Reference / Paper · 1912
Design and Performance of Magnets in the Automatic Telephone
Thesis on the electromagnets of Strowger automatic switches.
Reference / Paper · 1912
Thesis on the electromagnets of Strowger automatic switches.
HAYS & THVEDT ?3esign and Performance of Magnets in the Automatic Telephone 9 .:. 2 trNtvrov I CJCitNOIS xftB rably ^^^^^^ THE UNIVERSITY 3#K OF ILLINOIS LIBRARY i I'M % HS3 m 1 Kill .<fe?:?m Digitized by the Internet Archive in 2013 http://archive.org/details/designperformancOOhays DESIGN AND PERFORMANCE OF MAGNETS IN THE AUTOMATIC TELEPHONE BY DON LLEWELYN HAYS AND CHRISTIAN BERNHARD THVEDT THESIS FOR THE DEGREE OF BACHELOR OF SCIENCE IN ELECTRICAL ENGINEERING COLLEGE OF ENGINEERING UNIVERSITY OF ILLINOIS 1912 UNIVERSITY OF ILLINOIS May 28, i9(jL2 THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY DOTC LITOTLfH HAYS ENTITLED AIT 7) SHRISTIAH B2R3SEARB... IHTEBT.. DESIGN ANT) PERFORMANO B OF MAGNETS ._IB THE. JSMHAIIfl HBUS2HDBS IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF^ BACHELOR OP SCIENCE I ff 7h BL SCTN ICAL EEC IMBUING 7 Instructor in Charge APPROVED: HEAD OF DEPARTMENT OF ELECTRICAL ENGINEERING. UIUC The Automat io Telephone Switch.. TABLE F CON T E N T 3 THE DESIGN AND PERFORMANCE OF ELECTROMAGNETS IN THE AUTOMATIC TELEPHONE Frontispaeoe Obleot Introduction — pie Automatio Telephone Switch. Page Page 1 1 PART I Requirements of Design of Automatio Tel, Magnets Page 2 Action of Eleotro- magnet s Page 2 , Features of Eleotro-magnets to be considered Fage} ,4 Core and Armature Page 5,6 (a) Reluctance of Magnetic Path Page 4 (1) Shape of Armature Page 5 b) Shape of core Page 5,6 flo) Inf luenoe of Saturation Page 6. (d) Design with respeot to magnet io Leakage Page 6 Sparking, Its effect and Elimination .-Page 6,7 (a) Use of ondenser for Removing Sparking Page 6. c (b) Differential winding for removing Spark. Page 7, Time of uake and Break Page 7,8. . , . Diagram of Testing Devioe and electrical Connections P. PART II Character of Tests „ -PagelO Description of Testing Apparatus-—,— —PagelO, 11. Specifications for Standard magnet .mature Page 11. Description of Tests Made Pp. 12, 13,14 (a) Test No.l Lifting Power of Vertioal Magnets Fo r variable Current Page 12 (b) Test No. 2 Lifting Power of Magnets for Constant current .but variable Air GapPage 12,13 (o) Determination of Time of Make and Break by means of the Chronograph Pagel} ,14 (d) Measurement of mduotanoe of Magnets .-Page 14. . . — . , Tabulated Results and Graphs (a) Data obtained from Test No ,3 Curves plotted from Same (b) Data obtained from Test Ko.2 .Curve plotted from same Chronograph Reoord and Explanation of same Summary - ™ Suggestions— — „ , Pp. 15,16,1715 Page 15 Page 16 Page 17 Page lg Page 19 Page 20. Page 21. , (1) THE DESIGN AND PERFORMANCE OF MAGNETS IN THE AUTOMATIC TELEPHONE. OBJECT: To study and test magnets used in the automatio tele- phone, and to investigate means of probable improvement in the same. INTRODUCTION. In perhaps no eleotrioal devioe has the eleotroraagnet become so important or achieved suoh an extensive application as in the modern automatio telephone. While the development f the auto- matio telephone switch operated by electromagnets has extended a period of some twenty ov-;r r twenty-five years, yet comparatively little effort has been directed toward ascertaining the proper design of magnets most effeotive and efficient in telephone operation. The most up-to-date automatio telephone switoh now in use has six sets of magnets each set form in the operation f f whioh has a oertain function to per- the switoh. The prinoipal feature, however in such operation is the attractive force exerted n a movable arma- ture by the magnet. Hence, in order to asoertain the most effective design f magnets for suoh use described, this test and study is mainly concerned with this attractive force and its effect in producing desired results. (2) PART I. Design f Magnets in the Autoiaatio Telephone Switch. In general there are two classes of magnets used in the opera- tion of the automatic telephone awitoh; rapid aoting and alow acting magneta. of these perhapa the rapid acting magneta perform the most important function, and henoe will reoeive most attention. The requirement a f r the proper Working of auoh magnets are varied and may be stated under the following general heads _ t (1) Rapidity f aotion in response to a completed oirouit. Ability to exert a high degree of attractive foroe (2) for a given current in a brief period of time. Ability to quickly release upon breaking (3) f the cir- cuit. (4) Such design f magnets as to reduce the sparking at oontaot points in the circuit so far as possible. For rapid aotion it beoomes at once apparent that the magnet in use should possess but small residual magnetism and henoe should preferably be made should be that f pure Swedish iron well annealed. Its form f a short bar in order that the ends r poles f the magnet may exert a demagnetizing force upon the mass of the in- terior and thus quickly rid the magnet f residual magnetism. Where slow action is required, however, the core should be long and thus allow the residual magnetism present to be unef fected by the neutralizing action of the ends Sluggishness f action, as is f s the magnet. metime desired, may be ais increased by placing a copper or brass 3leeve over the core. Rapidity great extent f action, n the number n the other hand, is also dependent f turns and the nature t f o re used. a <3) An increase in the number f turns increases the induotanoe pro- duced as the ourrent builds up or dies down, and this in turn increases the time of energizing and also the time f demagnetizing whioh is enhanced by the use of a solid oore. It is olear that hihh induotanoe effect win be inoreased and henoe should be avoided in the design of rapidly acting magnets. In the design f magnets, it is well to keep in mind the follow- instil) Shape of magnet, (2) Quality f lr n (3) Reluctance used in c re. of Armature operated by the magnet. (^)Magnetio Leakage. In order that a minimum number f lines of force be lost due to leakage, it would seem f advantage to shape the magnet some- what according to the f assumed by the lines This form is that rra f force themselves. f a oirole(See Fig.na&b) Sharp turns and edges 3houid therefore be avoided. Figure la \ Figure lb. Sinoe it i3 of advantage to reduce the reluctance due to the air gap as far as possible, the shape of core and armature is of importance. The reluctance of air being taken as unity, we have R - (1) A U where R =Reiuctanoe of Magnetic Path. 1 = Length of Magnetic path. A s Sectional Area f Path. u ^Permeability f substance comprising path. Prom equation sectional area W A'» (l) it may be seen that an increase in or as reduces reluctance "R". oounter_3unk and the par J ; Thus if the f the armature in c re be line with the core be tapered as shown in Figure 2, it is clear that f r a given mag- net an increased attractive f rae will be produce^ due creased 3eotionai area. t the in- (5) Figure 2 The quality of iron is of importance as shown by permeablity "p" , the lines of foroe being greater the better the path. This Is evidently true for the armature as well as the magnet, Sinoe magnetic leakage tends to decrease the strength of a magnet, it is to be reckoned with in the design form f the magnet be f magnets. If the U shaped, by putting the legs f the U clos- er together the magnetic leakage will be increased s&nce the path thru the air is shortened. But, considering the path in the arma- ture, by increasing the air gap between the two U's al3o increases the percentage of leakage, for the reluctance of the path must neoessarily increase with the length some limiting position f the legs f path, tjence there must be f the will allow the least magnetic leakage. U shaped magnet which This position will of course depend upon the nature of the armature and core used, and can readily be determined experimentally for any given magnet by using a U shaped ne, the legs of which may bs moved any desired position relative t each other. This position, however, w uld not be fixed f r any given value bered that the saturation curve t f excitation, f it must be remem- ir n, which is used in such arma. tures as well as in o res, is n Q t a straight line, and that the re. (6) luotanoe f the tion. Thu3 useful path increases with the increase of excita- with low degree path being small ) the legs f saturation (reluotanoe of useful f the magnet should be plaoed farther apart than with a high degree of saturation, Henoe, for Oompaotne* and effectiveness, a comparatively high exoitati n w uld be preferred. in the design f magnets it is further important to fas as possible the path f the leakage flux, looate as m this way the forms, the magnet may be adjusted t ,give the pr per path the lines f force which otherwise w uld have been lost by leakage. A study f the distribution of the leakage flux may be made by means small magnetic needle suspended by a silk thread. by the needle w uld give the direction f the t f a. The direction assumed leakage flux. In the automatic telephone where innumerable contacts are used for make and break, sparking becomes a factor of considerable import. Vicious sparking is largely due to inductance. If an eleotr -magnet is connected in a circuit, inductance tends, as the circuit is breaking, to increase the E.M.F. in full thus causing current to be main- tained across the break. nature f inductance. This may be explained by a study f the It occurs upon a variation in exoitation,a voltage being set up due to this change in flux out ting the turns of wire. Sparking may be reduced in several ways, condenser action, it may be d ne t by the differential method, methods not disoussed here. r by various other While condensers may be used t or eliminate sparking, it is n t reduce praotioal economically as suoh a high oapaoity would be required to supply an entire telephone change. by ex- (7) The differential method consists in winding a o il in parallel with the winding f the eleotro-raagnet thus forming tw distinot windings insulated fr m eadh other and lying side by side in every turn f the winding. The tw terminals f the auxiliary winding are then connected together thus short circuiting the winding upon If n w a current be suddenly passed thru the exciting itself. Ooil,a momentary opposing current will also flow in the auxiliary ooil. The induotanoe ralize that f f the auxiliary coil will thus tend to neut- the other and hence eliminate sparking. The aux- iliary coil need not have as many turns as the exciting ooil of the magnet, but would nevertheless reduoe sparking considerably. The extra expense involved w uld not be very great, and it might be of value to adopt s me such method in the installation f a commercial telephone exchange. In the automatic telephone switch the percentage of make and break is about 50, that is, by dosing the electrical circuit and putting the magnets int operation, the time f attraction armature by the magnets and the time f release f the same bout the same. break is f importance The ratio of make t the economical and effeotive w rking f the switch. the f is If the a. in time during which the oircuit is made is only, say 10% and that for which it is broken is ninety percent, then it is found that olutohes lifting the shaft fore prevent of make t do n t the have time to operate and there- the switch from doing its duty. If n w the percent break is made ninety percent to ten peroent f break , $hen,too, it is found that the switch will n t w rk. T'nis, to a great extent, is due t the magnets, to the residual magnetism provided that the operation f the o re f f the switch is satis- factory, however, it would seem preferable and more eoonomioal to adjust for a 3horter period of make than of break in order eoonomize in the use of power, to ^he limiting time of make w uld in in any oase depend upon a number of faotors, and for any given set of magnet 3 would have to be found experimentally. (9) Figure 3. Eleotrioal Connections (GreerO B L" thumb sorews for adjusting armature to any desired height f r various sizes f magnets. «M« Balance Soales with whioh t weigh pan and oontents 11 1 ra 5° firmly to table or benoh. i . ? ?^ §i!SS wnioh »T» Sorew 1 threads with nuts "IT" t adiust length J c of air gap "2". (10) PART II. Teata on Electro-Magnets in Automatic Telephone SwitoU Aa haa been mentioned before, the moat up-to-date automatio telephone awitoh oontaina aix aeparate aeta of magnate. In an effort to determine the varioua aotiona characterizing the performance of the eleotro-magneta in the automatio telephone awitoh, teata were made, firat, to learn aa aoourately aa poaaible the lifting power of a aet of magna ta for varioua vaiuee of current paaaing thru the coil oonduotora; and aeoond, to aaoertain the power exerted by the magnet3 for varioua lengtha of air gap. In order to determine effectively the lifting power of magneta under varying oonditiona of voltage, current, and length of air gap. a aui table device had to be oonatruoted by meana of which the power of the magneta could be aoourately noted for 8uoh oonditiona, This device waa oonatruoted aa ahown in Pig. 3. AA represent a aet of eleotro-magneta held to frame B and adjustable along OG by the thumb aorewa S. The iron armature E moves in the direction indioated by the dotted arrow under the influence of the magnetized coil a. in the automatio awitoh, thia armature lifts the a el eo tor or connector ahaf t at point p. Inatead of using the shaft in t eating the lift- ing power of the magneta, a pan Q was auapended from the armature at point P into which weights oould be plaoed. Supports RR were neoess- ary to auatain the at rain upon B when the pan Q waa loaded. Air gap "z" waa determined by measuring D miorometer calipers, and lengths and and T> 2 by three in oh by finely oalipered soaies. Air gap "z« was adjusted to any deaired value by means of the thumb aorew T which engaged nut N in the lower part of the frame. Aa outlined in Figure 3 and aa just deaoribed, thia devioe per. f (11) mitted the measurement of the lifting power of the magnets by the measurement of the torque exerted upon the lever Ur (Figs. 3 and |) By placing weights upon the pan of suoh amount as to enable the mag- nets to barely lift the armature with its load, and then weighing the total (armature and pan together with the weights upon the pan) accurately upon the scales, the lift of the magnets under various conditions was readily determined* Figure k shows three views of a standard vertical magnet armature. Suoh armatures are made of No. 9 (Brown and Sharpe)Norway iron annealed, having a thickness of .Ilk- inoh. The armature, as is shown to a certain extent in Figure 3, moves about a round steel pin as a fuiorum. This pin fits closely into drilled holes wings of the armature as shown in Figure 4b. Figure 4. in (12) First among the testa made was that to determine the lifting power of the magnets for various values of ourrent. The results of this test are shown in tabulated form on and graphically in the page form of a ourve on page 16 (See Graph No.l) Under usual conditions of operation of automatio telephone switoh magnets, the ourrent is limited by the voltage Employed. This is neooessary owing to the magnets being wound with suoh size wire as to oonveniently oarry a oertain maximum ourrent for a time with- out overheating and possibly destroying the magnets. Tn obtaining data on page l5the magnet ooils were found to heat very rapidly when the higher values of ourrent were used. Graph No.l shows that more ourrent is required proportionally to the weight lifted for smaller than for larger weights. his is to be expeoted T when it is remembered that the inertia of the weights, pan and armature must be overcome. A second cause for a proportionally greater ourrent to lift the smaller weights upon the pan is that the friction of the armature upon the steel pin is greater at first. After overcoming the inertia and friction, muoh less ourrent compared to the weight lifted by the magnets is required for additional weight. Su£-j eoting the magnets to more usual and normal conditions, a test of the lifting power with variable air gap and constant currant at all times was made. T his value of current was that normally flowing thru the magnets when in operation on an automatic telephone system. Graph No. 2, plotted from data on page 17 shows the results obtained from the test. From the ourve, it is seen that the lifting power of the magnet is not decreasing in a direct proportion to the increase of air gap as might be expected. The magnets used seem to have reached their useful limit of attraction with the normal current flowing for a gap about .093 inoh. in length. That is, supposing that an armature of zero weight were aoted upon by the magnets under the given conditions, the armature would not be moved by the magnets until the length of air gap was out down to .093 inoh To inorease the ourrent, as was done in the previous test(graph Nol) and, as has been previously stated, is not permissible for the safety of the magnet ooils. «phis may be better realized when it is remember, ed that an ordinary telephone exchange contains thousands of such magnets, and that a ourrent of high value for only a few moments might not only mean the loss of many magnets but the tying up of the system. To determine the time of make and break referred to on page 7 as well as the ratio of the two, a ohronograph was used in oonneotion with the vertioal set of magnets previously described in the opera- tion of the shaft(See Figure 3). The data obtained concerning the time of make and break taken with the switch properly adjusted. was Referring to the ohrono- graph record (Figure 6 ) the distanoe between breaks in the blue line 3 represent a period of time of one seoond. The small waves on the red lines, on the other hand, are those for make and break the switoh. of As nearly as can be determined from the waves, it seems that the periods of make and break are equal. The actual dur- ation of any such period may be found by noting the distanoe between the breaks in the blue lines by means of a scale or compass. The same measurement is made in the breaks in the red line. The ratio between the two will give the period of make and break in terms of a fraotion of a seoond. Referring to Figure length of the three 6 and considering the red line W A", the waves marked is found to be 0.55 inoh. The (14) length between breaks on the blue line is found to be 2.02 inches, and the ratio between the two is 0.272 thus showing the time re- presented by a single wave to be one third of this or one-eleventh of a seoond. The induotanoe of a set of eleotro-magnot s as shown by diagram (See Figure 5) wa9 measured aooording to Anderson's modif ioationi of Maxwell's method of measuring induotanoe. An alternating current galvanometer was used in connection with this measurement. duotanoe of the coils was found to*. 375 henry. Standard coil. The in- 6 (15) DATA FOR GRAPH No. 1. I (ampa) wt.(gr) Lever Arm (inoh) Fig. 3 "HD" TS.o 9.00 9.10 9.30 8.90 9.00 9.06 0.170 0.170 0.155 0.150 0.150 0.159 oT5o 13.50 13.90 13.90 13.77 0.220 0.220 0.220 0.220 1.00 6l,0 60.0 60.5 0,220 0.270 0.2S0 0.277 1.50 2,75 M n — 60. 50 17.90 17.00 17.90 17.60 57.9 60.0 23.70 23.00 0.370 2,00 2.75 %370 n n O.360 0.367 n It 11 •• Average 72.5 70.0 69.9 69.5 69.5^ 64.0 62.9 5^.0 Average 60. 30 Average 60, Average — 59.57 23. So 23.50 (See Diagram No. 3) N till II n II 2.75 n 11 11 U. OF I. 6. S. FORM 3 (17) DATA FOR GRAPH No. 2, D Dp 1 , , Bj. 2.672 o.75o 2.72S 2.690 " 2.S5f6 2.913 2.634 Gram 3 Volts 2.750 0.0350 0.0214 « 0.0500 • 0.0720 « 0.0776 l4g2.2 1531.0 1331.0 59S.2 ttg.o 47.0 (Inoh) "HD" 2, §46 2.651 2.546 2.600 "z" (Inoh) "HJ n (inoh) "XM" Wt. Volt. Aoro ss Mags. 4 (Inoh) "XY" Length of Air Gap in * (Inoh) Cur, I in Amps. o.7« ^,7^ 0.7* 0.7s 0.76 it N 11 Hi (Ibr explanation of above nomenoiature see Diagram Pig. 3 Note: Letters marked thus "x" refer to Figure 3 ) (19) —CHRONOGRAPH RECORD-- The aooompaning chronograph reoord shows the ratio of time of make and break to the time of one seoond. The drum on whioh the record was taken re. volved in the direction indicated by the arrow. Length "AB W includes three wave3 (on red line) each of which represents a "make" and a "Break" In an eleotrio circuit, "CD" is 2.02 inches in length and, at the rate of travel of the drum, in - dioates one seoond. "AB", there- fore, whioh is 0.55 in. in length is 0.55 — 2.02 = 0.272 sec. For a single wave, therefore,the time of make and break is one third of this period, or O.272 _ 3.0 s O.0907 sec. That is, the time of make and break of the oirouit containing a oertain set of magnets is 0.0907 seoond. a SUMMARY It would seem that a more extensive test of magnets under a greater variety of conditions should be made in order to determine the best design of magnet for any given oonditions. Determinations should be made as to proportion of diameter and length of magnet, and effeot of dimensions upon the induotanoe of the same. As mentioned on Page k 9 it would seem advantageous to construct magnets and armatures of suoh design as shown in Figure 2 and, by experiments upon the same, to note the eff eotiveness of suoh magnets compared to those now in use, Thruout the tests performed upon the electro-magnets, the aim has been to determine the performance of magnets under as wide range of oonditions as possible. From Graph no, 1 it is found that a minimum current to be used in standard magnets tested should not be les3 than 0.15 ampere in order to obtain the greatest efficiency. For a oonstant ourrent of about .78 ampere passing thru the magnet 8, it was found that (See Graph No. 2 Page \t) beyond 0.05 Inch length of air gap, the weight lifted by the magnets was inversely proportional to the length of air gap. Below 0,05 inch air gap, very small increase in attractive effort is obtained. More extensive experiments might be made to investigate the operation of the switch for various ratios of make to break. (21) SUGGESTIONS The following suggestions, then, as to possible improvements on the eleotro-magnet have ooourred to the writers. A design of armature and oore similar to First: that shown on Page 5 would seem better than that in use at the present time. The same power of magnetic attraotion would be available for less current, Seoondt A further reduction in the length of magnet would be of advantage. Third: ,phe use of an auxiliary short oirouited coil as previously explained would be of advantage in the reduction of sparking* | Besides these points mentioned, indications have been made as to various experiments that ought to be made in order to deter- mine the best design of magnets for use in the automatic telephone* "'•V * Us? • * _ _> -*••« • • * V: Si" 1