat ‘ u7*' Pt. 18 gs.'w 2896960 Flat, Black deposit (as - ' An.. in gs. 3 2896960 Flat, Black deposit (11.9 4' Pt. 18 gs. ‘3’! 2837800 Flat, Blue}: deposit E50 - * An; 14 go. 3 2687800 Flat, Black deposit The material lost on the positive springs 39 to #5 was slightly less than the smsunt lost with springs 47 and 49. Test on Self-Interrupting Eatery Lineswitches: metal arcs at contacts and the resu;ting formnti n of neg- ative craters, particularly thJse as sh wn in the previbus Fig. 6, may depend on still other contributing causes than these out- lined above. Concurrent with the above tests data.wee obtained on the performance of veriOns combinations of the three contact metals; both riveted and welded, when Operated in the lineswiteh circuit Fig. 16. Condenser charge and discharge oscillogrums for one switch are given in the osciLIOgrem U-lYn to G-177 for one, one and two, one to £011 and one to six switches Opercting simultane- ously of! the same current supply; in interaction is indicated. Just as a number of these switches may Operate simultaneously in practice so were a number of switches never less than six cpereted in obtaining the date listed in Table 4. It is very difficult to describe the ccnditi n of the contacts, and conclusion and com- parisons can only be made berciereuce to tee original samples. -1”- . ' . . . . r o .. . - .. I l ‘. . I). n . . . I V, a I . a . — x . I . L , n A .I. . .. ,3: i c _ .t. . . 4 , y . .r‘l: ..".~ I ‘ a ... .'> ' ,1 .y..l ‘ n . Vb . , . 1 r 3. . u. u. . , n A r. , in c A I ‘4 r \i ,V c I . tr . . $3». , Iv A ie . x «V . . . e r . II I . v 4 n I | A I I- A \ 0 . r. . ..I¢ \ J fliers-photographs of some of these test Specimens are attached. The relation of electrolytic ectiin, thermal cinductivity, etc. between the veriaus contact metals on each other when Operated in such a circuit can perhaps be better correlated after further tests which are now in pregress are completed. ' In a general way the results obtained in this self-inter- rupting circuit are similar to those obtained with the pulsing type relay. That is, a large contact on the negative and a smell- er platinum contact on the positive spring result in a combination yielding long life, freedom from excessive sparking and very small weer. It is to be noted, however, that such tests as the shove because of the greater cuzrent, increased rapidity of Operntiin and heavier duty cause the deterioration to proceed at s greater rate than in the pulsing relay where the Operation through several million pulses requires from three to four weeks continuous 0p- oration. Con c 1118 ion 3: Metal arcs may produce a crater in either positive or negative electrodes or'contncts. The intensity of the spark and the ensuing rise of'tempereture at the positive contact form a Small bend of miltcn metal which gradually is enlsrged into a long point causing a correSponding depression in the negative contact. Such a combination of contacts causes irregular action due to sticking of the cintects and a deviation from the normal adjustment of the s rings and sir gap. By the eliminatiin of relay chattering, the sparking is materially reduced, and in particular its effect since during -15- (o ‘1 ii. chatterihg the en rent is interrupted at a high rste which does not allay the contacts to cool between interruptions. Chettering in a relay may be prevented by the adaptation of a suitable damper spring. Tie use of a large negative contest further increases the life of the contact combinatiJn. The best combiiuti n on the PUESng relhy within the seeps of centect metals tested cuuSiStS of e 318 gs. platinum pifiltlve cautsct, either riveted or Welded on the armature Spring, Operating with s 31h gs. silver negative contact rivet- ed on the buck Spring. in edditi n the relay is sluipped with a suitably tensioned damper Spring. P. G. AXDEYS lesesrch Engineer 1:31.11 5. 1921‘- O ‘l .a !~ 73 60 62 71 Sheet 1. P-~PIYU¥ flS G“. 9333?) ‘u. :‘Y"‘1T1 ) {Null}. r. .EImv- * (luxtect - uleee cauditien C ”test 0311- ditien -—”¢-- -- * no“. fat .4 .- o notffiwJ Irregwlar CzLe 5166300 art "’ (3} me Lmaoth Pen: evr‘ )" 1°"! .) .a J {1‘ 5...: H Even Cone 5553933 fxooth svuvex xelblsfi 233,44 3.13 9 .3313-.>th $312.11. Cam 6 ': a8. 1]. vv - I! "W": 112‘ In ‘~‘f ““7 1—" ,1 .‘a . ."Jh‘ '? L, W .. ML.“ .J’.‘ ‘79-. q ‘ naI vvn - I o?."' .. ‘I- Contact EOLt feat beipg rep~ated Fmoath sklllow Crater Smell Even ureter Smooth Cl‘i‘t t- E‘ 1' 03;.hllijw 1r- rc 33L; tar slim-L tor z , q .-- Vidal]. "JVUD "1 any " -‘ bl h b er CJMtect See- 30. So. ulse Ce ditiau ultiou 72 66 78 60 71 S3 65 77 "t. It. 75 no.5. inuuth {hellow Cm tc-r Swallow Irreg- ular Crater mnell even Creter ijouéas Smooth Crater 3mm 3 Lung Smooth 1‘ 1‘0 1.) 3 3595175 3261000 Deep Smooth Crater Bmoeth Jjuvcx gmgoth swell Cone {35:0 0 th 5“. . e l 1 0.3110 amooth Lven Snell Came Very Heap Crater Long reinted Cone Hun-m - ' ‘ g- . v k , I ‘i I 1 . 1 J u ; . ‘ A ‘V‘ O . .. .. .o...-... .4. n '1”. .4..." .-.-,.,.. .._ , r—... f SILVII‘. crm‘r i-flil‘IL «$18 372L031) POSITIVE + -, - élmp- * Contact - Contact 30. 30. uleee Condition Condition 77 78 #18 Pt. 3861000 Long related Deep, Smootho Cone Crater 79 80 #18 A.E. 3667250 Deep grater wedge Shaped ”V" Snaped POlut 81 82 #18 H.E. 3757225 Deep Round Large Round Crater Core *56 55 .414 5.31. 2362125 1201311 Shallow Small irregular Crater Cone ‘Gn 63 #14 A.E. 669930 Small Rough Smell Irregular 0 Crater Gone ‘65 66 #18 Pt. 350317 5 Very 13661) Long, Smooth Crater Prong ‘69 70 $14 237.13. 3362525 Very Deep Lost Crater .IL :2 c STAG? azrlt HELDBD r16 03. HSGABIVE - o + 31mp- — Coutact + Contact 30. Ho. ulees Condition Condition 7a 73 §18 Pt. 3562950 Small Even Smell Even Crater Gone 86 65 #18 e.z. 3725475 Smooth4F1ot, Smooth Flat Edges Black Edges Blackened 30 79 #16 A03. 3667250 23'6ng .3116de Deep V Shaped Point Crater ‘53 5a #18 Pt. 5296275 Smooth Shallow Smooth Convex Crater - ‘66 67 512+ «.12. 3104375. Flat Rough, Flat Rough Edges Black Granular ‘55 56 $1# A.B. 2362125 Small Irreg- Rough Shallow ular Cone Crater ‘63 61;, $11; A.E. 669900 Smll Irreg- Small Laugh ular Cone Crater ’Indinntas in; lean Rilvm- Contact TEiVQtnd o— H .'.. I‘ -.“' 0 ‘-__ o.- o.- ’1' B L 13 7’3; Sheet 3 5.21.1) CMTTACT “HIM. 716‘ 7:3:an POSITIVE + - - #Imp- 0 Contact - Contact Con- HO. HO. “1808 Condition dithDM 83 817 5718 Pt. 3605175 Smooth Even Smooth Crater Small Cone 85 66 718‘ 7.7;. 37251775 Smooth - rlot Smooth - Flat Edge Bulckened Edge Blaczened 87 $8 }14 W.E. 3757225 Rough-fledged Rough V shaped Shaped Edges urater. Edges Blackened BlacLened ’58 57' 714 27.13. 727677550 Eeerly Plat Very Stella Slightly Con- Crater, nearly Flat vex ‘67 68 7127 7.3-. 34117375 Flat-.aough- Flat-30.311 Granular Edges bhickened GULD COHTILC: :53?fo 5:18 Gil. :nlfjuDED BBSAIIVB - 4- + #Imp- - Contact * Contact E0. I50. - ulece Cozditlon Condition 76 75 718 Pt. 3166200 + Came off. to. trying again 82 81 #18 A.E. 3757225 Large Round Deep Round Crater . . Gone ‘51 52 #16 Pt. tot-37500 Last Irregular Conn ‘57 58 714 7.7:. 71766850 Shallov emtor Slightly ooh-“s nearly Plat vex, :eerly Elet ‘68 87 7716 7.75:. 3757225 Rough W" Rough 77’0ng shaped Crater shaped, Edge Black edges Blackened ‘70 69 51117. 1.8. 3362525 Lost Very deep crater I'Indicetee flu Gauge Gold Contact Riveted ~ ~ . r‘ - , ' ; ~ . ‘ l . a. « ‘ 7“ G c r a ‘ 'r" i Q . - . . f u . . . v .. \ O n K .. . ‘I I . 1 h _. ' L. -' ‘- ‘7 ‘4‘ a - ._ .., . - - - . . . ... . y . M I l. 7 .7. - . .. a. . '( . l" ' I 7 .. ' 7 . s M .~ .1-.. - p... A . . ,. . . . . . l I! ~‘ ”.1 t .- .. v Q \ . .3 .0‘1. I J \ . - ., . .1. ‘ -a-. .- .I“- .o-.- . -' v . . J . .1 *w . ,7 I \ -- C‘— - 4 \. - . 7 _ t .. , v . . . I" _. .0 f I u' . 3- m . _ u .- ,‘ ‘- _ .\ 1 u ‘\ . . a. . . 1 . ‘4-——— 0.04.1- — «J (—1—--- - _ 1‘4 0.0655— 2— —~H | l i ’ ' l I — — 7—— —~ I I 0-020" 0-065" | o-oas" _7_ "___L_ BLHNK FOR RH/ETED CONTHCTS. __ - ______%_ l | l l | *1 I I 5 .R BLHNK FOR WELDED CONTHCTS. FIG.1 NO-iB B43 GHUGE CONTHCTS ‘ 1 —_l CONTHLTS HND CONTRCT ENDURRNCE. 51.37537 *1 4A -———‘o-064-"— +7 BLHNK FOR RIVETED CONTHCTS. .\ comncr RNETED m seems \ \ SPRUIG. “ RIVETED. FIG.Z NO‘!4 8.175 GHUGE CONTRCTS- SH b—k“ comncrs 9ND CONTHCT ENDURHNQE. EL5‘53'5_ .‘ni' B: 0-5 fiMPs. v=0TO 0-0025 VOLTS RES.=Ioo“= (an 7.28;. .5? . V. Q '7” + --. Q . C5 _ F @3 (6;. E? .—. If e 'l'III+——' ~ " _' :W" _ H 1' 377:??? 2. (IR) Fl G . 3 D , e = !R - "5‘?” ‘f '1 ’WW” "TS HND CQMTHCT Emeuemgce w .-“.. —.o l .- . . n o--o. .__,___ -1- I l -'1 0-Obl“ l ' ‘2’ com: *4 I l I . __ 5‘ ‘ 0.0m" -‘_l_ , . 1". . . n , 0076—." —"‘ |F~‘ 0-078'~— «.24 . l 1 ”I I 7 'EPPROX. SIZE or NEGHTIVE CRHTER 4— Pesnnve POINT. 7‘ |'. \5 >3 )2 3 Si . 'l : .. F108 ' SPRRK QUENCHER Hostess certs, 1‘ . H99 212 . fl . ’ \ V , - — - @7777; c T e n N o c ON 1 5 c7 ‘2an‘ouaeuo‘27 iicNGmG . . f—tNTeezum-sn. t. _ Confad’f'a 4M (£2;fo Endurtnce. . Tr o- ‘ F C 0 n n ecf’or 6,439.: nichlagmi'TSSlfi‘W 5' m P, 'f '3‘ ” , 3 __-- $7.».me W m! l w .71---- -47. - - 3171...} _ . . “\QVKNVEW -xn fix fie DNQSDQM . N . . . ,, . . . _., . _ _ h . . . . ._ h _ . . . -- - H _ _ ._ q 7 . 1-07-7-LfiI-IIQ- If . L . . 14 ------1ldozvumw-_fi.ziufifi1 h -.-.-L-.----,- 1 _ .. W +3 . no _ H _ . ._. _ l P 1‘ L4. 0 H I” j . . d h- . .U 7 .. < . . . . _- 1e, _ . . . . .1 i 1.1%.} + l _ 7.7- - . l. . . ~ .9 . .. . i _t. U...w.s ._ rl t‘LWII 5 7' ll 1 4T. 10¢ rl twill; it-rl.4..1t|lllllll9§.lvlrt . . A Mi] 1. _- h . . 4-H- U- * . _ . u“ . . . M L . .. 2 ‘ tr 7 . o l O IIO 7.01 .1“! TI: .l.|o 0'09. llrnft‘lj‘ifiv‘ lllTl't‘lt-V 4‘1 ._ l . 1 . N M. H . . . e. LI F1 2 v . V .7 W L . . _ . . . _I. - . . _ h . w . M {Quay _l. a up 2 . _ . . . .an . _. 1 . o o . l . I .. 0.! «I .. - v 3.0 1 OllvdL fto o it .7.-o!q+ n 70L l Olul I'll. 7. . WI 1 . . _ _ _ . do 2% -. . :. tr > H 2.. d 2’ 'Q .2 ' . l C. ’A 1| 4.4 4 4 H 4 <1 h H . . . .. x ... . .. :7. A- . o ‘ n o : fi.. ,AZ 0 W...’ . . ... - w. . otl‘ o|¢1¢061 at t IM.5 170 Ilt. v. .1 1-0%... . 4 0 filo. vol 0" I?! J19-017|91.V¢o . _ . .. _ . e 1 t. . . . . . -. O . . . . L # ..~'.. . 4‘ - . l 41 _ . n1 a . . - ‘ n w W . . n 0 ~ .. . ‘0 M...-o.ofi.tn-._..Iu.o~ Yll-tol‘vl . o t .\ . - I. .0 101 «1'7! ‘00-“; c.-.w wofl 'OJfiI 140 u l _ 7 h w - h h .2 . a n... 47: .t 1 . . n . _ . _ _ . ... ... . . . . . . ..»o . . o . . . l - . I a . vll YVOV‘ATQ tbbl‘v .O“'lvo|. c ol 'l H 0 nl .1“ 1 O l 04!- I'EK . . ‘JT '1 . . 7H . a fi . .. . . . o _ I» . “f _ w . . w 4 fl .. _ 1— _ a _ w . a. . h c . . . a ........ A .. . . . . . n . . . . 707-0._v 7-7-6.79Ll ttr {viii-1.7.: .9 I o I In L0 . -1 I . Y ,io $3109.01 11: «I...JY¢‘I . . . . . “ . o n m . fi . N M.v v“. . . . A . . . . . r . . . ll. . u- 9 *T‘t' IfitO‘QJ llv nvll‘ 1i’T!O-.I|llu 4 4 . " m l l V . V v . _ . _ v. -. .. o . . . It -o. o Ilozflvslt {0.4.1Il' 1 . . _ _. _ . _ . . ” . .. . . . . . . — W .7 v . . 7. H 4 _1l . . I“ .. 7 . . _ m . . . l . , . . __ .—_- _.d-._.._ N -_ .4 _.- - 7 r ‘- o- W I « l. 7. 6: -| ‘6. 1‘. will: V - '...l.-..':g - - ‘ - 'I ;r , - __,f 9'. H F -9,- | ' '2 0-048" - "a-o'la' 2 2- , , ~--*—-*o 0'15 O J I H’HRRDX 3125 .OF NEG. 9 P05. CONTHCTS HFTER USE WITH D."'U”'IPER.S‘F’F2?W_S3'i F\‘G.l4. In 0 301972770» 0F DHMP'ER sPRw-G TO Renew: .~ FlGJJ .13 @Rw .N-G . /’/ "fie SPRING HGHINST DRMPER SRRWC- RG’WWS? b so .0 59 z o RRM e r D'RE 5.89M,- w. .2 -_._ - s IG. H Fl; G. IZ. .tm..u...k uni...“ h..u..m.#.z.@~wx .. ..o..m. 9. a2: $.22. .8. H - . .. . . .. . h .b _J J . C .._TL - “- . . . . - ,Wcr .-.. O - . . . . :1. H . 2 r T—Tm . . . . _ . .27..“ .... . T .w. L ._ .2. aka 02¢ .wusz 0.... 2270777772- If: T I $2 ij T 0 b C . 1 DH. _ . - .. w H c _ J 9 null 7 N ~ 0 EIL a.__I__o H. .. lTL. .(v... .fiqUhL.<3mfi$k Q§V Mzwwkht .WQQNQ .1 turn Qg‘h I .Whflfikk UN Nu QR“ N2 .Q Run. Hal .kkaV 63* watwdsw .bmixxnwnxha kVfik §§M stws “K buskxsb 11????Pfifififififififififififififififi 131...? 1 111.11.111.11111113111111111111111 11111111111111111111111:11 11111 .11 1111111 1“ N 1‘40! . . O. __ 1.1.11.1 1.11.1 11.11.11. _ ..1_. .... .... .. .. 1.1.11.3... ......m. 1.1.1.1... .4!“~u.| I‘I...O t C [5. f7. #157121. IJ'OQO I'm/Se 6:5. Foo. C/é. Pf. #49 74¢. Araoo I'm/bakes. ”:7. (:13, Fl. #95741 Ifoao im/su/ses. 1’05. 512/: M .6441. ‘ 6 l7. IDI‘. #73741. lioao lawful/505. 4/67. {- - (73 /3‘. Iiy‘c. 45240 I'm/4 a/se‘s. P015. 1a 5 finft/es, (In /_ f5: . /yr. .fenw’re . 6215'. PI. #45741. ”’67. Hana/u/u, flak/412', 19.7? ('26. ff #7574. m. (:27. fl/fifax. 76.6720 1'07/1/1“. as. Dem/er #r/;7_ (28, F7! #/€7¢_ 7655—20 ilk/babes, 097. “07" gr 626/70? 1 —-.¢".‘. 627. I")? #4974. . 75! 76 a z inky/sea (:30, Pf 4/3 w 751750 I'In/ba/J (5. 473,. “Mfivf J'firl'fif . ‘fiji-‘E- (3/. H #/£7¢ . /J‘/J740 I'M/(l 436‘. (‘32, B? ‘fiyt. 1.575744 I'mfiu/ses. a .u Icy . C344 Pt “5/574. (.35. Ff. #46741 ‘3’ fyeerg, h’afla/o/o #awa If IX 7. A/ey. (A I J" Jyeors, HoaaVu/ul Mafia/I; IV. 7. figs. .3 a" . {P 1a.. ... 1.1. 5. 5. . /“.. M. fl 11/ .1 .1 _ 0 m a .. m 1” h 1, M w m1 . 11.1 3. fl 4 . c 141 [ ‘,1-I.-... . \n 1 E“‘jl . -‘r‘. v “--fi-mfi‘. b ‘J‘W“<» " "g - f’v“‘a :J' ...-T ‘9‘- ”a 1. ‘Wv‘fl.-_’ Fruit»- C 47 .1 Pi". #4974. 5537800 {pr/u/ges, lpos. (5'0. flu. "V4ya. 2987500 [fife/K595. My. 1?:3. 47. #M-ya. 52762 7.5" I'In/ry/xd. F? "75’74. {17527.5- z'rnfia/st . - ..Pnil...’ I’l- F615 47. "/4741. 3508/75 I'm/w/ses. loos. 5’66. PF. 9017‘. area/75' xm/w/ses. ”by. nl.)l|1l.l1ll|ll l I. Ell-I... u“1"‘;. 113‘; 4427. 1315.08/75 I'mfia/Sed. H’.’ #787a1. R64. 1X. ‘ ' ‘1 U32 Q33 F1005 cut-492% v ~9i$ ..- 1 '1“. 1 t 1'. *9. v w.’__'... ' v _ o 1 I it! fin. .flllnkl. .ltflrl. 1.1.141!) I. . V. is.) .37! .4ulr. Iqu cliff HIHIIVV-IM-Uda. 4a.. a. z . fi 1 . w. ‘I I'll. lilw . . § v..|. ‘ HWJJK. 'll‘ A: . film". ....._.'1.':‘_‘_-_L_1 3! . 23“. .... a ‘1“, TATE ""iflifimgflnjmm t\mufimufgfixfiifixj‘uiflijmfif L