Ij

ECTRICAL

E

NGINEERING TEXTS

PRINCIPLES
t

^S

OF

DIRECT CURRENT MACHINES

BY

ALEXANDER

S.

LANGSDORF, M. M. E.

PROFESSOR OF ELECTRICAL ENGINEERING AND DEAN OF THE SCHOOLS OF ENGINEERING AND ARCHITECTURE, WASHINGTON UNIVERSITY; FELLOW, AMERICAN INSTITUTE OF ELECTRICAL ENGINEERS

FIRST EDITION

McGRAW-HILL BOOK COMPANY, INC. 239 WEST 39TH STREET, NEW YORK
6

BOUVERIE STREET, LONDON,

E. C.

1915

McGRAW-HiLL BOOK COMPANY,

INC.

THE MAFZiE PRESS YORK PA

Dedicates

TO

MY MOTHER

SARAH SUSS LANGSDORF

321585

in the. opinion of the author. manifested by many full of the students. This has been done to counteract the tendency. In the selection and arrangement of the material dealing with commutation. in addition to the usual description of ing typical windings. attention has been concenupon certain important features that are ordinarily dis- missed with little more than passing mention. Chapters VI and VII include a considerable a full amount istics of of material concerning the operating charactergenerators and motors. wherever possible. the treatment being largely new for depicting the and in Chapters graphical and including the use of three-dimensional diagrams mutual relationships among all of the variables . to look upon a mathematical -in itself. For example. Although the methods of the calculus have been freely used throughout the book. a conscious effort has been made to give special prominence to the physical concepts of which the equations are merely the short-hand expressions. are vital to a thorough grasp of the subject. care has been exercised to eliminate those minute details and excessive refinements that are more likely to confuse than to clarify. to this end. by a and copiously illustrated discussion of the physical facts problem and their relations to one another. the book will be found to contain in Chapter III derivation of the rules covering armature windings (followProfessor Arnold). an end complete solution of a problem as apparently without a due realization vii . Instead of attempting to touch the "high spots" in the whole field of trated direct-current engineering. the mathematical analysis has been preceded. VIII and IX there has been developed a much more extensive treatment of the important subject of commutation than has been heretofore easily accessible to students of the type for whom the book is intended.PREFACE This book has been prepared with the object of placing before junior and senior students of electrical engineering a reasonably complete treatment of the fundamental principles that underly the design and operation of all types of direct-current machinery. but which.

for the most part. Such originality as has been brought to bear. and who also assisted in the proof reading. Mo. for the reason that each instructor will naturally prepare a set to meet his own needs. 1915. Answers have not been given in the text.viii PREFACE first essential is a clearly thought out analysis of physical of this procedure. and the part of Chapter VIII in which the formulas for armature inductance are developed. That part of Chapter IV which deals with details of the calculation of the magnetization curve and of magnetic leakage. Clifford. S. who made helpful criticisms and suggestions after reading the original manuscript. but will be supplied upon request to those instructors who ask for them. August. with suitable ac- knowledgment in all essential cases. has been made to include as complete a set of problems as is desirable in studying the subject. and to the various manufacturers who have kindly contributed illustrations. Numerous wellknown texts have been freely drawn upon. the author desires to express his sincere thanks to Professor H. Louis. As an example chapters have. may be omitted without interfering with the continuity of treatment. In conclusion. LANGSDORF. It is not to be expected that a new book on direct currents can avoid including much material common to the large number of existing texts on the subject. has been exercised in selecting from the vast amount of available material those parts that seem most essential to an orderly presentation of the subject. in case design is taught as a separate course. attention is directed to the new material of Article 210 of Chapter XI. ALEXANDER WASHINGTON UNIVERSITY. of Harvard University. aside from that represented by the new matter already referred to. but all of them have been successfully solved in the author's classes. . been designed to prevent the practice of feeding figures into one end of a formula and ex- No attempt tracting the result (painlessly) from the other end. ST. E. The illustrative problems at the end of each of the first ten that the realities. Some of the problems at the end of Chapters VI and VII will be found to tax the reasoning powers of the best students.

3 Lines and Tubes of Force 3 5 6 6. t 43 43 . Magnetomotive Force 23.F. 12.M. Tractive Effort of 23 26 27 30 32 34 37 38 CHAPTER THE DYNAMO 31.M. 13. Field Intensity 19. 8.CONTENTS PAGE PREFACE TABLE OF SYMBOLS vii xv CHAPTER GENERAL LAWS AND DEFINITIONS 1. '. 8 8 9 10 11 . . .F Force Due to a Current in a Magnetic Field Unit Current. 5. Introductory 2. . Practical 17.. The Law of the 20 21 Magnetic Circuit 25. Kirchhoff's Laws 27. II Dynamo. Paramagnetic and Diamagnetic Substances Unit Magnet Pole -. 7. Magnetic Field. 3.* ' . General Expression for the Magnetic Potential Due to a Coil of any Shape at any Point 22. 29. 11. Self-induction 28.M. Unit Quantity Direction of the Force on a Conducting Wire Magnitude of Induced E. Flux Issuing from a Magnet Pole Magnetic Potential Equipotential Lines and Surfaces * Induced Currents and E. . Mutual Induction Energy Stored in a Magnetic Field Electromagnets 30.M. 4. Permeability 24. . 14. 9. . Units of Current and E. 12 12 Lenz's Law 14 15 15 16 17 19 16. I 1 1 . Direction of Induced E. Field Intensity . 15. 10.F . Magnetic Potential on the Axis of a Circular Coil 21.F Heating Due to a Current 18. Generator and Motor ix . Field Intensity Due to a Circular Coil on the Axis of a Solenoid 20. Applications of Law of Magnetic Circuit 26. 1 2 .

of an Armature Resistance of Armature Winding 40. General Rules 74. 68... 34. . Field Rheostats 'I . 44. 38. Bipolar and Multipolar 42. Self-excitation 54.. Direction of Rotation of 72 73 73 74 74 76 77 79 81 83 Motors .. Windings . Number of Conductors. Motor-generator.X CONTENTS PAGE 32. E. Shunt Excitation Compound Excitation 57.. . . 62..F General Case of the E... III 1 85 CHAPTER ARMATURE WINDINGS.. Effect of Distributed 37. 35.. Number Armature Paths 73. of Elementary Alternator Induced and Generated E. 70. 46..M. 33.F. 56.. 65.. . Pulsations Average E.M. . Interpole Machines 50. Elements and Commutator Segments Winding Pitch. 63.F. Field Excitation of Machine Dynamos 52.F. 66.. Polarity of Generators 60. of an Alternator Rectification of an Alternating E. I . Brush Holders and Rocker 47. . 67. . 39.M. Turbo-generators 46 48 50 52 52 55 56 57 58 59 60 62 65 66 66 68 70 71 49.F 36. Brushes. Separate Excitation 53. The Unipolar or Homopolar 51... Types of Armatures. General Considerations 69. Construction of Dynamos Machines 41. The Commutator 43. The Armature Core The Pole Cores and Pole Shoes The Yoke Ring . Types of Windings Ring and Drum Windings Winding Element Ring. Construction of Field 58.. 64.. 61.. 71.. Winding Magnitude of E. 45. Dynamotor 48. . General Rule for the Degree of Reentrancy 93 94 95 96 97 98 98 99 103 . o . Commutator Pitch and Slot Pitch .M. .. Field Displacement of * / 72. .F. 59... Series Excitation 55. 88 88 89 89 90 91 of Degree Reentrancy . . Lap and Wave Windings Number of Brush Sets Required Simplex and Multiplex Windings.M.M.

Predetermination of Flux Distribution in the Air-gap 154 155 CHAPTER 103. 143 145 147 148 149 152 102. VI 159 159 161 161 OPERATING CHARACTERISTICS OF GENERATORS Requirements Characteristic Curves 105. Length Required Required Required Required 123 for Teeth Armature Core the Pole Cores and Pole Shoes 124 128 . Shape of Magnetic Field Produced by Armature Current Approximate Distribution of the Resultant Field Demagnetizing Component of Cross Magnetization Excitation Required under Load Conditions Experimental Determination of Flux Distribution Potential Curve . . Ampere-turns 87. 92. . . Effect of Speed of Rotation on the External Characteristic 108. Details of Calculation of 83. Calculation of the Magnetization 81. IV 113 113 115 116 117 119 121 MAGNETIC LEAKAGE The Magnetization Curve Experimental Determination of Magnetization Curve 80. 100. Service 104. 79. xi PAGE Two-layer Windings Examples of Drum Windings 77. 99. 101. Equipotential Connections 104 107 107 CHAPTER THE MAGNETIZATION CURVE. 133 Magnetizing Action of Armature Commutation Components of Armature Reaction Cross-magnetizing and Demagnetizing Ampere-turns Cross-magnetizing and Demagnetizing Effect in Multipolar 133 135 137 138 139 94.128 129 129 for the Yoke The Coefficient of Dispersion CHAPTER V ARMATURE REACTION 90. Regulation 106. Corrected Expression for Demagnetizing Effect of Back Ampere' turns 96. Machines 95. 76. Ampere-turns 86. 97. Ampere-turns 89. 109.CONTENTS 75. Load Characteristic The Armature Characteristic . 78. 165 166 167 . 162 . Corrected Axial 85. 98. Characteristic Curves of Separately Excited Generator 107. 91. . . . . Curve Magnetic Leakage Magnetization Curve 82. 93. Correction to Pole Arc for the for the 84. Ampere-turns 88.

Counter E.. . 121. 127. Railway Controllers 139.M. X . .177 Form of Characteristic upon Resistance of Shunt 178 Analysis of Field Circuit 117. . Case of Simple Ring Winding Elementary Mathematical Relations Discussion of the General Equation Modified Form of Sparking Criterion Linear Commutation The Current Density at a Commutator Segment. . . Equation. -. 145. . 116. Regulation for Constant Current 114. . Characteristic 119. VIII 237 237 239 241 243 245 246 248 Fundamental Considerations Physical Basis of the Theory .F.. Division of . . General 143. Series-parallel Control 138. 199 128. Tirrill Regulator 184 185 185 186 191 195 CHAPTER MOTORS 125.170 170 171 173 112. Connection of Generators for Combined Output The Thury System . 124... of Commutation 142. Torque and Power The Starting of Motors .. . The Reversing Motor 133. '. 141. 122.. . Characteristic 111.. Characteristics of the Separately Excited 129. \ . . Regulation of Speed of Shunt Motors 135. 144.F. Characteristics of the Motor Motor Compound Wound Motor 132. 250 . 199 199 . VII -.xii CONTENTS PAGE 110. Shunt Generator . . Dependence Dependence of the of Form of the Characteristic upon Speed . Applications of the Series Motor 136.. Service 126. Parallel Operation of Generators 123. . 146. Load between Motors CHAPTER COMMUTATION 140. . Requirements Counter E. . Condition for Stable Operation 113. Three-wire Generators. Curves of the Series Generator 168 Dependence of the Form of the Characteristic upon Speed ' . . 202 204 208 209 211 214 215 217 221 225 227 229 235 137.. Characteristics of the Series 131. Characteristics of the 115. Cycle of Operation of Railway Motors . Starting of Differentially Wound Motors 134. The Series Curve Shunt Compound Generator 120.M. .. . General Case. Characteristics of the Shunt Motor 130. . . . 147. Approximate Mathematical of the Shunt Generator 179 181 Characteristics 118.

Allowable Operating Temperatures 183. The Ohmic Losses The Core Losses Mechanical Losses 174. Successive Phases of Short-circuit in Coils of a Slot 152. Calculation of the Self-inductance. 168. Duration of Short-circuit 154. All-day Efficiency 181. Calculation of the Mutual Inductance. . Sources of Loss 171. L. Simultaneous Commutation Brushes 254 254 256 259 263 263 Effect of Wide . in Slotted 156.F Pulsations of Commutating . 167. M 157.. Winding of Commutating Poles Effect of Commutating Poles upon Coil Inductance Compounding Effect of Commutating Poles . Heating of Railway Motors Maximum Efficiency .. . Commutating Poles . Efficiency of Conversion. Condition for Maximum 321 324 325 326 327 332 Efficiency 179. Summary True of Losses 304 304 304 306 314 315 316 Efficiency.. 166.. .. 158. Field . 169. . xiii PAGE Brush Examples Simultaneous Commutation of Adjacent Coils . Additional Losses 175. RATING AND HEATING 170. The Stray Power Loss of Efficiency with 178. 264 268 273 276 280 281 282 283 Sparking Constants 160. . Reaction of Short-circuit Current upon Main Field 161... '* . Variation of Local Current Density at the 149..M. Commutating Devices Commutation in Machines having no Auxiliary Devices .. v . Selective 153. The Armature Flux Theory CHAPTER IX COMPENSATION OF ARMATURE REACTION AND IMPROVEMENT OF COM286 MUTATION 162.. Rating and Capacity 182. Commutation in Wave Windings of Several Coils.. Location of Point of 180. Armatures . 172.. 173. . . 176. . 159. Principle of Compensation 163. Variation Load. Further 150.. Electrical and Me317 318 chanical Efficiency 177. The Commutating E. ..-'. : 286 287 290 292 297 298 300 301 CHAPTER X EFFICIENCY.CONTENTS 148.. 155... 151. Compensating Devices 164. . 165...

. Temperature Specifications of Electric Power Club Output Equation Heating and Cooling Curves Heating of the Armature . 185.^. . 203... \ .. .... 193. v . The Rosenberg Train Lighting Generator ..... .. .. . Generator Field Regulation 205.F.. Regulation by Means of Armature Reaction 207.. 191... ... The Entz System The Bijur System . 201.. -357 358 .. -. . Rating of Enclosed Motors 334 334 336 340 343 345 345 CHAPTER XI BOOSTERS AND BALANCERS. Voltage Regulation in Train Lighting Systems.. 196. 208.. System .. 187.M. 189.*. 186. ... Train Lighting . . 200. Auxiliary Control of Boosters 197.. 202. 188... . 195.... . Field and Line Regulation 206. Balancers .. Operation of Rosenberg Machine as a Motor 209.. "... . Reversible Booster .. Resistance Regulation 204. The Wagner Auto mobile Lighting Generator . Modification of the Rosenberg Type of Generator 210.. 192. Boosters TRAIN LIGHTING SYSTEMS . 347 347 347 348 350 353 354 354 355 . 194.'.. . 359 360 362 362 364 366 366 373 374 376 395 INDEX . 199. Heating of the Field Coils Heating of the Commutator 190.xiv CONTENTS PAGE 184. .. . The Series Booster The Shunt Booster The Constant-current .. '. or Non-reversible Booster . . The Hubbard Counter E./. . . . 198.

by the use of the double prime ("). Unless otherwise indiInch units are distinguished in the text cated. air-gap per cm. teeth per cm. pole core per cm. armature core per cm. In a few instances the same symbol has been used to represent more than one quantity.. though such cases are widely separated in the text.. are not separately listed.) A PAGE a constant length number of armature circuits constant in Froelich's equation radiating surface a x cross-section of tube of flux at at a at c at g at 8 att at y amp-turns amp-turns amp-turns amp-turns amp-turns amp-turns amp-turns per cm per cm.. and the meaning may readily be determined from the lettering of the accompanying illustrations or from the context.TABLE OF SYMBOLS (The figures refer to the page on which the symbol is first introduced. symbols formed from those given below by the mere addition of subscripts or primes for the purpose of distinguishing between quantities having the same general meaning.. cm cross-section of armature core total area of brush contacts A Ag c cross-section of pole cores cross-section of air-gap cross-section of pole shoe A 8 At cross-section of teeth under pole A y cross-section of yoke AT amp-turns per pair of poles ATa amp-turns required for armature core ATarm armature amp-turns ATC amp-turns required for pole core (A T) d demagnetizing amp-turns per pair of poles ATg amp-turns required for air-gap 39 57 179 340 156 117 116 116 116 116 116 116 4 116 315 116 116 116 116 116 116 116 300 116 138 116 XV . yoke A Aa Ab area.. pole shoes per cm. metric units are implied.. sq.

. . .. . b' constant in Froelich's equation brush width corrected pole arc .. . flux density in teeth B't apparent flux density in teeth t B B s flux density in pole shoes -... .. corrected arc of commutating pole .. b c peripheral b'i width of pole core .116 . . .. . .. 116 125 116 By Bo flux density in field yoke intensity in axis of commutation. '.<. .xvi TABLE OF SYMBOLS PAGE A Ti amp-turns c required for two interpole cores ATig amp-turns A Tn amp-turns ATa amp-turns A T amp-turns ATy amp-turns t required for two interpole air-gaps required for two sets of teeth opposite interpoles required for pole shoes required for teeth . required for yoke 300 300 300 116 116 116 B b pole arc 119 179 241 119 131 . 299 122 125 ba width of slot 308 123 270 B flux density 24 B' amplitude of flux pulsations at pole face 313 Ba flux density in armature core 116 field intensity in axis of commutation due to armature current 280 Bc flux density in pole core 116 field intensity at pole tip due to armature cross-field 141 B g flux density in air-gap 116 299 Bi g flux density in air-gap due to interpoles bv bt width of tooth at tip b' t width of tooth at root width b width of ventilating duct of slot opening . -. . 309 34 D d diameter diameter of armature d c diameter of pole core dcom diameter of commutator 47 146 131 281 17 D distance or length .. . no-load 279 C c C constant constant '. .

m. . . . Rosenberg generator . at end of period of commutation . ... . terminal or line voltage .f. .f.m.m.f. .. no-load .m.f.......m. ( E E ET t commutating commutating e.. in abvolts .m....f.. in volts commutating e max e.m. .. .m.. 3 100 306 315 39 H h radial depth of armature core under teeth circular mils per ampere 128 335 131 h c radial length of pole core h s radial length of pole shoe 130 depth ht depth h' 2 depth h 3 depth h* depth hi of armature coil of straight part of slot above lower upper coil coil H H e of straight part of slot above of inclined part of tooth-tip of tooth overhang 269 269 270 270 270 3 18 field intensity field intensity at field intensity at end of solenoid Ho center of coil or solenoid 17 I i current in amperes current in line current in short-circuited coil total 31 ia armature current 77 243 57 . of self-induction e.f ci field axis. in short-circuited coil e.. maximum reactance voltage . .f. c. e. 36 242 248 79 247 389 15 E E Ea ECO e. 12 counter e. .m. ... .. ..f. in volts es .f Eb e. .f.f..m.m.TABLE OF SYMBOLS xvii E PAGE e e. .. .. .... F force in dynes .. .. 77 369 242 279 280 77 245 F f force in dynes field step in terms of single pole pitch frequency of magnetic reversals coefficient of friction . ... load .. commutating e. er drop in shunt field rheostat average e. . N . . ...... generated in EC commutating e. at beginning of period of commutation .f..-" > -.m..

. L 3 self-inductance of element due to end-connection leakage 345 271 299 116 79 116 116 130 131 33 269 82 270 270 270 271 M m strength of magnet pole 3 . . short-circuit current io = field current extra component of short-circuit current ia/a current per armature path / current in amperes 7 current in abamperes is ix shunt 250 77 250 237 11 10 K k constant . 3 * . la lc length of magnetic path in armature total length of wire on armature length of magnetic circuit in pole core axial length of pole core axial length of commutator total length of end-connections of element . . .XVill TABLE OF SYMBOLS PAGE (ia) ib if ii armature current. . . . .. . . . . . . . . mean length of turn length of magnetic path in teeth length of magnetic path in yoke distance between adjacent pole shoes distance between inner surfaces of pole cores self -inductance in henries . 119 119 116 305 116 131 Icom I/ I'i la It . . separately excited machine linear 319 369 162 component.. pole face loss . ' 3 lamination factor K correction factor. ly li lz L Lb L s self-inductance of coil edge at bottom of slot self-inductance of shunt field winding Lt self -inductance of coil edge at top of slot LI self-inductance of element due to slot leakage L 2 self-inductance of element due to tooth-tip leakage . Rosenberg generator field current.. corrected axial length of interpoles length of magnetic path in pole shoes .. .. constant ratio of iron section to air section under a pole 125 313 34 125 L I I' length axial length of core corrected length of core . no-load exciting current. .'. .

copper loss. turns per pair of poles. field P C m. . .... Ibs.-. . .i constant part of power loss P ca eddy current loss.. . in.' ./ N n revolutions per minute number of coil edges per element number of coil edges carrying reversed current in neutral zone .. . .. 36 '_ .... armature core loss .. . - number of poles sq.. ... series machine na turns per pair of poles... ... P p constant distance 11 . n/ turns per pair of poles.. . 98 displacement in terms of commutator segments mutual-induction . . M\z coefficient of mutual-induction between coils in adjacent slots.. . .' . 13 123 ... . teeth Pha hysteresis loss.. . P bf P ca P cc brush friction loss copper loss. .. 275 Mi 3 coefficient of mutual-induction between coils separated by 276 one slot M coefficient of . . . separately excited machine ". .. 103 146 Q Q heat generated per sec. ... ... kg-cal quantity of electricity. . Pet eddy current loss. of armature periphery .. .. .. ... commutator PC/ copper loss. .. . .... p e brush pressure.'. ... PIU hysteresis loss.. per P power permeance summation of losses 50 315 - . armature core . armature .'.. . 47 98 144 162 168 79 280 125 206 14 N number of turns of coil or circuit . teeth Ph + e combined hysteresis and eddy current loss in pole face stray power loss Pp PS Po rated output 317 315 305 306 305 322 310 312 306 308 322 313 319 324 Q q degree of reentrancy amp-conductors per cm. .. ..TABLE OF SYMBOLS field xix PAGE . . . shunt machine nsc number of coils simultaneously short-circuited by brush n v number of ventilating ducts n ideal no-load speed of motor : . .. . .. .. abcoulombs 336 12 .... ...

potential velocity volume of core 47 306 Vt volume of tooth 308 W w watts radiated . deg. . .XX TABLE OF SYMBOLS R PAGE r distance. seconds torque 122 305 309 123 136 201 V v velocity 13 vc peripheral velocity of armature peripheral velocity of commutator 281 264 6 V energy. .. Ri resistance of commutator lead 244 244 244 S s distance 13 of number specific sa winding sections coil section diagonal of rectangular heat of 55 271 cross-section of armature conductor S number commutator segments 336 305 96 T t time in seconds tooth pitch 13 tv T working temperature of armature. . . . radius resistance of shunt around series field r' resistance of resistance of resistance of ra resistance of 77 r s resistance of 3 172 armature and starting rheostat 201 armature including brushes and brush contacts. C thickness of laminations distance center to center of ventilating ducts period of commutation. 340 .. . 57 series field winding 79 162 separately excited field winding shunt field winding 77 : R Ra R c resistance in ohms on armature 15 total resistance of all wire 57 Rb resistance of entire brush contact resistance of short-circuited coil \ . work.

... ... \ . energy ... .... ... of cooling . amp-turns required for double air-gap................ ...........' 6 129 Y y commutator pitch yi 2/ 2 back pitch front pitch . 119 241 383 139 7 y specific resistance of core material ................ brush lead ......... ............... .... ........... ........... .... ................. ... .. 4 138 .. . angle angle of brush displacement from neutral coefficient ............ two sets of teeth and armature core . ............ ........... W loss at brush contact .. 116 ........TABLE OF SYMBOLS work...... supplement ................ WR loss in rheostat . constant .............. .......... 5 ....... ... 310 383 length of single air-gap . 269 47 201 101 X 10 a a.. ....... . weight of core . 97 97 97 z number of Z number a 60 of conductors per coil edge armature conductors ... .... ..................... . '... xxi PAGE W c 12 ............ ..... ............................. 306 250 228 X x variable distance X ....... i ....... constant ... . 336 383 /3 angle subtended by pole arc width 0' of commutator segment of double angle of constant ... ...................... Z .......... . .....

.. .. X ....... -. coefficient of dispersion of interpoles . ....... .... ..- temperature....*.... . .. .... .. 331 X number of flux linkages A Carter sparking criterion . .. 122 corrected length of single air-gap under commutating pole 301 length of tube of flux in air-gap 156 relative shift of segments with respect to brushes ..... .. .... ...... . ..... 39 122 ... p .. .... .. .xxii TABLE OF SYMBOLS PAGE 5' 5'i 5X A corrected length of single air-gap .. . ... ..... . .. . 20 267 AI permeability .... ..... . .. .. . ..... .. 7 ...... e 164 6 eddy current constant ...... .. . . . ... ... ../ ... \ .......... 306 317 variable angle rise of .... 79 a intensity of magnetization correction factor for fringing flux . 257 ... .. .. 335 p specific resistance of copper...'.. ' ......... .. . . ... .... .. .... . 312 77 hysteresis constant efficiency ..... .. deg........... .. ... w ... ." . ... .. .. C.... . . 118 299 output coefficient ... ..... ..... .... . Ae brush drop ... 24 V v coefficient of dispersion or vi leakage coefficient .. .... .. .

........... 3 .......... ....... end-connection leakage flux lateral surfaces of pole cores ................... total flux per pole ................. ..... armature Rosenberg generator . inner surfaces of pole cores ............ total flux due to interpole ............. ......... ............. leakage flux flux......... tooth-tip. .......... 10 118 130 268 130 268 131 268 131 14 4 ?4 flux.. ...* 4> 47 369 369 299 299 118 ratio of pole arc to pole pitch ... ......... .... leakage flux ..... ># flux............. \$ flux .... .............. >... 142 solid angle 19 .. p leakage ....... ..... f> 2 angle leakage flux per pole leakage flux.... >6 field flux.... inner surfaces of pole shoes slot leakage flux leakage flux............... Rosenberg generator . .... lateral surfaces of pole shoes ....... useful flux per pole .............TABLE OF SYMBOLS T xxiii PAGE T pole pitch ............ <p 50 <f> <P]_ 3>i .................. .. useful flux due to interpole ..................

.

The north-seeking end of the the north or positive pole. extended treatment of these basic principles the student is referred to the numerous texts in which the subject is treated in detail. the design and the operating characteristics of electrical machinery depends upon a thorough understanding of a few fundamental physical facts concerning the properties of electricity and magnetism and of the formulation of these facts as laws or definitions. such a magnet is freely suspended at its center of gravity. Field. Experiment shows that if two such magnets are placed near each other poles of like sign will repel each other and poles of unlike sign will attract each other. The magnetic fields produced by natural magnets 1 like the . The object of this chapter is to present in condensed form those facts. In the case of a magnet of elongated form. and the south-seeking magnet end is is called called the south or negative pole. and these are referred to as the poles of the magnet. measurable mechanical forces which act upon magnetic substances or upon electric currents in the field.PRINCIPLES OF DIRECT CURRENT MACHINES CHAPTER I GENERAL LAWS AND DEFINITIONS A clear conception of the theory underlying 1. If earth's field only. Magnetic stances. Paramagnetic and Diamagnetic SubThe space surrounding a magnet is called a magnetic The existence of a magnetic field is manifested by the field. laws and definitions which are immediately For a more applicable to the theory of direct-current machines. Introductory. it will turn until its polar axis coincides with the magnetic meridian at the point of support provided it is subjected to the influence of the extremities. the force due to it is greatest near its 2.

are used in practice to the virtual exclusion of all others. but its intensity generally varies from point In speaking of the pole of a magnet it should be under- is no one point at which the magnetism is actually but the conception of concentrated point poles is concentrated. iron and steel in particular. its other pole being so far removed as to be acted upon with little or no force. it will become magnetized by induction in such a way that will the induced pole adjacent to the nearest inducing pole have a polarity opposite in sign to that of the inducing pole. so that such a magnet approximates fairly well the condition of concentrated point poles. other. they are usually called simply magnetic non-magnetic in its properties. made permanent magnets. slim magnet. of the magnet. those of the latter class are called diamagnetic substances. originally unmagnetized. Materials of the former class are called paramagnetic substances. that is. The powerful magnetic fields required in generators and motors are always produced by the magnetizing action of an electric current. Every magnetized body exhibits the of polarity. Unit Magnet Pole. Other substances. This property is also possessed by nickel and cobalt but to a less extent than in the Experiment 'shows that iron and case of iron. 3. net steel are attracted by a magwhen placed in the field of the latter. In particular. useful for purposes of computation even though the idea is In the case of a long. which is assumed to be neutral or Since paramagnetic substances. If into a magnetic a magnetic substance. if one stood that there pole of such a magnet is placed in a magnetic field. phenomenon the simultaneous existence of One polarity cannot exist without the poles of opposite sign.2 PRINCIPLES OF DIRECT-CURRENT MACHINES and by artificially lodestone. the . like a knitting artificial. The classification of # material as paramagnetic or diamagnetic depends upon the nature of the medium in which it is immersed. are not sufficiently powerful for practical purposes except that permanent magnets are used in small magneto-generators and in some types of measuring instruments. The medium used as the standard of reference is air. The magnetized condition obtains throughout the entire mass to point. of which bismuth is the most prominent example. substances. the needle. are repelled by a magnet. is brought field. magnetism acts as though it were mostly concentrated at or near the ends.

and equal to unity. / is likewise unity. field intensity (in air) is called the gauss. The magnetic field at a given point is measured by the force in dynes which acts upon a unit magnet pole placed at that point. dyne upon an equal pole the be an attraction if they are of opposite sign. in air. and the force between them is found to be ra dynes. It is represented by the symbol H. and ra'. or with a force of mH dynes upon a point pole of strength m units. If a unit pole is placed 1 cm. distance in centimeters. For purposes of quantitative measurement. a and m! separated by distance r the force between two poles m is then /-fc If force is 5 ? and a) pole- measured in dynes. Two magnet poles of strength placed 1 cm. both poles being The force will be a repulsion if the two unit poles are of in air. the surrounding medium being air.. the strength of a magnet pole is measured by the force in dynes with which it acts upon (or is acted upon by) a unit pole at a distance of 1 cm. . apart. in a If a unit magnet pole it will is moved about magnetic field. strength in r are all terms of the unit defined above. respectively. In 1800 Coulomb discovered the fact that the force of attrac- tion or repulsion between two magnet poles is inversely proporIn general. in accordance with this definition. m of mm' dynes. will then act upon each other with a force pole has a strength of m units. hence k = 1. it will In other words. (2) Field Intensity.. a unit magnet pole is defined as a point pole of such strength that it will exert a force of 1 at a distance of 1 cm. A field is of unit intensity at a particular point when it acts upon a unit magnet pole This unit at that point with a force of one dyne. and the forces acting upon it can then be studied. tional to the square of the distance between them. A field intensity of H gausses then means a field which will act upon a unit pole with a force of H dynes. m'. = mm' 3. Lines and Tubes of Force. away from a pole of unknown strength.dynes intensity of a . or / 4. the force acting upon in general .GENERAL LAWS AND DEFINITIONS 3 magnet will behave as though it consisted of a single isolated pole. 5. then when m. it is assumed that the second same sign.

is AH cos a (3) the angle between H and the normal to the plane sec\$ = field. and the and a is the angle between The total flux across an area is usually expressed in lines. it is ized is by a finite of actual lines of force in any magnetic convenient to picture the field as characternumber of conventionalized lines of force. for if they did. such lines in any magnetic field. therefore implying that a magnetic pole placed at that point would simultaneously experience more than one force The a condition that is clearly impossible. in which case the total number of lines. of cross-section taken at right angles to the In a uniform magnetic field the force is direction of the field. crossing an area A is \$ = If AH field. the area is not at right angles to the the flux crossing it is where a tion. positive direction of a field or of a line of force free north pole is that in which a would move. but this is seldom is where dA used. cm. Although the number field is infinite. since there is an infinite number of points which do not lie on one and the same line of force. At each point in the length is the force can be represented by a line whose proportional to the magnitude of the force and whose If direction coincides with that of the force. It is also clear that lines of force cannot intersect. .dA H H (4) intensity at the element.4 PRINCIPLES OF DIRECT-CURRENT MACHINES in vary magnitude and direction from point to field point. or the magnetic flux. everywhere the same and in the same direction. in air. each of the intersecting curves would have a different tangent at the point of intersection. curves are now their tangents are at each point in the direction of the force at that point. the International Electrical Congress of 1900 adopted the name maxwell to represent unit flux. In the case of a non-uniform '\$ = f H cos a. This H accomplished by representing a field of intensity H. is the field a differential element of the surface. by lines per sq. normal to the element. such curves are lines drawn in such a manner that Obviously there will be an infinite number of of magnetic force.

and the entire bundle is The total flux across all sections of a tube the same. p.dA' = fH"cosa".) units. is It (Trans. I. from a pole of m m units is H = -^ gausses. from a pole of strength m units. is Tube of force. 2. from a sphere having as center and a radius r. This fact can be otherwise stated by saying that the flux across zero. If a unit magnet pole placed at a distance of r cm. (>or maxwells) 4?r lines. however. elements of a tubular surface. intensity Field the field will around magnet pole. will be repelled with a force of it f This is r = m dynes i 2 equivalent to saying that the r field intensity at a distance of cm. This can be done by suitable changes in the definitions of some of the funrecurs in Its The quantity electric and problems. of force is 1. and attempts to eliminate it from these equations have been proposed. called a tube of force. and the direction of is Fio. Since the locus of m all points distant r cm. and since lines of force cannot intersect. to note that'47r appears in the results . The outer lines of such a bundle constitute the respectively. FIG. 130. 2. 1 (5) flux from a unit magnet pole 4?r therefore. no flux can cross the walls of the tube. many of the equations that apply to magnetic presence in certain of these equations has been objected to by some scientists because of its incommensurable nature. Vol. Fig. the field intensity will be the same at all points on its surface. or fH'cosa'. damental 1904. the gphere be at each point along the radius to The total flux across the surface of j g then TYl \$ = The 1 AH = 4?rr 2 X -3 = 47rm lines is. as in Fig. 1. International Electrical Congress. the walls of a tube of force 6. interesting. is Flux Issuing from a Magnet Pole. that point.GENERAL LAWK AND DEFINITIONS A bundle of lines of force threading through a given area will converge or diverge as the field intensity increases or decreases.dA" for of the by hypothesis the longitudinal walls tube are made up of lines of force.

If m' = 171 1.. as is clearly If the evident from the above evaluation of the flux issuing from a pole.dx dV = mm'. the expression becomes > r which represents the potential energy of a unit magnet pole placed cm. for this is given . Determination of potential energy of two magnet poles. is r . with a separation of r cm. Let Fig. as m'. 3 represent two magnet poles m and m units. It represents also the amount of work or energy required to bring one pole from an infinite distance into the presence of the other. dx FIG. ergs 2 . The an entire amount of the infinite distance.6 7. it will inevitably reappear in others. separated by a distance r of strengths 777 777 x cm. from the pole. quantity 4ir is eliminated from some expressions. from a pole m. 3. move a distance dx under the influence of then work will be done to the extent of . Let this force. work done in separating the two poles from an initial separation of r cm. and therefore represents the stored or potential energy of the two poles in each other's presence. or it is the magnetic potential due to a pole m at a distance of r cm. because it is an inherent function of the geometry of space. PRINCIPLES OF DIRECT-CURRENT MACHINES Magnetic Potential. to V= Cmm ~^dx j r = (6) Since no work has been done upon the system by any outside agency during this process. Each will repel the other with a force of 3- dynes. one of the poles. . the work represented by the expression must have come from the system itself. respectively. by or the same as before.

and the direction of this force will be displaced 9. Fig. from the elementary path ds by an angle the distance ds will be The work done over dV = and the total -s 2 ds cos 6 = r r2 -5 dr will work in going from PI to 2 m . energy dissipated by the separation of the poles under the influence of their mutually repelling forces. . and Pz the final position of the unit pole. m. the 77? force will be -^. Repulsions become If. expressions for the forces and potentials will be reversed in sign. . . . above discussion. and 7*2 is the magnetic the magnetic Hence the work done by the agency producing upon a unit magnet which moves from one point to pole potential at P%. This case is entirely analogous to that of two heavy particles of ordinary matter which attract each other with . and is independent of the path followed in the travel from one point to the other. Determination of FIG. work Tr . provided the poles are free to move. and let the path between them be any curve whatsoever. At any general point on the curve. ri* But 7*1 = n-* mm P2 be (7) is the magnetic potential at PI. the magnetic potential Ot the terminal point of the travel is higher than that of the starting point. the pole -f is replaced by an of or all of the equal pole opposite sign. in the m attractions and work done by the system becomes work done upon the system. 4. from m. field another in the field is simply the differ- ence of magnetic potential between the points. 4. be the initial. and the work so performed reappears as into If such a system is left creased potential energy of the system. difference of magnetic potential between two points in a must be done by an external agency to produce the motion of the unit testing pole.GENERAL LAWK AND DEFINITIONS The work required to 7 move a unit magnet pole from one point to another in a magnetic field can be calculated as follows: Let Pi. distant r cm. will the stored be itself.

there will be a current flow if the field is steady and the ring is rotated around a diameter so as alternately to include and exclude the magnetic flux. and again if the field is steady and the ring is given a motion of translation parallel to itself from a region where the field has a certain intensity to a region . connecting points of equal potential. or linked with. Linear (or curvilinear) elements of such a surface. Induced Currents and E. tential finite energy In the case of attracting magnet poles the magnetic pois the work required to move one pole to an in- gravitational from the other. for if they did not there would be a component of force acting along the tangent to the equipotential surface at the point of intersection. The locus of all points in a magnetic field which have the same magnetic potential is called an equipotential surface. 8. which is contrary to the assumption. This phenomenon duction. On the one hand the work is done by moving magnet poles in a magnetic field. It was discovered by is Faraday in 1831 that a closed conductor which threaded by. from point to point in an equipotential surface or line. a magnetic field will have a current induced in it when the strength of the field is altered. again. on the other hand by moving gravitating matter in a gravitational field. 9. 5 is threaded by lines of magnetic force FIG. Conducting cir- cuit in magnetic field.F. consequently work would be required to move the pole along the surface. and called electromagnetic inmay be demonstrated in a is number of different ways. are called distance No work is required to carry a magnet pole equipotential lines. a current will flow in the ring.8 PRINCIPLES OF DIRECT-CURRENT MACHINES a force proportional to the product of the masses of the particles and inversely proportional to the square of the distance between them. in the case of attracting masses the gravitational potential energy is also the work required to move one of them to an infinite distance from the other.M. that the lines of force must intersect the equipotential surfaces at right angles. It follows. Equipotential Lines and Surfaces. therefore. if the stationary closed conducting ring of Fig. whose number changes from instant to instant. 5. Thus.

therefore. which 1 in turn produces FIG. and let H. It should be borne in mind that the flow of current in the various cases mentioned above is dependent upon the condition that The the circuit be closed.M. of magnetic its is in process of displacement. H II II II induced current is a secondeffect of the ary phenomenon. what actually occurs is a displacement of electricity along the conductor in a direction mutually perpendicular to the direction of the field and to the direction of motion of the conductor. Strictly. as for instance.M. but currents of this sort are called displacement currents to distinguish them from the dynamic currents ordinarily dealt with in direct-current circuits. and a (dynamic) current if the circuit be closed. constitutes a true current.F. Direction of Induced E. FIG. which in turn gives rise to a difference of electrical potential between the terminals of the circuit. hand tion. the wire ab move to the Development of E.F.m. when the circuit is open. rule. will result a displacement of electricity along the wire ab. ient A conven- method of determining the direction itself is 1 If an electrical circuit that is not closed upon "cut" by lines force. a current will flow around the closed circuit abed.GENERAL LAWS AND DEFINITIONS 9 where the intensity is different. . a positive charge appearing at b H and a negative charge means that there will be a if at a. but the difference electrical and of electrical potential may exist inde- pendently of the current. let For instance. There lines of force at right angles to the plane of the rails. Fleming's rightGenerator ac- 10. but if the ring be given a simple motion parallel to itself in a field of uniform intensity there will be no induced current. the primary changing magnetic field being to induce an e. G. 6. Fig.f. conductor cd. 14) between a then the rails are joined by a b. 7. the current. This difference of potential (Art. the While the electricity (consisting of electrons) movement primary circuit is effect of a changing magnetic flux upon a conducting electrical to produce a displacement current. Fleming's Right-hand Rule. represent magnetic right along the rails SS'. as described above.

m. 12). directed downward. the pole m will be acted upon also by an equal force. as illustrated in Fig. is closed known thumb. Force Due to a Current in a Magnetic Field. the thumb in the direction of the motion of the wire. like the three axes of space coordinates.f. which is as follows: Hold the forefinger and middle finger of the right hand mutually is of the induced e. m be a magnet pole of strength m units. Force duo to current in a magnetic field. The defi- nition of unit current in the absolute electromagnetic system of units is based upon Oersted's discovery that a current of electricity will deflect a compass needle in its neighborhood. 1 This relation is known as the law of Biot-Savart.f. 11. Fig. 8. then the middle finger will point in the direction of the induced e. tion of the lines of force. versely. see Art. 8a. which is carrying a current of / absolute units (abamperes. element dl will The force acting on the then be df = -j / dl sin (p dynes 1 (8) where gent is <p is the angle between the radius vector r and the tan- T to the wire at the element. 7 point the forefinger in the direc.10 PRINCIPLES OF DIRECT-CURRENT MACHINES and of the resulting current if the circuit as Fleming's rule. . This : fact form by Laplace as follows Let dl.m. and let into mathematical was put df * 00 FIG. The direction of this force 7 . represent an element of a wire ab. perpendicular to one another. perpendicular to the plane through r and T and under the Conconditions shown in the figure -will be directed upward.

86. from the axis of the wire. It follows from equation that the absolute unit of current may be defined as a current of such strength that if it flows in a straight wire 1 cm. long carrying a current of 7 abamperes is placed in in a uniform field of intensity such a manner that its length is H perpendicular to the lines of force (Fig. -Zi " Tv sin = (sin ai -f sin 2) (12) 12. the pole being at a distance p cm. in a magnetic field. Hence df If r is perpendicular to dl. : = 2mp I dl 2ml y <p 2j _ := 27 To^ (H) where I If is the current in amperes (see Art.r if mldl sin <p m = 1. that if a straight wire I cm. Unit Quantity. (9) Unit Current. (9) / If = Hll dynes <p the axis of the wire makes an angle with the direction of the flux (Fig. 16). 9. is the wire of finite length and m = P 1. Fig. 96). 9a). therefore.GENERAL LAWS AND DEFINITIONS In equation (8). in which sin <p <p case It follows. Thus. -. long placed perpendicular to the lines of force of a uniform magnetic field of . = HI dl sin = 1. at a distance p from a length carrying 7 abamperes. by a current I flowing in a wire of infinite length. it will be acted upon by a force Force due to a current FIG. Jo / becomes the straight wire of infinite P #. the force becomes / = Hll sin <p (10) Equation (8) serves to determine the force exerted on a mag- net pole m. or field intensity. 11 -y is the field intensity at the element due to m.

the work done is first point W = EQ ergs . Unit quantity of electricity (in the same absolute electromagnetic system of units) may then be defined as that amount of electricity which will pass a given cross-section of a conductor in one second when the current strength is one abampere. abcoulombs are moved from one point in a circuit to another by point whose electrical potential differs from that of the E abvolts. (generator action) except that the left hand is used instead of the right hand. thumb It will trical circuit when unit work (the erg) is expended in mov- ing unit quantity of electricity (the abcoulomb) from the one This unit is called the abvolt. Magnitude of Induced E. This absolute unit of current is called the abampere. then. It is important to be able to determine the direction of the force acting on a wire for any given set of This can be done by means of Fleming's left-hand rule: Hold the thumb. forefinger and midconditions. 14. dle finger of the left hand mutually perpendicular to one another. as shown in Fig. This unit of quantity is called the abcoulomb. Point the forefinger in the direction of the lines of force and the middle finger in the direction of the current. Q point to the other. Equation (10) shows that whenever a wire carrying a current lies in a magnetic field there is a force exerted upon it amounting to HIl\$m<p dynes. the wire will experience a side thrust of 1 dyne.m. Unit difference of electrical potential in the absolute electromagnetic system is said to exist between two points in an electrical field or in an elec- then the wire. will point in the direction of the force on the be noted that this rule is the same as Fleming's rule for the direction of the induced e.F. 10.M.f. Fleming's Left-hand Rule. If. This action is utilized in the electric motor.12 PRINCIPLES OF DIRECT-CURRENT MACHINES unit intensity. 13. Direction of the Force on a Conducting Wire. which consists essentially of a number of wires carrying currents and arranged to rotate in a powerful magnetic field.

the power. 11.f. of. Consider now the circuit of Fig. per second and the energy developed during a time dt will be dW = The current / acting toward the left. (in absolute units) is equal to the cut per second.GENERAL LAWS AND DEFINITIONS If this 13 work is is done in a time t seconds. or rate of doing work. in wire the will be There second.f. number of lines of force . Hence the e. 11. E generated H d ( FIG. say.m. in accordance with Faraday's law. 12). and let the wire ab of = ds/dt cm. will and a current^ of / The be set power developed will be El ergs abamperes up. El dt ergs I in the wire ab will produce a thrust of = Hll dynes hence the work done in overcoming this force through a distance ds will be dW = HU ds and by the principle of the conservation of energ}^ Eldt = Hll or ds E = HI But the expression Hlv lines of force cut per ds/dt = Hlv abvolts (14) is numerically equal to the number of second by the moving wire. abvolts. move to the right with a velocity v an e. per length I cm. P = since Q/t - W r t O = = #7 El t ergs per second (13) = I (Art. Moving conductor in a magnetic field.m.

The flux originally linked with circuit aibicd is 0i = His i and in the final position the flux enclosed is 02 = Hl(si + = s) is The change and if of flux during the movement 01 = this 02 t His is change occurs in ~T seconds the average rate of change n. in words. and as the wire moves to the right in such a way as to enclose more and more flux current has such a direction that in the circuit. the circuit linked with the flux has turns. is equal to If the the average rate of change of flux linked with the circuit.m. will simply vary from instant to instant the rigidly true. . the induced current opposes the action which pro- duces it. equation (14) still holds e. the absolute e. the resulting variable e. and by supposition the velocity was constant.f. t/Vaverage Hi S ~ & average which.m.f. the effect of the current is to oppose this increase.m. the current is an force/ produces opposing so directed that it tends to produce a magnetic field directed upward. the induced its reaction against the magnetic = /'. Moreover. rate of change of flux is not uniform.f. 11 shows that when the external force /'. in such a manner that the equation is continuously satisfied. induced by a change in the flux will be at any instant which If N E = N 15. Let the wire start from position aibi and move to position uniform.f. In other words. But if the field is not if the velocity is variable. is wire ab field moved to the right by the Inspection of Fig.m.14 PRINCIPLES OF DIRECT-CURRENT MACHINES it In the above discussion netic field was tacitly assumed that the mag- swept across by the moving wire was uniform. will at any instant be given by is a general form of the equation expressing Faraday's law. This is known as Lenz's law. in opposition to the inducing field H. and a 2 &2. abvolts (15a) Lenz's Law. states that the average induced e.

there or volts. also the induced oppose the inducing action.m. results P = El = I 2 R watts. or PR X 10 7 ergs per second which represents the power absorbed in the resistance R.m. the difference electrical potential between the terminals of the resistor is E such that I IR. respectively. is taken 10 times as large as the absolute unit. energy reappears as heat in the Since 10 7 ergs . Hence. time t seconds the energy supplied to the circuit -is In a W and this = Pt = PRt X 10 7 ergs resistor. watts of / of Heating Due to a Current. But 10 7 ergs per second are equivalent to 1 watt.8 volts is (17) Since the power in a direct-current circuit P = El when ergs per second E tively. The Joule.M. it follows that a positive increof flux should be regarded as producing a negative e. 10 8 abvolts are equivalent to 1 volt.f.f. that is. the abvolt. Practical abampere.m.F. since the latter is incon- The veniently small. When a current amperes flows through a resistance of R ohms. = E/R. so that P = El 17. from equation (16) E = -N~X (it 10. The practical unit which the strength of a current is expressed is the ampere. respecbecomes (E P = X 10 8 ) (^) = El X 10 7 ergs per second when E and / are expressed in volts and amperes. called the volt. therefore.f.GENERAL LAWS AND DEFINITIONS 15 e.. Since the induced current and. which has such magnitude that 10 amperes are equivalent to 1 16. 8 practical unit of e. this expression and 7 are expressed in abvolts and abamperes. in accordance with Ohm's Multiplying both sides of the last equation by 7. be written should Consequently equation (15a) d E = abvolts (16) ment -N ^ in Units of Current and E. E = law.

The axial component CLtL of df is = Nidi 2 -~^-~: r +x 2 COS a. by equation (8). 1 Resolv- dF ing this force into spectively parallel ular to. and perpendiccoil. it will be of all the perpendicular components is zero since for sum each element of the rise to coil which gives a perpendicular component of force in one direction. 12.16 PRINCIPLES OF DIRECT-CURRENT MACHINES is are equivalent to 1 joule. that 18. and the pole is acted upon by an equal force in the opposite direction.19 joules in the metric system. and 1 joule per second to 1 watt. or 4. 12. acting in the direction indicated in the figure. with the thumb of the current flow. the above expressions can be written equivalent W The heating readily calculated = I Rt p = El = 2 joules I 2 R watts (18) effect of a current flowing through a resistor can be when it is remembered that the mechanical is 4. = Nrldl 7-. Let P. Field Intensity Due to a Circular Coil.19 X 10 7 ergs of work are required to raise the temperature of 1 gram of water 1C. The ringers will then point in the direction of the . represent a unit magnet pole on the axis of a plane circular coil of turns and radius r cm. the axis of the clear that the components reto. Fig. rule: direction of these lines produced by a current will tend produced by. and link with. from the plane of the Then the coil. 4. (Joule's) equivalent of heat is. Field intensity axis of coil. N force acting on an element dl of the coil is. 2 (r r + x ^z )^ 2 1 A positive magnet pole in the field to move along the The lines of force that are current. there is a diaFIG. lines of force. the can be easily found by the following in the direction Grasp the wire with the right hand. and let the current in the latter be 7 abamperes. on metrically opposite element which an produces equal and opposite com- ponent. Let P be x cm.19 joules.

P on the axis. the field intensity due to this elementary ring at the point P is. N D Consider an elementary section of the solenoid dx. HQ from which it 2irNI _2TT_N1 r ~10 r (20) follows that the absolute unit of current (abam- pere) may be denned as that current which. this . will act upon a unit magnet pole 19. from the center of the solenoid.D (21) + ^ where v^ + D = 27TJV/ l2 becomes At the center of the solenoid. radius. The element may be considered as a plane circular ]V" coil of -rdx turns. from P. JTJ _ ' and the total field intensity is then H ZrNIf* '-~- C^ +D _dx x*)* ro ~r . Let Fig. by equation (19).GENERAL LAWS AND DEFINITIONS Hence 17 (19) -f- At the center of the coil. distant x cm. 13 turns uniformly distributed over the represent a solenoid of It is desired to find the field intensity at a point length I cm./-&~(i- 27T^/ 10 2+ D -'2- 2. Field Intensity at the center with a force of 2w dynes. where x = 0. on the Axis of a Solenoid. when flowing in a circular coil of one turn and 1 cm. distant cm. 0.

H with r. -.- = 0. where e D = -=-. value of tie HQ is a trifle less f than ~T7^~> namely.9984^7.18 PRINCIPLES OF DIRECT-CURRENT MACHINES which reduces to i~o if I is (22) large compared with r. 14. JVTur FIG.-/ > an d ~~ = 1.04. = H = ZirNI 101 / or half as great as at the center. Variation of field intensity along axis of solenoid. Field intensity on axis of solenoid. The FIG. i. At the ends of the solenoid. 13. 14 is large compared shows the variation is of H its along the axis of a solenoid f whose length twenty-five times radius. provided Fig.. instead 7j7 -- of Q/ be observed that axis. will -.e. and that . H is very nearly constant over the greater part of the it falls off abruptly near the ends.

the unit (positive) pole would be Magnetic potential on axis of circular coil. or toward the To move the pole to the right over a distance dx there must be expended T7 dV . and the total work required to move the unit pole out to from a point distant D cm. H It has . at the plane of the jfZ any distance x from coil. = Hdx -- -- (r /2 ergs of infinity work. 15. 15. 18. Magnetic Potential on the Axis of a Circular Coil. or m m V m = wmNI (25) . the work done times as great as the above amount. greater the divergence the more rapidly will 20. left. rent flowing as indicated in Fig. 2 (r + = 2irNl(l - ~j^= / Vr + D=) 2 2 cos a) (23) where a is P by the coil. the semi-angle of the right cone subtended at the point But 2ir(l cos a) = a> is the solid angle at the vertex of the cone. is given by 4> | / n\ 3Z this being the force in dynes that would act upon a unit magnet pole With the curplaced at the point. hence the field is nearly uniform will be practically constant. urged to the coil. and the Bide of the H decrease. from the coil is V= 2<irNIr 2 ( ** . near the ends of the solenoid and the lines diverge in the manner indicated in Fig. 18 that the field intensity on the axis of a circular coil.GENERAL LAWS AND DEFINITIONS The tion of 19 physical interpretation of these facts concerning the variaalong the axis is as follows For some distance on either H : middle section of the solenoid the lines of force inside the winding are nearly parallel.been shown in Art. hence V = If a>NI (24) would have been the test pole had been of strength units. FIG.

maxwells hence cow = CO 4rr that passes represents that part of the total flux due to the pole or links the turns of the coil. through. with. For let Fig. m N Vm = <f. V becomes the work required to bring the unit pole from up to the point in question. that the resulting equations (24) and (25) are general. however. The above expression for the potential energy of a magnet pole in the presence of a current was derived by assuming a circular coil and allowing the magnet pole to move along the axis of the coil. Magnetic potential due to coil of any shape. The expression when the current flow is as indicated. is called the number of flux linkages. is the magnetic potential at a on the axis of the coil. it represents the work regeneral point to a move unit quired pole from the point out to an infinite distance. in other words. the potential field energy of a current in a magnetic produced by some other agency is the product of the current (in abamperes) and the number of flux linkages. linking the coil is Then the flux emanating from m and = com maxwells Now allow the pole to move in such a manner that the solid angle . Therefore. and let co tended by in represent the solid angle subit at a point pole placed m any general position. Equation (25) can be put into a more convenient form. If the current is reversed. the product of the flux and the number of turns with which it links. Or. It can be shown. as follows: The total flux emanating from a pole of strength m units is 21.20 PRINCIPLES OF DIRECT-CURRENT MACHINES V. +m FIG. equation (24). 16 represent a coil of any shape which may or may not be plane.NI = \I ergs (26) where X = <pN. 16. an infinite distance General Expression for the Magnetic Potential Due to a Coil of any Shape at any Point.

17. The flux linked with = and there will m coil be induced in the an e. E = . making a total of 4irNI ergs to carry the pole once around a closed path linking the magnetic potential at of ivith the coil. None the flux emanating from the pole will pass through the coil. that is. Magnetomotive Force. subtended at Q by the coil is 27r. Q is then 2irNJ. and let a unit magnet pole be placed at a point outside P in the plane of the coil its but of boundary.f .m. carrying a current of I abamperes.JV~ abvolts If this change of flux occurs while a current of / abamperes is flowing in the coil. a further amount of work equal to 'ZirNI ergs must be expended. 17 represent the side and 22.) of the . and the difference P and Q. is lirNI ergs.GENERAL LAWS AND DEFINITIONS changes the coil 21 magnitude by will change by in dco in a time dt. magnetic potential P is Now let the unit pole be carried along any path to a point Q infinitely close to the The solid angle plane of the coil.f. the work done is given by dV = -EJdt = and the tance total (GO NIdtp = mN I do) ergs infinite dis(co work required to bring the pole from an = 0) to a point near the coil = co') is = mNI or the _ /** ' f' mta'NI = <pNI = \I same as equations (25) and (26) . as Q'P.f. If the pole is now carried back to P along any path which requires the pole to be threaded completely through the coil. Closed path linking with coil. This is The unit of m. Let Fig. FIG. coil. the amount of between magnetic potential work required to carry the pole from P to Q. front elevations of a plane coil of any configuration. (m. and therefore at the zero.m. called the magnetomotive force is called the gilbert.m.

18 INI = which states that the m. the Fig.m. the the total is field intensity is not constant along the that amount of work to describe the complete path of Fig. if the field intensity were constant all along the the work required to carry the unit pole from end to end of the solenoid would be force X distance = H l = ^ NI axis. coil of Fig. is riot a/orce. 18. l = TNI = m.f.f . so However. the 12 is pulled out into the form of the solenoid of work required to carry a unit magnet pole around a all closed path threading through of the N turns is 47rNl = fev7 FIG. but work per unit of elecpole. \ Hdl (27) is the line integral of the magnetic.22 It PRINCIPLES OF DIRECT-CURRENT MACHINES should be particularly noted that m.it is of the nature of work per unit magnet to e.. quantity (see Art.f.m.f. 18. Magnetic path linking solenoid. 14). But the noid is field intensity at the middle p'oint of the axis of the sole47T NI so that H which means that axis.m. which trical If It is exactly analogous is likewise not a force.m. . force.

represents to the same scale the work required to carry the magnet pole along the path from end to end of the solenoid. 14 represents the work required to carry a unit magnet pole through the solenoid from one end to the other. 13 is The field intensity at the center of the 47T 16 NI f gausses and lel if the lines of force passed straight through the solenoid paralto the axis. 19. while the area of the rec- tangle enclosing the curve represents the work required to carry the unit pole once around a closed path linking all the turns of the solenoid. as in Fig. cm. Yt\ _Lu maxwells ~i~A I Now I let the solenoid be provided with an iron core of length sq. solenoid of Fig. The resultant field intensity at the center is then _ 4T 10 NI I " ~ _2ro_ <Z/2) _ == 47T NI / Sm Z 10 2 . strengths bar. and were uniformly distributed over the area of cross-section A. cm.. Solenoid will wound on iron core. 19. therefore. Permeability. the flux across any section would be (f) = AH A = 4?r NI . and cross-section induced poles of FIG. 23. + m and m then be developed at the ends of the and each of these poles will give rise to a field intensity at the center of the solenoid equal to 777^2 an(^ opposite in direction to (I/ Z) H.GENERAL LAWS AND DEFINITIONS 23 It is interesting to note that the area under the curve of Fig. but outside of its windings. that the area above the curve. but inside the rectangle. It follows.

24 PRINCIPLES OF DIRECT-CURRENT MACHINES In other words. This demagnetizing or "end effect " becomes negligibly small if the solenoid and core are long. the induced poles exert a demagnetizing effect upon the field which produces them. From the pole + m there will emanate 4?rm lines of force. 20 shows a number of such curves as used in practice. or the more it is susceptible to magnetization. Curves showing the relation between B and H are called normal B-H curves. the lines of force and the 4?rm \$ = Assuming the flux 3> AH + to be uniformly distributed over the crossis section A.r= MH + 4T _ (29) = g 1 (30) It is called the permeability of the material of the core. total flux consists of the original lines of induction. the ratio of the strength of the induced pole to the intensity of the inducing field. The better the material from a magnetic standpoint. not as lines of force. so that it is 5 in terms of H. therefore. and will be neglected in the remainder of this discussion. Its magnitude is dependent upon the ratio -rjf or 77. Fig. is the ratio of the flux density in the material to the intensity of the inducing field. The relation must be found experimentally for each material. hence the ratio -. for several different kinds of iron and steel. There is no known relation /z impossible to express either or . or HA Inside the iron. all of m through the surroundwhich find their way back to the pole These lines of force may be assumed to be continued ing air. ^ is called the between m and H. but as lines of induction. field. that is. the greater will be the strength of the induced *??? pole for a given inducing susceptibility. or magnetization curves. through the iron core back to the starting point. and is therefore a measure of the specific magnetic conductance of the material. since the former exist only in the external non-magnetic medium. the flux density B where == + M 4 = = ffl+4.

n is less than unity. for if the value of the former quantity. In practical calculations it is = -y- = -^-H = O. even in H one and the same material. It follows that ampere-turns per cm. corresponding to a given value of B.Uaits FIG. 20.SH to deal with more convenient am- pere-turns per cm. in air and iron. and NI/l expresses the number of ampere-turns per cm. and in diamagnetic substances. Magnetization curves. magnetization curves are commonly drawn with B Consequently.GENERAL LAWS AND DEFINITIONS 25 In certain kinds of iron the permeability ju reaches very large values 2000 to 3000 but its value varies as varies. In paramagnetic substances like is always greater than unity. ji = 1. than with H. nickel and cobalt. (lines per sq. the total excitation (in ampere-turns) is simply the product of ampere turns per cm. In the equation 5 100 # = 4-jrNI Jn ~y~> 500 600 700 00 tne term NI represents the number ) Ampere Turns 200 300 400 per Inch (Upper Curves 900 1000 1100 1200 1300 1400 2'0 30 40 50 60 7 100 //mC. other non-magnetic substances.G. and the length of the circuit. //. can be found. of ampere-turns of the exciting winding.S. .

26 PRINCIPLES OF DIRECT-CURRENT MACHINES cm." The unit in which reluctance is measured is called the oersted. moreover. The total through the core then be 4?r \$ = or AB A r> NI The numerator solenoid. H= will TT^JI where flux I is the mean length of the core in centimeters. the curves are drawn with ordinates B X (2. Equation (31) is of the form flux = m. 21. FIG. 19.SH (= ampere-turns per 24.m. in. inch). is of equation (31) is is the m. If (= ampere-turns per cm.) as English units are employed.f.m. hence the reference to permeability (Art. netic circuit. the permeability p appears in the expression for the reluctance in exactly the same manner as does the specific conductance in the expression for electrical resistance. The demagnetizing effect of the ends of the core of Fig. 21. of the and the denominator the reluctance of the magIt will be noted that the expression for reluctance for the resistance of of the same form as that it is cuit. as in Fig. for proportional to the length an electrical cirand inversely proportional to the cross-section. The Law of the Magnetic Cir- Magnetic Reluctance. 23) as "specific magnetic conductance. can be eliminated by bending the core into a closed ring The value of form.f. and with abscissas 2. reluctance .54) 2 (= lines per sq.) as ordinates and with 0. referred to in the preceding section. cuit. H and will be Ring core.). will then be uniform around the entire circular axis of the coil.8# abscissas.54 X O.

maxwells = . Applications of Law of Magnetic Magnetic cir- be joined in series. called permeance. A reluctances in series will have a total reluctance given by and the resultant flux through the circuit will be UlAi V-lA 2 HnA n where NI is the total circuit as a whole. number of ampere-turns acting upon the This equation can be written and in this form each term of the right-hand member represents the m. resistance volts ^r amperes = ohms is r The reciprocal of reluctance. like electric circuits.GENERAL LAWS AND DEFINITIONS or. is then .m.f. and the solution of problems involving any of these combinations is in every case carried out by methods that are the exact analogues of those used in the corresponding eleccuits. or in series-parallel. circuit consisting of a number of methods to be employed 1.f.gilberts - oersteds 3- which corresponds term by term with Ohm's law of the circuit electric current or = - e.f. in 27 terms of the units themselves. Thus. parts a and b represent typical and in each case the analogous electrical circuit The following examples will serve to illustrate the in the solution of ordinary problems. The total m.m.m. 22. or -y. may trical circuits. Circuit. magnetic is circuits indicated. SERIES CIRCUITS. 25. in Fig. (in gilberts) required to maintain the flux through the corresponding portion of the circuit. in parallel.

. in Fig. The number of ampere-turns required to maintain the flux through the air-gap.m.28 PRINCIPLES OF DIRECT-CURRENT MACHINES merely the sum of the m.(where all Y~ quantities are in metric units) or (32) NI = 0. 4?r where /* = 1. Rn . may be found from the relation . In most cases arising in practice it is not necessary to compute the reluctances of the several parts of a circuit in accordance with the above equations. = = length of path in cast iron 26.fs.000 maxwells on the assumption that is made of cast iron.15 = 2060.000 lines per sq. Zi mean path follows quadrants of circles. required by the individual parts.125 B = and in air-gap = j = 40. . . From ampere-turns required by the core is 79 X 26. In the analogous electrical circuit / = - E Rl + R% + 2 . in. corresponds to an excitation Hence the number of of 79 ampere-turns per inch length of core. length of path in air flux density in iron lines per sq.3133 X 40.125 = la = 0. 0.f.m. Assume that the mean path of the lines of force follows the center of gravity of the cross-section and that at the corners the /.000 the curve for cast iron in Fig. A more direct solution is possible through the use of curves like those of Fig.000 X M = 1567 total excitation for the entire circuit is 2060 + 1567 = . . in. + 4) + 2* - 0. . required to maintain the current through the circuit is the sum of the potential drops in each part of the circuit. 22a.15 in. = 2(6 in.3133 X lines per sq. X air-gap in inches Hence the ampere-turns for the air-gap are NI = and the 3627 ampere-turns.8BI = 0. and E = IR 1+ IR + . 20. +IR n or the total e. in. let it be required to find the number of ampere-turns necessary to the core produce a total flux of 160. 20 it is found that a flux of density 40. B = H = NI -T. Thus.

226. In the case of the parallel magnetic circuits. Typical magnetic and it does not have that electric circuits. circuit the entire battery e. 22. acts equally on each of the parcircuits. but that the right-hand circuit. for example. FIG. the flux in each part is to be if the parts Suppose. circuit of Fig. and for each value circuit will and indeed compute the corresponding total number of ampere-turns. It is again 160. Just as in the corresponding electrical are connected in parallel. Assuming that the flux through the left- hand branch maxwells. This can be done by trial. that the left-hand circuit of Fig. the actual flux corresponding to the latter can then be read from the curve. then to find that value of flux through the cast-steel circuit which will require 3627 ampere-turns for its maintenance. Plot flux and ampere-turns. is made of cast steel instead of cast iron. 22a. of the exciting circuit act on each of the magnetic computed as though the other part were not present.GENERAL LAWS AND DEFINITIONS 2. value because the reluctance of The problem is the cast steel is less than that of the cast iron. are exactly alike the flux will be the same in each.f.000 maxwells. the reluctance of the cast-steel part of the circuit is so small compared with that of the air-gap that a first approximation to the final result may be found by assuming that the reluctance of the path .000 coil must supply an excitation of 3627 amperedoes not follow that the flux through the cast-steel be 160.m. so does the entire m. the turns. allel electric PARALLEL AND SERIES-PARALLEL CIRCUITS. In the case under discussion the total flux through the cast-steel circuit way found to be approximately 303. circuits.m.200 maxwells. extending the computations far enough so that a curve may be drawn that will include within its range the given number of ampere-turns. 226 is exactly the same as that of Fig. 22a. 29 two magnetic In Fig. as follows: Assume a series of values of the total flux. is in this In such a case as that illustrated by this particular problem. though having identical dimensions.f. 226 each of the type illustrated in Fig.

and each conductor is to be given a symbol to indicate the magnitude of the current flowing in it.200. would be given by through the consumed in 3627 = 0. Then any is . and an arrow to indicate the assumed direction of the current flow. in. The algebraic sum of the currents at is any junction of the conductors in the 2. in.f. It is then necessary to select a would require 1635 ampere-turns. charge of electricity at the junction would steadily change and its potential would change correspondingly. no effect of this kind has ever been observed. network zero. that is directed positively around the loop to be given the positive sign.3133 X B" X H or B" = 92.f. 20 shows that at this flux density the steel circuit would 3627 + 1635 = 5262 ampere-turns. and that the entire excitation is maintaining the flux through the air-gap. so that the entire require smaller value of B". The solution of problems involving the flow of current in networks of conductors depends upon two experimental facts. known as Kirchhoff's Laws: 26.m. repeating the calculations until the given number of ampere-turns is included in the range of trial values. or considerably more than the available number. The algebraic sum is of the potential drops around any closed loop in the network these zero. Let it be agreed that the clockwise direction around any closed loop of the network be considered to be the positive direction through the circuit (though cases may arise when it might be more convenient to select the counter-clockwise direction as the positive one).m. the of all the currents leaving that point. On this basis the flux density in lines per sq. it may in that case be considered A curto produce a rise of potential in the positive direction.800 lines per sq.30 PRINCIPLES OF DIRECT-CURRENT MACHINES steel is negligible. 1. shall active e. is The first of two laws a statement of the fact that the sum of all the currents entering a junction point is equal to the sum If this were not so. In applying the second law it is convenient to make a diagram of the network and to give to each active e. Reference to the curves of Fig. KirchhofFs Laws. and \$ = 371. (such as that from a battery or dynamo) an appropriate symbol to indicate its magnitude and an arrow to indicate the direction in which it acts.

have each a resistance of 0. a and 6.m. hence. which are supposed to con^^a Ohm resistor of r 0. Each a virtual internal resistance of 0.GENERAL LAWS AND DEFINITIONS rent of i 31 amperes flowing in the positive direction through a ohms will produce a fall or drop of potential of ir volts. the second law. heaters. directed as shown. Each half of the battery has an A and B are the loads e.2 ohm c and have resistances of 6 and 8 ohms. Fig. |jB|80hros Ohm-=-[115Volts I<IG. which must. tion the number of lines of induction coming up to the juncmust be equal to the number leaving it.1 a sist of lamps.3. .84.31. referring to the upper loop of the + And 115 - 4(0.wire system.1 0.5) = From these three independent equations it is easily found that = 17. From the first law.5 ohm.f. the sum of the currents entering point P must equal the sum of the currents leaving it. KirchhofFs laws are applicable to the magnetic circuit as well as the electric circuit.1 + + 6) + + 0. respectively. accordingly.m. The meaning of the negative sign of 4 is that that current actually flows in a direction opposite to the assumed direction.f.47 and 4 = . be treated as negative. 23 represents a storage battery connected to " a three. Thus at any junction in a magnetic circuit. from the assumed directions of current flow From diagram. or _T other devices which do not oimiiiit 113 y lts develop a counter e. Example.Net-work ot conductors. wire c has a resistance of 0. for the reason that . half of the battery has It is required to find the current in each of the supply lines. 4 = 13. the neutral" wire c being connected to the middle point of the battery.2 ohm.1 ohm. ~y~ .5 + - 0. The outer wires.24 = referring to the lower loop + ia 115 - 0.24 4(8 0.. 0. of 115 volts. and the neutral 5 _=. 23.

circuit the flux is and the reluctance R. is directed outward from the plane of the paper. which tends to main- panding line of force H is filament of the wire shown . Self-induction.m. the same line of reasoning will show that an initial decrease of current induces an e. the drop of magnetic < potential is &R. Relatively. or in opposition to the direction of the current flow. in any lines of induction are to KirchhofTs first closed magnetic circuit. 24a.m.f.f. the effect is the same as though the filament were moving radially inward so that if Fleming's (right-hand) rule is applied. of reversed direction. Conversely.f. The whole effect is in accord with Lenz's law.fs. with due attention to the sign of each term. which tends to retard the change in current. . The lines of force thus expanding cut across the wire manner indicated in Fig. 27. here. in the about to cut across the longitudinal at P. This is equivalent law of the electric circuit.m.. The txon a pond. the motion of the line of force at this point being radially outward. and the summation of all such drops must then be equal to the summation of all the active m. Fig. the wire will be surrounded by magnetic lines of force as shown. When a straight wire. the flux will increase proportionally from zero. which represents a crosssection of the wire in (a) when viewed from the left. 246. Again. carries a current in the direction indicated.32 PRINCIPLES OF DIRECT-CURRENT MACHINES always closed loops. like spreading ripples FIG. however. the original change in the current strength which produced the change in the flux immediately calls into existence an opposing e. 24. the algebraic sum of the drops of magnetic If in any part of the closed magnetic potential must be zero. and may be thought of as issuing from the center of the wire and expanding outward.m. (6) (a) Lines of force surrounding a conductor. it is found that the induced e. As the current increases from zero to any arbitrary value.

as for instance. and is dependent upon the shape.m. the self-inductance from equation is (34) that square of the number the flux. and a mean length of magnetic path of I cm.m.f. in those coils of a direct-current generator or motor which are undergoing commutation (Chap. of self-induction. The mean path is to be taken as passing through the center of gravity of the cross-section On passing a current of i amperes through the coil of the core.m. induced in it by a current . Its magnitude is of very great importance in all electrical circuits in which the current is proportional to the of turns linked with changing in strength. Let Fig.f.m. cm. a cross-section of is A on a core having sq.GENERAL LAWS AND DEFINITIONS tain the current at its 33 self- original strength. where T = 4ir ~ N nA 2 TO r~ The quantity L is called the coefficient of self-induction or the self-inductance of the circuit.. and in the practical system of units is measured in terms of a unit called the It is evident henry. called the e. This e. From equation (33) it is seen that the self-inductance L of a circuit is numerically equal to the e. size and material of the magnetic circuit.f. being induced. 25 represent a coil of wire wound constant permeability ^. there will be produced a flux N' * = 10 "T = 4*Ni A To~T" and a change of current di will produce a change of flux of N_di _ 4rr This change of flux will then induce an e. VIII).f.

= Ij . - *i = Jgtfrfl n Mi^-i = Ci# 1*1 (36) A part of this flux. of mutual induction in the second circuit. The product the number of turns with which of flux per ampere by this flux links is called the num- ber of flux linkages per ampere. a circuit has a self-inductance of 1 henry if a change of of 1 current of ampere per second induces an e. so that. as well as upon the rate of change of current in the inducing circuit. Mutual Induction. the self-inductance 8 is equal to the number of flux linkages per ampere. or 1 the of lines of force produced ampere. multiplied by the number of turns which this flux links. a change in the current strength in first circuit will induce an e. respectively. are so placed with respect to each other that the magnetic field due to a current in one of these links in whole 28. and divided by 10 8 .f.m. changing at the rate of 1 ampere per second 1 (~r.34 PRINCIPLES OF DIRECT-CURRENT MACHINES is which that is. turns. or w = Xi*i = KiCiNiii (37) . It is clear that the magnitude of this will depend upon the geometrical shapes and relative positions of the two circuits. Equation (35) says that the product of the numby ber of lines produced by 1 ampere. Thus from equation or L = in this N^ X 10number 8 (35) equation -p is numerically equal to the rate of change it is of flux with current. Let a current of ii amperes in the first circuit produce a flux 3>i such that 47r A7 . is equal to the coefficient of self-induction. briefly. volt.m. If two circuits of Ni and N% the or in part with the other.f. The self-inductance (33) may also be defined in another way. divided by 10 .

it follows . sijice the potential energy of the system can have but one value. due to unit current in the and from (42) which represents the number of flux linkages with the first circuit due to unit current in the second. so that number of linkages with the second circuit is X 21 N = NWI = K&iNiN&i is (38) i2 and if the second circuit traversed by a circuit of amperes its potential energy in the presence of the first circuit is. so that the is number of linkages with the first circuit Xi 2 = Nw = KiCtNiN&a (42) The potential energy of the is first circuit in the presence of the second Fi2 = X 12 ~= Ho KtC^Niite ergs (43) But Vzi must be equal to V\.GENERAL LAWS AND DEFINITIONS 35 the total (where K\ ^1) will link with the second circuit of 2 turns. from (44).'. . KiCiNiNs = K 2 C 2N 1 N 2 (44) From (38) or it is the number of flux linkages with the second circuit first circuit. Hence. 21. the current total flux 4?r i2 K&NiNMi circuit will ergs (39) in the second produce a 3>2=r-.i= Ho Similarly. V 21 = X. by Art. 2 C 2 N zi z (40) which a part <p z = K 2 3> 2 = K 2 C 2 N 2i2 (41) (where total K 2 ^ 1) will link with the first circuit of Ni turns.= of 10 N* .

8 = KtCtNiNt X . as unit current in the latter will produce in the former. the e.C 2 AV X 10. When the current in circuit No. there duced in circuit No. e. It is obviously of the same nature as self-inductance. (47) M 2 = KiKsLiLi (49) . or the mutual inductance. will be in- = - N^ X 10- = . 2 changes.-* X be written ID-" From (44). divided by 10 efficient of mutual induction. From (45) it follows also that the mutual inductance of two circuits is numerically equal to the e. 1 is clear (36) that the self-inductance of Li = ^^ X 10^i 8 = Ci#i 2 X 10. 1 an e. 2 is x 10 ~8 = ~ KiCiNiN x 10 ~8 and when the current in circuit No.m. from equation No. induced in circuit No.f. and is measured in henries. 10~ 8 (46) the number of flux linkages with one circuit due to unit current 8 This is called the co(the ampere) in the other.KtCtNtN.8 (48) (46). 1 changes.m.f. of the two circuits.8 (47) and from equation (40) that L2 = Hence. these equations 62 may = - M di.f.m. and (45) where M is = KiCiNiNi X 10.36 PRINCIPLES OF DIRECT-CURRENT MACHINES that unit current in one circuit will produce the same number of linkages in the other. induced in one of them when the current in the other changes at the rate of 1 It is circuit ampere per second. from N ^X and 10~ 8 (48) .

m.f. Energy Stored in a Magnetic Field. ing tion coil portance as a factor in the commutation process in direct-current machines. the primary and the secondary.GENERAL LAWS AND DEFINITIONS 37 If the circuits are so related that there is no leakage of flux between them. work required to therefore. K. if all of the flux produced by one circuit links with all of the turns of the other. e = L di at -yr VOUS If the current is i amperes at the moment when the rate of change effect of current is -j- amperes per second. = and K 2 = 1 M or the mutual inductance of two perfectly coupled circuits is a mean proportional between their respective self-inductances. both of which consist of an iron core upon which are wound two coils. A coil or circuit of self-inductance L henries carrying a variable current will have induced in it an e. The factor \/KiK 2 is sometimes called the coefficient of coupling.f. Mutual Induction is also of ime. insulated from the core and from each An interrupted or alternating current in one winding sets other. The phenomenon of mutual induction is utilized in the inducand in the alternating-current transformer. a periodically varying flux which in turn induces an alternatup in the other winding. from zero to a value W = j Lidi = ^Li 2 joules (50) This energy is not lost. but is stored in the magnetic field. the power required to is the change of current ( e)i = Li time -T7 watts and the work done in the ( dt is dW = The total e}idt = Lidi joules raise the current amount i of is. and may be recovered by allowing the magnetic field to collapse to . that is.m. 29.

: is from (39) and (43) W and by = Ho tfiCitfitf 2*1*2 = Ho #2Catfitf 2*1*2 ergs (46) this becomes 10 8)*i* a W = Ho(M X = Miiia X 10 7 = Mi iiz joules ergs If the two circuits magnetize in opposite mutual potential energy is evidently reversed the stored energy of the system is directions. it is therefore analogous to the mass of a ity. of the electric circuit the current i is the quantity of electricity that passes a given point in a second.. Miii Zj can be derived as follows The potential energy of one circuit in the presence of the other derivation of the .38 PRINCIPLES OF DIRECT-CURRENT MACHINES It is this energy which appears in the spark or arc zero value. there is stored in the system an amount of energy W = JLii! 2 + KL 2t 2 2 + Miiit joules (51) provided the two currents magnetize in the same direction. 26. It is instructive to compare equation (50) with the equation This is of the form for the kinetic energy of a moving body. be con- sidered as the kinetic energy of electricity in motion. on the axis of a cylindrical bar magnet of radius r cm. or electrical velocity. Fig. formed on opening an inductive circuit. from the end of the bar magnet. and is analogous to velocself -inductance L represents a sort of electrical inertia. assumed to be uniformly distributed over the end surface of the . let the strength of the pole of the magnet be m units. operates to resist any change in the current flow. W where = y% mv 2 In the case ra is the mass of the body and v its velocity. When two circuits of self-inductances LI and L 2 have a mutual inductance equal to M. The it since mechanical system. Let a unit magnet pole be placed at the point P. and distant a cm. so that W = %Liii 2 + JL 2 2 *2 ~ Mitt (52) 30. Tractive Effort of Electromagnets. The energy J^Li 2 may. therefore. The first two terms of equation (51) is obvious from (50) the last term. their in sign.

from (56). the force which it will exert upon the unit pole at P is and the total force F = where is due to the entire pole C r xdx j of the magnet cos 0) is 27rcm ^p^p = 27r<7(l - (54) P by the end of the magnet. 39 The pole strength per unit area. 26. then Considering an annular element of radius x and width dx on the end surface of the magnet.X vdA and the total attraction between .and a. and if the intensities of magnetization of the adjacent surfaces are + a. if a is small relatively to the dimensions of the end area of the magnet. or the intensity of is magnetization. the attraction of one of the magnets upon an elementary magnet pole of area dA on the other will be dF = 2iro. in which case or F = If 27TO- (55) two bar magnets are placed end to end with a very small separation. cos0 approaches zero as a limit. <r =-r> and since the flux issuing from a pole it m units is \$ = 47rm. the semi-angle of the right cone subtended at the point If the distance a is made very small.GENERAL LAWS AND DEFINITIONS cylinder. 2 <J> BA 2 -8T . FIG. respectively. Field intensity on axis of bar magnet. follows that a = T~X\ <j> whence. adjacent poles will be F = From equation of "77? 2w(7 2 A dynes (56) (53).

tion of the coil 9.? 6. storage battery having a short-time discharge rating of 10. distant from. about a vertical axis passing through its middle point? How much work must be done to turn magnet B through 180 deg.f generated in the coil if the intensity and direction of the earth's magnetic field have the values given in Problem 1? (b) What is the maximum e. If the total intensity of the earth's magnetic field is 0.m. are placed with their axes in the same horizontal straight line. A circular coil of an angle value given in Problem poles? 6. A lie that horizontal wire 30 cm. 3. PROBLEMS 20 cm. center to center. The acceleration of gravity is 980 cm. each 20 cm.m. the north pole of B. ? 4. the south pole of A being adjacent to. a cross-section of 1 in. Find the force exerted by one magnet on the other. with the plane of the magnetic meridian. (a) What is the average e. How much work must be done to turn magnet B of Problem 2 through 90 deg. A trolley feeder running in an east and west direction. carries a continuous current of 300 amp.f. faces being placed in parallel vertical planes. apart. has its maximum value? A coil of insulated is a current mounted wire having a resistance of 200 ohms and carrying in a glass tube through which is passed a stream of .40 This is PRINCIPLES OF DIRECT-CURRENT MACHINES the fundamental equation underlying the design of tract- ive or lifting electromagnets. the bus-bars are spaced 6 in. - 1. per sec. what is the force on a length of the feeder between What is its direction? A is amp. 7. generated in the coil. is placed in a vertical plane that makes 45 deg. and indicate its direction. and what is the posi- A concentrated circular coil of on a horizontal axis pointing east . per sec. per sec. Assuming that the current may be considered to be concentrated at the center of crosssection.g. and the intensity and What direction of the earth's magnetic field are as given in Problem 1. when the e. what is the total flux that passes through the coil? 2.000 connected to the switchboard by copper bus-bars which have . by 10 in. the 10-in.. long falls freely between frictionless guides in a plane perpendicular to the magnetic meridian plane.f.42 gauss and the angle of dip is 62 deg. Solve the preceding problem on the assumption that the current in each bus-bar flows uniformly along the vertical plane through the central Y axis.s. its fall? 10 turns and radius 10 cm. Two identical slender bar magnets. and 5 cm. in diam. is revolved and west at the rate of 10 rev. and carried on poles 100 ft.. units at their ends. flowing from east to west.of the third second of 8. is the difference of potential between the ends of the wire at the end .m. long and having concentrated poles of 200 c. If the intensity of the earth's magnetic field has the 1. what must be the distance between supporting brackets in order that the bus-bars may not deflect more than in. A and B.

cm. but owing to different con- ditions of charging. What is the strength of the current? 10..000 maxwells. Find (a) the number of ampere-turns to produce a flux of 320. in 0. Their internal resistances are.g. distant. is placed so that its plane lies in the magnetic meridian. flowing in a counter-clockwise direction when viewed from the magnet. there is placed a circular coil of one turn having a radius of 25 cm. 22a. respectively. and direction of the earth's magnetic field being as specified in Problem 1. about a vertical axis through the center of the coil. of 50 volts. How much work must be done to carry the bar magnet along the axis to a symmetrical position on the other side of the coil? 12. and the initial temperature of the water is 20 C. The net thickness of the core is only 90 per cent. would be induced if the current and flux are reduced to zero. rectangular coil.15 ohm. Haw many ampere-turns are required to produce a flux of 30. battery. the centers of both coils lying on a line perpendicular to the line joining their centers a current of 10 amp.m.000 maxwells. 13. Three sections of a storage battery are connected in parallel with each other and supply current to a circuit whose resistance is 2 ohms. the second 48 volts and the third 45 volts. flows in the must flow through the circular coil in not be deflected? What must be the the two coils? long having concentrated poles of 200 c. by 60 cm. (d) the permeability of the steel. Each section of the battery contains 24 cells.. 0. A slender bar magnet 20 cm.000 maxwells? Compute the permeability and reluctance of the ring.as an e. 14.s. The temperature of the surroundThe is 30 C. sides horizontal. A cast-iron ring has a circular cross-section of 1 inch diameter and a intensity of 10 in. 0.m. per min.f.. their planes. units at its ends is placed on the axis of a one turn circular coil of The north pole of the magnet lies nearest to the plane of the radius 25 cm. Find the current supplied by each section of the current would flow through each section if the external circuit were disconnected? What 16. (c) the reluctance of the air-gap. compute the amount of work required to turn the coil through 90 deg. amp. current is then adjusted until the temperature of the outflowing water is constant at 40 C. The coil carries a current of 20 coil and is originally 10 cm. 15.1.GENERAL LAWS AND DEFINITIONS ing air 41 water at the rate of 500 cu.01 sec.. at a uniform rate. to it.f. and 40 cm. If the current has such a value that it produces a total flux of 30. (6) the reluctance of the sheet steel. away from it. of the gross thickness because of scale and air spaces between the punchings. is Midway between them and on If placed a small compass needle. A circular coil of 10 turns and radius of 20 cm. one of them h. how much current order that the compass needle may relative directions of the currents in 11. carrying a current of The 10 amp.12 and 0. what e. A rectangular coil 30 cm. and having a single turn is placed Parallel in the magnetic meridian plane with its 60 cm. Compute the self -inductance of the cast-iron ring of Problem 13 assuming that the winding has 500 turns. A magnetic circuit made of sheet steel punchings is built up to the mean diameter dimensions of Fig.? .

. When the two coils are connected in series so that they magnetize in the same direction the self -inductance is found to be 0. when they magnetize in opposite directions the self-inductance is 0. PRINCIPLES OF DIRECT-CURRENT MACHINES Two circular coils. amp. (c) the amount of work required to turn coil B through 180 deg.3 henry. assuming that they are connected in series and are carrying a current of 50 the other. in pounds. and a current in B produces a flux of which 90 per cent. starting from the position in which the two coils magnetize in the same direction. Find (a) the self-inductance of A and B] (b) the mutual inductance of A and B'.05 henry. links with A. 18. parts. are mounted concentrically. If the cast-iron ring of Problem 13 is split into two semicircular what is the pull. A A and B. A produces a flux of which 70 per cent. required to separate the two halves? is What they have been separated before? the magnitude of the pull between the two parts of the ring when He -> the excitation remaining the same as m .42 17. current in one inside links with B.

of the seat of # = where I Hlv ab volts (1) is the length of the wire in centimeters. and when In other words. H is the intensity of the field through which it is moving. a dynamo is a reversible machine. used for the second called a motor. say. the circuit of these active conductors external receiver circuit. or inversely. . motor action recurrent from some external source is sent through is On a set of conductors located in a magnetic field.fs. machine. so that Current will be produced when e. a driving force must be applied to the conductor and work must be done at the rate of Fv 43 = Hllv = El ergs per second. When used for the first it is purpose it is called a generator. or a dynamo.m. may be defined as a machine for the conversion of mechanical energy into electrical energy. and v is its velocity in centimeters per second in a direction perpendicular to that of the field and to flow a current upon Ohm's law. hence. in accordance with its own of. I abamperes. to maintain the action. Dynamo. A dynamo-electric 31. Every generator consists of a conductor.m. On closing the circuit there will length. for the conversion of electrical energy into mechanical energy. the word dynamo is a generic term which includes the other two. In the case of generator action. The conductor will then be acted upon by a force of F = HU dynes (2) in a direction opposite to its motion. being capable of operation either as generator or motor. are induced in them. each conductor an induced e.CHAPTER II THE DYNAMO Generator and Motor.f. the value of which will depend the resistance of the circuit as a whole. sults is completed through an when the other hand. or set of conductors. subjected to the influence of a varying magnetic field.

A direct current. so that it IF acted upon by a lateral thrust of F = Hll motion of the wire intensity results. that all of the energy supplied reappears as useful energy after the conversion process has fact this condition been completed. As a matter of never realized in practice.f.f. and (2) direct-current machines. The armature of a dynamo is the part in which the e. according to the nature of the e.m.f. type produces a current through the external circuit which flows in one direction only.m. In the case of motor action. and current produced they are (1) alternating -current machines. . is called an alternator and produces an e. Either may be the rotating member if the armature rotates. vary in magnitude from . the one structure with respect to the other is most easily obtained by making one or the other rotate.m. which acts alternately in opposite directions. The field member is the The relative motion of part which produces the magnetic field.f. so that when the armature circuit is completed the current in the cirThe second cuit flows first in one direction and then in the other. and the electrical source of supply at the rate of is El = magnitude done by Hlvl = Fv ergs per second. v dynes of the field and under the influence there is H and velocity induced in the wire an e. There are two distinct types of dynamo-electric machines.44 PRINCIPLES OF DIRECT -CURRENT MACHINES. is generated in the case of a generator. the energy supplied must be greater than that usefully converted by an amount equal to the loss of energy inevitable in the conversion. the machine is called a revolving armature machine. E = flow there Hlv abvolts in a direction opposite to the current. may. namely. when used as a generator. of sufficient to balance this counter-generated e. each conductor is caused to carry a current of / abamperes.m. however. while if the field rotates it is called a revolving field is . machine. The first type. though characterized by constancy of direction.m. so that in general the two have concentric cylindrical forms. To maintain work the current must be impressed an e. or the part which carries the working current in the case of a motor.f. In this discussion ideal conditions have been tacitly assumed.f..m.

28 represents diagrammatically an 8-pole .m. FIG.THE DYNAMO 45 instant to instant.fs. field that passes across from pole and S are the extremities of the field structhat the pole pieces ture. that is. . front to back. in the latter case the cur- rent is said to be a continuous current. 27. tion r 2 will be negative. Fig.m. in other words. alternating-current machines usually have more it and that touching will than the two poles shown in Fig. Fig. In the position shown in the to the insulated slip-rings n.f. abed. while the e. In practice. it may be pulsating. the current is said to be a direct current. and that the excitation of the magnets is effected by a direct N N the current from some suitable source circulating in coils wound on The ends of the armature coil are attached field structure. form alternating-current generator or motor is the simplest dynamo.m. in series in such manner that the e. rotating in a magnetic The of It is understood to pole S. 28. stant in magnitude as well as in direction. in cd will be directed from back to front. 27. directed from will wire ab have generated figure. or it may be conIn the former case. Multipolar revolving field alternator. 27. After half a revolube seen that the polarity of the terminals reverses.f. they are The winding consists of a number of coils connected multipolar. FIG. so that each terminal is alternately of opposite polarity. Elementary alternator. Reduced to the most elementary type. it consists of a loop of wire. of the individual coils add together. r 2 in it an e. the collecting brush touching ring r\ will therefore be positive.

motors consist a wire. 50. 32. 30.or bar-wound armature arranged to rotate between inwardly projecting poles of alternate polarity. 31. E. whose armature winding consists . but so far as FIG. 28.f. changes its therefore.m. FJG. 44. all standard forms of direct-current generators and FIG. It is the function of the commutator to convert this internal alternating e. Consider first the elementary alternator of Fig. Fig. in an alternator in order to understand thoroughly what is happening in the case of the direct-current machine. Each of the armature conductors is. hence it is important to analyze the development of the e. 29. Elementary alternator. every direct-current machine (with the exception of the homopolar machine) is essentially an alternating-current machine.m.f. With the exception of the homopolar machine described in Art. the armature winding itself is concerned. 30. the seat of an alternating e.f.f.m.46 PRINCIPLES OF DIRECT-CURRENT MACHINES revolving field machine with the winding of the stationary armature arranged in eight slots. Alternator armature in a general position. in the manner illustrated in Fig.F. of Elementary Alternator. into a uni-directional e. Developed armature winding of of alternator of Fig.m. in the external circuit. 29 is a development of this particular type of winding as it would appear if the cylindrical surface of the armature were rolled out into a plane.M. which direction each time the conductor moves from the influence of one pole to that of the adjacent pole.

per conductor is then = Hlv X 10.m. (3) and at any general . /sec. and that the field in the gap is uniform and of intensity H Further. when the coil has moved 6. degrees from the vertical (Fig.f. in either of the following (a) ways 7? : The peripheral velocity of the conductors is V = ird ^ cm.f. as indicated by the light horizontal lines. 31).. <l>.m. S are so shaped that the magnetic flux issuing from them passes straight across the air-gap between the pole shoes and armature core.m.M. varies from instant to . the component of velocity perpendicular to the direction of the flux v is 71 = V sin = ird 7. Time variation of E.F. so that 27rJV<J>~sin t)U = 7rZ<J>^sin uU X 10~ 8 X 10~ 8 (6) This equation shows that the e.f. let the armature rotate with a speed of n revogausses. of elementary alternator. The instantaneous e e. that passes from pole to pole. may then be found lutions per minute.8 = irdlH -^sinfl bU it is X 10~ 8 and for the entire Z e conductors = TrdlHZ i sin = X 10~ 8 (5) But dlH e is the entire flux.instant. 32. Let it be assumed that the poles N. The generated e.THE DYNAMO of a concentrated coil 47 having Z conductors (or N= v -^ on z turns) the external periphery of the amature core A.sin (4) FIG.

in the direction of .F. in one revolution each conductor cuts 2\$ lines. is found by dividing the area of one loop of the curve of Fig. This is equivcilent Induced and Generated E. period from 9 = corresponding to a half revolution of the armature. divided 8 Now. 8 I. induced in a conductor is .f. is therefore identical 2\$>Z ~X (8).f.8 = 2irN3> ~X 10~ 8 (7) to 6 = IT. the average e. to the assumption that the intensity of the field at each point on the periphery of the armature. equal to the number of lines of force cut per second. 2\$Z lines per revolution and 2\$Z^ Eaver = 33. Chap.8 which (c) is the same as equation (6) above.f.m. the instantaneous e. by 10 hence the entire 71 Z conductors cut second.8 > (6) In the general position of Fig. tn*b flux linked with the coil is 9 hence. 32. 31. hence = 2wN3> ) sme X 1Q. shown in Fig. 10~ 8 = 4N ^Q Q X 10. By definition. 32 by the base.f. during the or Eaver - M ede = Z3> ~ X 10~ 8 j sin OdO 7? X x 10~ 8 (8) 10. lines per The average e. The average e.m.8 .M. = \$ cos (9) is by equation e (17). which is with equation In the preceding article the direction of the flux in the air-gap was assumed to be everywhere parallel to the axis of the pole pieces.48 PRINCIPLES OF DIRECT-CURRENT MACHINES The graph is instant according to a sine law. sine curve of this equation is a whose maximum value Emax = as irZ^^Q X 10.m. But -rr is the angular velocity. and this equal to 2x . = - N~ X 10~ at e = N\$sin6>^ X at is 10~ 8 .m.

It is important to note that in the first instance the total flux linked with the coil is zero.f.m. further instance of the distinction between these two views of the matter. the former being due to the generated motion of a conductor across magnetic lines of force. 33. would be a maximum. 31).m. is zero.f. so that if the armature coil were stationary in this position. has reached the dotted position. in the position shown by the full lines (Fig. due to a given change in polar flux. generated in the conductors is a maximum. 33).f. since half of the one direction and the other half in the opposite direcit is advantageous . flux acts in tion. 33. is H. the latter As a to a change in the total flux linked with a closed coil. is proportional to the sine of the angle between the particular radius considered and the axis of reference (the vertical Such a distribution of flux is said to be sinuline in Fig.to distinguish between and induced e. For this reason 4 . whereas if the coil were stationary in this position the induced e. e.f.f.. Distribution of field intensity at armature periphery. opposite the center of the pole. as indicated in Fig. cutting the field at the greatest rate. the strength of field at each point of the periphery will be represented by the ordinate of the sine curve whose maximum the coil is When each value.m.THE DYNAMO 49 the radius. as shown by the equality of the hatched coil positive and negative areas. and the When the e. a change in the magnitude of the polar flux would not induce e.m.f. soidal.m.f. in the coil. is zero. the e. coil edge is FIG.m. it will be noted that in the dotted position of the coil the generated e.. If the armature surface is developed into a plane.m. since all of the flux through the coil is then in the same direction.

w < - if"?. a sinusoidal flux distribution.m.M. < e.f. 34. the flux distribution may depart considerably from the sinusoidal. The average e. let the armature have a diameter of d cm. however. the armature surface be represented by curve further. The discuswas based upon the assumption of a bipolar field structure. divided by thus each conductor in one revolution cuts lines per pole. is number of lines of force cut per second. and let the speed of rotation be n revolutions per minute. except for a change in scale. Let Fig.f.. edx Jo = C Hldx I = .m. PRINCIPLES OF DIRECT-CURRENT MACHINES General Case of the E. of an Alternator.50 34.8 (ID T where is the flux per pole. will then be a curve which is the same as that showing the flux distribution. a coil spanning the arc from center to center of poles. The instantaneous e.F. . 34 represent a partial development of an alternator having p poles (like Fig. H length of I cm.f. and let the distribution of flux at FIG. 32 pole pitch. 28). and the coil spread may be greater or less than the sion of Art. in a direction parallel to the shaft. Multipolar alternator. that is. per conductor is E f .>f d\ Hldx X Jo w 10~ 8 = PQQ& X 10.f. The last result might Jo have been anticipated Trom the fact that the average equal to the 10 8 . Generally. there is more than a single pair of poles. non-sinusoidal flux distribution.V. and a full-pitch armature coil.m. the conductors have an active . generated in each conductor 10- is HIV X Hlird 6Q X 10" (10) and the graph of this e.m.

m.F.THE DYNAMO 77 51 or p\$ lines per revolution. 29 so that the coil spread is the same as the pole pitch. If H is the is first \$ is the integral of the function.f. derivative of the flux function.f.m. 35.f.f. curves. hence P^QQ lines per second. in coil of fractional pitch. that of coil-edge a following curve a. in wound with Z conductors. It is interesting to note that conversely. 35 that the maximum e. and that of coil-edge b following curve b. The total instantaneous e. Fig.m. in all conductors will be simultaneously in the same phase FIG.fs. 34. of the two sides of the coil will differ in phase. the e.f. But if the coil spread differs from the pole pitch. Eaver = pQZX 10- 8 (12) which becomes identical with equation (8) if p = 2.f.m. of the coil is obtained by adding the ordinates of the individual e. whose variation from instant to instant is represented graphically by a curve identical (except for a change of scale) with the curve If the conductors are arranged as in of flux distribution. is the total average e. 29. Fig. will be simply times that of a single conductor. Z ous e. Fig. as indicated by coil ab.M. It is evident from Fig.m. E. H the armature series as in Fig. each conductor is the seat of an e.m. of such a . Fig. and the total instantaneous e.f.m. all connected in such a manner that the coils are of full pitch. the instantaneof the variation. 35. As pointed out above.m. 34.

that is. coil of the armature commutator as to tionary brushes. then.f. cally identical. or curFlG 37 Rectification of alternating rent varying as in Fig. respectreversal of the e. voltage will vary in the manner shown in Fig. and 38 are therefore electri- and the latter the same give pulsating e.m. Elementary flux distribution is sinusoidal. An e. the brush direct-current generator. 37. coil and negative The polarity.f. 6 2 . and staare mounted so 61. Let the Z peripheral conductors of the coil of Fig.M. 36. 38. The large amplitude of the pulsation of the e.m. 37 is due to the fact that the entire armature winding of Figs. it will be uniform in direction. in Fig. but there are two each having -^ turns. 36. The part of the winding inside the ring core plays no part in generating The windings of Figs. the plane of the brushes being coincident with that through the shaft and the polar axis. as the former. but pulsating between zero and a maximum value. or Z turns in all.f. If. 27 are connected. and means must be found to make it more nearly continuous.m. still 17 Z peripheral conductors. Effect of Distributed - Winding. brushes 6 2 and 61 will always be of positive ively. respectively.. 36 or 38 is inactive twice during each revolution. If the terminals the elementary alternator of Fig.m. 36 be wound as in Fig. make sliding contact with the re- volving commutator segments. will 36.f. if the winding can be so disposed that .F. Here there are coils. to the two insulated segments of a 35.52 PRINCIPLES OF DIREC T-C URREN T is less MA CHINES "short-chord" winding the same a and than that of a full-pitch winding of number of conductors.f.m. since it cuts no lines of force. C. of the takes place coincidently with the passage of the brushes across the gaps between the segments. Rectification of c of an Alternating E. the FIG. 36 e. as in Fig. 37 is not desirable.

(A sinusoidal flux distribution is assumed. in turn. There are now four pulsations . similar to FIG. FIG. The resultant brush voltage will be obtained by adding the ordinates of these two component curves. the voltage at the brushes is equal to that of either half alone. as shown. 40. 38.. generated in the coils shows that the brushes must FIG.) Similarly. Since the winding consists of two equal halves in parallel.fs. 37. A study of the directions of the e. Considering the particular half-winding made up of sections A and B.m. section .THE DYNAMO small sections of sations will sections is 53 it undergo commutation successively. now be placed in a plane perpendicular to that through the shaft and polar axis.m. B generates a wave exactly like that of section A but the two waves differ in phase by 90 deg. but with the Z conductors in four equidistant groups of concentrated coils conarranged sufficiently large. 39 be a diagrammatic of sketch of the armature Fig.M. 38. 38. that of Fig. it will be observed that section A generates a wave of e. but of only one-fourth the amplitude since it has but one-fourth as many conductors as the coil of Fig. Ring winding with four sections. and that the entire winding is now equivalent to two equal halves connected in parallel. 40. E.f. 39. Each half. generated in four-coil winding. is made up of a pair of winding sections connected in series. nected to a four part commutator. as shown in Fig. the pulbecome insignificant when the number of such winding Let Fig. Elementary ring- wound armature.F.

The resultant and their Ring winding with eight sections. X FIG. as the number of .m. 41. if series. E. made up in as in Fig. as shown in Fig.F. D. The average voltage at the brushes will mum be Eaver = \$Z or one-half as great as given 10-' (8) since (13) by equation only Z/2 con- ductors are in Similarly. is brush voltage.M. but the range from minimum to maxiis much reduced. It follows. the winding is of two halves in parallel. 42. 41. The physical explanation of this fact may be traced to the circumstance that the winding section cut out from the armature circuit during its passage under a brush plays a smaller and sections smaller part in its effect is increased.54 PRINCIPLES OF DIRECT-CURRENT MACHINES instead of the original two. from its neighbor. made up of eight pulsations is still less than in the case of the four-section winding. each half containing four sections series. then. and the maximum e. B.f. of any one section is sections. generated in eight-coil winding. Each of these four A. in place of the original two. that by subdividing the winding and correspondingly increasing the number of commutator segments.m. the entire Z V conductors are grouped in eight equidistant concentrated sections. 42. 10~ 8 . upon the total e. amplitude 27T FIG. the amplitude of the pulsations may be reduced to any desired extent.f. C. generates a wave differing in phase by 45 deg. connected to an eight-part commutator.

8 .f. deg. . + sin(7r \ o / J J = 5 3>Z ~ cotan OLf S TT X STT 10~ 8 (14) B - F -- '4sin + sin -+. 55 of E. + / - TT \ - "1 m*(.M. in these sections are. = -- 3>Z ^ sin X 10~ 8 = E'max sin 6 -f es = # wax ' sin(0 + + 2- J e s/2 = E' max sin 6 - -. respectively. a -7 s minimum 6 value of occurs when 8 = and a maximum when = s deg. and assume also that the magnetic As before. ) J cosec (15) .m. Assume a ring winding peripheral conductors divided into s sections having Z/s turns each. Fig. . Magnitude consisting of Z and let n be the number of revolutions per minute.THE DYNAMO 37. Substituting these values of 6 in the equation E= ei + e + e +.m. Pulsations. let <i> be the total field has a sinusoidal distribution. X 10.m.0 and 42 of will show that a complete pul- sation occurs in an interval e. + es/2 = #'max I sin + + sin (0 + -H o \ / it follows that Emin E max r sin \ - L o + sin -o + . Each winding section will then generate a sinusoidal e. and the e. . 41. 4. whose maximum value is .l = E' max sin Investigation of Figs.f. 2 3 .fs. There are s/2 flux per pole ^ sections in series in each of the parallel paths through the arma- ture winding.. .\$Z X 10~ 8 volts.F.

is table: In other words.f.) Consequently EMr = -i7T/ S J it +/ f""< c\ 62 " e. Average E. is 7T 7T cotan - The percentage of the minimum to maximum in terms minimum cosec s s^ 7T X 100 cotan - and the magnitude of shown in the following this quantity. the fluctuations are quite insignificant.M. (See Figs. the average e. for various values of s. With the type of machine premised in the foregoing article.F. 40 and 42. and then decreases symmetric- minimum when = s .m. may be found by observing that the e. varies from a minimum to a maximum 38. of an Armature. with the winding divided into 30 or more sections (the field structure being bipolar).f. value between ally to the 6 = and value = . /2 )de ~ X 10 " sn J sin X 10poles (16) In the type of machine thus far considered there were but two and only two paths for the current from brush to brush .56 PRINCIPLES OF DIRECT-CURRENT MACHINES variation from value.m.

THE DYNAMO inside the 57 armature winding. or is p\$ Alines per second.. may be any even integer. so that the aver- age e. the entire number oU of conductors is divided into a groups connected in parallel. the poten- difference zero. per group of conductors. the e..f. tions of of Fig. is. the actual resistance of the arma= ra ohms.m. and the number of parallel paths through the armature may be any other even integer. so that there are Z/a conductors in series per group. provided all paths have the same resistance. a. FlG : 43. of the armature as a whole.f. which are subjected to the inductive action of one pole and the other half to the influence of a pole . of a direct-current machine. If the total resistance of all the wire on the armature is R a ohms. p. the total armature current i a will divide equally between them. The average value of the generated e. and since all of these a paths are connected parallel. ture.m.. in such a case may be easily calculated as follows: Each conductor cuts p\$ lines per revolution. Brushes displaced of opposite polarity. of course. . ture from proper position.m. Resistance of Armature Winding. as will be explained in detail in the chapter on armature windings (Chap. and.f. 39.per conductor X 10~ 8 volts.. . is between them will be Each path through the armamade up of conductors half of . therefore. will be R a /a* m The drop of potential due to the entire current i a flowing through the resistance ra or ia ra volts. III). as measured between brushes. is X x a 60 X 10 s (17) This is the general equation for the generated e.m. ously if the brushes of the armature Thus. In an armature having a paths. provided the brushes are so placed that the winding sec- any one group are simultaneunder the influence of one pole. equal to the drop of . the resistance per path will then be R a /a ohms. 43 are so placed that of commu- tation takes place in coils opposite the middle tial the pole shoes. But in general the number of poles.f.

X Ra /a = ia ra 40. The dynamo consists essen- tially of an electrical circuit and a magnetic circuit placed in inductive relation to each other. The magnetic . or ia /a potential through volts. Construction of Dynamos. The electric circuit consists of the armature winding and the commutator.58 PRINCIPLES OF DIRECT-CURRENT MACHINES any one of the paths.

45. brush-holders and the rocker-arm. III. cannot exceed definite limits without im- pairing the operating characteristics of the machine. however. are described in the following articles. with the exception of the armature winding. 45 shows a semi-enclosed type. Bipolar of simplicity and Multipolar Machines. though not at all regularly. and Fig. Although for the sake much is on the assumption seldom used except in machines of the preceding discussion has been based of a bipolar field structure. Further.THE DYNAMO circuit is 59 made up of the yoke. when carrying current. this type of field of the smallest size. in (in direct-current general. the armature spider and the bearings. pole cores and pole shoes. FIG. the choice of the number of poles depends upon the consideration that the magnetic reaction of the armature. Semi-enclosed motor (Sprague). The principal structural features of the various parts of the machine. 41. . The actual number of poles generally varies from four to a maximum machines) of twenty to twenty-four. Totally enclosed motor (Sprague). FIG. Fig. Fig. The annular space between the revolving armature and the stationary field structure is called the air-gap. Other parts of the machine are the field winding. the brushes. 46. 46 a totally enclosed motor. 44 shows a common arrangement of these parts in the open type jpf construction. the number the with The capacity. and the armature core. increasing of the the of choice the number explanation principles underlying of poles in any given case must be deferred to a later section. The subject of armature windings is taken up in detail in Chap.

60 PRINCIPLES OF DIRECT-CURRENT MACHINES with an armature core of given dimensions. as the number of poles is increased beyond two. The optimum limit occurs peripheral spread of the pole faces is approximately to the axial length of the pole face. and the copper segments are then wedged together to form a rigid circular arch by means of cap-screws tapped radially through the outer steel ring. of segments into a rigid structure is an interesting one. A compact field frame equal is advantageous in that comparatively little of the field flux leaks when the from pole to pole without entering the armature core. separated from one another by the mica insulation. 42. such as mica. and with pole pieces that cover a definite percentage of the armature surface. The completed commutator must be giVen a high voltage test to insure the thorough insulation of each segment from the others and from the . The insulation between the commutator and the supporting hub consists of molded mica cones and cylinders. after which the auxiliary steel clamping rings are removed and the external surface turned to true cylindrical form. 47. The V-shaped grooves are then turned out and the commutator spider bolted into place. The process of assembling a large number material. Construction of commutator. 47. The segments. the field frame becomes more compact. are placed around the inner periphery of a sectored steel ring. up to a certain limit. as in Fig. The Commutator. The commutator is built up of wedgeof shaped segments drop-forged or hard-drawn copper insulated from one another by accurately gauged thin sheets of insulating FIG.

in high The material must be so selected that its voltage machines. for two reasons: first. working temperature of the commutator is from 95 C. there must be at least 60 segments between adjacent brushes of opposite polarity. 600-volt railway generator is to have not more than 10 volts between adjacent segments. between segments are imposed by the requirements of sparkless commutation. if a 6-pole.06 in.THE DYNAMO spider. Commutators this used because it meets requirement. second. for the latter must withstand the full terminal voltage of the machine while the former is only called upon to withstand The average voltage the smaller voltage between segments. The minimum diameter of the commutator is then determined if the minimum peripheral width of a segment is known. and they determine the minimum number of segments in the completed commutator. The thickness of the insulation between segments varies from 0. depending upon the current per brush arm. Commutators must be designed to have a sufficient amount of exposed peripheral surface to radiate the heat caused by brush rate of wear friction The perloss due to brush contact resistance.. largely are sometimes built in such manner that the insulation does not come quite flush with the surface. in low voltage machines up to about 0. this minimum width is rarely less than %g in. . and from 20 to 25 volts in the These limiting values of average voltage case of railway motors. 1 The to 130 C. American Institute of Electrical Engineers. blistering or wear. between adjacent segments should not exceed 10 to 15 volts in lighting and railway generators. 1914. because some allowance in the radial depth of the segments must be made to permit turning down the surface in case of pitting. to be as 61 The insulation between adjacent segments does not have heavy as that between the segments and the commutator spider. is Amber mica is the same as that of the copper bars.02 in. thereby obviating the necessity of selecting the material for a definite rate of wear. or not less than 360 segments in the entire commutator. 1 In the case of turbo-generators running at See Standardization Rules. For example. because the taper of the segments would result in too thin a section at the inner periphery if a smaller external width were used. design must provide sufficient mechanical strength to withstand and the missible the centrifugal force.

thick.f. to secure sufficient radiating surface the commutator must have a considerable axial length. per minute. not exceed 8000 43. Accordingly. greatest. If the core were solid of as might be thought very large number of metallic filaments running parallel to the armature conductors and all connected together. sketched in Fig.f. Fig. but the loss due to them decreases as the square of the thickness of the sheets._. points perpendicular to the direction of the generated e. or. 48. the sheet steel ordinarily used in armature cores is an inch or two.PRINCIPLES OF DIRECT-CURRENT MACHINES high speed the diameter peripheral velocity shall is limited by the consideration that the ft. but generally the oxide. . on one side of each disk. hence. 0. would result in excessive heating and loss of power. Eddy current paths in core. EDDY CURRENTS. Armatures of the now . 48. if solid armature and currents would circulate in the mass of the core in the manner . the core must be laminated in such a manner as to preserve the continuity of the flux path and to break up the The plane of the laminations must be at all current paths. To prevent springing of the segments are in held steel they place by rings shrunk over the segments. 57. The armature core not only carries the magnetic flux from pole to pole.. parallel to the direction of the flux and to the direction of motion. The Armature Core. e. and insulated thoroughly therefrom..m. or scale that forms on the sheets. unchecked. is relied upon to provide the necessary insula- core is tion. as shown in Fig. The e. by Fleming's rule. 44. but revolves through it in exactly the same manner as the conductors of the armature winding.m. at those points.014 in. which. the armature built up of thin sheet steel punchings insulated from each sometimes the insulation consists of a coating of varnish other.m. will obviously be greatest near the surface where the peripheral velocity and the active component of the flux are likewise To minimize these eddy or Foucault currents. in machines of the usual radial pole type. in some designs a layer of paper is inserted at intervals of Laminating the core does not completely eliminate eddy currents.. in such a case each filament would it made up of a be the seat of a generated FIG.f..

are supported by lock nuts screwed directly to the shaft. had cores built up of concentric hoops. 44). One-piece armature punching. Ventilating ducts through the core are formed by spacing pieces placed at intervals of from 2 to 4 the axis of the core. 49. 44. 50. in larger machines the end plates are held together by bolts passing through the laminations. 50. 49. as in . in. in the case of small machines. the joints between segments are staggered from layer to layer in order to preserve the continuity of the magnetic circuit. and the end plates are shaped to provide a support end connections of the armature winding (see Fig. Lamination of disk armature. with active conductors arranged radially. or. then punched by a special punch press which Core punchings of this sort are generally keyed directly to the shaft. of along The spacing pieces are generally made by .THE DYNAMO 63 obsolete disk type. first Core punchings up to a diameter of about 16 made in one piece.Fig. Fig. and are sometimes provided with holes near the shaft to form longitudinal ventilating passages. The disks are are generally blanked out N N Disk Machine Spirally Laminated Armature FIG. Cores of large diameter are built up of segments which are at- and the slots are cuts one or more slots at a time. means in. tached to the spider by means of a dove-tail joint. The core punchings are held together by end flanges which. as in Fig. therefrom. more practically. but insulated for the FIG. of thin strap iron wound as a flat spiral.

Armature core assembly. to a punching of heavy sheet steel. as illustrated in Fig. showing spacing pieces. or by wooden or fiber wedges driven into the recesses at the tips of the teeth. Where 5 semi-closed slots are used. the coils ? T FIG. The spacing pieces should be so vary in width from as to the teeth as well as the body of the core.64 PRINCIPLES OF DIRECT-CURRENT MACHINES riveting brass strips. or they may be made by pressing spherical depressions into a thick punching. Fig. slots. except in the case of very small machines. The embedding of the armature winding in the slots serves a . 51. on edge. Y % FIG. The coils are held in place in open slots either by steel or bronze bandingwires. The ventilating ducts to in. designed support in order to prevent vibration and humming. may coil be formed on a winding but the wires of each side of a must be slipped into the slot one at a time. 51. Typical shapes of teeth and jig. special machines. formed coils that can be readily and slots for directSmooth core armatures are used only in Open slots with parallel walls are generally slipped into place. 52. used. 52 illustrates typical forms of teeth current machines. for the reason that they permit the use of insulated.

the poles usually have a circular cross-section because this results in minimum length and weight of the copper wire in the field winding. to the first following tooth. The flux density in the body of the pole core.000 lines per sq. The average flux density in the air-gap should not exceed 62. be transferred from this tooth-tip to the The line of force holds on. until the increasing tension causes it to snap back suddenly to the next tooth. is less than it would be in a smooth-core construction having the same amount of armature copper. increased area is secured by means of pole shoes bolted or dove- by means of projecting tips or horns punched integrally with the sheets composing a laminated pole. hence the This pole faces must have greater area than the pole cores. since the latter in large measure passes around the slots by way of the teeth. by reason of the motion of the armature. or . Solid poles are commonly bolted to the yoke. it seems surprising that the fundamental equation for the generated e. is the same for a slotted armature as for a smooth-core armature. that a line of force which at a given instant crosses the air-gap from the pole face to a given tooth tip. Laminated poles of course require a rectangular cross-section. At first sight. or distance from the pole face to the iron of the armature core. Laminated poles in be secured either a dovetail may place by joint or may be cast into the yoke.m. increased velocity of cutting of the lines of force by the conductors exactly compensates for the reduced value of the field intensity in the slot. erally The The Pole Cores and Pole Shoes. must later. therefore. the air-gap..f. and the armature conductors are supported by the teeth when subjected to the tangential forces caused by the reaction of the armaWhen the armature conductors ture current upon the field flux. It must be remembered. however. 110. in. The made of cast steel. is considerably greater running as high as than can be eco- nomically produced in the air-gap. When cast steel pole cores are genis used. and so reduces the amount of field copper necessary to produce the flux. The pole faces or shoes are almost always lamitailed to the core in the case of solid poles. in slots the thus embedded are they are apparently shielded from the inductive effect of the field flux.THE DYNAMO 65 double function. as it were.000 lines per sq. 44. in. tooth in the manner of a stretched elastic thread..

Carbon brushes are . These pole-face eddy current losses will obviously be reduced by so proportioning the dimensions Tufting of flux at and slots as to prevent appreciable lack of uniformity in the distribution of the flux along the pole face. except in the case of low-voltage machines when they may consist of copper or copper gauze. or whenever saving in weight is important. the flux passing between the pole face and armature core tends to tuft opposite the teeth. integral part of the bed plate in larger sizes it is cast separately. It is made of cast iron in small machines and of cast steel in larger sizes. The yoke is that part of the field structure which carries the flux from pole to pole. then in the other. though the laminae of the pole shoes do not have to be made as thin as those of the armature core. Brush Holders and Rocker Ring. The yoke is usually split on a horizontal diameter for convenience in assembling and In machines of moderate size the yoke is cast as an repairing. The connection between the revolving armature and the external circuit is made through the brushes^ which are usually made of graphitic carbon.f. and as the teeth move across the pole-face these tufts are drawn tangentially in the direction of rotation until the increasing tension along the lines of force causes them to drop back to the next following tooth. in order to reduce the loss nated. Brushes. directed parallel to the shaft. thereby inducing an alternating e. 53. 46. as shown in Fig. 45. but with lugs for bolting to the bed plate. and at the same time serves as a mechanical support for the pole cores. To mini- mize the flow of current the pole tips of * face must therefore be laminated in planes parallel to those of the armature laminations. The tufts of flux are therefore continu- ously swaying back and forth. FIG. and if the pole face is considered as built up of thin filaments. even and heating due to eddy currents set up in the pole faces by the armature teeth. for.66 PRINCIPLES OF DIRECT-CURRENT MACHINES when solid poles are used. 53. as at P in Fig. . The determining factors in this proportioning are the ratio of slot opening to airof teeth gap and the length of the air-gap itself. 53.m. The Yoke. each of the filaments p will tufts first in be cut by these swaying one direction.

must be such that there construction of the brush holders be no vibration of the brushes. The brushes are commonly set at a trailing angle with respect to the direction of rotation. The width of the brush from three to five times that several armature a so width of commutator the segment. which. but insulated from. but in such a manner that the springs do not carry any current. the total axial length of the brushes constituting a set is determined by the consideration that there must be a contact area every 30 to 50 amperes to be carried by the set. though in machines designed to run in both directions. and should allow the brush to slide freely in order that the brush may follow irregularities in the commutator surface.THE DYNAMO made 67 of varying degrees of hardness to suit the requirements of commutation. such as railway motors. as illustrated in Fig. takes on a polished surface of dark brown color. The individual brushes of a set must not be too large in cross-section. The subdivision of the set offers the additional advantage of allowing the individual brushes to be trimmed one at a time without interfering with the operation of the machine when under load. as discussed in a later chapter. otherwise there would be of 1 sq. The individual brushes are supported in metal brush holders which are in turn supported by studs attached to. when fitted with brushes of the proper composition. Tfye brushes are held against the commutator surface by adjustable springs attached to the holder. though this current density may be exceeded in the case of interpole machines. they are set radially. 54. The brush holders serve as guides for the brushes. Single brushes are used only in the case of machines of small current output. When the tangential width of the brush has been decided upon. for brush difficulty in making and maintaining a good contact over its entire contact surface. short-circuited.5 to 2 pounds will The . The carbon brush must coils are simultaneously in limit the current the short-circuited resistance to have sufficient in sparking when result below that which would a value coils to in the tangential direction is generally the short-circuits are opened. in. The graphite in the brush serves to partially lubricate the commutator. The tension of the springs is adjusted until the brush presses against the commutator with a force of from 1. this being a common cause of sparking. the rocker ring.

Dynamotor. or vice versa. but increases the sliding friction and. of contact area. results in increasing loss of power and heating of the commutator. Motor-generator. 55. a purpose motor-generator is used. this limit does Increasing the brush pressure above not materially lower the contact resistance. direct connected to each other. that the voltage of the generator end may be controlled independently of that of the motor end of the outfit. motor-generator set consists of two separate machines. a motor at some other and a generator. 54. and mounted on a common bed plate. The carbon is generally copper-plated at its outer end to insure good contact. PRINCIPLES OF DIRECT -CURRENT MACHINES in. as illustrated in Fig.68 per sq. Motor-generators are also used to convert direct current into alternating curThis type of machine has the advantage rent. higher or lower than the first. a As ordinarily constructed. It is frequently necessary to convert direct current at one voltage into direct current For this voltage. FIG. The over-all efficiency of the set is equal to the product of the efficiencies of . called a pig-tail. therefore. Brush holder and rocker ring. which tached to the outer end of the brush by means of a metal at- band clamped tightly around the carbon. The connection between the brush and the brush holder is made through a flexiis ble lead of braided copper wire. 47.

called a dynamotor. to convert the current from one voltage to another. 31).f. 55. each with its own commutator and brushes. windings. thereby causing motor action. it is combine the two into a single unit. The truth of this statement can be seen from the following reasoning: If the e. as in a motor-generator set. This type of machine is built in small sizes only. current may be introduced into one of the windings. providing the armature of such a machine with two separate possible to I FIG. while the other winding will then generate an e. It is open to the objection that the voltage at the generator terminals cannot be independently regulated.f. Motor-generator set. a By having single field structure and a single armature core.m. in general 69 The power rating of the motor must be sufficiently greater than that of the generator to allow for the losses that occur in the double transformation of the energy.THE DYNAMO the motor and generator. if there were < .m. impressed upon the motor terminals is E mj the rotation of the armature through the field flux will generate in the motor armature winding an approximately equal and opposite voltage (Art. Instead of using two separate machines. but is fixed by the voltage impressed upon the motor terminals.

the generator e. and is practically MB* : FIG. 48. Turbo -generators.m. from trouble due to armature reaction. this counter e. The high rotative speed calls for special design to .70 PRINCIPLES OF DIRECT-CURRENT MACHINES no losses in the motor. 56. is Eg or = a~ g 60 X 10 8 E (18) The disadvantage of the fixed ratio of voltage transformation is by the reduced cost of construction made possible by the armature and field structure. Generators for direct connection to steam turbines must be designed for high speed of rotation since the steam turbine develops its maximum efficiency under this condition.m.f. Turbo-generator set. Em = am X 60 X 10 8 Since the generator winding rotates through the same field as the motor winding and at the same speed. would be equal to hence by equation (17) Em . The dynamotor has in addisingle offset free tion a higher efficiency than a motor-generator.f.

Fig. 56 represents a 300-kw. manufactured by the General Electric This machine has a double commutator. 125-kw. A full discussion of the function of The interpoles. 1500-r.THE DYNAMO 71 mutation. 57. Commutator of high-speed generator. of the proper magnitude and direction to reverse the current in the short time required for the segments to pass across the brush. are small poles placed midway between the main poles they are wound with coils through which the armature current. also called auxiliary poles or commutating poles. Fig. interpoles must be deferred to a later chapter.p. idly and must be adjusted with great care. whose function it is to generate in the coils undergoing commutation an e. machine.f. is made to flow.m. and the interpoles are clearly seen between the main poles. 49. Interpoles are used in machines where sparkless . Interpole Machines. 57 shows the steel rings around the commutator of a 125-volt. 49) FIG. withstand the centrifugal forces and to provide sparkless comThe end-connections of the armature winding are held in place by metal end-shells in place of the usual banding ing and the commutator segments are prevented from springby a steel ring or rings shrunk over them.m. 125-volt. or a fractional part thereof. To provide for satisfactory commutation these machines are provided with wires. Turbo-generators require a high grade of brushes to insure The brushes wear down quite rapsatisfactory commutation.. set Company. interpoles (Art. .

In the centrifugal pump.72 PRINCIPLES OF DIRECT-CURRENT MACHINES commutation would otherwise be difficult or impossible of attainment. as in turbo-generators. on the other hand. cm. the length I is likewise limited by such mechanical features as rigidity.f. which are then rectified by the commutator. the axial length of the active part of the cylinder I cm. machine consists of a conductor so disposed in a magnetic field that the cutting of the lines of force is continuously in one direcThe armature tion. and its peripheral velocity v cm. The armature rotates in a magnetic . it is a true continuous-current machine. the developed pressure acts continuously in one direction. shown in section in Fig. freedom tion that .m. Homopolar or acyclic generator. variable speed motors. The Unipolar or Homopolar Machine. etc. the latter plays much the same part as the valves of a double-acting reciprocating pump. the generated e. 58. 50. In the type of armature described in the preceding sections. produced by the exciting winding F. Load FIG. or homopolar. this acyclic generator. or In principle.m. and the electrical analogue of the centrifugal pump is found in the so-called unipolar. will be e = Blv X 10~ 8 volts. The maximum e. obtainable with this type of machine is determined mainly by the consideraB and v may not exceed definite limits. The lines of force pass radially across the air-gap all around its periphery. and upon whose edges the two sets of brushes BI and B 2 make sliding contact. 58.. consists of a metal cylinder A of low resistance. per second.. the path of the flux being indicated by the dashed line. insulated from the shaft.f.m. If the intensity of the magnetic field in the air-gap is B lines field per sq. the individual coils have generated in them alternating e.fs. thereby obviating the necessity for the rectifying valves.

p. briefly. Unfortunately.). 1905. 2 B. At the high rate of rotation required for any reasonable value of e. 1912. such machines are called Their use is magneto-electric machines or.m. Separate Excitation.E. field excitation may be recognized: The following Separate excitation f Series excitation Self excitation \ Shunt excitation I Compound excitation field excitation 52. at 500 volts and 3000 r. per second I 60 (about sq. Field Excitation of Dynamos. Such a machine has been built by the General Electric Co. field 1 In this type of the winding is traversed by a current supplied from a source J.500 (100. Noeggerath.f.E.000 lines per ing good brush contact. . Trans. difficulty is experienced in securThus if B = 15. depends upon the motion of the armature inductors through a magnetic field. A.. E.000 ft. and v = 5000 cm.p. G. June.I.I. is types of accomplished by means of electromagnets. however. running at 1200 r. just as high pressures may be obtained with the latter by using several stages. Because of the fact that the armature consists of a single conductor of large crosssection.THE DYNAMO 73 from vibration. the magnetizing action of the large armature current when the machine is under load so weakens the field produced by the exciting coil F that the voltage drops considerably.m.f. in. and 260 volts. Trans. now confined to small machines intended for ringing call-bells ductors in in small telephone systems.5 volts. and the 2 Westinghouse Electric and Manufacturing Company has built one for 2000 kw. Jan. (about 10. The analogy between the homopolar machine and the centrif- ugal pump suggests the idea that. e = 46.E. 51. for gas-engine igniters and for testing The field excitation of all other generators and motors purposes. In every dynamo-electric machine the generation of the armature e. the machine is adapted for heavy currents at relatively low voltage.). per minute)... A..m. In the earliest types of machines this magnetic field was produced by permanent magnets.. higher voltages may be obtained with the former machine by using several inseries. Lamme. 2 ft. magnetos.E. = cm.m. etc. 1 for 300 kw.

of The use electromagnets. depending upon the connections of the field winding to the other parts of the classes series excitation. the current taken by the external circuit passes through the field 54. thus causing the excited machine. FIG. field was made to traverse the winding. was introduced by Wilde in 1862. Self-excitation. S. A represents the armature All of the of a four-pole machine. Connections of series generator. whereby the armature FIG. 60. compound excitation. and of low. machine to develop its own magnetic field. current. field structure S.74 PRINCIPLES OF DIRECT-CURRENT MACHINES itself. 59 represents the connections of such a machine. separately excited for the production of field. In Fig. Fig. winding and the armature. shunt excitation and these are circuit. in whole or in Diagram of separately part. in series. N. and N. Series Excitation. A great step in advance was made in 1867 when Werner Siemens the magnetic discovered the principle of selfexcitation. diagrammatically external to the machine The most prominent examples 53. such as a storage battery or another of this type are altergenerator. 60. 59.voltage directcertain kinds nating-current generators current generators used for electroplating. Self-excited machines may be divided into three classes. since all these parts of the circuit are The arrows indicate the direction of the current in the .

watts in this part of the circuit. or. In order that the machine self-excite. the generated e. there occurs a loss of power equal to i r f . The field winding of series machines consists of relatively few turns of coarse wire.f.m. and so on. Further. in the case of a new machine. a small current will flow.m. then. thereby devel- may oping more e. Obviously.f. a small e. Assuming.THE DYNAMO 75 case of generator action and for clockwise direction of rotation of the armature. that if the reversed the machine will refuse to " terminals are build up" as described above. it is obvious that such a machine when running on open circuit (the receiving circuit disconnected) will develop only the small e. this being determined by the degree of saturation of the field magnet and by the resist- ance of the circuit.m. This current will further excite the field structure.f. and. likewise increases. delivered by the machine to the receiver circuit must flow through the resistance. however.f. From the above description of the process of building up of a series generator. this loss must be kept as small as possible in order that the efficiency of the dynamo .f. that residual magnetism is present. it is necessary that there be some residual magnetism in the field poles due to previous operation.m. hence there can be no current in the circuit.m. For in this case the generated e.. It is important to note. This gradual increase of both e. will send a current through the circuit in such a direction as to neutralize the remanent magnetism. as the external resistance is lowered.f. will be generated when the armature is rotated. 2 r f of the field winding. and current will continue until a condition of equilibrium is reached. produced by sending current through the field winding from some suitable external source. If the field structure of such a machine is originally unmagnetized. though not in general proportionally.m. if the resistance of the circuit exceeds a the resultant flow of current may be insufficient to produce the requisite magnetizing force. upon closing the external circuit through the load. i. rotation of the armature cannot generate e. and a still greater current. critical value. Since the entire current.m.f. caused by residual magnetism. and that with increasing current. in a manner that will be discussed fully in field the chapter on operating characteristics.

unlike one of the series type. therefore. field must be kept as small as possible. Shunt Excitation.m. but only to the extent that variations of the external resistance affect the brush voltage. 61a represents the connections in a simple diagrammatic It is evident that the exciting current manner. that the external circuit is closed through a considerable resistance so that a small load current. therefore. shunt winding and field regulating rheostat is precisely the same as that of a series On open circuit. 60. now. when no current is being supplied to the receiver circuit. i. It is clear. which was originally equal to generator. but provided with a shunt field winding. . and moderate length. it is usually because of one or the other of the reasons discussed in the preceding section.f. now depends FIG. pose. if it fails to do so. at the brushes it is and upon the ohmic resistance of not dependent upon the resistance of the receiver circuit in the same sense as in the previous case. build up just as a series generator does under load conditions. 61. If the external circuit is entirely disconnected. that Supis. 61 shows the same armature and frame as Fig. Fig. it may current follows that r/ cross-section 55. develops full terminal voltage at open circuit. that a shunt generator.76 PRINCIPLES OF DIRECT-CURRENT MACHINES not be seriously impaired. is drawn from the The armature current. a shunt generator will generator. upon the e. and since the magnitude of the i is fixed by considerations of the load to be supplied. the the field winding. Connections of a shunt generator. hence the conclusion that the wire of the field winding must have large Fig. remaining connection between armature.

is the seat of a counter-generated e. ET but the armature current is i a = Ta . VI.THE DYNAMO is 77 i).m. The field current is given by is = 56. and the like. will be E is t . the shunt will field current. A complete discussion is given in Chap. i and the shunt field current i8 .m. In some of the most important applications of direct-current machinery. As the load current becomes greater and greater the terminal voltage therefore becomes less and less. and the generated e. this in turn results in a decrease of the exciting current. and the Ts loss in the winding be 2 is r s E = - 2 t . Compound Excitation. now becomes . It is clear that the drop of voltage will be minimized if the resistance of the armature is kept low. street-railway operation. for the reason that the armature. Since the center of the . it follows that r s must be as large as is feasible (in order to keep down i s and the loss of power) hence the use of wire of small cross-section and great length. it is necessary to maintain a constant difference of potential between the supply mains no matter what the load may be. The field winding of a shunt machine consists of many turns of fine wire. i a . in the case of generator action.(i rent through the ohmic resistance of the armature is to cause a + drop of terminal voltage. where Ea is 7*8 the counter e. is given by the equation ia = i + + is (19) is In the case of motor action the relation i obviously (20) = ia is It should be remembered that the armature and field currents of a shunt motor do not divide in the inverse ratio of their respective resistances. The drop of voltage under load conditions is also affected by armature reaction and by the degree of saturation of the magnetic circuit.f. for the following reason t : If the terminal voltage of the machine power is E volts. since rs E t fixed by other considerations. when running.f. The relation between the armature current i a the line current .f. i a and consequently also of the magnetic flux . such as incandescent lighting. Tji ET . and the effect of this increased curalone.m.

in accordance with Fig. the classification of under- In the short-shunt compound-wound generator the relation . with increasing current. or the voltage may remain very nearly confield (a) Short Shunt FIG. this characteristic are called compound windings.f. stant for . The shunt winding of itself would produce a " drooping" characteristic.m. combinations of shunt and series field nections made Conwindings. 62 and diagrammatic ally in Figs. 63a and 636. one in which falls the terminal e. FIG. those of Fig.78 load is PRINCIPLES OF DIRECT-CURRENT MACHINES usually at a distance from the generator. in order to compensate for the drop Field windings adapted to give of potential in the supply mains. illustrated in They are Fig. By properly proportioning them. as in a flat-compounded it may fall at a greater or lesser rate than with the rise to shunt winding alone. it follows that the potential difference between the generator terminals should rise as the external current increases. all machine or permissible values of current. as explained in the preceding section. 636 in a long-shunt winding. giving compounded machines. the voltagecurrent curve may rise. hence the resultant effect will depend upon the relative magnitudes and directions of the magnetizing actions of the two windings. in which case the machine is said to be over-compounded. Connections of a compound generator. 62. but the series wind- ing contributes field excitation which increases with increasing current. 63. 63a result in a shortshunt winding. () Long Shunt Diagrammatic scheme of connections of compound machines. that is.

Ea .f. ia . and from the dimensions of the pole core.m. line current and shunt-field cur- given by ia = i + is The terminal e. IV. of E t. E t.f.m. varying from 10 per cent. windings of shunt. (24) In designing the field 57. to 20 per cent. are related by the equation Ea = E + and the shunt-field current i. The object of the field rheostat is to permit an increase of i 8 by cutting out a part or all of the regulating resistance.. series and compound machines. Assuming that the number of ampere-turns per pole and the dimensions of the pole core are known.f. Construction of Field Windings. + z> (23) f. The resistance of the winding per pole is then .THE DYNAMO between armature current rent i. The calculation of these latter quantities . depends upon principles that are discussed in detail in Chap. thereby raising the generated e. and the generated t e. is 79 i. the selection pf the correct number of turns and the cross-section of the conductors follows from a knowledge of the number of ampere-turns per pole required to produce the flux <f>.. ^-^ = il+i* Ts fs In the long-shunt compound-wound generator these relations become = i + is E a = E + ir. the determination of the size of wire to be used is as follows Let = current in shunt winding is E = terminal voltage at no load : t er = voltage consumed in regulating rheostat.m. ia t . *>/ + tr (21) is given by (22) .

eter of the core plus four times the winding depth.) ture of the winding (75 na It = number shunt turns per pair of poles mean length of a turn of A = If area of cross-section of conductor. .. + The winding depth must not exceed a definite limit. if the pole core is circular. As a check on the calculations. lengths are expressed in feet and cross-sections in circular = 12. If the cross-section the pole core rectangular.80 PRINCIPLES OF DIRECT-CURRENT MACHINES where p = specific resistance of copper at the working temperaC. 64. . of diameter d c It ir(d c winding depth). circular mils. Hence mils. found by assuming a depth of winding of from is 3 in. in. p A = The mean length of . of either round or rectangular section. of exposed radiating surface. (6) FIG. . otherwise the heat generated in the interior of the core cannot be readily conducted to the surface. Sometimes they are wound on . will be approximately equal to the perim- () . is (26) of a turn. Shunt coils are usually made of cotton-covered wires. Ventilated field coils.6 at 75 C. 1 to lt lt. . it must be ascertained that the power lost in the coil (t*r/) does not exceed approximately two-thirds of a watt per sq.

The coils motors are usually impregnated with then taped and varnished. 64. air of series-wound railway FIG. 65. edge-wound copper strap coils. number of turns per coil is an integer. 81 are metal frames arranged to slip onto the pole cores sometimes they wound on removable winding forms. by distance pieces In order that connections may be easily made between the coils of adjoining poles. Interpole machine. 65. the coils being held in shape by suitable insulating materials and dipped in. are frequently made with a double wall to allow the circulation of between pole core and winding. so that the plus one-half. To this end a variable resist- . of compound and interpole machines are often made of copper strap. they with.THE DYNAMO . wound on edge. the terminals of the coils are brought out on opposite sides. or painted When metal frames are used. moisture-proof varnish. as shown in Fig. as shown in Fig. Field Rheostats. 58. the current in the shunt-field winding must be under control. The series coils insulating compound. To permit regulation of the voltage of shunt and compound generators. the turns being separated of insulating material.

66. For instance. Diagram of connections of field rheostat. tance of the shunt winding. QY field rheostat. for example. the . 61 and 62. If the field circuit is abruptly broken. i Field rheostats are generally arranged to be mounted on the back of the switchboard. This resistance arranged in the manner shown in brought out from the high resistance wire or ribbon composing the resistor to a series of insulated studs over which moves an adis Fig. The terminals justable contact arm. . if. where L s is the inducble. the energy may FIG.. 67 represents a made by the General Electric Company. field Fig. stored in the field is 4800 joules. FIG. is inserted in series with the shunt winding. back connection. with the regulating handle on the front of the board. taps being i i f that clockwise rotation of the regulathandle increases the resistance in circuit and so throttles the ing current in the manner of an ordinary valve. as shown diagrammatically in Figs. are always brought out in such a way . this energy will have to be dissipated in the arc formed on breaking the circuit. 67.82 PRINCIPLES OF DIRECT-CURRENT MACHINES ance. rheostat Field rheostats for machines of large capacity are made of cast-iron grids. 66. The inductance have a value of several hundred henries. In shunt and compound generators of large capacity the energy stored in the magnetic field is very considera2 t amounting to %L a i. . as shown in Fi fiS FIG. if L s = 600 and i s = 4. Field rheostat. Large field rheostat. 68.

wound generators illustrated in Figs. Polarity of Generators. In this case the arc would be very destructive. The machines are identical except that the remanent magnetism of the second is reversed with respect to that of the first. the polarity of the machine is determined by that of the remanent magnetism. respectively. In order that a self-exciting generator of any of the types already described may be operative. 59. field current is accomplished by allowing the field windings to discharge through a field discharge resistance. 69. connected in the manner shown in Fig. 69. it is necessary that there be some remanent magnetism in the field system further. For example. To obviate this danger the field current must be gradually reduced In large machines the reduction of before breaking the circuit. In other words. 1 Exciting Circuit Discharge Eesistance Carbon Break Contacts Field OOOOOOQO FIG. 70a and 706. and the average voltage induced by the collapse of the magnetic field would be L s ~ = 600 X 8 = 4800 volts.THE DYNAMO current were 83 made to disappear in one-half second. that the initial flow of current through the exciting winding have such a direction that it will strengthen the remanent field. consider the conditions existing in the shunt. Diagram of connections of field discharge resistance. With the connections as shown. is . the average rate of energy dissipation would be 4800 -f.J/ = 9600 watts. and the high induced voltage would be likely to puncture the insulation of the winding. In both cases the machine will build if the direction of rotation plockwise. but with the polarity of the terminals of the one opposite to that of the other.

Thus. There are therefore four elements which such a machine : The effect the polarity of sense of the windings of armature and pole pieces. Effect of reversal of residual magnetism. annulment of the remanent magnetism would again be the result. Left- when there is a definite between them. the armature winding is right-handed. a change any one of these four elements will cause . in the machine to lose its residual magnetism. other conditions similar role. 71). armature windings had been used remaining the same. and the order of connections of the field winding terminals to the armature terminals. that is. Finally. With a given remanent magnetism. a change in an odd number of the four elements armature. the machine will operate only relation FIG. In Figs.84 PRINCIPLES OF DIRECT-CURRENT MACPIINES counter-clockwise rotation would set up a field current which would wipe out the remanent magnetism. respectively. Assuming that the 71. it is wound around the core in the manner of a right-handed ing FIG. 70a and 706. a right-handed armature rotating clockwise in a given field flux will yield the same brush polarity es a left-handed armature rotating counter-clockwise in the same held. but a change in any two of them will not affect the operation. but with counterclockwise rotation the machines would again become self-excitif the terminals of the field winding are interchanged. 70. the direction of rotation. In general. handed ring wound conditions for operation are satisfied. If left-handed (Fig. it is clear that the direction of the winding around the poles plays a screw thread.

operates at praca sort of " Flux FIG.m. tend to reverse its direction of rotation and so buckle the connecting rod of the .f. Again. while a change in an even number of them will not affect the operation. The same types of field 60. windings and connections as are used for generators find equal conditions if Series motors.f. the shaded half of the armature circle representing a belt of current flowing into the plane of the paper and the unshaded half representing current of opposite direction. Diagrammatic sketch of series generator.. of series motor. VII). namely. just as the shunt generator terminal voltage (within limits of driven at constant speed. the shunt motor. when supplied with constant terminal e. its direction of rotation will be reversed as may be seen by applying (the left-handed) Fleming's rule. or.f. if overloaded. moved. If this machine is now connected to mains of the polarity indicated in Fig. speed characteristic of the series motor adapts this machine to street railway and hoisting service. 73. is this characteristic of variable speed at constant terminal mirror" image of the series generator charThe acteristic. e. Diagrammatic sketch FIG. fall off rapidly in speed as the load " increases. tically constant speed at all delivers a nearly constant machine capacity) when creasing load. if loads. when supplied with constant terminal e. variable voltage at constant speed. 72. Direction of Rotation of Motors.m. to put it in another way. race" as the load is re- application in the case of motors.. and is operated as a motor.THE DYNAMO will disturb 85 they were previously correct. Let Fig.m. This means that a series generator supplying a network fed by other generators may. An over- compounded generator used as a motor will rise in speed with in- supplied from constant potential mains (see Chap. 73. 72 represent diagrammatically a series machine used as a generator.

(6) a value equal to the average e.86 PRINCIPLES OF DIRECT-CURRENT MACHINES The fundamental reason for the reversal of driving engine.f. direction of rotation is that both armature current and field current reverse simultaneously.m.f.? What are the positions of the coil with respect to the direction of the field when the instantaneous e. PROBLEMS ing 50 cm. 74. one used as a generator and the other as a motor. the direction of rotation would remain unaltered. If maximum . If only one of these were reversed. the motor. polarity of the line supplya shunt motor is direction of rotation will not the ing reversed.m. 73. about an axis passing through a diagonal of the square.f. the direction of rotation will be the same as before. If the coil is in a uniform magnetic field of intensity 200 gausses. square.m. The direction of rotation of a shunt motor can be reversed by reversing the connections of either the armature winding or the field winding separately. as in Fig. Showing direction of rotation of shunt generator and motor. tion is the same in both. 100 turns is wound on a wooden frame measurrotated at a uniform speed of 1000 r. what are the maximum and average values of the generated e.p. whose direction is at right angles to the axis of rotation. FIG. In the case of two identical shunt machines. however. 74. but will continue to run as a motor. is supplied with current from mains of reversed To polarity. the direction of rotaing. Fig.m. has (a) 1. A shunt generator supplying a netnot reverse its direction of rotation if its will. A concentrated coil of The coil is its value. reverse the direction of rotation of a motor it is necessary to reverse the connections of either the armature or the field wind- not those of both. if the Further. what 2. If. mover is or disconnected shut prime down. work be affected since both field and armature current reverse simultaneously.? the square coil of Problem 1 is replaced by a circular coil of the same number of turns and a diameter such that it encloses the same area. therefore.

76 ohms per thousand feet at the working temperature.f. under standstill conditions. 18 wire which has a resistance of 7. from instant to instant. what is the brush friction loss. and a contact arc of 0. wire. Find the number of turns per coil.m. when running without load. per sq.? Problems 1 and 2. If the as in Problem number of armature conductors and the speed are the same 3. all conductors being conture. brushes. of 120 volts? 6.volt mains. what are the average of the generated e. of No. At what speed must the armature rotate to develop an e. 16 B.m. distributed sinusoidally around the periphery of the stationary armaEach of the 8 slots contains 20 conductors. by imWhat are the resistances pressing 11 volts upon the armature terminals. the field current being supplied from 115. If the total amount of wire on the armature of Problem 5 consists of 500 ft. X nected in series.. The commutator of a machine which runs at 650 r. each set consisting of three carbon of 18 in.f.5 amp. an armature current of 1.f.4 ohms. 41 has 400 conductors distributed uniformly on its periphery. what are the average and maximum values of the generated e.5 in.5 X 10 lines per pole.f. . The contact pressure is 1.? 4.m.f. and what is the counter-generated e. and S. Compare the voltage per 10 6 lines per 3.m. The alternator of Fig. and rotates in a 4-pole field structure that 6 produces a flux of 1.p. each individual brush has a width of 1. expressed in watts? 8.f. how much resistance must be put in series with the field winding to produce normal field excitation? 9.m.085 ohms per thousand feet at 75 F. Each field coil is wound on a cylindrical bobbin and has a mean diameter of 7 in.m.25 in.wound generator has a normal rating of 110 volts and 10 amp. when the machine is running under no-load conditions? 10. has a diameter There are four sets of brushes. It is found by experiment that the full-load armature current of 25 amp. which has a resistance of 4.5 Ib.m. 28 has a field flux of 4 pole. The field structure of the motor of the preceding problem has 4 poles and the winding is made of No.p. and. If the coefficient of friction of carbon on copper is 0.4 amp.3. can be made to flow through the armature. If and maximum values the speed of rotation is 900 r.m.. and they are so shaped that the flux of 4 10 6 lines per pole crosses the air-gap along uniformly distributed radial X lines. in. of the field and armature windings. 5. If the machine is to be used as a separately excited generator. A small series. The 8-pole alternator of Fig. and its field winding has a resistance of 1. A ring-wound armature like Fig..THE DYNAMO will 87 unit length of wire in be the average e.? Construct a curve showing the variation of the e. of brush contact area. A 220-volt shunt motor takes a field current of 1. what is the resistance of the armature measured between brushes? 7. 28 is provided with pole shoes that cover twothirds of the armature surface.

at the same time stretching the coil circumferentially until the spread of the coil pitch. bipolar machines the end connections run across the flat ends of the core and join conductors which are nearly diametrically In multipolar machines they join conductors separated opposite. considered as a whole. wound on the outer surface in a direction parallel to the shaft. The characteristic feature of ring windings is which do not cut that there are conducting wires inside the ring lines of force. RING ARMATURES. The ring armature is one in which a ring-shaped core is wound with a number of coils. but this feature is not essential to the definition. etc.m.f. therefore. be divided into three classes according to the shape of the core and the disposition of the winding upon it.fs. 3. and which do not. or elements. The only reason for having any openIn ing in the core at all is to permit ventilation and cooling. by an interval approximate^ equal e. 39. The drum armature may be thought of as evolved from the ring type by moving the inner connections of the winding elements to the outer surface. is connected to another active wire by means of connecting wires which do not thread through the core. classes are: 1.CHAPTER III ARMATURE WINDINGS may Types of Armatures. 2. Each active wire. 38. each of which winds in and out around the core in helical fashion. In these windings the coils are usually connected successively to each other so as to form a continuous circuit. These three 61. con- tribute to the e.m. so that the may be additive. dead wire The drum armature was developed to reduce the amount of in a ring winding. Armatures. is approximately a pole 88 . DRUM ARMATURES. in the conductors thus connected to the pole pitch. DISK ARMATURES. as in Figs.

a way 1 add together to produce the desired total e. for bothdirectand alternating-current machines. 1 CLOSED-COIL WINDINGS Ring and Drum Windings. open-coil and closed-coil windings. and can be thoroughly insulated before being Of the three types is slipped into place.m. (equation 17. because it obviates the hand winding required in ring armatures. S.fs. belong to one or the other of the two types. Disk are seldom armatures Fig. See Dynamo Electric Machinery. therefore. The use of open-coil windings is at present confined to alternating-current machines and need not. The disk revolves between a of armatures described. in a closed-coil winding. . Thompson. be considered here. An open-coil is in one conductor and tracing with winding which. lying on the outer cylindrical surof are face the core.m. the starting point will finally be reached after having passed through all. 75. 62. In designing the armature of a F^. Open-coil windings were at one time used to a large extent in direct-current series arc-lighting generators. All armature windings. the drum armature used practically to the exclusion of the others.m. so that used in modern practice. will Chap.f. they may be wound on formers. since the coils are wholly outside the core.-Ring wound armature.f. number of pairs of poles of opposite on both faces of the disk are active (see the wires signs. or some sub-multiple.ARMATURE WINDINGS The disk armature differs 89 from the other two types in that the of instead active conductors.. generator or motor. the number of armature conductors is determined by the fundamental equation 63. II). Types of Windings. starting any . The problem is then so to connect the various conductors that their individual e. such as the Brush and Thomson-Hous- ton machines. a dead-end is finally reached whereas. of the conductors. disposed radially on the flat sides of a disk. also. and in such that the winding as a whole will be at all times symmetri- cal with respect to the brushes. P. or winding jigs. for the e. " " progressively through the winding. 49).

wave winding. 76. by rolling out the cylindrical surface the armature core into a plane. Drum armature. beginning at a commutator segment. 76 of and 79. which. a ring armature. inclusive.there are shown three distinct types of a 4-pole machine having twenty-two Fig.90 PRINCIPLES OF DIRECT-CURRENT MACHINES closed-coil windings for In Figs. 76 to 79. Winding Element. It will be evident n J8. 78 Figs. It will be seen that in each case the winding consists of a number of identical elements which are Developedjtap winding. FIG. may be defined as that portion of a winding. 75 is active conductors. Drum armature. which are derived from and 77. respectively. 75. The arrangement of ^he winding of the last two is made clearer by resorting to the use FIG. 76 and in the of the 77 are case drum armatures. 64. ' \ An element Figs. 77. ^ . ends at the next commutator segment encountered in tracing through the winding. while Figs. and 77 . developed diagrams of Figs. lap winding. 76.

. 78 and 79 in the former. as a rule. the number of turns should be a minimum. railway motors for instance. 76 and 77. and wave types of closed-coil windings. Figs. an element may consist more than two active conductors. Ring. peded. to the ring. hence its self -inductance must be kept as small as possible in order that the reversal of the current may not be im- FIG. 76. . other. while in the latter they progress continuously in fashion around the periphery of the armature. Winding elements having more than one turn. frequently have as many as five turns per element. 75. 75. Small machines for relatively high voltage. Fig. and as the coefficient of self-induction increases as the square of the number of turns. only one turn per element for the purpose of improving commutation. but with two turns each. the successive elements lap back over each 65. instead of one. Every time an element passes through the neutral zone of the magnetic field the current which it has been carrying must be reversed in direction. Lap and Wave Windings. wave . Developed wave winding. The three windings of and 77 belong. for best results. The derivation of the terms lap and wave winding will be evident from an inspection of Figs. for instance. but FIG. 80 represents elements of windings similar to those of Figs. respectively. of more than one i.e. lap. in machines of large capacity there is. 80. 79. or unity.ARMATURE WINDINGS at once that of 91 turn.

. while two will suffice in the case of the wave winding. 75.f.l)st element. and the beginning of the (x + l)st element under the same pole as the beginning of the xih element. the same e.m. although the end of the xih con(x -f- lies nects to the beginning of the (x -\. but is separated from it by a double pole-pitch. the latter is not under the same pole as the beginning of the xih element. In a simple lap winding. will be generated with only half the number of conductors required by a ring or lap winding. Lap and wave windings are often referred to as parallel and series windings. These two facts in conjunction explain the reason for the use of wave windings in the case of direct-current railways motors. Furthermore. respectively. = wave winding shown in the }v 3>Zn 8 a60X10 or. where the combination of the cramped space and the moderately high voltage require a minimum number of conductors.fs.m. generated on the two sides add together. however.m. so that the e. the winding (a Other things being equal. considerations of accessibility for inspection and repairs limit the number of brush sets to two.f. each of these paths will carry one-fourth of the entire current supplied to the external circuit in the case of generator action. An examination of the directions of the current flow in Figs. 77. therefore. the end of any element. though there are four poles as in the other machines. what amounts to the same thing. and 77 will show that in the case of the first two diagrams there are four separate and distinct paths for the current through the winding (a = 4). 76. and no less important. there are only two paths through action. in a wave winding.92 PRINCIPLES OF DIRECT-CURRENT MACHINES be noted that in both lap and wave windings the two element are subjected to the influence of adjacent poles of opposite polarity. the diagrams show that four brushes are required in the cases of the ring and lap windings. or supplied from the line in the case of motor In Fig. the diagram will generate twice the e. on the other hand. say the xih. connects to the beginning of the It will sides of a coil or l)st element. of either of the other two in accordance with the fundamental equation 2).

But if brushes b and c are retained. the number of number neutral zones and consequent reversals is FIG. Since the number of permissible brush In sets may in all cases be the same as the number of poles. it will be observed that brushes . lap windings and in simple ring windings of the type shown in Fig. brush a in Fig. brushes will suffice irrespective of the number of poles. 81. and c may be omitted. as b and c (provided that a corresponding pair of negative brushes are removed at the same time). will short-circuit three elements in series when it spans two adjacent commutator segments. in which case the remaining brush. though p brushes may be used. Thus. 81. b and c are connected not only by the external conductor A but also by the short-circuited elements shown in heavy lines. will suffice to collect the current. equivalent to additional dead conductors joining the three brushes. these elements are in the neutral zone. b. in Fig. b and c may be omitted. any two of the three positive brushes a. 81. 93 Number of rent in an element Brush Sets Required. which shows a wave winding for a 6-pole machine having equal to the of poles. Fig. showing elements short-circuited by brush. for it will be observed that brushes a.f. two poles. hence conductor A and any two of the brushes a. it follows that the may be short-circuited by a brush at each such reversal. Inasmuch as the curmust undergo commutation once for each pas- element sage of the element through a neutral zone. The three elements thus short-circuited by brush a are shown in heavy lines. thirty-two armature conductors. consequently have little or no e. for example. Six-pole wave winding. therefore. 81 also makes it clear why two brushes instead of six.m. generated in them and are.ARMATURE WINDINGS 66. the number of brush sets must be equal to the number of But in wave windings. 75.

82 and those preceding it.82. etc. as indicated FIG. in Fig. to distinguish them from a simplex winding. In general. PRINCIPLES OF DIRECT-CURRENT MACHINES and which are connected together to form one terminal machine.m. of the This reduces the e. 82. Degree of Reentrancy. of to the there is independnothing multiplication ously. mutator segments be noted that the interleaving of the comcomponent windings requires the use of brushes of sufficient width to collect the current from each pair of circuits at a neutral point. together with economy in the use of material. quadruplex. 83 shows two independent It will of the . Windings of the type of Fig. in tracing through its entire closed course. and only once. Simplex and Multiplex Windings. by placing two independent windings on the same armature core. identical ring-wound machines are connected in parallel as indicated inFig. c. The arrangement illustrated in Fig. 83 are called duplex windings as distinguished from Obvithe simplex windings of Fig. in which each conductor must be passed through once. 83. windings. If conditions. both of the lap and wave varieties. prevent ent windings so as to form triplex. subjected to the magnetizing action of a single field structure. may be treated in the same way as has here been described for the case of ring windings. thereby improving commutation 67. segments bricated. 6. operate in pairs to short-circuit single elements. Armatures in parallel.the combined current output will bedouble that of either machine separately.94 a. of the Here both the winding elements and the commutator " independent windings are sandwiched" or imThe same result might also be secured by using two independent commutators.. Drum windings.f. windings of this sort are called multiplex. The same result may be two attained. of self-induction to one-third of the value that would otherwise have to be handled. one at each end.

84. there is now but a single closure. uniformly spaced. ture from brush to brush. FIG. independent windings. as in Fig.ARMATURE WINDINGS 95 windings each containing twelve elements. as in Fig. 84. but a study of the direction of current flow. and that the remaining twenty-three elements. that one of the twenty-four elements is omitted. quintuplex. but the former The doubly reentrant while the latter is singly reentrant. separately closed windings on the armature. FIGS. therefore. Suppose. degree of reentrancy of a winding is. Arnold of The Karlsruhe. is both Figs. or twenty-four in all. singly It should be understood that all of these conclusions apply with equal force to lap and wave drum windings. Instead of having two are connected alternately. just as in Fig. 83. it is possible to design windings as triplex. numerically equal to the number of independent. reentrant. 83 and 84 illustrate duplex windings. 83 and Duplex armature windings. 13 12 FIG. now. . 83. triplex. reentrant. 84. 68. each closed upon itself. Thus. The first systematic analysis of the relations to be satisfied in order that a symmetrical closed winding might result was the work of Professor E. following derivation of the fundamental formulas is based upon that of Professor Arnold. etc. 83. indicated by the arrow heads. will reveal the interesting fact that there are still four paths through the armaIn other words. triply singly reentrant. who published the result of his studies in 1891. General Considerations. the ring type having been used in the above discussion solely for the sake of simplicity.

In drum windings. in accordance with the definition of an element. the number of peripheral conductors may also be odd. but the number of elements in ring windings is usually made even. of Conductors. while drum windings have as a rule two active coil edges per element. in ring windings the number of commuta} Number equal to the number of active coil edges. (6) coil edges. and 81 are: How does one know in advance the number of coil edges to be stepped over in joining the end of one bundle of wires to the beginning of the next? Thus. and irrespective of the number of turns per element. Elements. of connection do equally well? the effect of changing the total number of coil edges from 32 to some other number? The answer to these and related questions is implicitly involved in a general equation etc. all kinds of closed windings this equation . and Commutator SegWithout regard to the number of turns per element. no matter whether S is even or odd. ring windings usually have only one active coil edge per element. integer. Elements having four active . and more particularly in simplex windings a multiple of the number of poles in order that each branch path of the armature may be at all' times identical with all of the others. ments. the order is 1-6-11-16. would not some other order And what would be covering 69. 79. 78. Further. Consequently.. while in the number of commutator segments is usually equal to half the number of coil edges. in ring windand in which S is odd. in which case the number of conductors will be even. in Fig.96 PRINCIPLES OF DIRECT-CURRENT MACHINES Probably the first questions that will present themselves to the student examining diagrams like those of Figs. there must be as many commutator segments. S as there are elements. 81. is derived in a succeeding article. S must of course be an tor segments is drum windings i i i i i i i i i i (a) FIG. ings which have one turn per element be either even or odd therefore. the number of conductors must be even. 85. but it may .

or pulley. end of No. the terminals of the elements are connected to segments which differ numerically by 5. is be designated by yi. In slotted armatures the number of slots spanned by a element is called the slot pitch. 78 be observed that the back. Winding Pitch. the winding is left-handed. The algebraic sum of the front and back pitches is a measure . Similarly. in Fig. Z. In 70. Commutator Pitch and Slot Pitch. in Fig. and that the front. where each element has four active edges. 1 is connected to the corresponding end of coil edge No. end of coil edge No. is 7.or left-handed ac- cording to whether one arrives at a segment to the right or left. 8. 85. so that the quantity S is the factor of importance. This numerical difference between the terminal segments of an differ element is called the cojnmutator pitch.ARMATURE WINDINGS 97 Since the 'number of active conductors. it will in this pitch. 78. Fig. 81. in Fig. will or 2/2. depending upon whether y\ is numerically greater or less than 7/2In other words. In Fig. 79 both front and back pitches are positive + and equal to 5. while the front pitch. Thus. wave windings are right. respectively. 79. 78. commutator in tracing through the circuit. or commutator. whose numbers by unity. y. in Fig. where p is the number of poles. 8. respectively. 3. it follows that the study of the arrangement of conductors may be reduced to one involving the order of connections of the elements to the com- mutator. the winding is right-handed if it progresses clockwise from seg- ment to segment of the On the other hand. 8 is connected The number of coil edges passed over to the front end of No. This can be accomplished in the manner indicated in Fig. the beginning and end of each element are connected to adjacent commutator segments. way is called the winding pitch. In certain drum windings it is desirable to reduce the number of commutator segments to a value smaller than that which corresponds to one segment for each pair of active coil sides. of the starting point after tracing through p/2 elements. Again. thus. the back which 5. must be a simple multiple of the number of commutator segments. if one faces the armature at the commutator end. Coil or Lap windings are right-handed or left-handed.

of the total winding. if handed. Thus. but differing from the latter by an amount which is again a measure of the field displacement. n n show that the terminals Reference to Figs. while simplex wave windings (series) have but two paths.98 PRINCIPLES OF DIRECT-CURRENT MACHINES advance or retreat per element in tracing through the In the case of lap windings. m and 81 plainly show. as Figs. m = y = 1. of each element of a winding are cont nected to segments which do not occupy corresponding positions with respect to the polar axes. 2y = ny or ( 3) y ^ ' n = 1 2 4 .). etc. obviously. 79 Figs. M> J4> integer. the double pole pitch case is S -*- ~> & or 2S/p. Expressed in terms of commutator segments. There is a field displacement between them which may be expressed in terms of the number of commutator segments. m is necessarily an may be fractional (H. while if it is negative the winding is left-handed. while in = y = 2. 83 and 84. but in wave windings determines whether the winding is right is positive the winding is rightor left-handed. 78. so that in this y = y+m m 2Sf (5) In lap windings and in ordinary ring windings. the terminals of an element are separated by an interval approximately equal to a double pole pitch. . by which the terminals fail to occupy homologous positions. 2y = while in that of yi-yz = yz 2y (1) wave windings 2y = yi + = 2y (2) the factor 2 arising from the circumstance that each element has two coil sides. ra. The sign of m m that simplex lap and ring (or parallel) windings always have as many paths through the armature as there are poles. It has already been noted 72. In cases where there are more than two coil edges per element. In wave windings. Number of Armature Paths. say n. Field Displacement. irrespective of . 78 and 79 will 71. in Fig.

This may easily be seen by referring to a simple ring winding like Fig. but it is equally true for lap and wave windings. the total number of paths must be which or is necessarily integral. doubled the number of paths. In the process of tracing through one path there will have been encountered a certain number. since it will have been observed from Figs. say a segment in contact with a negative brush. it will be noted that in tracing through the winding from any arbitrary starting point. P General Rules. one complete path will have been passed over when the successive increments of the field displacement ( m) have brought the total displacement to a value equal to a pole pitch. S' of commutator segments (not necessarily an integral number). The total dist placement is then mS' = P or o 7 = mp (6) Since S' segments have been encountered per path. 75. It remains to determine the relation ing between the number of paths and the winding and commutator pitches. that is. In the first place. 83 and 84 that the change of y from 1 to 2 changed the winding from simplex to duplex. S/p. has In the case of ordinary lap windings it now been shown v that r -*-it is (9) while in the case of y= wave windings 2/i -+ -?y 2 28 - 2 = 2Sa - do) . hence a = mp (8) m= 73.ARMATURE WINDINGS the 99 exist- number of poles. to each of which there corresponds a displacement m.

so that / therefore. the number- ing may be so arranged that the even numbers will constitute the outgoing group while the odd numbers will comprise all of the return group. Quite generally. lap-winding element always occupies the same field zone as the coil beginning. segment must path through edge. which represents a double pole pitch and expresses the fact that the terminals of a wave element having two active term edges are separated by approximately that amount. HO DD DC 79 1 necessarily even. therefore. bered. If. LAP OR PARALLEL WINDINGS. 86. An examination of the all formula . This means that even-numbered coil edges will be connected to odd-numbered coil edges at both ends. 86.100 It will PRINCIPLES OF DIRECT-CURRENT MACHINES be noticed that these two expressions for y differ by the 2/S/p. -**-- <> From this general equation there may be deduced certain convenient rules for determining the order of connections of the coil edges. 1. and in general the numerical coefficient of S/p represents the field step. 856 the term 2S/p would be replaced by 4S/p. of the elements in terms of the number of single The end of a pole pitches included between their terminals. thereby fixing the design OQQQQQQOOO I TN 135 r^ 5 (a) TN i (6) winding elements. 8 10 12 a another have return Standard numbering of coil edges. Since each coil side proceeding outwardly from (d) a commutator coil 246 FIG. has been pointed out in a previous section that the number It of the of 7 coil edges (nS) is of a drum winding then. = for lap windings. This is a general rule for front and back pitches must be odd. drum windings provided the numbering is carried out in accordance with the system indicated in Fig.shows that there are no restrictions upon the . In a winding like that of Fig. /. the coil sides are numhalf of them will bear even numbers and the other half odd numbers.

which represents a simplex wave winding for a 6-pole machine. is not unlimited as in lap windings. Since m = ~ and 7. not 2S essential that the average pitch approximate so far as mere closure is concerned.for most commercial windings number It is clear that the choice of S. and therefore of the of active conductors. respectively. of It is the connected sides will not be effectively additive. Thus. etc. purposely made As an example of these rules. In the great majority of commercial windings there are only two coil sides per element (n = 2). 2. 101 which may. . WAVE OR SERIES WINDINGS. p = 6. the would be an exaggerated form of = y. hence y = 32 + -2 . In Fig. 81. be even or odd. in a simplex lap winding the ends of an element are connected to adjacent segments. as otherwise the e. The general formula P reduces to y of this type. from which mains the further condition that both y\ and 2S> y% must not differ too greatly from the pole pitch. in a duplex winding they are separated by one segment. or 5 and 3. = ~ . for instance.ARMATURE WINDINGS number of elements.fs. and in certain so-called chord windings or is fractional pitch windings the average pitch larger or smaller than this value. 2S = 32. accordingly. it follows that in an m-plex lap winding the commutator pitch equals the degree of multiplicity. a = 2. so that yi - y* = 2y= CL 2 = p 2m it follows that the pitches must differ by twice the There redegree of multiplicity in addition to their being odd. >. 78 y Z = 22 = m = S = a - U y2 9 it p = 4 a = 2/2 4 = \ yl = = 2 yl = 7 = 5 Had the pitches been close.m. it may 2y be observed that in Fig. made but winding would chord winding.

II). the following pairs of pitch values are pos- etc. 86. a = 2. that is not a part of the winding. though values and 3 would result in a closed chord winding. Since 2/1 and 2/2 must be odd. 24 Slotl 68 Slot 2 62 64 66 68 Slot 16 Slot 17 FIG. 86c. Summar= 260. each slot being of of accommodate four coil sides. 87. it will be seen from Fig. Taking 2S = 258. The restriction upon the number of elements in wave windings frequently causes the use of "dummy coils. 2S 9fiO 4. = 5 or take y select = 5. For if the coil edges are numbered in accordance with Fig.102 PRINCIPLES OF DIRECT-CURRENT MACHINES 5%. 87 that the back ends of .2 -^-y = But since y must be an integer. p = 4. Therefore. for example. it follows that there must will satisfy the equation 258 (2S = 262 be one element. Chap. it is necessary to design a simplex 4-pole wave winding to be placed on an armature core having 65 sufficient size to slots. consisting of two conductors. it is put in simply to fill up the space in the two slots which contain only three active conductors each. and of course this value of Z must accord with the fundamental equation (17. 2/i = ?/2 = 5.2 y = Z^ = 64 or 65. whence izing. Since the pitches must approximate of 7 we must 2S = 5J/3." Suppose. 258 -). The latter value of y being impossible. and further -- - 9 = y. This means that Z = 260. the value of is 2S nearest is to 260 that impracticable because the maximum number of coil edges that can be placed in the slots is 260). Numbering of coil edges. assuming each coil edge to consist of a single conductor. Practical considerations in this particular case dictate the use of 2/1 = 65. in the manner Fig.

ARMATURE WINDINGS

103

conductors 1 and 3 may then be joined to conductors 66 and 68, respectively, thereby allowing the conductors to be insulated together in pairs and facilitating the placing of the bundles in the
slots.

In wave windings the

field

displacement

is

given by

m=

->

so that after tracing through
circuit of

p/2 elements, corresponding to one the periphery, the total displacement is X- = s Zi Z p

Therefore, in simplex windings (a = 2) the end of the p/2ih element connects to a segment adjacent to the starting segment; in duplex windings it connects to the next

commutator segments.

but one, etc. The ordinary wave wind3. SERIES -PARALLEL WINDINGS. = 2) through the armature irreing results in but two paths (a
spective of the

number

of poles.

But

it is

possible to secure

any

multiple of this number of paths by a suitable choice of S in the Wave windings having more than two paths general formula.
are called series-parallel windings. Thus, if a 6-pole armature has 154 conductors wound to form 77 elements, it may be arranged as a 4-circuit (duplex) wave winding; substituting

/

=

2,

S =

77, a

=
25.

4,

and p = 6

in the equation y

=

>

there results y
74.

=

General Rule for the Degree of Reentrancy.

If,

in the

general formula,

P
the two sides of the equation have a

common

factor

q,

we have

y
q

.
p

or y

,

.
p
is

i
really

which means that the original winding

o

of q

independent windings, each of which has S'
,

=

- elements and a
f

commutator pitch of y' the latter counted with respect to the S segments. That is, the winding will be multiplex and multiply reentrant of the qth degree in the event that y and S have a

104

PRINCIPLES OF DIRECT-CURRENT MACHINES
factor
</;

common

it

will

be singly reentrant
(/

if

y and

S

are prime

to each other.

In ordinary duplex wave windings

=

2)
2)

yfrom which
factor,
it

2S

4

2 (S

~-

follows that
also be

if

y

is

S must

even, that o _j_ 2

is,

contains 2 as a

even because

must be an integer and
is even;

p

is

always an even number.

This leads to the simple rule that a

duplex wave winding is doubly reentrant if y corollary that it is singly reentrant if y is odd.

and

to the

In triplex wave windings in which/

=

2,

2S6 . 2JS_
P p

3)

Suppose now that y contains 3 as a factor, where x is an integer; then from (13)
3z-

in

which case y

=

3z,

= S

3

Therefore, since ^- xis integral,

S must be

a multiple of
triplex

3,

hence

the winding
reentrant
if

is

triply reentrant.

Hence a
3,
3.

will be triply reentrant if y is

a multiple of

and

it will

wave winding be singly

y

is

not a multiple of

In the case of quadruplex wave windings, however, such simSuch windings plifications of the general rule are not possible.

may be singly, doubly, or quadruply reentrant. Thus, if / = 2, a = 8, and p = 6, S = 79 leads to a singly reentrant winding in which y = 25; S = 82 results in a doubly reentrant winding, y = 26;and>S = 80 gives quadruple reentrancy, y = 28. An inspection of any of the wind75. Two-layer Windings.
ings of Figs. 78, 79, 81, etc., will show that the end connections of successive conductors proceed alternately in opposite directions.
If all of the conductors lay in the same cylindrical surface, as in the case of smooth core armatures, the crossing of the end connections would make the actual winding process difficult of exe-

cution.

But where

slotted armatures are used,

if

the conductors

ARMATURE WINDINGS
are arranged in
all

105

two layers, the end connections of the upper layer proceed in one direction while the end connections of the lower layer, at the same end of the armature, may all proceed in Since the upper and lower the opposite direction, as in Fig. 88,

may

FIG. 88.

Partially

wound drum armature

(lap winding).

FIG. 89.

Samples

of

winding elements.

layers include
tively,

all the odd and even numbered coil sides, respecconductors in the top layer must connect to others in the lower layer, the transition being effected by the peculiar bend in

the coil

shown

at B, Fig. 89.

106

PRINCIPLES OF DIRECT-CURRENT MACHINES

It is easy to recognize an armature as lap or wave wound, when the conductors are made of bars or strips of copper, by observing the relative directions of the top end-connections at the two ends

(

a) Lap Wound
Armature of

(

I ) Wave Wound
Armature
in lap

FIG. 90.

Showing direction

end connections

and wave windings.

FIG. 91.

Duplex lap winding, singly reentrant. = +17, 7/2 = -13. 2/i

Z =

62,

S =

31, y

=

2,

Thus, if the top end-connections, when produced, meet at or near the center of the core, as in Fig. 90a, the winding is a lap winding; whereas if the top end-connections are parallel, as in Fig. 906, the winding is a wave winding.
of the armature.

ARMATURE WINDINGS
76.

.

107

Examples
clearly

of

Drum

Windings.

show

how

a slight change in the

Figs. 91 to 94, inclusive, number of coil edges will

change the winding from single reentrancy to multiple reentrancy. Thus, in Figs. 91 and 92, although both windings (duplex lap) have identical pitches, the former, with 62 coil edges, is singly reentrant, while the latter, with 64 coil edges, is doubly reentrant.
Electrically, however, these windings

have identical properties,

FIG. 92.

Duplex lap winding, doubly reentrant. y = +17, y z = -13.
l

Z =

64,

8 =

32, y

=

2,

except for the slight difference in e.m.f. due to the different number of active conductors. Similar remarks apply to Figs. 93

and

which show singly- and doubly-reentrant duplex wave Note that in Fig. 93 the pitches (y, yi and y z ) equal windings. but that 17, pitches of 15 would also work out correctly; and that in Fig. 94 it would also be possible to design the winding with
94,

pitches of y

= 14, y = 15, y 2 = 13. 77. Equipotential Connections. Consider
like

a

parallel-wound

armature

Fig. 95, which represents diagrammatically a

108

PRINCIPLES OF DIRECT-CURRENT MACHINES

winding for an 8-pole machine. The eight parallel paths through the armature from terminal to terminal are shown somewhat more clearly in their relations to one another in the diagram of Fig. 96. It will at once appear that if each path is to carry its proportionate share of the total armature current, each path must at all times generate the same e.m.f. and have the same
resistance as

the other paths. In any case, the current will divide between the eight paths in accordance with KirchhofTs
all

FIG. 93.

Duplex wave winding, singly reentrant.

Z =

=

y*

=

64,

S =

32, y

=

yi

17.

laws for divided

circuits,

namely:

(1)

the

summation

of all the

potential differences in each closed circuit must be zero; (2) the sum of all currents meeting at a point must be zero. If for any

reason the e.m.f. generated in one path is greater than in another, for instance, if that of circuit 3-2' is greater than that of 3-3', the brushes 2' and 3' will not have the same potential and an
equalizing current will flow in the lead joining brushes 2' and 3'. Even very small differences of potential may give rise to internal

ARMATURE WINDINGS

109

equalizing currents of large magnitude, owing to the low resistance of the circuits, so that excessive heating of the winding and sparking at the brushes may result if preventive measures are not

employed.

The causes
as follows
1.
:

of possible unequal e.m.fs. in the various paths are

The armature may not be

the pole shoes, due to natural

exactly centered with respect to irregularities in construction or to

FIG. 94.

Duplex wave winding,
y

=

16, yi

=

doubly reentrant. = 15. 17, y t

Z =

60,

S =

30,

wear of the bearings. The air-gap is consequently not uniform, and some of the poles therefore carry more flux than others,
thereby generating more e.m.f. in the coils subject to their influence than is generated in coils under. the weaker poles. This
cause

armature

importance in lap and ring windings, where each is at any one time under the influence of one pole only- in wave windings each path is simultaneously acted
is

of

circuit

110

PRINCIPLES OF DIRECT-CURRENT MACHINES
all of

upon by

the poles, hence this type of winding

is

free

from

the disturbing effect of non-uniform polar flux. 2. The poles may not all be identical in construction, so that
their fluxes

may
+

differ

even

if

joints

is uniform. Thus, the between the poles and the yoke, or between the pole cores and

the air-gap

the shoes, may not all be equally good, or the magnetizing effect of

rH
2 (TJOOOOOOOOOOOOOOOOOOO
1

OOOOOOOQOOOQOOOOOOQ 2/

^
2'

j)OQJ)QQ^!QOOOOOa(!_0(!

3'

ti
FIG. 95.

Parallel-wound armature

with equalizer connections.

FIG. 96. Diagrammatic scheme of connections of armature of Fig. 95.

the

field

coils are
3.

windings may differ, especially in cases where the field connected in parallel instead of in series. The armature circuits may be unsymmetrical, due to a
is

choice of

the

number of elements which number of paths. In this
it

not an exact multiple of

connection

should be ob-

served that in multiplex windings there is always dissym-

metry between the circuits, due to the fact that the

FIG. 97.

ments

Short-circuiting of eleof multiplex winding.

FIG. 98. Equalizer connections in parallel winding.

brushes short-circuit an unequal number of elements of the component windings (Fig. 97). This kind of asymmetry will give rise
to equalizing currents even if the individually generated e.m.fs. are equal, because the circuits have slightly different resistances.

The

equalizing currents are a source of loss because of the

When the equipotential connections were first introduced by . c potential. c. are always under the influence of corresponding parts of Should irregularities exist. and a'.ARMATURE WINDINGS extra heating caused 111 by them. etc. d'. This effect can be minimized by striving for the greatest possible degree of magnetic and electrical symmetry. it is clear from Fig.. occasionally referred to as the equalizing rings. The equalizing current is an alternating one. 99. To obviate the remaining difficulty of sparking at tions. will flow extra current of course through these connections. 98. under ideal conditions. would at all times have the f Thus. Equalizing rings of large lap-wound armature. the commutator. the points a. b. to reduce the usual number of brush Thus. relieving the brushes of the and preventing sparking. FIG. same . b'. 99. d. The equipotential connections. 98 that two brushes might be . equalizing current poles of like sign. resort is had to the use of equipotential connecwhich are low-resistance conductors joining points in the winding which. Mordey they were intended sets. are sometimes placed between the commutator and the armature core under the end connections in large drum armatures they are generally mounted on the exposed side of the core as in Fig. in Fig.

at 125 volts and 1200 r.m. a mil being the thousandth part of an inch). using the kind of diagram shown in Fig. A 6-pole drum armature has 450 coil edges. Assume that each element has two active coil edges. The armature core of a 4-pole 110-volt generator has 63 slots each of> accommodate 4 conductors. and the pole arc is 70 per cent. Show position of brushes and indicate 1. Make a drawing of a simplex wave winding for a 4-pole drum armature having 46 coil edges.p. and there are 47 commutator seg ments. 6. and indicate 3. 7. of one complete turn is 37. that are possible give front. line cuits. The brushes is that all of and being sufficient to introduce an unbalancing of the In present practice the number of brush sets is not duced when equalizing connections are used. the specific resistance of copper at the running temperature is 12 ohms per circular mil-ft. 93. 91 Use pitches of 13 and 11.. the length of armature core is 5 in.000 circular mils (a circular mil is the area of a circle 1 mil. since the commutator segments that would be touched by three of the brushes of one polarity are already connected to the fourth brush by the equalizdifficulty arising from the use of only two the armature circuits are not identically situated with respect to the line terminals. cir- re- PROBLEMS of a simplex lap winding for a 4-pole drum arma ture having 48 coil edges. back and commutator pitches and degree of reentrancy ii each case.m. of the pole pitch. up to and including triplex windings.. how many parallel paths are there through the armature? Find the type of the winding-. A 6-pole duplex wave winding has 127 elements each having two turns. Use front and back pitches of +11. Knowing that the flux density under the pole faces does not*exceed 55. Maka a drawing + direction of current flow in the winding. and the winding and commutatoi sufficient size to . how many conductors are necessary. The armature has 47 j slots. If the flux per pole is ap6 proximately 10 lines and the speed of the machine is 1300 r. The diameter of the armature is 12 in. pitches. Make a table showing the order of connections' of the coil edges of th( winding of Problem 2 if the front pitch is + 11 and the back pitch + 13 4.. in.p. Find all possible lap and wave windings. the extra resistance o the equalizing connections between the remote armature paths ing connections.112 PRINCIPLES OF DIRECT-CURRENT MACHINES expected to take care of the entire current.. what is the resistance of the armature measured betweer brushes? The mean length which has a diameter of If . using the kind of diagram shown in Fig. each containing 4 conductors.000 lines per sq. Show position of brushes direction of current flow in the winding.5 in.. and the conductors have a cross-section of 60. A 4-pole generator has a rating of 15 kw. 6. 2. and how must they be wound? State pitches.

f. is dependent upon the magnetizing effect of the field winding and the reluctance of the magnetic circuit in accordance with the relation Flux Since = \$= p a 60 m. is a function of the field excitation. case. the result teristic is Field Ampere-Turns that the no-load charac- FIG. The dashed lines in Figs. or the If the magnetic circuit reluctance. follows that Zn a 60 X 10 8 reluctance which means that of this function is curve.CHAPTER IV THE MAGNETIZATION CURVE. 100.f.f. the activity of the armature is dependent upon the magnitude of the magnetic flux. Every dynamo consists of an electrical circuit interlinked with a magnetic circuit. The armature winding is the electrical circuit in which the e.f. The Magnetization Curve.m. droops below the straight line Magnetization curve of a dynamo.. 100. as indicated in Fig. the graph called the magnetization curve. in turn. form.m. E had constant characteristic the no-load line would be a straight origin. 101 and 102 represent the mean paths of the flux in typical forms of bipolar and multipolar ma8 113 . the flux does not bear a constant ratio to the m. MAGNETIC LEAKAGE 78.m. reluctance E = it X 10 8 m.m. or the saturation no-load characteristic. through the but since the permeability of the iron of the magnetic circuit falls off as the flux increases. is produced in the case of a generator and in which the workIn either ing current produces the torque in the case of a motor. and the latter.

two pole cores. the will largely form of the magnetization curve determine its operating charac- measure Conversely. is to be used as a generator or as a motor. ing on every alternate pole would then magnetize all the poles equally. the form of the magnetization curve. consisting of a single air-gap. 102. FIG.114 chines. PRINCIPLES OF DIRECT-CURRENT MACHINES These lines are so drawn that they pass through the centers of gravity of the sections of the tubes of induction. Fig. It will be observed that a complete path or magnetic circuit. Magnetic circuits of a multipolar machine. and half of the connecting circuit through the armature core and the yoke. two pole includes the armature core. FIG. Magnetic circuit of bipolar machine. fundamental is of this curve of that the determination apparent the In the a machine case of importance. 101. therefore. two sets of teeth. a given set of specifications will in large It is. such as C'. or in terms of ampere-turns per pair of poles. one set of teeth. The magnetization curve is of great importance. that required to maintain a magnetic circuit such as abed. therefore. Field excitation is generally expressed in terms of the excitation per pole the flux in number of ampere-turns per pole. air-gaps. and the connecting yoke. one pole shoe and core. 102. The is. Fig. Whether the machine teristics. hence the excitation required to drive the flux through in A a complete magnetic circuit is the excitation per pair of poles. 102. magnetizing winding P on one pole will set up the same flux each of the paths C and C" (assuming perfect symmetry of A similar windconstruction) since these paths are in parallel. magnetizacompleted fix . two shoes.

this can be done by reducing the speed.f. Experimental determmation of magnetization curve. source. measure E dithe armature must be to without current. Source of Current .m..~r ' m. E would be directly prois portional to the speed. 103. possible that current This may is. controlled an ammeter A. culated when the dimensions of the machine it 115 can also be cal- and the nature are known. reluctance \ a 60 X 10 8 = kn X e function of field ampere(1) .m. for it flow in those elements of the arma- ture winding which are short-circuited by the brushes. the speed being kept constant throughout. It is best. field (1) E is the e. and. rectly. 103. FIG. current A by a variable resistance R and measured by The procedure then consists of varying the by means of R. it is necessary to cut down the e.m.. with a given field excitation. To reduce the disturbing effect of these short-circuit currents to a minimum. which gives rise to them. and these short-circuit currents may easily reach values of sufficient mag- nitude to react upon the flux and so affect the generated e. . it must be on open circuit..m.In the above equation generated in the armature by rotation through the flux produced by the therefore. and taking a reading of voltmeter V for each setting of A. therefore.-. as in Fig. . Experimental Determination of Magnetization Curve. current. . . to determine the .. turns clear that it is only necessary to run the machine at a constant speed n (driving it with a motor or other suitable prime mover) and to observe a series of simultaneous pairs of values of E and ampere-turns. The machine must then be separately excited during this test. supphed to the field winding from..f.THE MAGNETIZATION CURVE tion curve can be determined experimentally.. construction in its used the materials 79.. . i . '<. . current being .f. a suitable external -. Since of E it is p - Zn ^7rr. Inspection of equation (1) indicates that with a fixed value of field excitation the generated e. that is.m.f. not quite true. however.f.-.

104. the following symbols will be used: The total ampere-turns per pair of poles will then be (2) AT = ATa + 2AT + 2AT + 2AT + 2AT + AT y t a S C Since for any path x TJ * _ ~ 10 lx . and then to multiply the observed voltage by the ratio of rated speed to the speed actually used. difference between the two curves due to the hysteresis of the iron part FIG. 80.116 PRINCIPLES OF DIRECT-CURRENT MACHINES magnetization curve at a speed considerably below the rated speed. teresis Effect of hys- on magnetization of the curve. VIII. The effect of the current in the short-circuited coils of the armature is discussed in more detail in Chap. The observed readings when plotted take the The is of Fig. To analyze the method of computing the coordinates of points on the magnetization curve. hysteresis being to that property of iron (or other magnetic substance) by virtue magnetic name given of which the induced magnetism lags behind changes in the magnetizing force. 104. of the magnetization curve obtained experimentally if the exciting current in the field winding is first The form is not the same gradually increased from zero to a max- imum and then gradually this maximum back again form Field Ampere-Turns reduced from to zero. the circuit. Calculation of the Magnetization Curve.

the coordinates being plotted in terms of metric and also of English units.f. Bx each part. this value of then fixing the flux density. of ampere-turns for the length l x units). . 102.THE MAGNETIZATION CURVE (all 117 the requisite number quantities being expressed in c.'.f. c circuit. is But 10 j AT X = TT H x = 7 dtx = ampere-turns per cm. as determined by the equation \$ < = p 6n . so that circuit when the values for the various parts of the have been determined. 20 shows magnetization curves for the usual commercial materials.m. of the circuit. The flux per pole. Fig. .L + 2att. may be designated the useful flux. 25 per cent. an appreciable part of winding does not penetrate " it leaking" across from .s.S + B 2at s .f.m.AT = ata.g. say. whence AT = the Sok.m. But the entire flux produced by the magnetizing action of the field the armature.l s of at + 2at c . ranging from about a quarter of full-load value to.lt + 2at a .8H). x.l c + at y l y (3) The magnetization (B-H) curves usually plotted in magnetic materials are (the latter being simply terms of B and of Q. <. Having determined B X) at x is found from Fig. in excess of that value. the corresponding values of at read from the curve and substituted in the expression for be may AT. since it is this flux which is involved in the production of the generated e. the ex- AT The general method of calculating the coordinates of points on the magnetization curve then consists of assuming a number of values of the generated e. To each value of e. becomes pression for If English units are used. summation being extended over a complete closed such as C. thus selected there will correspond a value of TTT flux. Magnetic Leakage. E. \$. 81. Fig.= in X 60 X 10 8 . 20.Z.

The ceeding true value of v can then be calculated and the tentative computations modified in case the discrepancy is sufficiently large to warrant a readjustment.118 PRINCIPLES OF DIRECT-CURRENT MACHINES and in general between all points which have between them a difference of magnetic potential.. minimum reluctance of the main magnetic circuit and maximum reluctance of leakage paths. the Since the leak- age flux must traverse the poles and yokes. etc. accordingly. therefore. The ratio called the leakage coefficient.f. <p. and in machines of dispersion.m. v = 1 + ^ ^ increases more or less with increasing excitation. is very nearly proportional to the m. of v in accordance with previous experience. and the dimensions are themselves dependent upon << and. as the saturation of the iron is increased. this means. the coefficient oj It is always greater than unity.m. increases the total flux from \$ to pole to pole. Methods for the calculation of the coefficient v will be given section. the usual radial multipolar type ranges from about 1. also upon v. conseIt is therefore necessary in such a case to quently. practically.1 to 1. In general. The leakage being mainly in air. cannot be determined until the dimensions have been decided upon. hence the necessity of keeping down leakage as much as possible. the cross-section of these parts must be sufficiently large to carry it as well as the useful flux. prothen to the calculation of \$ and the dimensions. is larger values corresponding to small machines.25. The conditions to be satisfied to attain this end are. This "leakage flux. or preferably.. while \$ is less and less proportional to the m. machine under all conditions. The magnitude of the coefficient of dispersion for a given is not constant flux. the interpolar spaces being wide and of small section. assume a value .f." <p. a compact magnetic circuit made up of short poles. t a function of the dimensions of the machine. in a v subsequent is For the present it will suffice to state that This introduces a difficulty in a new design because the flux densities.

accurate determination of the excitation consumed in the The air- therefore. g would be very simple. the flux crossed the air-gap along radial lines. of predominant importance. as illustrated in the side and AT The flux always distributes itself in such a elevation. subtended the quantity If by the pole shoe 3 usually between 0. gap is. FIG. Two different cases arise in practice (a) smooth core armatures. angle subtended by the is pole-pitch P The angle 2-jr .THE MAGNETIZATION CURVE 82. of Calculation of Magnetization great permeability of iron as compared with air is responsible for the fact that the reluctance of the air-gap often constitutes from 70 to AMPERE-TURNS REQUIRED FOR THE AIR-GAP. the determination of Bg FIG. 106. 106. and (6) slotted armatures. as indicated in Fig. The f . In a machine having p poles. The 00 per cent. the is . Fig. The spreading of the way flux is from b to b equivalent to an increase in the length of the polar arc and an increase in the axial length from I to I'. 119 Curve. : (a) Smooth Core Armatures. of polar embrace. Details 1. the flux out spreads beyond the pole tips. mean flux density in the gap a is then b'l' B - (6) . Spread of flux at flanks of pole shoes. Fringing flux at pole tips. 105.7 of X 100 2~7 i g called the per cent.55 and 0. actually. and there is a further spreading at the flanks. 105. however. of the entire reluctance of the magnetic circuit. that the total reluctance is a minimum.

3133B fl "5" (8) For practical purposes it is sufficiently accurate to take b' as the average of the polar arc 6. Obviously A' g = bl. the inch is taken as the unit of length. . the clearance (5) being the both cases. the fact that the flux tends to tuft at the tips of the teeth. by way of the slots. be represented by l a and I. In other words. and this effect may be allowed for either by assuming d to have been increased to a larger value. Fringing flux at tooth tip. as indicated in It is clear that a given difference of magnetic potential between the pole face and the armature core will produce less flux when slots are present than when the armature sursmooth. respectively. b' . In this case the calculation of Bg is complicated by J. /' may be taken as V = I + 25 (10) and armature core (between If these lengths are not equal. 107. face is game in by assuming a contraction of the pole arc b to a smaller value. the slots increase the gap reluctance. let them heads) are the same. 107.54) = 0. or FIG. and of the arc on the armature subtended by the angle and increased by 25 on each side. it follows that g AT = all 0. and that more or less of it enters the teeth Fig.8 (2%^" X 2.120 PRINCIPLES OF DIRECT-CURRENT MACHINES in air and since B = H . then b'l' in the above equation for B g should be replaced by an area A g such that in case the axial lengths of pole shoes Ag = ~ ~ (11) where A' g is the area of the pole shoe and A" g is the area on the armature core threaded by the flux. (12) J i and / Ck (13) (b) Slotted Armatures. AT = g 0.8B g 5 If (7) dimensions being in centimeters. that is. d / _l J + 5) + 25 (9) Similarly.

to the depressions opposite the slots. there will be dips in the curve of axial flux distribution corresponding FIG. Peripheral and axial distribution of intensity. Correction to Pole Arc. /. as in Fig. The effect of the increased gap length at the pole tips is to produce a fringing flux in the inter- (a) FIG. and the fringing so that the two It at the flanks effects is tend to 83. 109. Chamfered and eccentric pole shoes. 108a. 109. improve commutation it is common to chamfer the pole tips. or to make the cylindrical surfaces of the armature and pole face eccentric. . FIGS. I. W. just as in Fig.THE MAGNETIZATION CURVE The problem gap is 121 To is further complicated by the fact that the airfrequently not of uniform length over the entire pole face. 109. there is a fringing field at the ends of the core. 108. Fig. Ordinates of this curve represent the radial shown Similarly. 1086. component of flux armature on the The periphery. has been shown by F. 109 FIG. 110. field and 110. density at corresponding points ripples at the crest of the curve are caused by the slots and teeth. Fig. and if ventilating ducts are provided. polar space. as shown by the flux distribution curve of Fig. The extra reluctance due to the ventilating ducts is equivalent to a reduction in the axial length equivalent to an increase in neutralize each other. 110.

it follows that 1 Air-gap Induction. XXXVIII. where (14) FIG. 112. Vol. World. 1901. the effect of the slots is taken into account by reducing the pole arc instead of lengthening the gap. Correction factor.122 PRINCIPLES OF DIRECT-CURRENT MACHINES Carter 1 that the presence of slots increases the effective length of the air-gap from 5 to d f . If. Elec. p. <r. 884. . 111. however. Dimensions of teeth and slots. and where t ba = = tooth pitch width of slot opening and 0" = ~ 2 4 6 8 b 10 12 14 16 18 20 "T FIG.

as in r an elementary tube of width dx and breadth b Assume that the made up of quadrants of The permeance of Fig. Vol. 114. FIGS. 114. flux. so that the equivalent axial length V = l'i + J' (17) (T hH Pole Shoe rp FIG.THE MAGNETIZATION CURVE 123 ratio Values of the factor a are plotted in Fig. Corrected Axial Length. the correction takes the . 274. 113 FIG. is 1 Die Clleiohstrommasehine. The reluctance due to the ven84. 112 using as argument can be further corrected to take account of the flux which enters the sides of the core from the flanks l' z by the dotted lines in Fig. Correction of axial length due to fringing of The value of 1' 2 may be estimated as follows: lines of force of the fringing flux are circles and of straight lines. I. as indicated form of is an additional length. 112 in terms of the bs = slot opening - d clearance fringing field at the pole tips 6. and where a is to be found from Fig. but it is generally unnecessary to introduce such refined calculations. T is equivalent to an increase but this effect is generally offset by the increased 1 gap space at the tips. 113). 113. 113. p. Arnold has given a method for computing the increased length of pole due to fringing. tilating ducts may be considered as reducing the axial length I to The in the value of : - I ' l '> = l -~t~ -r- w The length l'\ where t v is the distance between centers of the ventilating ducts and b v is the width of the duct (Fig. and 114. 2nd ed. of the pole shoes. .

and airin centimeters). When the teeth are not highly saturated their permeance is so considerable that the flux passing down the slots is relatively insignificant and may be neglected.96 to 1. If. is limits C P = But the permeance of force of length 5 I r 2 b'dx -f- I = J TTX 2. 85. or AT = g 0.3133 B" g 8" (B" ) if flux density is given in lines per sq. the corrected value of flux found Having density in the air-gap is it is \$ B = and therefore VV Q. b' and I'. The same difference of magnetic potential that maintains the flux through the teeth also produces a certain amount of flux through the slots. and air-gap in inches (6"). -b' log e . it were as- sumed that the entire flux per pole passed through the teeth .8B 5 AT = g if gap metric units are used (flux density in lines per sq.86. but in many machines the iron of the teeth is purposely worked at high flux density in order to limit the effect of armature reaction (see Chap.56logi to 5 times (l 6. Generally sufficiently accurate to take /' 2 = 1. and cross-section hence ^ and = 2j ZY = For values of r 1. +y) /' 2 (18) from 1 varies from 0. 6 + 5 ?rr TT is to be made equivalent to that of a tube b'l' 2 . and in such cases the permeance of the teeth is decreased to permeance becomes comparable with such an extent that the slot it. Ampere-turns Required for the Teeth.56. since the two paths are in parallel. then. permeance of all the tubes between the on both sides of the core. V). cm.124 PRINCIPLES OF DIRECT-CURRENT MACHINES entire end the x = and x = r. in..

the relation of the actual to the apparent density are (1) that the total flux per pole is equal to the sum of the flux in the iron of the teeth slots.X cross-section of iron = f \$a cross-section of air -. 115. t. -. and cross-section of air = bs l + + b t . shows the relation between B and at Thus. 115. for commercial sheet steel. Therefore (19) __ \$ = \$. and the ampere-turns per unit length corresponding to this apparent density might be greatly in excess because of the flatness of the magnetization The actual tooth density.9). of K it is possible to a series of simultaneous values of B t t M t t. Dimensions of teeth between laminae. cross-section of iron = b t (l n v b v )k (21) where k is the lamination factor (usually about 0. space and (2) that the magnitudes of the flux in the iron and in the air are proportional to the permeventilating ducts.n v b v + b t (l n v b v )(l k) (22) equations (21) given set of dimensions.+ ^ air \$ iron . u. The remaining . B' The conditions that determine of the true value curve at high saturation. actual tooth density B Referring to Fig. in Fig. the resultant tooth density would be higher than it is in reality.K (ZO) ir where /x is the permeability of the iron corresponding to the t . and in the air of the and insuFIG. B must therefore be distinguished from the apparent density.. From and (22) K = may be determined for a It follows that 3>air \$ \$ir on = Qiron \$> ^ron K _t*K = #J ~ x^v B t compute from this equation and B' by assuming values for B finding the corresponding values of ju from the magnetization curve of the core material. 116. and substituting in equation (23). curve For a given value t . as determined by test. t. lation ances of the respective paths.= M .THE MAGNETIZATION CURVE 125 immediately under the pole (with an allowance for the spread of the flux at the pole tips).

and the labor of preparing such curves as those of Fig.126 PRINCIPLES OF DIRECT-CURRENT MACHINES t curves give B' for various values of K. B corresponding to B' = PR. 116. 116). directly from the magnetization curve (M. as follows: From the relation possible to find t t. . B.5. cm. B from B' for any value of K. It is The above method of determining K 150 M 140 K= Q Magnetizat on Ciir a 130 120 110 100 100 200 300 400 500 600 700 800 900 1000 1100 1200 Ampere-Turns per Inch FIG. when t t Thus 0. we have R' B> R Bt R ' H (24) K provided B is expressed in lines per sq. 116 is in itself very tedious. QR is t the value of K= Bt when B' is known has the disadvantage that differ from may any of the values for which curves have been prepared. Fig.

Then. If B is in lines per sq. is. and assume for the present that B and To the left of the origin lay off are plotted to the same scale. Bf value of B' since PQ for a given lay off OS on the axis of ordinates equal to this given value and through draw a line parallel to until it intersects curve S in MN C a point P. in. ON must be made equal to 2. and lay off is any value ON equal to unity to the scale of K. suitable modifications must be made in the construction. where A' Q X ^4r IU D o = 3. PR = OS = That B't. if in H OM of K. and at in ampere-turns per in. If FIG. B is plotted to represent lines per sq.19 ~ to the scale of -DO = number of ampere-turns per inch per unit length of hori- zontal axis .DO K. 117. and at in am- pere-turns per cm. and if OR is drawn parallel to MN the There- intercept fore. per unit length to -77^ IU. OM -r = = will tan a.. the length to the scale of ON must be made equal cm. 117. cm.54 K.. Ordinate PQ is then the actual tooth density (B t) and OQ is the correspond- ing value of H. where of ampere-turns per of hori- A = number zontal axis B = number of gausses per unit length of vertical axis. same and N since magnetization curves are usually drawn in terms of B and at. Graphical relation bein- tween apparent and actual tooth duction. Since B and plotted to the H are never scale. corresponding to H = OQ. let C represent the magnetization curve plotted terms of B and H. any convenient scale to represent values of K. QR equal t H A .THE MAGNETIZATION CURVE 127 In Fig.

]16. of lines per sq. is parabolic. It does not immediately follow that AT = at -l t) because the tapering of the teeth results in an increasing density from the tip t t from point to point along the length to the root. . of the core below the roots of the teeth carries half of the useful flux per pole. nor does it follow that the value of at to be used is that corresponding to the flux If values density at the middle of the tooth. on of . Variae xci tation along tooth axis. .128 B'Q PRINCIPLES OF DIRECT-CURRENT MACHINES = number axis. *--2z: If the radial depth ofr the iron under the teeth is h. assuming that the curve mean ordinate. and the cor- responding values of at found from Fig. 101 and 102 that the iron for the Armature Core. per unit length of vertical Otherwise the construction is the same as before. in general result. by Simpson's rule. consequently at changes in value of the t tooth. being greatest near the yokes and gradually decreasing toward the gap as more and more of the leakage Shoes. the + 4a< + at. is at. in. of B t are computed for a number of points along the length of the tooth. ' . ata la (29) Ampere-turns Required for the Pole Cores and Pole from section to pole cores and pole shoes varies section. 2 whence AT = t (at t ) mean -l t (26) It is 86. whence Ba value ATa = 87. Evidently the true value of at t is the mean ordinate of the curve. Therefore. H (27) Aa = To the value of kh(l - nv bv ) (28) of ata thus determined there corresponds a ampere-turns per unit length. 118 will t. . ti FIG. a curve like Fig. 118. The flux carried by the . clear Ampere-turns Required from Figs. are plotted.

by assigning simple geometrical forms to the leakage paths. depending upon the type of machine.THE MAGNETIZATION CURVE flux is 129 shunted is off. Since the greater part of the leakage flux takes place through air. by the yoke is either equal to v\$ or Y^v <. or it is a equivalent to g t X = AT + 2AT + 2AT ampere-turns. Then. We have then. Fig. hence AT = C at e 'l c and AT = S at a -ls (31) The flux carried 88. B = c V *& and -jAC B =^~ " s V*& (30) to which values of flux density there correspond unit excitations of atc and at s . (32) y to which value there corresponds at y ampere-turns per unit length.f. acting between the tips of adjacent poles is that required to drive the useful flux across the double air-gap. the stray flux in each of them is equal to the m. two sets of teeth. the reluctance of the path in the iron may be neglected.f. paths associated with the main flux \$ but which do riot close mapped the armature. respectively. 102 the second case. divided by the reluctance. Ampere-turns Required for the Yoke. the m. (34) . usually. 101 illustrates the first case. If the leakage paths are correctly out. For instance. that the flux but it may be assumed without sensible error uniform and equal to v\$. the latter is representative of most modern machines.m. The leakage flux <p that enters into the equation includes the flux in all and originating through in the exciting winding. and the armature. and a fair degree of accuracy attained. The calculation can be simplified. and AT = y aty-ly (33) 89.m. It must be remembered also that all of the leakage paths are not acted upon by the same difference of magnetic potential. The Coefficient of Dispersion. and Fig.

Leakage between Inner Surface of Pole Shoes..2X-^. Leakage between Lateral Surfaces of Pole Shoes. = = ^X J. and the total leakage flux per (f> . by etc. pole is <p = fyi + 4<p 2 + 2^3 + 4^4 (35) FIG. <pi <p\. the lines of force are of Assume that Zi.(lengths expressed in inches) <p%.130 PRINCIPLES OF DIRECT-CURRENT MACHINES then points on adjacent pole cores that are each half way between the yoke and the shoe will have between them a difference of magnetic potential approximately equivalent to ^X ampere- turns. Let Fig. . Paths of leakage flux. made up of straight lines of length and quadrants of circles of radius x.^ =OAXh 76 s log e 2 T (37) (lengths expressed in centimeters) or s Iogi (l + ^f- ) (lengths expressed in inches) . 4*_. 119 represent a development of a portion of a multiThe leakage flux in any one pole P is represented polar machine. 119. the dashed lines 2 <^i.U (lengths expressed in centimeters) ^i or <pi (36) 3.

<ps <p^. This is approximately the case when the machine has numerous poles.(lengths expressed in centimeters) 1U Z Is (38). or <ps = 1.THE MAGNETIZATION CURVE Leakage between Inner Surfaces of Pole Cores.6X -y-^ (lengths expressed in inches) . The leakage paths may be assumed to be made up of straight lines of length Is and of quadrants of circles. then it is readily shown that _47T te lc hc -I (I'sjmax ('s/mt'n -. 131 = f -f. let (Is) min and (ls)m ax represent their minimum and maximum separations.f.m. I [\^)max \^)max or _ O <ps (I'SJmin (lengths expressed in centimeters) . OV vc ' vc I "V^max |_ (I )max _ 1 "j (Is) min (lengths expressed in inches) Leakage between Lateral Faces of Pole Cores. respectively. acting on each elementary tube of force is 47T (lengths expressed in centimeters) or (41) I I* (lengths expressed in inches) The coefficient of dispersion is f then = l+_ 1 = - 1 + - = + - (function of frame dimensions) (42) . If the poles are considerably inclined to each other. at the pole shoes and yoke.89d c (39) The above expression for p 3 is derived on the assumption that the axes ot the pole cores are parallel. If sumed bc = lc = ~ \/^ = 0. (40) o. The average m. of diameter d c they may be asto have been replaced by a square pole of equal crossIn that case section. the pole cores are round. <p.

in. the coefficient of dispersion of the machine specified in Problem . in the magnetic field produced by a skunt-field current of 18 amp. in.m.7 12 in.132 If PRINCIPLES OF DIRECT-CURRENT MACHINES the flux < the coefficient were directly proportional to the excitation X. there is no magnetic leakage. assuming that the' coefficient of dispersion is 1.375 : in. in. of winding elements 450 COMMUTATOR: Diameter 3. in. would be constant. YOKE (cast steel) : Radial depth Axial width 4 . Width Depth of slot of slot in. the ratio v is is not constant. 1. 56 1 80 . but since X includes the excitation required to drive the flux through the teeth. Assume that ARMATURE . Compute 2. 15 5 in.17. Diameter 13 in. in. 35 in. 2. POLE CORES (cast steel) Ratio of pole arc to pole pitch of pole cores Radial length of core 0. (of sheet steel) : External diameter of core Gross length of core No. hence more or less variable. of ventilating ducts Width of each duct Radial depth of core' below teeth No. ARMATURE WINDING: Type Total number of conductors Simplex lap 900 6 Conductors per slot No. It generally increases as the load on the machine increases. and these v j- are frequently highly saturated. AIR-GAP (clearance) 0. and that the rated speed is 375 r. 4 % 6% 150 0. of slots 46 12 in.p. Construct the magnetization curve of a 6-pole machine which has the following dimensions. PROBLEMS arid The shunt-field winding of a 10-pole machine has 550 turns per pole 6 produces a flux of 9 X lT) lines per pole when the exciting current ia Find the inductance of the shunt circuit and the energy stored 18 amp.

it was tacitly assumed that Under this no-load condition the the armature was currentless. Armature conductors on opposite sides of the geometrical neutral axis will then If. magnetizing action of the armature is called armature reaction. is the geometrical neutral axis.m.CHAPTER V ARMATURE REACTION 90. the winding elements lying in the neutral axis will be short-circuited by the brushes at the moment when they are not cutting lines of force. since the reluctance there than in the interpolar gap. the armature current and the field excitation exist simultaneously. If the field excitation is now removed and the armature is supplied with current from some external source. the flux in a bipolar machine. For the sake of simplicity. 121. Magnetic poles will be developed in the line of the brush axis.fs. In the foregoing dis- dynamo. Magnetizing Action of Armature. magnitude and distribution of the magnetic flux are dependent only upon the excitation due to the field winding and upon the shape and materials of the frame. due to the field excitation will be symmetrically distributed in the manner illustrated in Fig. let us examine first the conditions If the armature is currentless. and the resultant flux can then 133 . hence the short-circuit is harmless. the brush axis coincides with the geometrical neutral. But. is much less Under load conditions. The line ab. there will result a magnetic field whose distribution is approximately as shown in Fig. drawn through the center of the shaft at right angles to the polar axis. sitely directed e. Most of the flux will be concentrated in the region covered by the pole shoes. 120. be the seat of oppounder no-load conditions. cussion of the behavior of a the current in the armature conductors gives rise to an independent excitation which alters both the magnitude and dis- This tribution of the flux produced by the field winding alone. under load conditions. it is an axis of symmetry of the flux under no-load conditions.

netic field.f. Distribution of mag- armature currentless.m. the effect is the same as though the flux had been twisted or skewed in the direction of rotation in the case of the generator.m. 120..) . position a'b''. which serves equally well for the cases of generator and motor action. and this condition is. 120 and 121 as components. of course. and that these m. B and B''. FIG. not satisfied in the presence of iron cores.m. It is not exactly true that the resultant field is made up of the What actually happens separate fields of Figs. which in turn produces the resultant flux. especially if the iron is worked at a flux density at or near the knee of the magnetization curve.m.134 PRINCIPLES OF DIRECT-CURRENT MACHINES fields be thought of as compounded of the two shown separately in Figs. 122. 121. then combine to form a resultant The composition of the m. at least as a first approximation. It will be observed that in the case of the generator the field is strengthf ened at the trailing tips. is that the windings of the field structure and of the armature each produce a definite m...m. is being generated. 1 The form of the resultant field is shown in Fig. 120 and 121. the brushes (assumed in the axis ab) short-circuit elements which are cutting lines of force and in which an active e. whereas in the case of the motor the exact reverse is Moreover. and in the opposite direction in the case of the motor.fs. still Currents of large magnitude may therefore flow in such elements 1 NOTE. field magnets not excited. the neutral axis (that in which the winding elements are not cutting lines of force) has been shifted to the true. separate fields would give correct results only if the flux were at all points proportional to the m.f.f. FIG. Magnetic field due to armature current. (See 98.f. . As a to be result of the shift of the neutral axis. A and A and weakened at the leading tips.

121) moves with them in such a way that the resultant polarization of the armature coincides with the brush The net result is that the resultant field tends to skew more and more as the brushes are moved toward the neutral axis. due to conductors between b and a. Fortunately. 122. the remedy for both troubles is to shift the brushes until they are in (or near) the neutral axis. however. the former because it reduces the life of the commutator and lowers the efficiency.m. opposite in sign to that of the e. and the rupture of this circuit. Obviously. Distribution of magnetic field under load conditions. as the commutator segments pass from under the brush. field (Fig. but when the brushes are shifted. for of the Z/2 conductors in series. so that further skewing becomes insignificant even pass the neutral.f. sarily reduces the available the armature axis. and the brush axis may examine the commutation somewhat more in phenomena occurring during Commutation. those between b and b will generate an e. It is desirable at this point to . 122. Furthermore.f.ARMATURE REACTION 135 because of the low resistance of the circuit which includes the short-circuited winding elements and the brush contacts.. the piling up of the flux in the pole tips A and A' ultimately results in their saturation.f. Both of these effects are f f Indicates Current Flowing from Observer Indicates Current Flowing Toward Observer FIG. objectionable. say on the lefthand side of the armature of Fig. will cause sparking and perhaps blistering of the commutator.m. the machine will not develop its full e. 91.m. the latter because it unneces- output of the machine.

armature current commutation. b and c.m. Reversal of during FIG. Ideal variation of current in coil undergoing com- mutation. same .f. a. Now.m.136 PRINCIPLES OF DIRECT-CURRENT MACHINES detail than has yet been done.m. to the equal and opposite value existing in conductors on the right. it follows that the short-circuited coil must be under the influence of the pole in advance of the neutral axis. of a ring winding operat- ing as a generator. 124. then. the initial 124. Fig. exactly in the neutral axis.f. and that during the b position its current must change from the value existing in conductors to the left of the brush.f. the effect of selfinduction would not be opposed. 123. of self-induction could not be completed in the time T. of self-induction must be the as the final direction of the current (see coil c. This is represented are. ively. 123 represents three elements. in order to settle in a general way the conditions that must be satisfied by the brush position. Since the direction of the e. and in order to counteract this tendency. in the ideal case the current would change uniformly from the initial to the final value in exactly the time required for the element to pass under the brush. Fig. the were brushes If. respect- diagrammatically in Fig. in the direction of rotation.m. This change cannot occur instantaneously. or the period of commutation. the self-inductance of the element undergoing commuta- tion tends to keep the current at its original value and in the original direction. is short-circuited in the manner of coil b. 123).f. and the reversal of the current e. and T is the duration of the short-circuit. where + i and i and final values of the current in the element. It is evident that element a will occupy successively the positions of b and c.f. during which time the element 1 FIG. required to cancel the e. the must be balanced by an opposing e.m. no e. would be generated in the short-circuited coil.

f. OF. a g of an identical generator. Components called the angle of backward lead. will It may. is seen from this figure that the brush displacement. magnetomotive force may tion of current flow in armature . Resolving OA into the components OC and CA. a m of a motor slightly less than the displacement. 92. 122). the angle is called the angle of brush lead. 121 supplied with constant current from some external source. as in When motor. OA be represented by a line OF (the direcand field windings being taken the same as in Fig. 126. the armature m. in generators Axis of commutation and motors.m. with respect to the neushown in Fig. . The demagnetizing action of the armature is a direct consequence of the rocking of the brushes to the position most favorable for commutation. armature M.ARMATURE REACTION The tral axis. OF). If the brushes are rocked forward or backward. of the it is Imagine the brushes armature of Fig. Components of therefore. It is . 125 for both generator and motor.f. which directly opposes the main excitation. and a demagnetization due to CA.m. and it will remain constant in magnitude. l<--y | / ' / Generator 7 /^Brush Axis Brush Axis FIG. the field being unexcited. FIG. when in the direction opposite to the rotation. 125. If the fields are now excited. . their 126. in line with the brushes. Fig. of Armature Reaction. as in a of rotation. is 137 position of the axis of commutation.F. it is seen that the armature magnetizing action is equivalent to a cross-magnetization due to OC (so called because it acts across the main m.M. be represented by a line of constant length. the angular displacement of the brushes is in the direction a generator. follow.

Similarly the demagnetizing component becomes magnetif the brushes of a motor are given a forward lead. therefore.m. 2a. 127 may be considered as made up of two distinct " belts" of conductors. izing. Were it not for the fact that a negative brush lead affects commutation unfavorably.f. the armature reaction might be purposely exaggerated to such an extent as to self-excite the fields. Cross-magnetizing and Demagnetizing Ampere-turns.. 360 2 360 . when grouped in pairs in the manner indicated constitute a by the horizontal lines. but in general it requires special auxiliary devices to take care of the commutation erator for train difficulties. XI). since their effect is in direct opposition to the main exciting winding. number of turns whose magnetizing effect is directly across that of the main exciting The rewinding. This feature is taken advantage of in the Rosenberg type of genlighting (see Chap. Tho resolution of the armature m. grouped in vertical pairs. they are called the cross-magnetizing turns. 127. constitute the demagnetizing turns. FIG. but in both cases the commutator will spark viciously. that is. is in Fig. 93. maining conductors. OA. those between CB.138 PRINCIPLES OF DIRECT-CURRENT MACHINES It will be clear from Fig. into com- ponents not a strictly accurate it is proceeding. that in a bipolar machine the demagnetizing turns per pair of poles are equal to the number of armature conductors within the double angle of lead. and those between AD CA and and BD. 126. AT d. equal this number multiplied by the current per conductor. 126 that if the brushes of a generator have a backward lead. conductors. and the demagnetizing (or back) ampere-turns per pair of poles. It follows. the armature will assist in magnetizing the field. than quantitative. The former Cross-magnetizing and demagnetizing belts of conductors. namely. qualitative rather But it leads directly to the conclusion that the entire armature winding of Fig.

Reduction of demagnetizing action caused pitch of winding. however. the entire magnetizing action of the armature is across the main field. for if the brushes are originally in the geometrical neutral axis. windings are generally so shaped that the brushes are opposite the middle of the poles when the sides of the coil undergoing commutation are in the geometrical neutral. FIG. Cross-magnetizing and Demagnetizing Effect in MultiIn the foregoing discussion of the case of bipolar machines. the cross-magnetizing ampere-turns are given 139 by (2) AT = C 360 180 2 720 where ' = - 2a It is important to realize that the demagnetizing turns are a consequence of the cross-magnetizing action of the armature.ARMATURE REACTION Similarly. An extension of the principles developed for the case of the bipolar machine leads to the generalization that all of the conductors lying within the double angle of lead have a demagnetizing . 94. The shifting of the brushes to a position near the resultant neutral axis then brings the demagnetizing turns into existence. the brush axis and the axis of commutation coin- and no distinction was made between them. It must be remembered. polar Machines. a ring-wound armature was tacitly assumed. by fractional Accordingly. -128. that the end connections of lap and wave cided. thereby causing a distorted resultant field.

conductors occupying the space between pole tips carry currents which are partly in one direction and partly in the other. that is = = Z = ra = p a 4 4 80 1 15 and y 2 = 13. Demagnetizing belt of conductors in multipolar machine. On tracing through the winding a of which is shown in Fig. 129) : . thereby For example.140 effect PRINCIPLES OF DIRECT-CURRENT MACHINES upon the field. / \ FIG. found that the current in the interpolar region is alternately in Take y\ = opposite directions. 128. Omitting from consideration the special case of short-chord windings. while the remaining conductors produce a This conclusion holds accurately for all cross-magnetization. it will be portion diagram. the number of demagnetizing ampere-turns per pair of poles can be determined as follows (see Fig. assume a 4-pole simplex lap winding having 80 conductors. windings of the ring type. 129. and for lap and wave windings of full But in short-chord windings it will be found that the pitch. partially neutralizing the demagnetizing effect.

But though all of the main conductors outside of the double angle of lead contribute to this effect. be confined to the ^p. attention n r? may. hence the remainder.2A T t CL \.2a.f.p and. . I will produce a cross-field whose intensity at the tips is = where 5' Tn oTfri a 10L360 I " 2A L t 7777 (4) = d t is 'equal to the slots.V/ Cross field in multipolar machine.ARMATURE REACTION The belts 141 total number of conductors lying within the demagnetizing is ~. is ^ FIG. or ^ J-V/ l_OO\y 0-7. gap length corrected to take account of the effect of the The resultant pole tip densities will then be /D (B g D B \ c) at the I leading . .m. upon a path so largely in air that the flux due to them is negligible.m. due to these conductors gilberts.. those which are not in the angle ft subtended by the pole exert their m.f. The m. n (2A T t ) gilberts. equal to and this acts 1 J. ) \ \ ( . therefore.f. con- ductors under the pole. as in Fig. the pair of poles is number i' /2 of demagnetizing ampere-turns per AT ' 360 A " ?p p/2 it ' a T 180a (3) The cross-magnetizing ampere-turns produtee-disttfrtion of the field by strengthening the field at one tip of the pole and weakening it at the other. pole h tip of a * [generator) \ trailing J [ motor k . 130. therefore..m. the teeth consume a m. upon a path C whose reluctance is mainly due to the double air-gap and two sets of teeth. . 130.

1 field ampere-turns per pole number of (8) The factor ordinarily 1.25 to 2)B Substituting this relation in the expression for B and transposing C c. from which possible to compute the length of air-gap necesreversal of the field at the commutating tip.m. in practice. it follows that B g = (1. ] \ .6 B. lf\ to sary prevent the clearance.. 1 it is (1. it follows that B c <B g . ing the term AT we may write t . /? = 360\i/ - .m. f \ [ generator e 1 \ .65. Further.000. neglect- (1. Bg . cm.) and since B g is usually between 6000 and 10. Substituting these it will be found that relations. B g . The above formula also leads to a relation which serves as an approximate criterion for a successful machine. .B = 2000 to 3000 (lines per sq. of the original field.f.8 to 0.25 to 2) ~^= 1. of self-induction. the ratio of pole arc to pole pitch.142 PRINCIPLES OF DIRECT-CURRENT MACHINES and (Eg +B c) at the [ i trailing . the formula gives an idea of the extent of chamfer to be given to the pole tips. 5. .7 to 0.where ^. is is usually about 0. pole * tip of a .f. Thus.26 to 2) . s S 1.3' = 2ATg (6) or (7) Now. c Generally.7 in direct-current machines. and since the direction of the "commutating field" must be that Now.in order to generate in the short-circuited coil an e. [ leading J motor since commutation takes place at the weakened pole tip. . A T g will vary 0. and Z total a the from armature ampere-turns. .9.9 of the ampere-turns per pole.1 is it will an upper limit that is seldom found have a value of from 0. has been fixed. of the proper direction to balance the e. Armature ampere-turns per pole ^1.

Corrected Expression for 143 Demagnetizing Effect of Back Ampere-turns. that the reversed currents all lie in a zone to the left of A 2 the extent ot this zone depending upon the differ. which. (a) Lap Windings. the current in the coil sides immediately to the right 1 of. Fractional pitch lap winding.e. and including. is a consequence of the fact that the winding sketched in Fig.ARMATURE REACTION 95. It follows. In the second place. i. The number corresponding 2S . y\ the words "right" and "left" would have to be interchanged. AZ will all be in the same direction. the first coil is edge carrying re- versed current is (y\ -f 1). therefore. however short the chording may be. 131 ?/i y.. FIG. there will be zones of reversed currents to the left of all the coil edges. 131 is a development " chorded" (short-chord) lap winding emof that portion of a It is rebraced in a span slightly greater than the pole-pitch. In the first place. Similarly. ence between y\ and the pole pitch. to the last one in the group will evidently be-~-'=yi 1 This is If the winding were left-handed. If coil edge A i is numbered 1 . etc. or downward in the figure. are connected to commutator segments touched by the brushes. 131. the second (yi + 3). 7/2right-handed. like A i and A 2. all of the coil sides lying to the left of A 2 would carry current in the same direction (vertically upward in the figure). =yo7r^ due to the fact that the coils in the neutral zone carry currents which are not all in the same direction. it will be evident that if the winding were of quired to determine the reduction in the value of ATd 2S full-pitch. Fig. > < .

the total reduction will be that due to 2n bundles. It should be noted. extreme chording would not be used in practice. therefore vitiate the above correction. 132). ATd will be less than the com- puted value by Zb It ( tt \ p 2/1+1) ampere-turns (10) should be noted that extreme chording fall may cause some of the n reversed coil sides to lead. If 2/1 is the back pitch of the (6) Wave Windings. and mutator end). outside of the double angle of But such and. there are n of such reversed bundles. Summarizing these 2/i yi T/i ?/i 2/i + + + + + l 3 5 (2k /2 Si corresponds to the 1st corresponds to the 2nd 1) corresponds to the 3rd corresponds to the kih \ (~ - 2/1) corresponds to the nth In other words. would result in Full-pitch winding 2/2 > 2/iuniform opposition of direction of current on . obviously. 2 2/2 is the front pitch (at the com- _ 2/i_+J/2 _ 2S +a p the positive sign of a indicating that the winding is right-handed. Bi(nce each element contains ^a turns. where 2n-l= or 25 - 2/1 Since the current in each of these n coil edges balances the demagnetizing effect of the current in n bundles whose direction is normal. The extent to which 2/1 falls short of the pole pitch is then a measure of the chording. winding (at the pulley end). (Fig.144 PRINCIPLES OF DIRECT-CURRENT MACHINES results. then. hence the correction may be safely used. that neither the formula for A Td nor for the correction due to chording takes account of the number of coil edges in the neutral zone which are short-circuited by the brushes during commutation. further.

since y\ is necessarily odd. (2n + ~n FIG. y = = T/ (2n (12) is 2n y *~~ hence the number of ampere-turns to be deducted from (ATd) if y is odd. In other words. 145 in contact A2 . (13) 96. be yi 1). whence group is y y y is odd. If there are n such conductors. Due to chording. 2/S Now. the second is (yi + 3). 1. 132. the current 10 . however. which are with the brushes. the last even - 1 = 2/1 + (2n - 1) (11) if y is even. 132. the number of the nth conductor will All of these conductors bear even numbers. Fig. Fractional pitch wave winding. y - the a -. The current in the armature conductors lying to one side of the commutated coil has a direction opposite to that of the current in the conductors on the other side. reversed current (y t + 1). If but which may number or in the be taken as equal to y. number of conductors (coil sides) per pole pitch is which is a mixed number. the first conductor is which carries etc. Shape of Magnetic Field Produced by Armature Current.ARMATURE REACTION either side of coil sides like AI.

f. P.146 is PRINCIPLES OF DIRECT-CURRENT MACHINES distributed around the periphery in a series of alternately directed bands or belts.f. is -^ 2x where d the diameter of the arma- FIG. Peripheral distribution of armature m. the armature flux would all points directly proportional to the m. ture.f. it is the current reverses. acting on tube 4?r jTj P is^ r a 10 *d 2# = q - 2x where q = Z i . ia being the total ~~ ' armature current. is the number of ampere-conductors . Calculation of armature flux. to the armature current will then be represented by the ordinates This is of the broken line. them..m. the m.m. 133 which represents a development of a The peripheral distribution of m. indicated in Fig.f. 134. Since each conductor carries amperes.m. be at If and the surface the pole shoes completely surrounded the armature of the latter were smooth. due 4-pole generator. The m.m. The number of conductors rr surrounded by the elementary tube is of flux. 134.f. and is a maximum at the points where zero at the points midway between Rotation FIG.m. equal in number to the number of poles. Fig. 133. since in such a case the reluctance would be constant all around the gap (neglect- ing the reluctance of the flux paths in the iron in comparison with those in the air).

In case (6) there is distortion . F means that the flux added to the trailing tip (generator action being assumed) is exactly equal to the flux removed from the leading tip. between the pole tips but nearer the leading under the middle of the poles. 6.m.f. 97. 135. FIG. curve the resultant of F and A. 135. 136 represent the effect of the armature field modifying the magnitude and distribution of the resultant magnetic field for three positions of the brushes.m. as shown in Fig.f. curve because of the practically uniform reluctance. and curve R alone. tip. The diagrams are drawn for the cases of commutation: A (a) (6) (c) midway between pole tips. but between the pole tips the reluctance increases at a much greater rate than the m. Distribution of armature flux.85 In the usual case of machines having pole shoes separated from each other by an intervening air-space. hence the armature flux density will be small at the point midway between them. in Approximate Distribution of the Resultant Field. Under the pole shoe it will closely follow the m. In each diagram curve F shows the flux distribution due to the field excitation is is the flux curve due to the armature. the flux distribution curve is not similar in form to the curve of m.. 147 x The is flux density at a point distant from the center of the pole then xq_ (14) 0.f. hence the flux per pole remains unaltered.A RMA TVRE RE A CTION per unit length of periphery.m. a. In case (a) the distortion of the magnetic The symmetry of curve A with respect to field is clearly shown. Parts and c of Fig.

a field produced by the m. Distribution of resultant flux for various brush positions. the preceding article the shape of the resultant field R was determined on the theory that it may be considered as made up of two components: one. In case (c) there is no distortion. a field produced by the armature m. the other.f. but only demagnetization. 136. FIG. as might be expected from the fact that the brushes have been shifted to such an extent as to eliminate all the cross ampere-turns and to raise the back ampere-turns to a maximum value.m.f. As a matter of fact. as it is and clear that the flux A under a pole is more subtractive than additive. as the following illustrative analogy will show: .148 PRINCIPLES OF DIRECT-CURRENT MACHINES also a resultant demagnetization. of the field winding acting alone. In 98. this theory is not strictly acting alone. correct.m. Demagnetizing Component of Cross Magnetization.

Amp.ARMATURE REACTION 149 Imagine a rod of cast iron acted upon simultaneously by compressive and tensile stresses. though not 99.fs. m. Therefore. corresponding to zero brush lead and an entire absence of back am- m. is analogous to Hence we must conclude that stress.f. curve 137 represent the magnetization curve of Fig. The be the terminal e. at no-load . 136. therefore. OM Excitation required under load conditions. Turns per Pair of Poles under Load Conditions. hence the absurdity of the first method.f. this effect is usually it is very pronounced.m. even when commutation takes place midway between the poles. But it is quite clear from the assumed equality of the stresses that the resultant stress and. that the increased to form a single resultant. the resultant deformation are both zero.m. provided the tensile stress is beyond the elastic limit. taken in connection with diagrams a and b of Fig. the only correct procedure is first to combine the several m. at lowers the flux on that side this because of the fact that the permeability of the iron of the pole shoe and armature teeth decreases with increasing magnetizing force.m. on the side flux N M pere-turns.f. on this basis there would be a resultant elongation. and from the latter determine the and magnitude of the resultant flux. is then (AT) ampere-turns per pair of poles. Excitation Required Let FIG. the excitation required to maintain and let OE Q excitation required to generate this e. at rated load. In the case of the magnetic circuit.f. the elongation due to the tension would considerably exceed the shortening due to the compression. When the armature delivers current to the load. and flux to deformation.m.m. there is still a resultant demagnetizing action due to the cross ampere-turns. It will be clear from the above considerations.f. distribution of the pole shoe cannot raise the resultant on that side to the same extent as the diminished m. nevertheless appreciable. 137. of a generator. and suppose that these stresses are If we assume that the stresses act independently in equal.m. deforming the rod.

two sets of teeth. .. of It to an open-circuit E 2 volts. Fig. by v(AT) d that is. if necessary.f. paths. corresponding e. but be observed that the following parts are common to both paths: the double air-gap. R7 7 ' ampere-turns acting upon this path is o^-- half of which oppose B.f. the pole cores and yoke in F. C and F. 137. to take care of chording) amount to (AT) d the field excitation must be in. namely. caused by the flow through the resistance of the armature. or drop in potential. (b) If the armature demagnetiz- ing ampere-turns per pair of poles (corrected. at These two excitations act on different magnetic paths. and.m.f. creased over and above (AT). must -be greater ia ra . 138. the total generated e. the main excitation at pole tip A. Referring to Fig. the armature core The remaining parts of the two (in part). and that the path of the cross field is The number of cross as shown by the line marked C. must be greater than (AT) compensate (a) : The ohmic of the current (b) drop. under load conditions is to remain the same as at no load. and the transverse path through the pole in C. 138.150 PRINCIPLES OF DIRECT-CURRENT MACHINES in order to constant terminal e.. the other half reinforcing it. (a) If the terminal e. FIG. and the pole shoes.m. (c) The demagnetizing component the armature cross ampere-turns. than EQ by the drop is where ra the resistance of the armature.m. to the value (AT) 2 . constitute such relatively small percentages of the respective total reluctances that there is no it will appreciable error in assuming the two paths identical. has been shown that the cross field is due to the con(c) ductors lying under a pole. of Comparison of paths main and cross field. there- . the excitation required to generate EI volts is (AT). effect of the The demagnetizing armature back ampereof turns. E\.f.m.

) It follows from the fact that the magnetization curve (Fig. except at the tips of the poles. that the REDACTION 151 two excitations may be combined to determine the (It resultant excitation. the excitation must . In order that the total flux still may have the magnitude which will yield the desired e.f. 140. whence the resultant diminution of the total flux per pole will be proportional to the excess of area abd over area bee. this d' number laid off to scale ampere-turns is to be on either side of (AT) 2. and therefore. more or two that understood will be readily act unless resultant a form to they upon general be combined one and the same magnetic circuit. just as forces cannot be combined to form a resultant force unless they act upon the same body. . The curve of resultant flux distribution 'in the air-gap (R Fig. In Fig. pole of poles that weakening to as due be taken must at the effect other.m. E . 139. Fig. Thus. strengthening M @Z i - 360'fl ampere-turns at each of tip. f'(AT) 2 9 9' creased to (AT) 3 on the curve the point Excitation under load conditions. also.m. 137. produced a flux distribution as represented by curve A the ordinates of which were taken to be proportional to the m. 140. being so selected that the following condition area f'd'e'g' OM is satisfied (see Fig.m. to the abscissas with per pair ampere-turns equal 137) is drawn the and one the at effect the tip. in cannot m.ARMATURE fore. the resultant flux. let curve M.f.m.. t 136a) was obtained by assuming that the armature m. be inb' f FIG. A greater degree of accuracy in the construction of curve A is possible through the application of the theory involved in Fig. . demagnetizing The demagnetizing effect of the cross-turns is obviously represented by the difference between (AT) 3 and (AT) 2 total flux will then be the effect . therefore.fs.f. 139) : + area 6 W area a'b'd' = area/cfegr (15) The same as though there were no due to the cross-magnetizing ampere-turns.

Wrap the strip around the commutator and fasten it to the brush studs so that the commutator may rotate within it without holes binding. 136a). Ampere Turns per Pair of Poles O X FIG. the measurement of this e. The free ends of a pair of leads connected to a low- . Take a narrow strip of tough paper (sheet fiber or press-board) whose length is equal to the periphery of the commutator.f. FIG.m.m. Double pilot brush. two sets of teeth. generated in an armature conductor is proportional to the radial component of the flux density at the point occupied by the conductor at the moment in question (see Art. will provide data for the calculation of the flux distribution.wound armature (Fig. Fig. Determination of flux distribution due to armature m. curve abc will show the distribution of flux under the pole due to the armature cross magnetization for the reason that curve abc relates to that part of the magnetic circuit acted upon by the cross-turns. 32). that all the turns of the element (if there be more than one turn per element) are simultaneously in a field of the same intensity. Fig.f. and along its axis drill a series of small whose spacing is the same as that of the commutator bars. and C be that part of it required by the double air-gap.m. 141. 140. Consider the case of a simplex ring. Since the instantaneous e. in other words. 140 (see curve A'.f. Experimental Determination of Flux Distribution. 100. and the intervening portion of the armature core. When the flux per pole is 0<I>. The straight part of curve A. 136a. 141) provided with such a large number of commutator segments that the turns of each element may be assumed to be concentrated in a radial plane. should then be replaced by a curve similar in shape to abc of Fig.152 PRINCIPLES OF DIRECT-CURRENT MACHINES let be the magnetization characteristic of the machine.

101. If the pilot brush (P. it follows that the first derivative of the function which represents the potential curve will represent the curve of flux distribution. Fig. 142 the ordinates coils.m. Variation of voltage per element. e 3 . e. that the expression It is evident. of the dotted . 143. pilot FIG.. e\.154 PRINCIPLES OF DIRECT-CURRENT MACHINES In Fig. be equal to the reading of a voltmeter one of whose terminals connected to the main brush (B) and the 'other to a single FIG. a potential curve of the kind shown in full line in Fig.fs. of this curve is the sum of the ordinates curve (which is the same as that of Fig. brush separated from B by an angle equivalent to the spread of n coils. 143) is moved around the periphery of the commutator and voltmeter readings (corrected for drop of potential if current is flowing) are taken at various points. represent the e. e z . 144 will result. generated in individual therefore. Potential Curve. 142. + will is 4-6. Determination of potential curve. etc. provided the winding is divided into a large Since any ordinate. 142) lying to the left of e.

Fig. W. A. 509 (1912). . 2d ed. E. Arnold. p.m. 136).I. pp.. but all of them. Vol. Carter. Vol.ARMATURE REACTION number coil 155 of elements. is derived by assuming a simple shape of pole core and pole face that is not ordinarily used in practice. B. C. 206. 136). R.f. Trans. VIII. The change cause of 102. the slope of the potential proportional to the e. E. XV. For determining the curve of field flux distribution.E. in F. p. to be able to predetermine the curve of flux distribution due to the field excitation alone (curve ble.E. Trans. XXX. S. Vol. XVII. (A. Several methods for determining these curves have been developed. A. II. 884 (1901). p. p.m. P. XXX.E.I. XXXVIII. Predetermination of Flux Distribution in the Air-gap. p. and 1019 (1909). Electrical World. p. Part 3. Lehmann. XXXI. is curve at any point In other words. and Carter's method. Elektrotechnische Zeitschrift. Lamme.E. Vol. are approximate." Vol.E. the method J W. 461. 91 and in detail in Chap. Part I. G. Goldsborough. XVI. T. Trans. 2362 (1911). and also the curve of flux distribution due to the armature m. It is therefore occasionally desiragreater designing a new machine. 7th ed.I.. 996 F. 320. Moore. t Relation between voltage per element and potential curve. "Die Gleichstrommaschine.f. "Dynamo Electric Machinery.. explained in a preliminary manner in Art. I. except that of Carter. Vol.E. 515. p. Vol. Fig. in the distribution of the air-gap flux due to the is magnetic reaction of the armature current its effect very important be- upon the commutatmg characteristics of the as machine. Vol. p. 144. i FIG. Vol. 679. though math1 ematically correct. Thompson. A. generated in the corresponding to that point.

f. FIG.f. This leaves much to one's judg- ment. it is also true that the flux will distribute itself in such a is manner a minimum. then.m. values of B x can then be found from . for a given m. or.156 PRINCIPLES OF DIRECT-CURRENT MACHINES of the recommended by Arnold involves mapping out the paths lines of force in the air-gap and in the interpolar spaces in the manner illustrated in Fig..m. more than surface. Under the have a permeance surface will be b x /8 x . between the pole face and armature If. what is the same thing. 145. where the air-gap (5) is flux the uniform. 146. FIG. Distribution of main field. Curve of flux distribution. that one will be most nearly correct which yields the largest total flux. but some guidance is afforded by the consideration that the lines are substantially perpendicular to the surfaces of the pole faces and of the armature where they enter or leave the iron. that the total reluctance one trial is made. density (B g) will also be uniform and inversely at any other point a tube of force of length d x to 5 proportional and mean section b x (taking a unit length along the core) will . hence the flux density at the armature bx 8 6X " Br = If m. central part of the pole. 145. d 6 X \$x ax ax (16) B g is taken as 100 per cent. the total flux will be a maximum.

Curves R' and R" represent portions of similar curves for adjoining poles (of opposite polarity).F.). Thus. but the same general method is applicable. bx 157 results and ax and the computed . 148.f. The area FIG. and the ordinates are multiplied by the corresponding ordinates of the curve of armature m. 146. 147 are indicated the paths of the lines of force FIG. 136) is more difficult than in the case of the field flux. multiplied by V (the corrected length of armature core) must then equal 3>. (curve A. Magnetic lines of force due to armature current.m. the resultant values will give permeance curve P P . emanating from the armature. At any distance x from the brush the permeance of the tube of force of unit depth parallel x to the shaft will be equal to . The determination of the curves of flux distribution due to the armature m. of one loop of curve F. Fig. Calculation of flux distribution due to armature m.f. (M. R' and R" gives the desired curve. in Fig. Fig. 148. 147.ARMATURE REACTION the scaled values of d x . so that the resultant of R. F.f.M. and if Ox the peripheral distribution of of is plotted. when plotted along the developed armature surface will determine a curve like R. as curve of Fig.m.m.

f. tit. measured between brushes.) it is recommended that the mean path of the flux issuing from the axis of commutation be taken as the arc of a circle extending to the middle point of the pole core. When the machine is deliverIV. The armature resistance of the machine specified in Problem 2.m. ? in Problem 2. field intensity at the pole tips due to the cross-field of the armature of the machine referred to in Problem 1.000 lines per sq. in. Chap. a voltmeter connected to a double pilot brush spanning adjacent commutator segments gives a reading of What is the field intensity in the region occupied by the winding element that is connected to the commutator segments touched by the pilot brushes? 10.158 PRINCIPLES OF DIRECT-CURRENT MACHINES the curve of armature flux (A). amp. in the paper by Lamme (loc. magnetizing action? 4. has a brush lead of How many ampere-turns per to compensate the demagnetizing action of the armature is when pole are necessary the machine delivering 600 2.67 volts. PROBLEMS 1. what additional field excitation is necessary to compensate the demagnetizing component of crossis 3. IV.016 ohm. shown at B in the figure. The field due to the armature m. six The machine specified commutator segments. Chap. 'ing an armature current of 600 amp.. and intersecting the surfaces of armature and pole core at right angles. has greatest influence in the axis of commutation. is 0. Make the calculations on the assumption that the field excita- Compute the tion sufficient to give an average air-gap density of 50. . Under the conditions assumed in Problem 2. when it is delivering 600 amp.

motors. or in series-parallel. an individual load unit in it cannot be disconnected . and street railways operating with constant difference of potential between trolley and rail.CHAPTER VI OPERATING CHARACTERISTICS OF GENERATORS 103. and in such a system any individual load unit can be disconnected (6) 'oOOOO FIG. without interfering with the operation of the remaining units. where each lamp requires the same this current. 159 If system is to be cut out of service. Parallel. as examples of this class of service may be mentioned the use of incandescent lamps for interior lighting. 149. or other translating devices supplied with electrical energy from a distribution circuit may be connected to the supply mains in parallel. as shown diagrammatically in Fig. series (O and series-parallel circuits. and c. b. Series connection (b) is used principally in arc-lighting and in series-incandescent flighting of streets and alleys. Parallel connection (a) is used when the individual units con- stituting the load on the system are designed to operate with a constant difference of potential between their terminals. 149a. The lamps. Service Requirements. in series.

must be Ne volts and the power consumed will be Ne i watts. but the difference of potential between the supply lines will be large at the generator end (G) and will gradually decrease as the distance frorn the generator increases. are connected in series. the e.m. The parallel system requires a much greater weight of copper in the line than the series system. or much too high for safety in indoor service. each consuming approximately 50 volts. Thus. if there are lamps each requiring e volts and i amperes. since the current is everywhere the same. and gradually tapering any considerable power limits the use of series circuits to outdoor service. again neglecting the line loss. once the principles underlying the series and the parallel systems are understood. In the first case (parallel system) the line conductor must have a cross-section at the supply end capable of carrying in the line. Although the parallel system of distribution is ordinarily called the constant-potential system. the total voltage consumed by the lamps will be 6250 volts. In the second case (series system) the line conductor will be of uniform cross-section from end to end. 1496 above. the total current N supply is Ni amperes. but this disadvantage is offset by the fact that the high voltage required in a series circuit Ni amperes. If in a constant potential (parallel) system there are lamps (or other load units) each taking i amperes. it will be readily apparent of that the difference of potential between the conductors will vary more or less from point to point. A common example of series-parallel connection is found in the lighting circuits of a trolley car. adding to this the voltage consumed in overcoming the resistance of the line. in order to preserve continuity of service in the remaining units. is N off as the end of the line is approached. but must be short-circuited by a "jumper" connection. and the power consumed. becoming less as the distance . where several strings of five 110-volt lamps in series are each connected across the 550-volt supply circuit. where E is the line voltage. required at the generator will be of the order of 7000 to 8000 volts. as in Fig. neglecting the loss Ni E watts. Series-parallel distribution (c) is merely a combination of the other two.m. if 125 arc lamps. In a constant current (series) circuit. and requires no special consideration.f.f. the total impressed e.160 as in PRINCIPLES OF DIRECT-CURRENT MACHINES the parallel system.

1 full-load voltage. divided by the 100. it becomes important to sufficient extent to 104. tage at load is generally different from that at no load. plotted as independent variable. 1 and multiplied by The new (1914) Standardization Rules define percentage regulation as change in the quantity occurring between the two loads to the value of the quantity at either one or the other load. taken as the normal value. In order to make this measure a perfectly definite one. In view of the various types of service requirements described above." Inasmuch as the full-load voltage is usually considered the normal voltage. Probably the simplest way to study and compare the several kinds of machines is to construct characteristic curves which show the relations between the variables involved in the For example. so that machines of different makes and sizes may be compared. because to the resistance of the line wires. 11 . the Standardization Rules of the American Institute of Electrical Engineers define the percentage voltage regulation (or difference between the full-load simply the regulation) as the and no-load voltages. irrespective of the current in the feeder circuit. provided the voltage of the generator is raised.OPERATING CHARACTERISTICS OF GENERATORS from the generator increases. Regulation. to ohmic resistance of the 161 drop of potential due This drop of potential due may be reduced to any desired amount by increasing the cross-section of the conductor. the full difference difference is called the voltage regulation. investigate the characteristic behavior of the usual types of generators in order to determine the kind of service to which each is adapted. plotted as dependent variable. istic of a generator is a curve showing the relation bet ween terminal voltage. In the case of a generator. the voltage at the lamps may be kept constant. it would he used as the divisor in obtaining the ''the percentage ratio of the percentage regulation. The between the two values is then a measure of the closeness with which the machine regulates for constant voltage. the terminal vol105. to a compensate the drop of potential in the line. but it is clear that a If limit is set by the rapidly increasing cost of the conductors. curves are discussed in following articles. the lamps (or other translating devices) are grouped at a distance from the generator. and external (line) Other characteristic current. as the load increases. the external characteroperation of the machine. hence the definition given above. Characteristic Curves.

terminal voltage resistance of ra = armature.f. generated in the armature winding and the exciting current in the field winding is given by the magnetization curve discussed in Chap.m.m.n) plane are straight lines whose slope increases as the distance of the section from the (E.f. Characteristic Curves of Separately Excited Generator. 150. E. is Since the generated e. its regulation would be computed in like manner by dividing the difference between full-load and no-load currents by the full-load current. The following symbols will be used: E = E = t generated e. engine. including brushes and brush contacts = R = i if resistance of field winding resistance of external load circuit = = = n/ n = current taken by load current in field winding field turns per pair of poles speed of rotation.p. Effect of variation of speed stant (by keeping .m. to (E. would be similarly defined 106. r. the relation between the e. there will result a surface of the kind illustrated in Sections of the surface cut by planes parallel to the Fig. upon magnetization curve.. The speed regu- lation of a motor.m. IV (see Fig. 100). No-load Conditions. will FIG. being driven at its Under no-load conditions. if a machine is designed to regulate for constant current.. speed. v^rmied p m f hp the directly proportional then.f. in terms of speeds at full-load and at no-load. etc. 150. n) reference plane increases . the armature rated speed.162 PRINCIPLES OF DIRECT-CURRENT MACHINES Similarly. turbine. given by the equation a it 60 if X if 10 8 follows that <S> is kept conconstant). ifj and n are plotted along three axes of coordinates. If. as an arc-light generator.

in. set with a angular The line current (which is here the same as the armature current) FIG. of contact area does not exceed 45 sq.f. in Fig. that is. the field by an lead most favorable for good commutation. excitation. Load Conditions. 151.OPERATING CHARACTERISTICS OF GENERATORS 163 Sections cut by planes at first rapidly. External Characteristic. and the brushes excited constant speed. but the resistance comprises not only the resistance of the of the brushes and their connections. Connections for determining no-load characteristic. provided the current per sq.m. but varies approximately inversely as the current. then more slowly. / E = Et ira (1) In this expression ra armature winding itself.). That is. are to the (E if} plane magnetization curves corparallel to the speed represented by the distance of the responding t secant plane from the origin of coordinates. the contact resistance is not constant. the total drop of potential at the stant contact surface between commutator and brushes is approximately constant. E by ira volts. 151. and is of the order of 2 volts with ordinary grades of carbon brushes. While the resistance of the armature is con- when the steady working temperature has been reached. the latter being consumed in the internal resistance of the will t armature. amperes (or 5 to 7 amperes per . cm. including the contact resistance between commutator and brushes. and the terminal voltage E will be less than the generated e. Separate vary as the external load resistance R is varied. let the With the connec- machine be driven at its rated tions shown a current ?'/.

Construction of external characteristic of separately excited generator. on open circuit (i = 0.). (10 to 25 amperes per sq.f. If the excitation is adjusted by means of the field rheostat (Fig.m. 152. the contact drop is of the order of 0.164 PRINCIPLES OF DIRECT-CURRENT MACHINES In the case of copper brushes. will be F G Q = OP and this will also be the terminal voltage when the machine is running . External ^Characteristic FIG. 151) until it has the value represented by the abscissa OF the generated e. R = <). 152 let O'M be the magnetization curve plotted with ampere-turns per pair of poles (n f if ) as abscissas and generated e. (E) as ordinates. E = Et ira - &e (2) where is ra ' is the.f. Strictly. therefore. in.m. .04 volt with current densities ranging from 65 to 160 amperes per sq. On R the terminal voltage drops below its no-load value. In Fig. cm. constant part of the armature resistance. this correction is ignored for the sake of simplicity. and Ae In what follows the drop over the variable contact resistance. (1) because the demagnetizing action of the armature reduces the useful . of such reclosing the external circuit through a load sistance that a current of i amperes is drawn from the machine. which are used only with lowvoltage machines.

by slightly increasing the angle 0. the construction using a value of field excitation such that the point GO falls on. V) and also due to the short-circuit currents The former may be taken in the coils undergoing commutation. and draw load current (i) any value of = = an such that at Draw i. magnetizing ampere-turns per pair of poles. (E). ir a at an angle of 45 deg. shows that the corresponding terminal The locus of all such points as P.. therefore.f. the difference. Speed of Rotation on the External CharacterFor a given value of the excitation and.m. (2) because . is voltage is A'P'. Draw the line citation. into account. the change being brought about by a change in the load resistance. This method is subject to small errors because of the fact that it neglects a possible demagnetizing effect due to cross magnetization (Chap. P'. Effect of istic. and FG 0V through the origin at an angle <p such that . then the external characteristic of the machine. etc. necessary. then the intercept HD = v = de. the knee of the magnetization curve. In Fig. t tan to the scale of the figure. If the load curvoltage corresponding to the current OA rent has a value represented by OA'. OD OA so that OH angle of the AH to the scale of the figure.f. E'. Join H and F Then OF is the resultant field exand draw DF parallel to HF = OK = generated e. It will be observed that the form of the external characteristic if P P is dependent upon the form of the magnetization curve O'M. OF The student will find it very instructive to run through . a similar construction. or . 152 let OA represent armature. and if <p = ra The intercept AS will AS is deducted from OK KP parallel to 0V). or slightly below. 165 and therefore drop in the also the generated e. then equal ira (by drawing be the terminal indicated in the figure. 107. to the scale of volts.OPERATING CHARACTERISTICS OF GENERATORS flux <f>. and <p. will = i.m.AP. There will be a different as well as upon the angles characteristic corresponding to each setting of the field excitation. The line 0V whose ordinates represent the ohmic drop in the armature is sometimes called the loss line. of the .

i . 153. and that the field excitation is so adjusted that the curThe I.166 flux. the curve becomes the familiar no- Now load characteristic or magnetization curve. The latter figure also curve of Fig.f. Fig. will be proportional to the speed. rent is maintained at its normal full-load value. By a load characteristic is meant a curve showing the relation between terminal voltage (as ordinates) and field excitation (as abscissas). Effect of variation of speed upon external characteristic of separately excited generator. If this load current happens to be zero. PRINCIPLES OF DIRECT-CURRENT MACHINES the generated e. case the machine must be replaced by a new curve whose same relation to those of the original curve that the new speed bears to the rated speed. Load Characteristic. OM. OP P. circuit external suppose that the resistance R (Fig. 151) of the is varied. 153.m. subject to the condition that the current supplied to the load is constant. shows that the locus of the external characteristics for varying values of speed (but with a fixed value of field-exciting current) a wedge-shaped surface. This is shown in the three-dimensional (diagram of Fig. 152 O'M ordinates bear the Po FIG. In is operated at a speed other than the rated speed. 154. is 108.

especially if the iron is saturated. for creasing excitation the de- component of cross-magnetization becomes magnetizing greater.f. n r t i i . required DF. it is clear that this feature will contribute to a further increase in FP. It is evident from Fig.OPERATING CHARACTERISTICS OF GENERATORS 167 terminal voltage will then be E t = IR. Finally.m. Const ruction of load characseparately excited generator. the load characteristic but no-load characteristic. but under the assumed condiOA by an amount = e. as by the constant length shown by the series of is not with in- ad Characteristic shaded triangles.f. but that r. 154 that if the terminal voltage is to be maintained constant for all values of the load current. represented in Fig. . equation (34). ~o eld Amp-Turns c G - K FlG 154. would OC.m. an increase in the current from = to i = I requires an increase of excitation from OC to OK . . remembering that the cient of dispersion. b should increase in magnitude as the curve rises. Further. Chap. where require an excitation tions the generated e. in other words. shifted downward and to the right remain constant in magnitude as is the same in shape as the DP. At no-load this terminal e. v. the excitation must be increased as the load increases. 109. the field excitation must be still further increased to OK. AB = must be greater than t GK = An ordinate through v(AT) d on the load characteristic corre- K then intersects A B (extended) in the point lies P. DF and GK = FP long as i = I. This construction strictly accurate. to yield a terminal voltage E = OA is DG. The Armature Characteristic. teristic. DF Ir that the where to be generated a is.m. IV). because of armature demagnetizing effect. coeffiit in- creases with increasing excitation (X increasing. corresponding to an excitation OG. is itself not constant. which point accordingly Since sponding to full-load current.f. 154 by the ordinate OA. i Thus.

155.m. curve II will result. 156. to the ordinates of curve III there be added the ordinates of the loss line. it follows that an increase of load causes an increase of excitation and therefore also of generated e. including brushes i ra = = and brush contacts = = n/ R = = 77 resistance of series field winding series field turns per pair of poles resistance of external circuit speed in r. Curves of the Series Generator. the field (Fig. Characteristic Characteristic. showing the relation between field excitation (as ordinate) and load current (as abscissa). 157. Connections of series of generator. then. External Let E = E = t generated e. determination ternal characteristic. terminal voltage current in the circuit resistance of armature winding. 155) though a better name would be Terminal Voltage = Constant Load Current vm/mwm FIG. (i) Since the same current winding. ex- 110. Fig. showing the relation .. under the condition of constant terminal voltage is commonly called the armature characteristic " regulation curve.p. The external characteristic will have the form of curve III." (Fig. If.f. Curve II is the internal characteristic. and the load circuit flows through the armature. or regulation curve.m.m.f. the speed of rotation being kept constant at its rated value. Armature characteristic FIG.168 PRINCIPLES OF DIRECT-CURRENT MACHINES The curve in order to maintain a terminal voltage equal to OA. 156).

current. .. and the length FB is equal to ira to the These two lengths. This leads to the following simple construction for obtaining the external characteristic from the given magnetization curve of the machine. will remain constant. is required to balance the demagnetizing component of armature reaction. curve II would be the magnetization curve of the machine. be seen that the angle size of the tri- PG and PGD must be altered from point DG are both proportional to the of to point. is then responsible for the generated e. draw 1 This I.m. AG. such as D'. therefore also the line OB. II The remainder. 157.f.. but in the actual machine. where armature reaction exists. a part. this process to find other points.ge e. Vol. 155. To find a point P' locate the point B. i.m.f. of the excitation OA required to produce terminal AP and generated volta. DG. scale of volts.m.f. on the external characteristic corresponding to current OA''. 1 i = OA The demagnetizing is given by effect AF aZi 180a corresponding to any current 1 " '' nf to the scale of current. the magnetization curve (I) is displaced from curve II in the manner indicated in the figure. it will Repeating O FIG. is the same construction given in Arnold's Die Gleichstrommaschine. Thus. 169 the internally generated This follows beoausfi. ir e. when laid off as in Fig. or OF. and the armature E = E If t (4) there were no armature reaction. A' F A ^ External characteristic of series generator. hence D is a point on the magnetization curve.OPERATING CHARACTERISTICS OF GENERATORS between current. But since their ratio. and consequently the slope DP.

Condition for Stable Operation. 158. surface with the base (n. 156. Corresponding to each magnetization curve there will be an external characteristic constructed as in Fig. it will usually be found that when the machine Referring to Fig. is driven at its . as OAoP.170 PRINCIPLES OF DIRECT-CURRENT MACHINES f A'B at an angle = tan" 1 . in Fig. tically until it intersects curve M in D to B'A' until P'. Thus. i) plane is a curve OA which shows the relation between speed and current when the machine is shortcircuited (E = t 0). 157. 158. 150 for the case of the separately excited generator. and the locus of all such external characteristics will be a surface indicated by the heavy lines. through B' draw B'D' verf . magnetization curves for various values of speed laid off along the speed axis n. the surface bounded by OD'D is the locus of the FIG. The intersection of this . Effect of variation of speed upon external characteristic of series generator. Variation of the speed of a series generator affects the magnetization curve in exactly the same manner illustrated in Fig. it intersects the ordinate through and draw D'P' parallel A' in the point 111. 112. Dependence of the Form of the Characteristic upon Speed.

R evidence of instability. external circuit. When R = R 2 . is It evident from Fig. tion of series generator. and these lines are characteristic of the Condition for stable operaFIG. the point of equilibrium will be at Thus." When the load resistance has been lowered slightly below this critical value. values of E t and i Since the points representing simultaneous must satisfy the characteristics of both gen- their intersection. . by voltage drop become greater. but there is no further . or Et R\ E = t iR of This is the line equation a straight through of the origin. Thus. Further reduction of will cause the current and e. 159. 156 that Ohm's law must hold for the external circuit.f to change. as Ri. correspond to OR 2 OR . The reason for this behavior will be evident from a consideration of Fig. being proportional R. or to build up. erator and external circuit. in which the curve represents the external characteristic of the generator.OPERATING CHARACTERISTICS OF GENERATORS 171 rated speed there is a critical value of the load resistance R above " which the machine will fail to generate. at points P 3 and 4 respectively.m. ORi. the terminal voltage and current will be represented by the coordinates of point PI.. but between the initial and final condition the machine is in a state of unstable electrical equilibrium. the line OR% coincides for a greater or lesser range with the external characteristic. 113. etc. P" 2 hence there is unstable equilibrium between the points P' 2 and Values of R such as R 3 and R^ will give stable equilibrium . P Inspection of the con- struction of Fig. . successively smaller values of R. the terminal voltage and current will at first rise rapidly and then more slowly until a condition of equilibrium is reached. 157 will series generator will action and internal show that the external characteristic of a droop more and more as the armature reIn fact. Regulation for Constant Current. when the external resistance is high. the slope the line to 3. 159.

thereby cutting down the generated e. 160.172 PRINCIPLES OF DIRECT-CURRENT MACHINES purposely exaggerating^. the external characteristic vertical line is in to be the AY. It is clear that a portion of the exciting current is then by-passed around the winding. and some of the older types of Brush arc-light generators were designed on this principle.f. and to the same scale lay off equal to any arbitrarily selected value of the resistance (r) of BC BC = the regulating shunt. Regulation for constant current. in which case the machine has an in- herent tendency to regulate for constant Such a characteristic is desirable current.m. Lay off AB to any convenient scale to the resistance (r/) equal FIG. for series arc circuits. as in Fig. in a simple 161. 160. Another method of regulating a FIG. 161. the characteristic may be made to bend over to such an extent as to become nearly r vertical. series Diagram showing connections of constant current generator. will AB will be fixed in magnitude while be variable. graphical solution. other words. Vertically upward from A set off AD = irf . The relation between the resistance of the regulating shunt and the corresponding terminal voltage of the machine is shown in Fig. with abscissas equal to the current in the field winding (instead let of ampere-turns). Every variation of the resistance of the load must then be accompanied by a corresponding change of the regulating shunt. of the series field winding. generator to make it deliver a constant current is to provide a variable resistance shunted around the series-field winding. he magnitudes of these two quantities. and OA is = i be the constant curthat the to rent machine required develop. manner Let OM be the magneti- zation curve.

Open-circuit Conditions. Characteristics of the Shunt Generator. HG = i Joining G and 0. . It will be evident from the above discussion that the voltage of a series generator can also be controlled by varying the position of the brushes. found by joining Finally. The total current.m.: if LM MP is the ohmic drop in the machine. magnetizing current of the armature. length KL 114. proceed as follows: Connect C and D and draw BF parallel to CD-. terminal voltage armature current external or line current shunt-field current E = E = t ia i is ra = = = = armature resistance . M and is N and drawing Other FP AQ the terminal voltage points corresponding to the shunt resistance r = BC.f. is drawn straight line and = i into two parts. and downward from A = constant.m. that if C and G are joined by a parallel to CG. and drawing HK OK will be the current through the series-field = v ^7. point parallel to P is MN. then DF = Thereand NF = i(r a + 7~r) i^r^r fore. such BH AG AH = parallel to winding.f. also off 173 set DN i AG = OA = = ir a = constant.OPERATING CHARACTERISTICS OF GENERATORS constant. OL will be the net excitation of the machine. may be found by exactly similar construction. thereby changing angle a and affecting the in Fig. therefore. or the equivalent deSetting off KL and GO. To find MP graphically. will divide between r/ and r in such a man- ner that ir + in = TJ ? i (5) and ^ ITS - (6) Hence. and the corresponding generated e. Let generated e. or ilra t -\ -- ). point ir H will divide r/ . 161. the terminal voltage will be E = LP = AQ. i.

the armature and shunt field constitute a simple series circuit identical with that of Fig. There is. however. 162. means t Since i8 = -p .p.m. Connections of shunt generator. therefore the observed readings of E t under the conditions assumed will not differ appreciably from E. The External Characteristic.f. 159. FIG. ia since both and ra are small. nal circuit open Thus. It is therefore easily seen that variation of the shunt-field rheostat will give rise to changes in and i a in the manner already discussed in the E t case of the series generator. Determination of external that the high resistance of the shunt-field winding will limit the flow of current (i a ) to values that are small compared with the current carrying capacity of the armature. The form of the external char- . 156.f. generally there will be a region of unstable equilibrium in the building up of the generated e. 162. hence the relation between E and i s will be closely represented by the magnetization curve of the machine. with the exter- E = E + t 2> a = E t (7) rent in the armature Moreover. this difference. fa the effect upon the terminal voltage of changing the value of r a will be given by a construction identical with that of Fig. the total generated e. as in Fig. When the load or receiver circuit of a shunt generator is disconnected. characteristic. including regulating rheostat R = resistance of external load circuit of poles n a = field turns per pair n = speed in r.174 PRINCIPLES OF DIRECT-CURRENT MACHINES re = shunt-neld resistance.m. the small flow of curnegligibly small armature reaction.m.

Suppose that the load has been increased to such a value that 1 Franklin and Esty. from equation (8) ' . R being so adjusted that a moderate current will flow.m. I. even were the excita- Now let tion to remain constant.e. it corre- fl>s sponds to the ORi. (12) the shunt machine. 162 evident that ' E = E + t iara / (8) = i + i. therefore. as in the separately excited generator. a decrease of terminal accompanied by a proportionate decrease of excitation. ORz. .OPERATING CHARACTERISTICS OF GENERATORS acteristic line t 175 current showing the relation between terminal voltage E and i can be determined by the following method: it is 1 From Fig. of Fig. The relation . etc. and the field excitan i s s tion. of.-.'*.m. M.f. Then. (ampere-turns per pair of poles). the terminal e. (9) (10) between the generated e. would fall because of the ohmic drop in the armature winding.f. 163. rs = - hence. the flux and the generated e.m. open circuit (R <). neglecting the small drop (i s r a ) in the armature. other things being equal. 159. let the resistance of the field -rheostat be . are reduced. proportional to r s . is given by the magIf the machine is running on netization curve.m. that is. thereby causing It is clear. the external circuit be closed. then being F QL] this will then be nearly equal to the terminal voltage on open circuit. = But. Vol. and that the t ...f E.f. Fig. The line ON is then the "field resistance" 77f line.>. in the case voltage since is is -p -- -. its slope being lines 4 = Tls^s J . that the greater a further reduction of E the load current the less will be the terminal voltage. when there is an appreciable load current flow- ing. drop of terminal voltage will be greater in the shunt machine than in the separately excited machine. Elements of Electrical Engineering. i. so adjusted that the excitation is represented by OF the generated e.

the ordinate FG would be the total generThrough then OF . and through K draw . Through B draw the horizontal line BK. and the generated E = BD ia r a = E . and. point B is found by drawing through C a line ON CB making the fixed angle <p with the horizontal.f. since iar a = OF. In other words. there- = E .f. CG would represent to scale the value of ia ra But since armature reaction does exist. the line the terminal voltage has fallen to the value problem is then to locate the point P on the horizontal such that P is a point on the external characteristic. ated e.E / t ) . 163. Fig. is is OD. corresponding to the excitation n s i s fore. CVP V draw the horizontal line VC intersecting 07V in C. therefore. 163. If there were represents to scale the new value of n s i s no armature reaction. the net excitation is less than OF by an amount FD. Actually.m.E = BD t HD tan V = = and BH = BH r<ij ~^n LtL ar ru aiia 180a 180ar -T~~ = vaL constant. the net excitation e.176 PRINCIPLES OF DIRECT-CURRENT MACHINES GV. ~ -BH R Construction of external characteristic of shunt generator.m. when a point C on the field resistance line has been fixed. where aZi n F F D FIG.

inferred from these facts that a shunt generator can be shortcircuited without danger.m. then. but this is not the case . but that eventually a critical point is reached beyond which a further lowering of the external resistance causes the current to decrease rapidly. tan \$ rs Hence AP = E and XP = t i are simultaneous values of terminal voltage and line current.OPERATING CHARACTERISTICS OF GENERATORS 111 KP that at an angle 6 with the horizontal. 163. such as P'. for the critical point at which the line current begins to decrease is generally far beyond the current-carrying capacity of the armature. Dependence of the Form Speed. becoming zero when the machine is dead short-circuited (R = 0) under . as illustrated in Fig. that KV tan0 _ *> _ hence P is a point on a curve whose ordinates are terminal voltage (Et) and whose abscissas are total armature current (ia ). A change in speed (r s remaining the same) 12 . It will be observed that the external current at first increases is lowered. = ra It follows. The diagram of Fig.f. and the winding will burn out before the current has had time to decrease to a safe value. this condition of complete short-circuit the external current is not zero but has a small value OS due to the fact that residual magnetism generates a small e. Corresponding values of line current (i) can then be found by subtracting OP at this can be done graphically by drawing the line an angle ^ such that tan ^ = rs to the scale of the figure. 115.except in very small machines. The terminal as the load resistance voltage falls steadily throughout the entire process. is - for it is easily seen that 0_F*. that is entirely consumed in It might be driving the current through the armature resistance. 163 of the Characteristic upon was drawn subject to the con- dition that both the speed and the resistance of the shunt circuit remain constant. this angle being so chosen tan 6 to the scale of the figure. Similar construction will then serve to locate additional points.

Effect of variation of speed upon external characteristic of shunt generator. if there were no residual magnetism. 164. 164. If the shunt field resistance has a constant value. In this figure the surf ace OOi M. etc. the locus of which has the peculiar tubular form shown in the diagram.178 PRINCIPLES OF DIRECT-CURRENT MACHINES will alter the form of^the characteristic. OZ. and to each magnetization curve there will correspond a characteristic LI.. curve ing on open circuit. the machine would fail to build up for any speed below a critical speed. but would intersect the speed axis in a point Z\ that is. drawn to the left of the speed axis. L'i. of The projection of this curve on the curve OL'iLi. t. will t. will give (E n) plane between terminal voltage and speed when the generator is operatIf there were no residual magnetism. is the locus of the magnetiza- tion curves for various values of speed. OL'L would not pass through the origin. the locus of the field resistance lines (ON) will be plane OOiJV. which shows the relation Dependence of Form of Characteristic upon Resistance Shunt Field Circuit. . If the speed of a shunt generator is kept constant and the resistance of the field circuit is varied bv moans 116. and the new relations between E ia and n can be most easily shown by a three-dimendonal diagram such as Fig. be a curve OL'L. and the intersection of this plane with the magnetization surface OO^M FIG.

with asymptotes as shown in Fig. Since the latter would necessarily involve a relation between B and for the iron comprising part of the magnetic cir- H and since such a relation is entirely unknown. A suitable choice of the constants ^ (13) a and b will make this hyperbola agree very well with the actual magnetization curve within the working range of the machine. The construction of the several characteristics is carried out in the manner described in connection with Fig. are the field resistance lines . hence an equation representing the relations between the variables E i and n must 117. this equation represents a hyperbola. but it cannot be made to follow the irregularities in the actual curve at low magnetizations. t. be a function of the equation representing the magnetization curve. V o FIG. 166. as follows: cuit. Effect of variation of shunt regulating resistance characteristic of shunt generator. Characteristics. ON 2 etc. corresponding to the setting of the rheostat. 165. the best that can be done is to represent the magnetization curve by an empirical equation originally due to Froelich.OPERATING CHARACTERISTICS OF GENERATORS of the regulating rheostat. E = where a and b are constants. If the speed is held constant. upon external Approximate Mathematical Analysis of Shunt Generator It will be evident from the preceding articles that the form of the external characteristic is in all cases dependent upon that of the magnetization curve. . 163. ONi. 165.. and n is the speed. OM is curve the magnetization corresponding to the speed at which the and machine is driven. and it does not take account of residual magnetism. the size istic will be affected in the manner 179 and shape of the charactershown in Fig.

we have the following relations (see equations 8. and ignoring armature reaction.(Vira + Vbr 2 s) ] (15) an equation of the second degree between the three variables E i and n. the surface is symmetrical with respect to the plane E which is an t br s ira (16) shown in the figure as ABC.e. (Vira . 166.Vbr 2 ] s) [an . small + E = t an an brs ira " an br8 ira )a 2 ir ar.. Empirical form of magnetization curve. is Solving for E and simplifying by assuming that r a r a = r s ) tliere is obtained compared with r s (i. Moreover.l - br s - ira This is t.180 PRINCIPLES OF DIRECT-CURRENT MACHINES Using Froelich's equation. . 11 and 13): am E= E whence t s E t E = t brs anE +E t (14) t FIG. 167) plane sections of which are conies or straight lines. hence it represents a surface (Fig. 9. r a t.

The drop in terminal voltage Long between no-load . the condition n t Inserting in the general equation for constant.OPERATING CHARACTERISTICS OF GENERATORS If in 181 tuted the i the general equation of the surface. From the former it appears the second. 118. there = 0. 167. Shunt Connection. fail to build up if residual mag- FIG. Idealized external characteristic surface of shunt generator. (15). it is seen that is substi- first of these br s and E = 0. hence OB is the critical speed when n = that E = t below which the machine would netism were not present. Characteristic Curve of the Compound Generator. br = OB. E = t an - t . that there are two values of the current (i) E which will reduce the radical to zero. (17) two equations represents open-circuit conditions. there will result the equation of the external characterIt is obvious istic corresponding to the chosen value of speed. hence the characteristic intersects the plane in ABC two points. hence the theoretical external characteristics are hyperbolas. one is of these values of current is the other (Vo^ + Between these values of current the radical becomes imaginary. short-circuit conditions.

As has been previously pointed out. Fig. istic of Connections of long-shunt compound wound generator. the series generator with the drooping characteristic of the shunt generator. at or near the center of distribution of the load. by the addition of a series field winding excited by the armature current. . 168. let F&i = AP be the terminal voltage corresponding to a value of ia = OA (the latter being supposed to be known). Starting as before with the magnetization curve. O'M. or OF. mines a point the intersection of this line with the magnetization curve deterL whose ordinate is (very nearly) the terminal voltage at no-load. the external characteristic can be constructed in a simple manner Let as follows: ON be the shunt-field resistance line. the increase in the voltage at the generator terminals being consumed by the resistance of the line. even overcompensated. or is to keep the line voltage constant at a distant point. will be less than this by an amount . The net excitation. The field excitation due to the shunt turns will then be given by OF 1} and the total field excitation will be OF 2 where FiF 2 = n fi a is the excitation supplied by the series turns.182 PRINCIPLES OF DIRECT-CURRENT MACHINES full-load inherent in a shunt generator can be compensated. (Fig. 169. the object of over-compounding and partially or wholly.i. the slope of the resulting curve depending upon the relative slopes of the components. 168) may In a general way the compound-wound generator be considered as combining the rising character- FIG. its equation being (18) = J(n. Assuming that we are dealing with an overcompounded machine. _- = demagnetizing ampere-turns per pair of poles.).

actually generated in the armature difference 183 drop in ing. Construction of external characteristic of compound generator. It follows. and the angles at the vertices of the triis In particular. is a point t . Summariz- FiF 2 = GiG 2 = n fia GH = M N (ra FIG. that all three sides of the triangle . The between FG and FiGi. 169.f. ing to any other value of armature current. as OA'. .m. then. or GH. it is desired to find the terminal voltage. draw A'J' AJ. and its length is proportional to i a Through the point G draw GJ parallel to ON. be the the armature and series field. correspondangle are also constant. therefore. therefore. is FG. and join J with A. Draw G\P' horizontally until it interparallel to M T. sects ordinate A'P* in the point P then the latter on the curve showing the relation between E and ia f . If.OPERATING CHARACTERISTICS OF GENERATORS hence the e. and through G' draw <7'(/' parallel to GGi. or i a (r a + r/). the point G'i being on the line ON. GGiG 2 are proportional to i a hence their ratios remain fixed no matter what the value of ^ a may be. A'P'. the slope of the side GGi constant. J being on the axis of ordinates (prolonged downward). must. . draw J'G' parallel to JG until it intersects curve in G'.

especially in large compounded ator. If it is desired to make is at full-load equal to that at no-load. A study of Fig. Short-shunt Connection. It will be observed that this method presupposes a knowledge The of the coordinates of at least one point on the characteristic. But at or near full-load the difference between i and i a will be relatively small. sary to subtract the value of ia . In practice it is quite common to design the series field windings of compound generators with a . . so that tan \f/ and P'X' are the values of i corresponding to ia equal to OA PX and OA'j respectively.184 PRINCIPLES OF DIRECT-CURRENT MACHINES desired to If it is show the relation i. machines. when the figure LGiG approaches it is triangular shape). It is clear frorn the construction that the characteristic of a flar- generator. like that of an over-compounded genetcannot be exactly a straight line because of the curvature of the magnetization curve. 169 is not strictly applicable. the curvature being considerable for small below the knee of the become more and more values of the load and less pronounced as the load increases. The Series Shunt. 119. . In this case the current through the series winding is i = ia is hence the construction of Fig. Then zontal. Thus. to make the full machine flat-compounded. that the terminal voltage to say. 169 shows that a reduction in the number of series turns will shorten GiG 2 and will make the characteristic more nearly horizontal. Considerations of this kind become important when the specifications of a machine call for a compounding that shall not depart from a linear relationship by more than a limited amount. If the point L is so placed that curve. To this end draw the line between E and i. the external characteristic will convex (from above). it becomes evident from the diagram that the external characteristic will approach a linear form more and more nearly as the magnetization curve flattens out (that is. so that the above method will give a very close approximation to correct results. chief value of the construction lies in the clearness with which it shows the intimate relation between the magnetization curve and the external characteristic. GGiG z (assumed to correspond to load conditions) will degenerate to Lgg 2 where g%gh = G 2 GH. it is necesfrom each corresponding value of OX at an angle with the horit \f/ = rs to the scale of the drawing.

1 The the FIG. the maximum voltage per commutator is about 3600 volts. Series generators connected in parallel (a) Series Generators. each pair being driven by a water wheel (or other prime mover). the prime when under load. To start the machine. The maximum line voltage in use at the Moutiers-Lyon plant is 57. for the reason that 1 Electrical World. . per pair of generators. or 5000 kw. Thury system. the switch being in the mover is brought up to normal speed. the switch being so arranged that when it is in the "off" position the generator is short-circuited. as in Fig. 294. and the switch thrown to the running position when the ammeter reads normal current. 170. a voltmeter. Vol. LXI. generators are grouped in pairs. the machine is in series with the line. 171 will not operate satisfactorily. Since all of the machines are in series potential of the circuit rises 122. 170. though a new installation projected for transmission from Trollhatten (Sweden) to Copenhagen a distance of 200 miles contemplates the use of a of line voltage of 90. and a switch.000 volts.connections. equipped with an Thury Each ammeter.600 volts. Parallel Operation of Generators. and when in the "on". 1913.186 PRINCIPLES OF DIRECT-CURRENT MACHINES alternators) for the supply of current at the receiving or disIn other words. as The individual distinguished from merely local distribution. though it is possible to design machines of this type to give 5000 volts at 500 amperes. starting is and stopping in generators is the system Diagram of . p. the from generator to generator. off position. as in Fig. the system is generally tributing end of the line. or running position. generator very simple. To shut down the machine this process is reversed. used for the transmission of power over considerable distances. In plants now operating. hence the machines must be carefully insulated from earth to prevent breakdown of the insulation.

173. two or more of them. Since the machines are in parallel. Division of load between shunt generators in parallel. Inasmuch the compound gen- erator partakes of the characteristics of both shunt and series generators. In this figure ordinates are plotted in volts and abscissas in per cent. Generators. therefore it OA FIG. of full-load current. . 173. if BO is A follows that the load to divide at all times in proportion to the ratings.188 PRINCIPLES OF DIRECT-CURRENT MACHINES first tage of the machine and a drop in the voltage of the second. portion Sup+ pose. whatever that may happen to be. for instance. will operate satisfactorily in parallel only when the series fields are provided . The original conditions will be restored. loads. their terminal voltages must necessarily be equal. 172. of full-load -as current. the characteristics. machine (a) will deliver cent. if over-compounded. But it is essential to good operation that the machines should divide the total load. of its rated current is such and machine that (b) OB all condition to be satisfied is OA = OB at per The per cent. when plotted in per cent. curve (a) senting the characteristic of one machine. hence if the load that the terminal voltage is OC. that two shunt generators that are to be connected in parallel have external characteristics as shown repre- FIG. curve (6) that of the other. assuming that the prime movers are properly governed. in in Fig. Shunt generators parallel. It is. must be (c) Compound identical. of course. not necessary that the two (or more) generators thus connected in parallel should have the same ratings. in proto their ratings.

Compound parallel. with the result that the entire load would be thrown on the other machine and so open its circuit-breaker also. the heavy equalizing current might open its circuit-breaker.OPERATING CHARACTERISTICS OF GENERATORS 189 This is a with the same equalizer connection shown in Fig. equalizing For the same reason. for livered the FIG. generators in readings would be affected by the equalizing current. a heavily loaded generator might be supplying an equalizing current of large magnitude to the other lightly loaded machines and at the same time the ammeter of the loaded machine would read low while that of the other machines would For this read high. reason the individual in the ammeters must be placed as at A. . if 174. It is clear that ammeters were connected as at A'j they would not indicate the true current actually deby the machines to the external circuit. 175. 171. if two machines in parallel are each delivering full-load cur- rent and one of them should suffer a momentary drop in speed. the equalizer is not necessary. that is. consequence of the rising characteristic. The diagrammatic scheme of connections of two compound generators in parallel is shown in Fig. Thus. if they are under-compounded. lead that connects to the side armature on the to the opposite connection. they should be placed in the same lead as the FIG. if incorrectly placed. But if the machines have drooping characteristics. 174. if single-pole circuit-breakers are used. Diagram of switch-board conneccompound generators in parallel. thus. that is. ammeters. tions.

their characteristics identical in the manner explained in connection with shunt machines. 175. equalizing connection. current. In street-railway generators built by one well-known company the series field in this case it is connected on the negative. therefore.190 PRINCIPLES OF DIRECT-CURRENT MACHINES of The complete switch-board connections two compound generators are shown in Fig. The main switch and the equalizer switch are usually combined in a triple-pole switch. and its full be unduly high because of the longer this machine will. not take share of the load. Neglect of this feature will result in a disproportionate For example. Moreover. since the series fields by virtue of the equalizer resistances of the series the connection. T^he series field of a compound generator may be connected to either the positive or the negative terminal of the armature. or grounded side. If two compound generators are to divide the load in propormust obviously be tion to their ratings. FIG. The process of paralleling a compound generator with one or more that are already running is as follows: The main switch of the incoming machine being open and its circuit-breaker closed. Diagram of connections of railway machines are at unequal distances from the switch-board. and proper division of the load is then secured. if the division of load. or zero. the main switch is closed. windings including the resistances of the are in parallel respective equalizing leads. but another circuitis . its load is shifted to the others by weakening its shunt-field current. must be inversely proportional to the rated currents of the two machines. if necessary. by further adjustment of the field rheostat. not sufficient to use one single-pole circuit-breaker (A) on the positive or feeder terminal. the prime mover is brought up to speed and the voltage of the incoming machine adjusted to equality with the bus-bar voltage by manipulation of the shunt-field rheostat. and the main switch is opened when the ammeter indicates a small. 176. To shut down a machine running in parallel with others. the resistance of the series field of the more remote machine will generator grounded supplying circuit.

that is. The individual lamps. no current would flow in the neutral. the lamps wires. In the early forms of three-wire systems. and the neutral might then be omitted. armature winding were to become grounded to the core. were exactly balanced. Economy in the use of copper and power dictates the selecand moderate current. designed to operate on the higher voltage of the system. Three-wire Generators. the three-wire system of distribution diagrammatically illustrated in Fig. 177 is extensively used. A later arrangement. this is sometimes done in 220. . Three-wire system. at higher voltages. the neutral being tapped into common junction. the shortcircuit current through the armature and series field would hold up the excitation and maintain the short-circuit without the possibility of protection by circuit-breaker A. 178. and larger motors. 177. 177). 176.OPERATING CHARACTERISTICS OF GENERATORS 191 breaker (E) must be put in the lead to the grounded bus. lamps designed for 110 to 115 volts are more efficient than those operating in distributing circuits for lighting tion of high voltage <o FIG. with a balancer set connected across the outside If the load on the two sides of the system. between neutral and outer wires.volt systems. are connected between the outer wires. as shown For if circuit-breaker B were not present and the in Fig. 115-volt lamp and To get the benefit of the high efficiency 110at the same time to obtain the copper economy of higher voltage. small motors and other translating devices are connected between the outer wires and the middle or neutral wire. shown in Fig. the splitting of the moderately high voltage between the outer wires was accomplished by using two generators their in series (Fig. 123. rather than low voltage and large current. consisted of a main two-wire generator wound for the voltage between the outside wires. two generators in series. but in incandescent lighting.

since always symmetrically related to those of 13 midway between the potenmidway between the potenthe potentials of C and D are A and B. will have a potential tials of C and D. FIG. respectively. so that the coil is D traversed alternating current which goes through one cycle (two alternations) per revolution per pair of poles. 181. points that occupy homologous posi- Y 1 I slots.OPERATING CHARACTERISTICS OF GENERATORS 193 browolsky. E. The difference between C and is alternating. Three-wire generator with auxiliary winding in of potential tions with respect to poles of opposite polarity. . and shown diagrammatically in Fig. Three-wire generator with two winding. this alternating current is small because of the large self-inductance due to the iron core by an on which the coil is wound. and therefore also tials of the brushes A and 5. 180 elec- degrees apart. wound on an iron core. A coil of wire. that is. 180. CED. is tapped into the main armature winding of the generator at the points trical C and D. coils tapped into armature of the coil. The middle point t FIG. 179.

or two balance coils. D and C". 181. tor. in the manner indicated in Fig. Three-wire generator with two auxiliary coils mounted externally. and the connection to the outside circuit is made through E FIG. 183. D' are 90 electrical . In machines of this kind built by the General Electric Company. connection to the armature winding being made in that case through two slip-rings. 180. a single slip-ring mounted on the main The Burke Electric which the coil CD is armature winding. the coil CD is wound on a core A that is mounted from the middle point inside the armature core.194 PRINCIPLES OF DIRECT-CURRENT MACHINES tral wire of tap brought out from the point E may then be used as the neua three-wire system. The alternating voltages between the points C. 182. builds a three-wire generator in Company wound in the same slots that carry the main Equalizcrs \ / ^ ^^ Series Field Shunt Windings not shown FIG. connected to the armature winding as in Fig. may be used. The balance coil CD may also be placed outside of the genera- shaft of the generator. Diagram of connections of compound three-wire generators in parallel.

with the result that incipient changes of voltage are immediately checked. below normal. it possible to maintain a steady voltage at the terminals generator irrespective of changes in the load or of fluctuations of speed. be compounded. the series field in must be two equal parts. An elementary diagram of connections of the regulator is shown in Fig.OPERATING CHARACTERISTICS OF GENERATORS degrees apart. mounted externally to the generator. on closing the regulator circuit the rheostat is short-circuited and the generator voltage rises. that is is. 184. of the greatest importance. When the exciter field rheostat. the other half in series with the other outer wire. but merely tends to do so. The regulator maintains the desired voltage by rapidly opening and closing a shunt circuit connected across the terminals of The field rheostat is so adjusted that disconnected the generator voltage is about 35 per cent. 182 shows the connections when two balance this is coils. the short-circuit around the rheostat is opened and the voltage again falls. 195 zero. are used. hence the main switch of a three-wire Electric If three-wire generators are to generator is usually constructed with four blades. It has been shown in preceding ar124. and also to compensate for line drop by in- may makes creasing the generator voltage as the load increases. The opening and closing of the by-pass around the exciter rheostat is accomplished by means of contacts on the armature of a differentially wound relay magnet of U . one of them is a maximum when the other Fig. If two or more three-wire generators are to be operated in parallel. The opening and closing of the short-circuit around the rheostat is so rapid that the voltage does not actually follow the changes of the field circuit resistance. and vice versa. as in Fig. is when the regulator the voltage reaches a predetermined value. Tirrill Regulator. the standard construction used by the Westinghouse and Manufacturing Company. half of the turns being in winding series with one of the outer wires. ticles how the voltage of shunt and compound generators may be regulated either shunt is field circuit or by manual adjustment of the rheostat in the by the automatic compounding effect of the In lighting circuits where steady voltage and automatic regulation series field winding. 183. accurate of voltage this device be obtained by the use of the Tirrill regulator. two equalizer connections must be used.

thus again inserting the rheostat in the field circuit. by the General Electric Company is shown in . The latter is wound with a directly across the bus-bars. pulled down. the action will cause the voltage to rise higher before the main contacts are opened than would otherwise be the case. 184. potential If the generator of the regulator is as follows: the coil of the main the current potential voltage falls. Diagram of connections at Tirrill regulator. A perspective view of a simple regulator built Fig. condenser shown in the figure is for the purpose of reducing sparking at the relay contacts. The degree of compounding its differential may be varied by means of the sliding contact on the compen- The sating shunt with which the current coil is in parallel.196 PRINCIPLES^QF DIRECT-CURRENT MACHINES One winding of the relay magnet is connected directly shape. thereby short-circuiting the field rheostat. voltage rises the armature of the main control magnet is again is main contacts are opened. and the relay magnet again energized. 185. through control magnet is weakened and the spring closes the main con- The operation tacts. magnet which is then demagnetized and the spring closes the reAs the lay contacts. If the current coil of the main control magnet is used. Current then flows through both windings of the relay Compensating Shunt FIG. thus giving a compounding action. the other winding is also connected across the bus-bars. but through a pair of main contacts actuated by the main control magnet. in series with a current-limiting resistor. across the main bus-bars. and a current potential coil connected coil (which may or may not be used) whose magnetizing action opposes that of the coil.

General Electric Co. from two to ten are employed. in such a case the main control line magnet would be actuated by the main bus-bar voltage and current.OPERATING CHARACTERISTICS OF GENERATORS 197 When several compound generators of moderate capacity are worked in parallel. and will compensate for line drops up to 15 per cent. the others performing a like function for the other machines. Regulators of this kind are suitable for two-wire or three-wire generators with shunt or compound windings. Of course. it is advisable to use sepaand to connect the regulators so that they act upon the exciter fields. part of them of . rate excitation In the case of very large machines. serving to short-circuit sections of the field rheostat of one generator. Regulators are also built for controlling the voltages two or more generators operating in parallel instead of using a single relay magnet. 185. and fluctuations regulator provided the load on the others will be equalized through the compound windings. a simple regulator may be connected to one of them and the others allowed to "trail. Voltage regulator. FIG." The generator with the will take the in load. .

and the electrical and magnetic quantities involved. make them especially suitable for that railway and in hoisting service. and are used for such purposes as driving line shafting. Counter E. Torque and Power. In the case of adjustable speed motors. II that when a current is sent into the winding of an armature which is under the influence of a magnetic field the individual conductors of the winding are subjected to a lateral thrust and that motion ensues. the impressed voltage remaining constant throughout. the motor drive. The immediate effect of this motion is to generate in the conductors an e.M.f 199 . in individual drives for machine tools. the speed can be fixed at any one of a large between a minimum and maximum value. whose direction is . These relations determine the operating or mechanical characteristics. Variable-speed motors include those types in which the speed is inherently variable.F. Constant-speed motors. etc. with constant impressed number of values and when so set will voltage. as in street Intelligent operation of motors involves a knowledge of the relations between speed. remain substantially constant for all loads within the limits of the machine's capacity.m. adjustable speed. which will be discussed for the different types of motors. there are three principal classes of service. in Chap. motors of this kind are used. torque (or turning moment). and variable speed. examples of this class are the series motor lative and the cumucharacteristics compound-wound motor. load (or output). maintain an approximately constant* speed at all loads when supplied from constant potential mains. fans. their speed class of service in which it is desirable to reduce the speed as the load increases. changing as the load changes.. It has been shown 126. Service Requirements.CHAPTER VII MOTORS In the industrial application of 125. for example. of which the shunt motor is an example. characterized by constant speed.

m.f.'. FIG." The effective development of torque in the case of a motor is dependent upon a proper space relation between the field flux and the armature current. there is an addidrop due to the resistance of the series field winding. Effect of brush position on the torque. the impressed upon the armature terminals must be consumed overcoming the back e. is called the "back e. 186a. the torque will be a maximum for a given current in the winding. opposite to that of the current.f. arc to the E =E t (1) where p a 60 X 10* tional In the series and long-shunt compound motor. in In the case of a separately excited or of a shunt motor. but if the axes of armature current and field flux are at right angles to each other. as in Fig. there is no resultant tendency to rotation. as in Fig.f.m.m. the brushes are so set that the axis of the armature current coincides with the axis of field flux. resistance of the armature winding . and the ohmic drop due and the brush contacts.m. e.200 PRINCIPLES OF DIRECT-CURRENT MACHINES This counter-generated e.m." or the "counter e.f. 186. If. hence E = Ea + t ia (ra + + r/) (2) In the case of the short-shunt t compound motor the ia r a relation is E = Ea + irf (3) . 1866.f. for instance.

and iron of this losses. windage. armature in watts. Not all developed power is useful power at the shaft or pulley. as explained later. respectively. Multiplying equation (4) t by ia . p = E a ia = 2ir~T T = watts or E a ia = Z'\$i a (10) The above unit of torque is inconvenient for practical application expressing torque in kilogram-meters. . there results za E ia = E a ia + The term i a r' 2 t V (8) E ia represents the power supplied to the armature. It follows. and pound. If amount P = T = total mechanical power. that E a i a is the is of mechanical power developed by the armature.MOTORS In general 201 E = t E. pound-feet. and the power dissipated as heat in the ohmic resistance of the armature circuit. + ia r' = I ^^ + and iS (4) or E .Ea t r' (7) an equation which is of importance in connection with the starting of motors. for some of it is lost in bearing and brush friction.i a r' ~*Z^t (5) where r' = resistance of armature circuits in series therewith and a 60 It is also seen that X 10 8 E . developed in the torque in dyne-centimeters -r- 10 7 (9) /. therefore. inches.

If the flux at starting has its normal full-load value. It should be carefully noted that in Fig. 1876 those of a shunt motor.p. and. by equation (11).48 la = 7. that the Et normal rf) small running resistance of the armature circuit (r a or r a must be increased during the starting period by the insertion of + a starting rheostat in order to limit the flow of current to a reasonable value. 1876 the rheostat is in series with the armature only. or even larger than. and Fig. 127. or more than ten times normal full-load current. the starting greater than. the resistance that must be put in series with the armature should be 220 0. the starting current.202 PRINCIPLES OF DIRECT-CURRENT MACHINES AO in7 = Q. But since a = E when at the the armature is stationary. Thus. therefore. the initial current would be 440 amperes. will then have to be equal to. a 10-h. 60 2.case of a series motor. ia it is clear from equation (7) that moment of starting = -7.\$75Z'&i a kg-m. If a mo tor is called upon to start a heavy load from rest.5 rated load. Fig. the starting torque may be as large as.r 1Q 7 980 X 453. its full-load value. the full-load running torque.05Z'<H a lb-ft. current may be smaller the greater the flux.6 X 30. The Starting of Motors. so the shunt field .volt shunt motor would take an armature current of approximately 40 amperes when carrying its 0. or perhaps somewhat Other things equal.5 ^7j- = 5 ohms is The resistance of the starting rheostat is usually so adjusted that the initial current torque. If the full voltage were impressed directly upon the armature. and is independent of speed. and would have an armature resistance of about ohm. 220. To limit the starting current to the -full-load value. (11) It is clear from these equations that the torque is dependent only upon the flux and the armature current. 187a somewhat greater than that giving full-load of the start- shows diagrammatically the connections ing rheostat in the .

for in that case the full line voltage . 189. start. and. If an ordinary rheostat of the kind illustrated in Fig. 187 were used in commercial installations.MOTORS winding receives the full line 203 voltage at all times. Motor starting rheostat with no-voltage and overload release. 188. since here the shunt field current is seriously reduced at the (a) Series (6) Shunt Correct (<?) Shunt Incorrect FIG. there would be danger of burn- FIG. 187c shows an incorrect set of connections. ing out the armature if. including the Fig. after an interruption of the service and the consequent stopping of the motor. starting period. the voltage should again be applied to the supply line. FIG. if the motor is unloaded. causing the speed to rise dangerously high. Connections of starting rheostats. 187. thereby reducing the flux and also the torque. Diagram of connections of starting rheostat having no-voltage and over load release.

impressed e. This behavior is due to the tendency of any motor to run at such a speed that the back e. therefore. as by a broken wire or loose connection and in the series motor by an accidental short circuiting of tht terminals of the series winding.. the speed will rise to a dangerously high value In other words.. impressed upon the armatur* l remains constant.204 PRINCIPLES OF DIRECT-CURRENT MACHINES of the ariria would be thrown directly across the low resistance ture (or armature and series field).f.f.f shall be nearly equal to the impressed e. It is seen small from equation (5) that if the flux <i> is reduced to value while the e. 188.m.m. inherently selfBoth equation (5) and Fig. if ia = 0. in other words.f. 190. corl- n that were it = E t ia ra ~7 not for the demagnetizing action of the armature cur- rent. For this reason most motor starting rheostats are pro vided with a "no-voltage release" which automatically restore the starting lever of the rheostat to the starting position when th line voltage is removed. the motor will "run away" and may wreck itself This contingency may arise in the case of a shunt motor if tht field circuit is opened. indicate that if ia = 0. This assumes that r a is constant . and Fig. 190 regulating as regards speed.m. n = -^7. current. quite frequently there is also an "over load release." which opens the circuit and automatically cuts ii the starting resistance if the current becomes excessive for ani reason. Actually. the denominator of the fraction would be constant and the speed would decrease slightly and uniformly with increasing values of ia as in Fig. the lowering of th flux demanding an increased speed. stant. When the motor is "running free" . Characteristics of the Separately Excited Motor. resulting in a very heav. there is no torque and no rotation. The separately excited motor with constant excitation is. Let it be assumed that both the (a) Speed Characteristics. 189 illustrates a starting rheostat of this type madi by the Ward Leonard Company. that the temperature of the armature is maintained at its normal running value. It follows E and the field exciting current are from the speed equation t. The connections of such a rheostat are shown in Fig.m. 128.

indeed. 191. equation shows at a glance that the effect of armature reaction.he point A in the figure. be varied through wide limits Thus.bo great an extent. if E E t It is obvious that the speed may t )y varying either 3> or ?an be raised by reducing E . 205 ture since sufficient is still some current through its armamust be supplied by the line to overcome power nternal losses due to windage. be the magnetization curve of the machine. are independently variable in the separately excited motor gives to this type its chief advantage. Strength above or below which here are serious commutation difficulties. E t is raised to . Variation of E t gives little or is no trouble so far is ^0?^ Approximate f speed separately excited concerned. provided the flux is originally adnormal value unless. there . since it reduces <. commutation usted to about its The form of the speed (6) Effect of Armature Reaction. thereby partially neutralizing the slowing-down effect of armature resistance and mproving the speed regulation. or both of them. a = and ia = 0. The speed n = pi ^ may be called the f .m. The armature of a motor may therefore . The fact that the field excitation and the armature impressed e. in which case also. Fig. unloaded). it is the speed that would be reached were no losses. unless special devices are used.f. friction. hysteresis and eddy curThe minimum value of armature current is indicated by ents. because field are limits to the r FIG. 190. The curve showing the between speed and armature current can be constructed in the following manner: Let O'G. she possible range of speed due to the adjustment of the excita.MOTORS that is.Ure of be designed magnetically more powerful than the armaan otherwise identical machine intended for use as a relation generator. the speed \$ or t )y increasing However. abscissas (drawn downward from O) representing ampere-turns per pair of poles ( = n/if ) and ordinates (drawn to the left of 0) . E S-r. will be to raise the speed. ion -here is rather restricted.here deal zero-load speed.

loUft then when ia = OA.f. such that tan dju V-IQK . Select any convenient scale of armature current along OA. draw OD making an angle with OA. t. Ea = E . and a scale to represent the impressed voltage along 0V. Construction of speed characteristic of separately excited motor. 191.iara = AB . n = E t OE OL . The back e.BC = AC Similarly. the value of \$Z' would be F G Q = OL and the ideal no-load speed is then . If the armature were currentless (i a = 0) as in the ideal noload condition. the intercept on the ordinate at . Draw EC E A EC EB be is BC = then ia ra = ohmic t drop in the armature. ra a straight line so that tan <p through the point to the scale of the figure.m. then for any value of i a} such as between and will OA. AD = CiZ PtoA~& loUtt = demagnetizing ampere-turns per pair of poles.206 PRINCIPLES OF DIRECT-CURRENT MACHINES representing values of <J>Z'. Assume that the field excitation is constant and equal to OF and that the voltage impressed upon the armature is likewise constant and equal to OE = E v . FIG.

From equation (11). the demagnetizing effect is given by AD. P' is a point corresponding to i a = OA'. ON N being on the ordinate through A. OA = OA. and draw QNo n ~ Similarly. It is clear that the speed curve cannot be exactly straight because of the curvature of the magnetization curve. points G and L.05FG. as follows: itself readily to a complete graphical solu- Select parallel to any convenient point L Q E. In precisely the same manner. T = 7.05 &Z'ia pound-feet Referring to Fig.05 OL.OA = when ia 7. This may be written OA constant which equation suggests the following construction for the curve showing the relation between the armature current and the torque: In Fig.MOTORS 207 At any other load. Proceed as before to locate Select any convenient constant length OR. = OM OL ON _ := OQ ON and are respectively it follows that be and made to n and equal to the speed by may proportional n. join OL OE ~ _ ON OQ L and M and draw QN parallel to LM. is AC = OM. 191. the \$Z' curve. P will be a point on the required curve. the latter a suitable choice of scale. across to P. 191.f. . and the line OD just as in Fig. then Q on the \$Z' Q axis. the torque is (c) The Torque Curve.m. Projecting Since OQ is constant. N But the diagram lends tion. joining D with F 0) and drawing AF parallel to DF the net excitation is reduced to OF and 3>Z' becomes FG = OL] at the same time the back e. Q P. 192 draw the axes of coordinates. this becomes T = 7. hence the speed is E t laTg _ OM OL ~~ S>Z' In this of ia way and the values of speed can be computed for various values results plotted to obtain the desired curve. as when i a = OA.

The 129. differs in no way from a separately excited machine with constant impressed e. and constant excitation. and having fixed field. its carried out in exactly the operating characteristics is. is almost constant for all values FIG.208 PRINCIPLES OF DIRECT-CURRENT MACHINES draw LR. concave downward. Construction of torque characteristic of separately excited motor. Characteristics of the determination of ceding article. 192. and P is then a The torque curve is slightly point on the desired torque curve. to be same manner as described in the pre- . SPEED AND TORQUE CHARACTERISTICS. By construction AP = OL ~ PL OA OR constant AP is proportional to the torque. resistance.f. A plain shunt motor operated on constant potential mains.m. and then draw OP perpendicular to LR (using the semicircle on OR as a construction line) until it intersects the ordinate through A in the point P. but to all intents and purposes it is a straight hence line of ia through the origin. since \$ within the working range. therefore. Shunt Motor. or may be made equal to the torque by a suitable choice of the length OR.

Characteristics of the Series Motor. 3> ia = = instance. the flux does not become zero because of residual while practically. are not independently variable. When the armature current is ia = OA. then will be the demagnetizing effect expressed in equivalent amperes instead of in ampere-turns per pair of poles. by gearing or direct connection. (a) for the Speed Characteristic.m. in railway motors. and AC will be the back e. a series motor must always be so installed as to be positively connected to its load. still it becomes so small that the speed reaches a magnetism.f. t. hoists. rolling mills. < the denominator increases. dangerously high value. if hence at no load the speed would be infinite. and the minimum load must be great enough to keep the speed within safe limits. Theoretically. less in The the shunt motor than possible range of speed variation is therefore it is in the separately excited motor. the field excitation (in 14 AD . . Fig. E and draw EC so that tan tp = (fa -h T f ) to the scale of the figure then BC will represent to the same scale the internal ohmic drop corresponding to ia = OA. never by belting.m. For this reason. as follows: Let O'G. Let OE represent to scale fication of the the constant impressed voltage. draw OD so that tan where n f the number of field turns per pair of poles. such is the case. 0. shows that the speed of the . for tion decreases. Assuming that the motor is to be operated on constant potential mains. as in the latter.f. its speed characteristic can be determined by a modi- methods described in the case of the separately excited motor. etc. be the curve which gives the relation between \$Z f and the exciting current (4). 193. . 0. is Also. speed Inspection of the general equation series motor must decrease quite for increases with with load the reason that rapidly increasing of the fraci In other the numerator while increasing a words. 130.MOTORS The 209 chief point of difference between the shunt and the sepamotor is that in the former the field excitation excited rately and the impressed e. assuming that E t remains constant.

j = OT-f tan Q = constant .L = . ~ r/) AC \$Z f HK == OM constant. the point F being found by drawing AF The demagnetizing at an angle of 45 deg. ON N Since FH = AD = and tan \f/ ^ tan 6 1 - OH = OF .FH = L . upon the Projecting ordinate at A. the resulting point P is a point on the speed-current curve. with the horizontal. Construction of speed and torque characteristics of series motor. effect is AD = FH. hence is proportional to the speed.210 PRINCIPLES OF DIRECT-CURRENT MACHINES amperes) is OF = OA. 193. and Selecting a point Q on the \$Z' axis such that drawing QN parallel to LM. hence the net excitation is OH and the corresponding value of <bZ' E t HK ig (r + is a = OL. _ OL OQ is OM OL ~~ ~ ON OQ FIG.

when the machine is used If as a generator. 209). that is. resulting in a differential effect. when ia has an appre- Compound Wound Motor. then. or are cumulative. a straight line. OR = ~ constant. 194. the shunt and series windings of a compound wound (long shunt) machine are so connected that their magnetizing effects cooperate. ally it is a curve of higher order. actu. Characteristics of the (a) General.05 &Zi a = 7. 131. follows that for r . Selecting a point R such that perpendicular to LR. and drawing OP APT OA whence OL OR PT is AP T is proportional to the torque. if the machine is used as a motor. T = when ia 7. as given by equation (11). and a point on the torque-current curve.OA pound-feet = OA. as ia = OA'.MOTORS it 211 any other current. 3> proportional to i a the torque would be proportional to (i a ) 2 . It will be observed that the torque curve of a series motor deviates considerably from the linear form due to the fact that the If the magnetization curve were flux varies with the current. curve of useful torque may be obtained from that of total torque by subtracting from the ordinates of the latter the "lost torque". the useful torque passes through zero value ciable value (see Fig. shunt and series motors refer to the total developed torque. the two windings will oppose each other. The actual torque at the pulley that would be measured by a brake test is less than the total torque by an amount which corresponds to the torque The required to overcome internal friction and iron losses. The torque curves in connection with the sepa- rately excited. lying between the linear and parabolic curves. This is illustrated diagrammatically in Fig. it is only necessary to draw A'H' parallel to AH in the process of locating P As before. If the machine . Lines such as OP T can readily be drawn perpendicular to as a diameter and drawing by constructing a semicircle on LR OR a line through and the point where discussed LR cuts the circle. the torque is (b) The Torque Characteristic.05 OL.and the curve would be a parabola.

212 is PRINCIPLES OF DIRECT-CURRENT MACHINES designed to over-compound as a generator. shown in Fig. The construction has been carried out in Fig.m. between this case and the one discussed is in connection with Fig. the fixed point R The curve is concave downward. the case is obtain without the series winding. In the same way a differentially wound generator. 191. characteristic when driven at constant speed. having a drooping e. for which the speed will be the same at full load as at no load. and hence a speed higher than would In general. construction. 195. 195. 194. . Relative directions of shunt and series exciting current in com- pound machine. that now tan v = ra + rf and aZ assuming that the resistance of the shunt field winding is constant. the differential motor action will be considerable. the speed rises with It is clear that there is a particular value of increasing load. resulting in a decided decrease of flux under load conditions. it but that it cannot be made absolutely constant at all loads (in the absence of special regulating devices) because of the curvature of the magnetization curve O'G.f. Differential The only difference Compounding. (b) (I) Construction of Speed Characteristic. is also . The torque curve being used in its due to the fact that the flux decreases with increasing current. becomes a cumulative-compound motor with a drooping speed characteristic when supplied with constant terminal voltage. lar to that of simi- a shunt motor with exaggerated armature demag- netizing effect. from which appears that if n/ is sufficiently large.( jlfiM^n Generator FIG.

O' \A i t FIG. 195.MOTORS 213 FIG. Construction of speed and torque curves of differentially compound motor. 196. Construction of speed and torque curves of cumulative compound motor. .

available for producing current. At the same time. manner by the arrangement shown diagrammatically in Fig. The cumulative compound motor has characteristics which are intermediate between those of the shunt and series motors. Counter E. The field of 2 from any con2 is excited by a constant current supplied venient source.f can be shown in a striking .m. in the circuit.m. Mij and the generator is electrically connected through the switch S and the regulating resistance R to the armature of an unloaded and separately excited motor. provided the resistance of the circuit is below the critical value. It differs from the latter especially speed curve now is in this. M tive e. the line it. A series generator.m. S the series generator will build up both and current. The existence of the counter-generated or back e. so that the acclosing the switch t On e. will rise nearly proportionally. = ra + r/ and therefore.f. 132. is the difference between the e. 196 is the same as in the previous cases for both the speed and torque curves. G.f. that its speed rises to a definite limit when full load is suddenly thrown off.F. the speed of 2 will rise and its back e.m.f.m. Connections for the experiment of the reversing motor. and the torque concave upward. FIG.M. 197. The falls considerably with increasing load. The Reversing Motor. 197. OD must be drawn afowe OA instead of below otherwise the construction shown in Fig. is driven at constant speed by a suitable motor.fs.214 (II) PRINCIPLES OF DIRECT-CURRENT MACHINES Cumulative Compounding tan (p Here. . instead of running away. M M . of generator and of motor Af 2 G .

m.m. the less will be the speed to produce a given back e.f.f. But as Mz has no driving power other than its energy of rotation. it very quickly comes to rest.m.f. the greater will be the 2 torque for a given armature current. Starting of Differentially Wound Motors. the e.f. of the generator also decreases.MOTORS The current it 215 therefore falls after rising to a certain value and with the generated e. moreover. circuit reverses (as M the generator. tendmake it run backward as a motor. and the more rapid will be will build at which motor M% moment of inertia of its M . the current. wound compound motors istic of 133. in the circuit. M proceeds the counter e. the current in the may be shown by the two-way ammeter A). therefore. the speed of Mz continues to rise because of its acquired momentum. it will be seen that the reversals of Mz will be more and more rapid.m. Meanwhile. it again stops.f.m. Differentially are seldom used in practice for the reason that in most cases the slightly drooping speed characterthe plain shunt motor meets the requirements of constant .m. the greater the excitation. The rate up in speed depends upon the armature and upon the torque. the greater the excitation of 2 The above reasoning will serve to explain why a series generator cannot be used to charge a storage battery. and MZ momentarily becomes a generator tending to drive G as a motor in opposition to motor Mi. Since the current through the circuit has been reversed. and its magnitude will depend upon the machines used in the experiment. For as the charging .f.f. over and over again. overpowers e. Finally.consequently as soon as MZ has come to rest G begins to build up again. . of the battery rises. the latter in turn depending upon the magnitude of the exciting current of The larger the excitation of Mz. will therefore continue to fall off until it becomes and then the battery discharges through the generator. of G decreases. but with polarMotor z then speeds ity opposite to that in the first instance. therefore. hence reducing the effective e.m. the process of picking up speed.f. of G. falls and as zero. The function of the resistance R is simply to prevent the cur- rent from reaching excessive values.m. and the entire cycle of changes is repeated. it The current ing to decreases. its back overpowers the generated e. up again in the reverse direction until its back e. the residual magnetism of G also reverses.

Normally the axis of commutation coincides with the axis of the interpoles.216 PRINCIPLES OF DIRECT-CURRENT MACHINES speed to a sufficient extent.m.f. 198. A similar state of affairs may arise in the case of shunt motors provided with interpoles if the brushes are not properly placed. and when the armature finally stops because of reduced torque or the dissipation of its energy of rotation. . but if the initial current flow is sufficiently great the series excitation may overpower that shunt winding. the differentially wound motor is subject to " racing" in case of heavy overload. The machine will then start up in the right direction. will rapidly speed up in the wrong direction. of the tation. cause the machine to become a generator and send current back to the line. due to the considerable reduction of field flux caused by the large current in the series field winding. the acquired momentum of the armature may even for a brief interval TIG. a heavy flow of current through the armature and series winding will result because there is now no counter e. Differential effect in interpole motor due to backward shift of brushes. if In that case the motor. In the meantime the shunt field current has been building up. may so impede the rise of the shunt field current that the current in the series winding. up in the wrong direction rheostat to the first Such motors are also liable to start on throwing the handle of the starting notch. and so bring about another reversal of roThis process may go on indefinitely unless the design constants of the machine are such that the successive impulses are damped out. but if the brushes are accidentally shifted backward. developing a considerable counand so reducing the cur- rent flow and the torque. inductance of that which builds up much more rapidly because of the small circuit. for the high inductance of the shunt winding.m. may overpower the magnetizing effect of the shunt winding. do not synchronize with the natural period of oscillation of the armature. due to the large number of turns. thereby ter e. unloaded.f. that is. Moreover. and so reverse the flux and the direction of rotation.

not in the main line or field circuit) a variable resistance. the method is inefficient because of the loss of power due to the flow of the armature current through the external resistor. It is not to be recommended in commercial installations. by varying the impressed voltage E and field control by varying \$. A motor used in this way has poor speed regulation. which includes the armature resistance r a '. Armatures Mechanically Connected FIG. 190 and 191). by varying the resistance voltage control. 198. A fourth method occasionally used involves changing Z' by using an armature having two windings and two commutators which may be connected either in series or in parallel. 199. This has the effect of imparting a pronounced drofp to the speed characteristic (Figs. Speed of Shunt Motors. that is. moreover.MOTORS 217 against the direction of rotation. namely. of the armature is In this method the effective resistance increased by connecting in series with it (but \\\ Baluucer. Inspection of the fundamental equation for the speed of a motor 134. the downward slope of the char- combined resistance of the armature winding and external resistor. t ] (a) Rheostatic Control. as in Fig. r'. Regulation of n E t ia r' reveals the fact that there are three principal methods for regulating the speed. acteristic being proportional to the . the speed will fluctuate between rather wide limits as the load changes. and so convert the machine into one having the charactera differentially istics of wound motor. but is frequently convenient in laboratory investigations and in special tests. Speed regulation of motor by means of voltage control. rheostatic control. the interpoles will produce a component of flux in opposition to that of the main poles.

etc. 200. generator makes it possible to impress upon the armature of the motor a number of different voltages. the smaller limit holding for large motors. mediate speeds may then be secured by adjusting the flux by means of a rheostat in series with the shunt field winding. The variation in speed between full-load and no-load with normal voltage will usually be between 2 and 10 limit for small motors. in order that the Supply may remain sub- FIG. giving six different speeds. namely. somewhat resembling railway motor controller. 180 or 220 volts. Where uniform gradation of speed in either direction is required. 199. 40. boring mills. Inter110. 70. means of a special controller. changed by an ordinary With the arrangement indicated in Fig. This method is extensively used for driving machine tools. It has the disadvantage of requiring a considerable investment in copper due to the extra wires of the distributing circuits. Diagram of connections of Ward Leonard system of speed control. as in the operation of the turrets of battleships or in steering by . or bus-bars by means of a balancer set. tially constant and will be approximately proportional to the impressed voltage. The armature connections are stantially constant. 150.218 (b) PRINCIPLES OF DIRECT-CURRENT MACHINES Subdividing the voltage of the main Voltage Control. such as lathes. 199 it is possible to impress six different voltages upon the motor. to each of which there will correspond a definite speed characteristic such as is illustrated in Fig.. as in Fig. the larger It should be understood that the motor field connections are such that the voltage impressed on the shunt winding is not changed when the armature is switched from field flux one circuit to another. For any given impressed voltage the speed will be substan191. per cent.

If the Field Control. is separately excited from the main supply lines and its armature is supplied from an auxiliary generator G. 201. The motor M Fig. smooth variation of the voltage impressed upon the motor. . instead of driving the generator G by a motor. but is naturally expensive because of the auxiliary (c) motor-generator set. This method is very effective. simplest and cheapest method of reguthe of a shunt motor is that in which the flux is lating speed varied by means of a rheostat in the shunt field circuit. tor is excited from the constant voltage supply line. any The field of the generaother form of prime mover may be used. in either direction. be- yond this point the field intensity at the pole tips becomes so weakened by armature reaction. 200.MOTORS electrically controlled rudders. by means of M f a reversing field rheostat. in this way it is possible to obtain a FIG. Sectional view of Lincoln adjustable speed motor. the 219 be used. all resistance of the rheostat being cut out. The machine operates normally with a nearly saturated magnetic circuit. the latter being driven which takes its power at constant speed by a shunt motor from the line. the speed may be approximately doubled by weakening the field current. Ward Leonard system may is to be regulated. and may be adjusted from zero to a maximum. whose speed . especially under load conditions.

.m. In this way the main field may be varied through a wide range without producing sparking.220 PRINCIPLES OF DIRECT-CURRENT MACHINES that commutation is seriously interfered with. They are generally provided with a controller which serves not only to start the motor. but also to vary its speed as desired. and to reverse its direction. Consequently this method is limited to those cases in which a very moderate range of speed will suffice. the interpoles always producing a field of the proper strength to reverse the current in the coils undergoing commutation. as in machine tool operation. 202.f. Lincoln adjustable speed motor driving pipe cutting machine ture. the winding of the auxiliary poles being so designed that the m. Interpole motors are used to a very large extent where variable speed is a necessity. The interpole motor affords means whereby a wide range of speed is or 6 to possible. a ratio of maximum to minimum speed of 5 The principle of the interpole 1 being fairly common. of the armature is either exactly balanced or else slightly overcompensated. made motor involves the neutralization of the armature reaction of the motor by placing auxiliary poles in the axis of commutation and exciting them by the same current that flows through the arma- FIG.

the motor here being geared to a pipe cutting machine. the speed can be varied by mechanical In the Lincoln devices which change the length of the air-gap. thereby weakening the field. therefore. Applications of the Series Motor. as load is makes this type of the motor its increased series speed . 203. Fig. so that as the armature is moved sideways by means of the handwheel the effective length of air-gap may be A range of speed of 10 to 1 is increased or decreased at will.. and Engineering Co. field) (and high speeds Fig. motor. and the commutating field is therefore sufficiently intense to prevent sparking at the left upper limit of the speed. as shown in Fig. Stow multi-speed motor. in the armature in section shown 201. are avoided weak by using interpoles. 202 shows a similar machine made by the Reliance Electric FIG.MOTORS The methods thus 221 far described effect the variation of speed by machine. speed adjustable core is conical. 203. But the adjustment flux and. of the electrical circuits of the Commutation difficulties at readily obtained in the smaller sizes. a of handwheel means The speed is increased by drawing the plungers away from the armature. The rapid drop in 135. Speed variation is obtained in the Stow multi-speed motor by which are moved in and out of the hollow pole cores by plungers and bevel gears. the thin shell of iron thus at the pole tips becomes saturated.

forced Electric Co. for since the constant if . & Mfg. 205. for example.) ventilation. In railway machine especially valuable j railways and hoisting service work. Box frame railway motor. Westinghouse Elec. 204. the additional power current taken by such a motor sive. istic like the shunt motor are seldom used for the reason that the in going up a steep grade is excesspeed of such a motor will remain substantially the impressed voltage is constant. (General FIG.222 of PRINCIPLES OF DIRECT-CURRENT MACHINES for traction purposes. motors having a constant speed character- FIG. as in street and in rolling mills. Split frame commutating pole railway motor. Co.

MOTORS 223 FIG. 204. FIG. 207. Commutator and brushes of motor of Fig. Brush holders of motor of Fig. 206. . 204.

224 PRINCIPLES OF DIRECT-CURRENT MACHINES demands a proportionally increased on the other hand. Characteristic curves of motor of Fig. to a certain extent. 204 illustrates a recent type of railway motor made by the General Electric Company. A too rapid cutting out of the starting resistance results in very heavy current. 208. while in the shunt motor the torque increases less than proportionately to the current. It is of the box frame. Electric Co. . exces- way and automobile and a wracking of the armature winding. hoisting and rolling required to climb the grade current. will slow down as motor. the load increases.) (General service. automatically preventing an excessive load. The series mill service are generally of the totally enclosed type. Fig. the motors must be waterproof and of rugged construction to withstand the rough usage to which they are subjected by reason of poor roadbed and improper handling of the starting controller. at the same time it develops a torque more than proportional to the current. 204. in particular. In rail- 100 FIG. Series motors for railway. commutating pole type with forced ventilation. and. the inlet and outlet for sive torque. automobile. tending to maintain a constant load on the" system.

The horse-power of a has little in motor rating railway significance determining its in Fig. Figs. Co. 208. 205 shows a split frame commutating pole motor made by the Westinghouse Electric and Manufacturing Company. 136. are shown suitability for a particular equipment. 205 in Fig.MOTORS 225 In the box the cooling air being at the pinion end of the frame. Elec. 206 and ^07 show the commutator and brush rigging of the motor illustrated 20 40 100 120 140 1GO 180 200 Amperes FIG. 204. Fig. & Mfg. (Westinghouse The characteristic curves of the motor of Fig.) 205. 204 in Fig. Cycle of Operation of Railway Motors. those of Fig. the nominal horse-power rating is defined as that load which the motor will carry for one 15 . 209. 20. frame type the armature may be removed from the frame through the opening at the commutator end.9 Characteristic curves of motor of Fig.

The torque per motor then be constant. The current has its largest value during the starting or acceleration period. Fig. After the time Ofri. at the commutator and 75 C.m.f. a\ l>i Time in Seconds d Speed-time and current-time curves of railway motor. but at a steadily decreasing rate.226 PRINCIPLES OF DIRECT-CURRENT MACHINES hour without exceeding a temperature rise of 90 C.f of rated value. including stops. and during this interval the current will decrease from the initial constant value Oe in the manner indicated by curve fg.E. as represented by the curved line ab. and the speed of the resulting car will increase uniformly.E. and the motor covers being arranged to secure maximum . hence the heating is largely dependent upon the number of stops in is any given schedule. hence the draw-bar pull and the acceleration will also be constant.I. the current is See 1914 Standardization Rules. 210. the speed will continue to increase. the motor being tested on a stand with an impressed e. .m. The cause e. as indicated by the line Oa. the resistance in series O FIG. 1 of the decreasing current is the increasing counter due to the rising speed. 210. will When the resistance is all out. When the car or train started. with the motors should be cut out step by step in such a manner that the current through each motor remains practically constant until all the resistance is out of the circuit. 1 ventilation without external blower. at any other normally accessible part. A. In any given case the motors must be so selected that they will not overheat and the heating depends in part upon the average value of the square of the current taken throughout the whole of the working period.

parallel control is a much more economical method than if each motor had its own starting rheostat. this is then cut out. and the two pairs are then connected in series-parallel just as though each pair were a single machine. motors were permanently in parallel with a single resistance for For example. The brakes are then applied and the speed rapidly falls from cci to zero. and after the resistance has been cut out. The slope of the line Oa is the acceleration of the car. the value ordinarily used varies from 1 to 2 miles per hour per second. the connections are quickly changed so that the motors themselves are in parallel with a resistance between them and the line. in seconds. assume a two-motor equipment with the following data: E = / r t = = = line or trolley voltage current per motor during acceleration period resistance of each motor duration of acceleration period. so that finally the motors are in parallel The elementary diadirectly across the full voltage of the line. the motors are usually connected in parallel in pairs. the At the moment such that motors being in series (Fig.MOTORS shut off. the starting rheostat must have a resistance of R\ ohms E or Ri = E - 2r . 21 la). Series-parallel Control. The series. In four-motor equip- resistance ments. 211. gear ratio and type of motor are known. or than if the starting purposes. The broken line Oabcd is called a speed-time curve. gram of connections is shown in Fig. In cars having a two-motor 137. and its area is proportional to the disin the tance traveled by the car in the time Od. of starting. equipment the motors and starting resistance are at first all conthis in nected in series. the speed accordingly falling manner shown by line be. and turn determines the draw-bar pull. torque and current when the weight of the car. 227 and the car allowed to coast.

I 2 R z t watt-seconds. and continue to take / amperes. But if the speed is to be increased at the original rate. the counter e. also increases uniformly. The resistance 2 must then have such a value that R = E ohms and the energy lost in the rheostat during the second half of the acceleration period will be w. and the motors then switched to the parallel position (Fig. but as the motor speeds up at a uniform rate under the assumption of constant current. series-parallel control. In order that there may be no break in the smoothness of the acceleration. During the first half of the acceleration period the energy lost in the rheostat is then watt-seconds. 211. . the in parallel FIG. each motor receives half of the line voltage and this condition may be maintained efficiently if it is desired to continue run- ning at reduced speed. 2116).228 PRINCIPLES OF DIRECT-CURRENT MACHINES loss in the rheostat at the first instant is The then at the rate of watts. as in Fig. and in order to keep the current constant the resistance PRi must be cut out at a uniform rate. at the first instant after the transition has been made the resistance #2 must consume E/2 volts since the remaining E/2 volts are taken up by the motors. each motor must 210. At the instant when all of the resistance Ri is out of circuit.m.f. All of the resistance should be out of circuit in a time ~ seconds. motors must be put and a new resistance R 2 inserted between them and the line. Elementary diagram of connections.

and have the characteristic feature of not breaking the power circuit during the transition from series to parallel . so that the current does not rema^i by constant throughout the. they are used for single motor railway equipments. and for cranes and hoists. The successive changes in the starting resistance and the change from sefies to parallel connection are accomplished by means of a controller. type R controllers are those in which rheostatic control is used. acceleration period. in positions of the conyoller in full parallel. controller varied changed from notch to notch the resistance i^ amounts. Thus. Fig. ipspecin these positions there because are called running points tively.MOTORS The total loss in the rheostat is then 229 W = WRi + WR2 R = ?*t(-T +Rz) = ~ HI*t(j r ] watt-seconds. mining locomotives with one controllers or two motors. 210' but in reality this represented by assumes a saw-tooth form lying partly above and partly hfelow is finite the desired constant value. 4s absolutely curve line the ef. in parallel. As the. Railway Controllers. as in Fig. Type K are designed for series-parallel operation of two or more series motors. now. all other positions are called resistance points except in the interval in which the transition from series The two which the motors are in full series and to parallel connection takes place. with a resistance of Rz ohms between them and the line.parallel arrangement. the motors had been originally 2116. riesor exactly twice as great as in the case of series-parallel 138. Controllers are commonly designated by characteristic letters which indicate the type to which they belong. If. The loss in the rheostat amperes to 'flow through each would then be t = /E - \ Pi(j watt-seconds rj control. the value of 3 would have to be r E R = 2I~2 hmS * in order to allow a current of / motor. is no loss in the rheostat. without the customary series.

Fig. Co. Westinghouse Elec. and include the feature of opening the power circuit during the transition period.230 PRINCIPLES OF DIRECT-CURRENT MACHINES connection. & Mfg. Type L controllers are also designed for series- parallel control of series motors. this type is now seldom used. and in addition allow the motors to run as generators for energizing magnetic brakes of the axle or track type. its function is to create a powerful magnetic field at the contacts between the stationary contact fingers and the segments on the controller spindle. 213 shows the succes- FIG. The oval shaped part near the middle of Fig. This field is so directed as to blow sents a out the arcs that form on breaking the circuit. 212 represents a K-10 controller made by the Westinghouse Electric and Manufacturing Co. instead of a single magnetic blow-out coil. Type B controllers have the usual power circuit connections. Fig. 212. K-10 controller. The diagram of .. sive stages of the connections. 214 repremore recent type of controller made by the General K Electric Co. there are individual blow-out coils for each contact. and Fig. 212 is a solenoid connected in the main power circuit.

It will be observed that this diagram differs from that of Fig. two handles. Motor Motor 2 [EQ-o-AA/Wo-AAAAr- FIG. one All controllers. ously in circuit. 213. such a manner that the reversing handle cannot be moved unless the main handle is in the "off" position. have for the usual operation of accelerating the car. in R These its motion. Successive stages of connections. in Fig. K-10 controller. with the exception of certain types. 213 in that during the transition period the latter involves short-circuiting and immediately thereafter open-circuiting one of the motors (or pair of motors) while in the latter both motors are continuNotch Controllers A' 10 and K\\ 1 Res. and the main handle cannot be moved the other for the reversal of the direction of two handles are mechanically interlocked .MOTORS 231 connections of this controller are shown in Fig. The system of transitional connections shown 215 is called the bridge control. 215.

Railway motor controller with individual blow-out Electric Co. called type M. the armature to with respect ings If the car is running and it is desired to reduce speed. or motor car and but where several motor cars and trailers are to be operated as a train. the controller handle should be turned quickly to the off position and then brought back again to the proper notch before the speed The reversing handle changes the direction position. coils. General has fallen too low. the multiple-unit type of control. The controller for this service carries only a . trailer. must be used. 214.232 PRINCIPLES OF DIRECT-CURRENT MACHINES is unless the reversing handle in either the forward or reverse of rotation field windterminals. A slow turning off is apt to draw destructive arcs at the contact fingers. of the motors by interchanging the connections of the FIG. perfectly feasible in the case of a single car. A characteristic feature of all of the controllers described above is that the main current This is passes directly through them.

and this current actuates electromagnets which operate contactors that control the main The contactors are usually mounted in current. 215. the auxiliary cir- cuit being extended from end to A system similar to the type end is of the train. waterproof iron cases under the car bodies.MOTORS 233 small auxiliary current supplied by the line. Bridge control. series-parallel system. serves to operate the contactors of all the motor cars in the train. . M also used in large single cars. called a master controller. In this system a FIG. single controller.

^rZ^ To Lights and Pu MU Tripping Sw Main Switch Trolley Terminal Auxiliary Contacts in Controller T~ FuauBos FIG. 216. FIG. Railway motor controller with auxiliary circuits.234 PRINCIPLES OF DIRECT-CURRENT MACHINES Fuse and Switch m TPT Kicking Coil . 217. Contactor with cover removed. .

thereby affecting the torque instead of the speed. The counter e. the manner discussed in parallel. is shown in Fig. so that its impressed voltage must also increase. 139. in which case the forward motor will speed up and spin the wheels. the torque. This type of control which the trolley voltage is feeder at times of heavy load for in such a case it the current passing through the auxiliary circuit . constant the speed would have to vary in direct proportion to To obviate this variation of speed. when connected in parallel to the same supply circuit. 217. as at the end of a long apt is possible that may be insuffi- Two or more shunt motors designed for the same voltage. An interesting case of unequal division of load between series wound motors of the car is is afforded rails. of the forward motor will increase as its speed rises.f. of full-load current. 173. and in particular to keep the speed constant. will develop approximately constant torque if the brushes are kept in a fixed position. 191) are identical in in connection with Fig. Series wound motors when connected in series in a constant- current circuit. therefore each will develop the If same the weight on the forward trucks is fairly light. more heavily loaded than the forward end on turnthe controller handle to the first notch.MOTORS 235 where very heavy current through the controller itself might be An elementary diagram of connection of such a objectionable. The same thing is true of series motors operating cient to operate the contactors. are identical. the same current will ing flow through both motors (or both pairs of motors) since they are in series with each other. and the contactors are shown in is not suitable on a system in to be low. and with their shafts rigidly coupled. VI) are provided with regulators which change the position of the brushes. controller Fig. Division of Load between Motors. plotted in terms of per cent. the series motors in the Thury constant-current system (see Chap. that is. if the speed curves. will divide the load in proportion to their capacities provided their speed-current curves (Fig. adhesion between the wheels and the rail may not be sufficient to prevent slipping.m. 216. but if the torque were the load. but any increase of the voltage on the forward . grade on slippery by the case of a car starting on an upAssume for example that the rear end .

the elements are successively switched from a path or circuit which the current has one direction to an adjoining circuit which the current has an opposite direction. f. = T. of the order of 0. The study of the commutation process therefore has for its object the determination of the conditions which will coil is zation winding element may be represented by a diagram such as Fig. 237 6. Fundamental Considerations. The period of commutation is of very brief duration. AB in the manner shown by such typical curves as a. in which ordinates Immerepresent values of current and abscissas the time. that the currents in all of the armature circuits are therefore. . As the commutator segments to which the terminals of the indi- vidual winding elements are connected pass under the brushes.CHAPTER VIII COMMUTATION 140. after the completion of the commutation it must have a value result in sparkless operation.002 second. and it may easily happen that the reversal of in in current is either retarded or unduly accelerated. e. or period of commutation. c. of the before the commutation diately beginning period AB. the current at the end of the period will tend to have a value which differs from that of the circuit to which the commutated about to be connected. in either case. 218. the current must be reduced from its original value to zero and then built up again to an equal value in the opposite direction. Each of the a parallel paths comprising the entire armature winding consists of Z/2a turns in series in each of which the current is i a /a amperes. and the result of the final equaliis a spark between the brush and the commutator segment. The time variation of the current in a assuming that the winding is symmetrical and. d. During this transition period.0005 to 0. the current in the coil under consideration has the value + i = /a /a. In the time interval i . the current may vary the same.

but it may involve such large localized current densities at the contact surface between as to lead to sparking and. Curve c indicates a uniform transition of the current from its initial to its final value. perto haps. its final value. Types of short-circuit current curves. with a zero rate of change at the end of the commutation period. commutator segment and brush Curve b represents a case in which the current comes to its FIG. Such a short-circuit cur- rent curve would generally result in satisfactory commutation. the contact surface the curve is a minimum. When this occurs the commuta- tion Linear commutation is very desiras will able. each of which corresponds to a definite set of physical condiThese curves are called the short-circuit current curves. appear later. value smoothly. Curve a shows that the current has been reversed too rapidly. so-called Curve d represents the is " one-half of a sine curve. 218. This will generally result in satisfinal factory commutation. for.238 PRINCIPLES OF DIRECT-CURRENT MACHINES tions. it gives rise to uniform current at the brush contact surface and the loss of power at density is said to be linear. such a condition being characterized as overcommutation. Here the current may reach its overreaching proper final value without a spark. glowing (incandescence) of the brush and certainly to excessive loss and heating and deterioration of the brush. . sinusoidal" commutation.

practice based upon more or less empirical rules has so far outstripped theory that manufacturers commonly guarantee sparkless operation between no-load and Notwithstanding this 50 per cent. Physical Basis of the Theory of Commutation. No account has here been taken of the effect of mechanical irregularities such as vibration of the brushes. Under such conditions sparkis. fact. It should be understood that these curves represent only the more important cases. the commutation characteristics cannot be predetermined with anything like the degree of accuracy that is possible in the calculation of the general performance characteristics. of self-induction due to the changing current in the coil. the armature winding being assumed to be symmetrical. Such mechanical defects will invariably produce sparking even though the magnetic and electrical conditions are otherwise perfect. overload with a fixed setting of the brushes. the self-induced e. that is. In practice. and in case the short-circuited coil is in inductive relation to one or . the curve is tangent to the drawn through B. Even though the final value of current may be correct. unevenness of the commutator surface. Vibration of the brushes causes the short-circuit current curves to take on a saw-tooth form. subject always to the condition that the initial and final values of current must be equal in magnitude and opposite in sign." that the current is not reversed with sufficient rapidity.m. The theory of commutation is much less advanced than that of other parts of the theory of direct-current machines. " Curve / shows undercommutation. the short-circuit current curves may assume an infinite variety of forms. It involves the -fact that the coil undergoing commutation has induced in it an e.f.COMMUTATION Curve e 239 change of current vertical line represents a limiting case in which the final rate of is infinite. just as in the case of overcommutation. that is to say.' The elementary theory of commutation is relatively simple and has been extensively discussed by numerous writers. 141. acting always in such a direction as to oppose the change of current.f.m. this condition may involve excessive current density under the brushes and hence possible glowing. ing would invariably result. etc.

f. acting in it. are afforded The ohmic drops at the transition surface between the commu- tator segments and the brushes. should in general be so directed as to neutralize the retarding effect of self-induction. Vol.f. it varies.f. the process is called resistance commutation commutation.fs. and in the short-circuited coils and their connecting leads are almost as important as the e. upon by any extraneous inasmuch as the self-induced e.m. in the coil. these factors will be . material and chemical structure of the contact surfaces. For these reasons the theory may be designated the " inductance" theory. A found very complete presentation of the effects of in Arnold's Die Gleichstrommaschine. to this end carbon brushes are employed. all of I.f. Examples of this type of machines by railway and hoisting motors in which the brushes are permanently set at the geometrical neutral because of the frequent reversal of direction of rotation. the temperature. in which case the process is referred to as voltage however. or the reversing field due to a commutating pole and that the rotation of the coil through this field produces a generated e. of field the fringing field near a pole tip. and upon the nature of the current itself (continuous. if there is no generated e. the direction of the current. it resembles that of the electric arc in many of its properties. is controlled only by the ohmic drops in the coil and at the brush contact surface. Resistance commutation is largely relied upon in machines having a fixed brush position and no special commutating devices such as interpoles. that is.m. As has been previously pointed out. There is also to be considered the fact that the short-circuited coil may be situated in a magnetic there will also be induced in it an e. due to self- and mutual-induction and to rotation through the magnetic field. moreover.m. the high contact resistance serving to keep the short-circuit current within reasonable limits.f. If.m. tact pressure 1 and the 1 relative velocity of the surfaces. this generated e. or pulsating) . the commutated coil is not acted field. The contact resistance between commutator and brushes is very much more complex in its nature than that of ordinary metallic conductors.m.m.240 PRINCIPLES OF DIRECT-CURRENT MACHINES coils more which are simultaneously undergoing commutation mutual induction. for it is dependent upon such factors as the current density. with the conalternating.

and breakdown in the form of sparking may be expected when the amount of the stored energy exceeds a critical value. For the sake of simplicity there will first be considered a simple ring winding in which the brush width b is equal to the width a commutator segment. On this basis. Case of Simple Ring Winding. From contact surface this standpoint the drop of potential across the is the sum of the counter e.f.m.f. The transition layer between commutator and brush is the seat of an energy storage.COMMUTATION It is . so that the reversing e. as shown in Fig. It has been explained in Chap. is not constant. and the true ohmic drop. Under this condition only one |8 of winding element will be short-circuited at a time.f.m. 16 appreciably larger.m. But during the commutation period the short-circuited coil moves through the fringing field from a position in which the generated e. the quotient obtained by dividing the observed drop by the current is then not a true resistance.m. but what may be called an effective resistance.the current across the between commutator and brush causes an sets ionization of the gaseous layer between them and that this up a counter e. similar to that encountered in the arc stream. made up of the true resistance plus a fictitious resistance equivalent in its effects to the counter e. as by the pilot brush method de- .m. has a certain value to another position in which the generated e. of sufficient magnitude to balance the retarding effect of the self-induced e. and the effect of the mutual induction of other coils is eliminated. THE INDUCTANCE THEORY 142. If the distribution of the flux in the air-gap is is determined experimentally.m. this is confirmed by an experi- ment of Professor Arnold's in which the current density passing from a carbon brush to a metal surface was raised until the brush glowed. in order that the fringing field at the leading pole tip may generate in the short-circuited element an e.f.f.m. 241 transition surface probable that the passage of . General Equation. 219.m.f. neither current density nor transition drop taken separately is a sufficient criterion of the sparking limit.f. but without producing sparking. VI that the axis of commutation must be slightly displaced from the neutral axis (in the direction of rotation in the case of generators).f.

will of flux distribution is between pole tips. become an integral part of the left-hand branch. 220. shows that immediately after has reached the brush. a part of the right-hand branch of the = ia /a.m. 220 represent. initial. at the beginning of the period. 219.m. t e. the current from the right- . FIG. FIG. and final stages of the commutation of a Parts a. Short-circuited element. coil C. of short-circuit of coil.f. the curve approximately linear for short distances therefore. in final position c. respectively. the commutating e. as a rule. the intermediate. in which coil C segment A2 is short-circuited. coil C has the through Similarly. The b position. All of the current and carries the current winding the brush from the two reaching adjoining paths must then pass lead 1. its current has coil is z' In position a the been fully reversed. closely represented by the function E = c e + kt (1) where t = time counted from the beginning of the commutation period h e = constant = commutating when = 0. Successive phases simple ring winding. V. and the combined current of the two paths must reach the brush by way of lead 2.f. and c of Fig.242 PRINCIPLES OF DIRECT-CURRENT MACHINES be found that. b. may be it scribed in Chap.

the contact resistances of areas A\ and RI = T . Taking the counter-clockwise we have L-jr= e. drop drop drop drop drop in coil C (positive) in lead 1 (positive) in contact area AI (positive) in contact area 2 (negative) A +iRc >T .f.m. . In the closed circuit consisting of coil C.f.m. of self-induction (negative) = = iiRi iiRi = iiR'i = izRi = Ec = iR c ohmic ohmic ohmic ohmic ohmic di dt in lead 2 (negative) commutating e._ \ t vJ _T__ J. and vice versa. 221. some arbitrarily assumed positive FIG. similarly.f.m. Current paths in short-circuited coil. R* t = Rb A T 2 are. direction as positive. in accordance with Kirchhoffs law. t Counting the time from the beginning of the short-circuit C. a drop of potential due to a current flowing in the positive direction is to be taken with the negative sign. 2.l/1) ~* pjy j. or in opposition to. is to be taken positive or negative if it acts in.244 PRINCIPLES OF DIRECT-CURRENT MACHINES let Now R = RI = Rb = c resistance of coil resistance of each C commutator lead of coil resistance of contact area of the entire brush. the to sign) must of all the potential drops (with due regard equal zero. (positive). 1 c t T -C C*0 RI +E = (4) . sum The sign of any e. respectively. respectively. commutator leads 1 and contact areas A\ and A 2 the two commutator segments and the brush. direction.

t'Qfl P J? ac -flfr-i 97? ' ^/t5?-o "f* xaf W ^ n U or A\ c?^ 1 / -^f RbT-L die Uber den Kurzschluss der Spulen und Vorgange bei der KommutaRiebesell. and i during the commutation process. when t = = iQ E = c e + 7? j T7 O'o hT = E T = 2Rbio i) i) V The value by T R --m i(*o + i b t = o or indeterminate. The complete results of the integration are not essential in most cases. since it is generally the end of the commutation period that is most important so far as sparking is concerned. however. 1905. of the expression (* _ } can.COMMUTATION which 245 may be written L where S+ iR + f'-^t <*'' + i} ~ ^T l (i - + & - o (5) R = R + 2R c This equation involves the justifiable assumption that the resistances of the commutator segments and of the brush are negligible. Kiel. be evaluated differentiating numerator and denominator separately with respect to the independent variable t. n o . Discussion of the General Equation. by Paul . 144. i = i Q) and when t = T.i = iQ. tion des Stromes eines Gleichstromankers. The complete integration of this differential equation has been worked out subject to certain conditions 1 and results in an equation of the form subject to the terminal conditions that when t = 0. At the last moment T. giving di T The di i + i t dt di dt di T ~TT at 1 general equation then reduces to T j j.

case. If may be deduced several impor- j~ = oo. the e. E T will generally be small. If. hence . the cornmutating e.. then.m. + > LJ 1 is in 3. The condition j-~ > 1 can be put into another form which has a simple physical . /di\ ET = in i (R + 274) will (9) which case the conditions for good commutation This state of affairs is be favor- able. iQ ^> and if (R + 274) > ET (7) ~r < . 1. 218. 7? shown in curve 145. as follows: 1. close to the neutral axis. i.f. E c and also its terminal value. Fig. If this were the infinite. E T< i Q (R 274). ET .m. that case of overcoil is commutation the final rate of change of current in the positive in sign.e. i (R + 274) < ET (8) Inasmuch as most machines are required to operate with the brushes in a fixed position. therefore.f. rate of change of current will be zero. of self-induction. Criterion. Modified V7 Form of Sparking 6.e. the final value of i -rr will be infinite. hence the numerator and denominator of equation the (6) must be of same sign. the rate of change of current should always be negative. 2. -j must in general 218a shows from in unity. 1. Inspection of Fig. i.246 PRINCIPLES OF DIRECT-CURRENT MACHINES From the last equation (6) there tant conclusions. (-77) \(lt J t = = T provided (R + 274) differs from di L-rr. most cases the The final criterion for satisfactory commutation.. since this condition of overreaching is to be avoided. would also be and sparking would 7? /77 result at the trailing edge of the differ brush. Therefore. Usually. 1.

that e r < 1 is the condition to be satisfied. for a given capacity of machine. or average e. The criterion R b T/L > Ishowsthat the brush contact resistance Rb and the time of commutation T must be large.f. This accounts for the fact that carbon brushes are ordinarily superior to metal brushes. 247 the inequality multiplying both sides of by we have L=e face r (10) The term 2i Q R b is the drop of potential at the brush contact surand is usually of the order of 1 volt with carbon brushes. 2i Q L/T. The self-inductance L the elements with a small can be kept within limits by designing number of turns. therefore. hence e r = 2i QL/T is the average reactance voltage. the inductance being proportional to the square of the number of turns. and that L must be kept small. It follows. a correspondingly large number of segments. . 222. though in railway motors the number is frequently 4 or 5. L can be kept down by limiting the axial length of the armature. Furthermore. of self-induction. and a small number of turns per segment and per element. since the former have the larger contact resistance. this means.COMMUTATION interpretation.m. The term 2i /T is the average rate of change of current during commutation. FIG. hence the possibility of using a large commutator. a relatively large diameter. in large machines the elements are designed with only a single turn. since the value of L is determined by the number of flux linkages per ampere of current in the coil. ordinarily the number of turns per element should not exceed 2 or 3. Sinusoidal commutation.

1 = . e max < 1. M. in case of linear commutation. 222. > from the previous discussion that if e r > 1. Linear Commutation. or e r = 2i QL/T < 1 is frequently ex- H. however. referred to . Hobart has proposed a method which assumes that the commutation curve (Fig. maximum ordinate being ? whose equation.248 PRINCIPLES OF DIRECT-CURRENT MACHINES It would appear at first glance that an increase in brush width would give good results because of the increased value of T. T -2t jT~ yT di _ ~ dt If) and 2i Q T some trans- Substituting these values in (5). 154). the commutating e.f. the current i at any instant t is given by . is I = 2-irt IQ COS om is The instantaneous value of the self-induced e. is the one-half of a sine curve of period 2T. curve c. 7T _ 7T Hence. 2i *0 ' 4 I = . Fig.f. there results after formation "? ~ % (T - 2t) ] (12) . 146. 218. or e max if er < 1. Thus. But The criterion R b T/L > 1. which 7T Hobart _ 2io j calls the reactance voltage. emax< 2. It follows 2. must be so adjusted that E T >i (R + 2R b ). origin O. 21Sd) pressed in a still different form. Equation (5) can be utilized to determine the conditions necessary for a uniform transition of the current from its initial to its final value.57.m. As usually stated.m. as indicated in Fig. this apparent advantage is offset because of the fact that the wide brush simultaneously short-circuits several additional coils whose mutual inductance is equivalent to an increase in the selfinductance of the original coil (see Art. then and is its maximum value.

COMMUTATION It follows 249 from (12) that t when c = = 0. hence the current density i\ is -r- = i ^1 It jV2 = 2iin r A = constant. t T. that .m. iz 2iv But the contact areas A } and ! A = 2 are given t by rA T t T where A is the total brush area. E = ET = -=jr + Rj In other words. in this case the actual non-linear current can be tation.f. when the current density is uniform. 221. Thus. j. (13) of the can also be shown that the ohmic loss due to the resistance brush contact is a minimum in the case of linear commu- For let it is. as shown by equation (12). that not be satisfied for other loads. An interesting consequence of linear commutation is that the current density at the brush contact to Fig. must not only vary as a linear function of the time. is not linear. without special corrective devices. ii = = 9 2i . the commutating e. 221. referring iz = IQ i and if T- 2t it follows that . Fig. E = c e '= t (^r IQ f . is constant. therefore. T-t j. be assumed that the short-circuit current in coil C. but it must to it is directly proportional conditions for perfect linear commutation were satisfied for one particular load they would iQ. also change with the load since if It follows.

loss is a = 0. Ri and 7? 2 . i and an extra It current. i xt where the latter may have any general form. 147. that there are no differences of potential along the brush contact (in the peripheral direction) to be equalized by a flow of current. m we find that W c =- 4i Q *R b + i f *R b i x *R b ~-minimum if i x from which it follows that the if the commutation is linear. thereby giving rise to non-linear short-circuit current curves and a non-uniform current density. or tend to exist.i io -\. It follows. The Current Density at a Commutator Segment.m. then. and remembering that it = 21 . that if such inequalities of potential do exist. therefore. extra currents will flow along the brush and complete their paths through the short-circuited coil or coils. generated in the . the potential differences which produce the extra currents are due to the fact that the e.e..250 PRINCIPLES OF DIRECT-CURRENT MACHINES t. General Case.ij -|- ix The contact resistances at the areas AI and A2 are. T and # = 2 Rb j and the ohmic loss at the contact areas is Substituting for rp ?'o _ ii.f. follows then that = and i\ ii + ix = ?o -f. Of course. respectively. iz . i. The uniform current density over the brush width that is characteristic of linear commutation means that the drop of potential across the contact surface is everywhere the same and. thought of as made up of a linear current.

Consider. b. segment emerges from the brush. Phase 1. The current iz crossing from segment ii S to the brush is = iz ~\~ io> since in this position coil A = is carrying current from the i . i\ Die Gleichstrommaschine. I. the iz is ti io i\ the upward 2. a case where the brush width is 3. 438. for example. and let it be assumed that the current density at any given instant is the same over the entire area of that part of the segment covered by the brush. 1 may be used.COMMUTATION 251 short-circuited coils differs in form. from It is conthat which would produce linear short-circuit current. 223. phases of its motion may be recognized 1st. : Three distinct segment is covered by the brush.e. wide brush. respectively. and c. instant to instant as it passes under the brush. parts a. 223. left- hand branch 1 circuit. due current. the segment approaches the brush. Vol. direction of current through the segment being taken as positive. sequently important to determine in what manner the distribution of current density is affected by a non-linear short-circuit For this purpose the following graphical method. Phase The current crossing from segment iz S to the brush is ii Phase 3. i. p. to Professor Arnold. Assume also that all the coils successively undergoing commutation have identical short-circuit current curves. The current crossing from segment S to the brush 2d.. the 3d. The current density at a particular segment S will then change from JL i\ iz =io *i *2 ii=-o *s a FIG. as a time function.5 times the width of a commutator segment. 2nd ed. as shown in Fig. . b c Successive phases of short-circuit.

then _ until it cuts the ~ 'Me MP ^ is a constant length. and the current i\ = io i\ etc ab = be M . Current density at a segment. 224. curves C A and C B represent the short-circuit current curves of coils A and B. respectively. ce is the current density corresponding to the abscissa XL and since MP . the commutator has moved to the left a distance Xi (during phase 1). Pd is proportional to the current density at segment S. The current density is proportional to io ii _ be Xi Me it Draw vertical the straight line Mb and produce through P in the point d. They are drawn in their correct time positions with respect to the edge 1 of the brush. f_i _ FIG. across segment S is iz the current in coil A is ab. in the posiphase of the motion of seg- ment S When just beginning.252 PRINCIPLES OF DIRECT-CURRENT MACHINES In Fig. 224. the line tion OO' being the shown is in the figure the first axis of their ordinates. Projecting the point d across to e.

kQ> or.COMMUTATION It is readily initial reversal of 253 apparent from this construction that a too rapid the current in a coil (overcommutation) may result in excessive current density at a segment as it passes under the brush. hence the current density it proportional ^. of the At the same time the length with the brush to -. During the third phase of the motion. represented by IN line Qhl. the current across segment S. Curves of current density at segments. mk. The point Z is found by drawing the straight Continuing this method for several points in each of the three phases of the process. 225. the intercept fg between curves iz ii.. current density have been drawn upward with respect to the . or after a travel indicated by x 3} the current across S is iz i\ = iz + ^'o = hk o' FIG. to the scale previously adopted. segment still is is in contact is kQ. the current density curve afiyd of Fig. given directly by This intercept is also proportional to the current density to the . or represented by x% is } after a travel of the commutator i.e. C A and C B same scale as ce. the ordinates representing is obtained. 225 For convenience. During the second phase.

nor does it remain constant at any given point. Average local current periodically. the density rises suddenly from dc to e the ec. the current density at a given point in the brush varies FIG. segment S" then comes under point 7. that the current and 229. This is the curve of average local current density or PQ of that corresponds to the curve of commutation the local from 226 that 224. and again falls off to the value gf. a'0'y'd' and a"/3'Y'\$". For instance. which is just about to make contact with segment S'" At this instant . is . 228. Local Current Density at the Brush. and thereafter.. 226. represents density at point 3 of the brush. the current density is not the same at the same instant all along the arc of con148. in other words. respectively.254 PRINCIPLES OF DIRECT-CURRENT MACHINES Similar curves. the variation of current density of segments S' and S". etc.. Further Examples. Variation of in the case of linear tact of the brush. the results when plotted give the curve 1'2'3'.. until S'" has moved from under the point. 225 consider the point 7 of the brush.8' of Fig.. 227. curves MN current of current density have been constructed are short-circuit current curves. readily apparent densities may differ considerably from the average current density of the brush as a whole. /3'. and to the same time scale as that of curve afiyb. Using the above methods. of course. It is Fig. as read from curve a". the current density of segment S"' is ab. etc. current density changes in accordance with curve bd. etc. 226. 149. /3". In precisely line the same way the saw-tooth with cusps at j8.. show axis OCX. since is not shown. The case of obvious from the previous analytical discussion. the same as bd. as well as from the geometry of the construction. Except commutation. the variation of current If the average current density is found for various points along the brush. The curve eg is. density at points along brush. Fig. it shown They linear commutation for several types of in Figs. the point e being on curve a". in Fig.

Curves of current density at segments and brushes. 227. 228. 229. 227. Short-Circuit Current Curve Axis coordinates FIG. . 228. FIGS.COMMUTATION Short-Circuit Current 255 Curve Segment Density Axis of Ordiuatea FIG. Short-Circuit Current Curve Segment Density FIG. 229.

256
density

PRINCIPLES OF DIRECT-CURRENT MACHINES
is

the same at

all

times both at the segments and at
in

each point of the brush.
Figs. 227
of current
is

and 228 represent cases

which the rate of change

tion period. ties at the commutator segments as they

too great during the initial stages of the commutaThis results in correspondingly great current densi-

come under the brush, near the heel. current In Fig. brush and high average density in the the current coils causes a of reversal 228 the overreaching of direction of current at the receding segments. Fig. 229 shows a case of undercommutation, with consequent reversal of direction at the heel of the brush and excessive densities near the end
of the

commutation period. 150. Simultaneous Commutation
as the process of

of

is

Coils.

Inas-

much

commutation

in a coil

affected

by the

FIG. 230.

Simultaneous short-circuit of elements of lap winding.

mutual induction

of neighboring short-circuited coils, it is important to be able to predetermine the number and relative positions of those coils in the same neutral zone which are simultaneously short-circuited. In the case of the simple ring winding heretofore considered, all coils in the same zone are short-circuited by a single brush if b is the brush width and (3 the width of a commutator segment, the ratio b/0 fixes the number of coils short-circuited at the same time. This ratio is generally
;

a mixed number, and the actual number of coils short-circuited will vary alternately between the two integers lying on either side of it. In lap and wave windings, however, the conditions
are as a rule not so simple, since in a given neutral zone some of the conductors are short-circuited by a brush of one polarity, others by a neighboring brush of opposite polarity, as illustrated

COMMUTATION
in

257

The diagram represents a duplex lap winding Fig. 230. having the following constants:

Z = y =
It is clear
1

122

S =
-

61

p
2/2

=

6
19

a ^

= =

12
2.5

m=

=

2

2/1

=

23

= ~

from the figure that in the position shown, conductors and 4 are simultaneously short-circuited; a moment earlier conductors 1, 3 and 4 were short-circuited. The successive combinations of short-circuited coils can be conveniently studied by means of the following graphical method, due to Professor
Arnold. 1
1.

Lap Windings.
be observed that in the winding here selected brushes

It will

BI and B% are not identically situated with respect to the segments of the commutator in contact with them. This is a consequence of the fact that S/p is not an integer. Coil edges 1, 3, 5, etc., drawn in full lines to indicate that they occupy the tops of the slots, are connected to commutator segments which are correspondingly numbered in the top row of The other sides of the same coils, whose numbers are figures. l + 2/i, 3 + 2/i, 5 + 2/1, etc., are connected to segments which are numbered 1', 3', 5', etc. (i.e., dropping the term 2/1 and priming the numeral) in the bottom row of figures. A coil will then be short-circuited when the brush BI is in contact with any A similar pair of segments which bear the same numbers. arrangement
is

Now,

coil

indicated in the case of coil edges 2, 4, etc. edge 2 is connected to one on the left which

is

1 from 1. Segment 2 separated by a pitch 2/1 from 2, and by 2/1 But is therefore separated from 1 by J^ (2/1 1) segments. brushes B% and BI are separated by S/p segments, hence the relative shift of segments in the vicinity of B 2 with respect to

those at

B

i

is

A
and
it
is

=
|

-

M(j/i

-

1)

(15)

is

toward the

left

when

A

is

negative, toward the right

positive.

The simultaneous
1

In the case considered in Fig. 230, action of the two brushes can now be studied
I, p.

A

=

when %.

Die Gleichstrommaschine, Vol.
17

354,

2nd

ed.

258

PRINCIPLES OF DIRECT-CURRENT MACHINES

of a diagram like Fig. 231; take a strip of paper cut to the width of the hatched area to represent the brush and slide it between the two commutators; when it touches segments simi-

by means

larly

numbered, the corresponding

coils will

be simultaneously

short-circuited

FIG. 231.

Diagram showing

coils

simultaneously short-circuited winding.

lap

2.

Wave

Fig.

Windings. 232 represents a portion of a duplex wave winding having
122
2/i

the following constants:

Z =
y

=
2/2

61

p
|

=
=

6
2.5

a

=

4

=

=

=

21
coil

In the position shown,
circuited.

edges 122,

1, 2, 3,

and 4 are short-

FIG. 232.

Simultaneous short-circuit of elements of wave winding.

the segments connected to coil edges 1, 3, 5, etc., with corresponding numbers, and the segments connected to the

Number

other sides of the same coils

1', 3', 5', etc.; similarly with respect to the other coils occupying the same neutral zone, as 2, 4, 6. Brushes AI and A 2 which are of the same polarity, are not similarly placed with respect to the segments in contact with
,

them.

The brushes

are separated

by 2S/p segments, while the

ends of an element are separated by y

=

2 Sf

m

segments,

COMMUTATION
where

259

m =

a/p; segments

1',

3',

5'

are therefore shifted with

respect to A 2 by an amount m(3, as compared with the relative The positions of segments 1, 3, 5 with respect to brush A\. is positive (as in the case illustrated), to shift is to the right if

m

the

The short-circuiting of these elements left negative. can then be shown by drawing two commutators one above the
if

m

is

other, as in the

upper part of Fig. 233. Obviously, brushes B\ and B 2 and the segments in contact with them, are related to each other in the same way as are AI and A 2 Now Bi is separated from AI by S/p segments, while
,
.

segment

2' is

separated from segment

1

by

^ (y\

1)

segments.

FIG. 233.

Diagram showing

coils

simultaneously short-circuited winding.

wave

The displacement

of

compared with that

of

segments 2', 4', etc., with respect to B\ as segments 1, 3, 5 to A i, is then
t

A
and
is

= |
is
if it is

-

y yi
2(

-

i)

A
are
is

=
If

+

to the right if %), to the left
in

A

positive (in the winding considered The complete relations negative.

shown

Fig. 233.

a strip of paper whose width is equal to that of -the brush moved across the fictitious commutators represented by AI,
,

A%, Bi

and B 2 it will touch a series of similarly numbered segments and the corresponding coils will be simultaneously shortcircuited.

151. Successive

Phases

of Short-circuit in Coils of a Slot

The method described above may be used

to investigate the order in which the coils occupying a given slot undergo commutation. Two distinct cases may be distinguished:

260
1.

PRINCIPLES OF DIRECT-CURRENT MACHINES

Coil edges lying in the same radial plane (one above the other) enter and leave short-circuit simultaneously.
2.

Coil edges lying in the
1.

same

and leave

short-circuit at different times.

If the coil edges are numbered in accordance with the in Art. 73 of Chap. Ill, and illustrated in described system and 2 of a two-layer winding will occupy the 1 coil sides Fig. 86, same radial plane, and so also will 3 and 4, 5 and 6, etc. Refer-

Case

ence to Figs. 231 and 233 shows, therefore, that if coil sides 1 and 2, 3 and 4, or, in general, any two in the same radial plane, are

and leave short-circuit simultaneously, there must be no displacement between the correspondingly numbered commutator segments; in other words, the condition to be satisfied
to enter
is

that

A =
or
2/i

-

|

X
1

(tfi

-

1)

=

=

~+

(16)

For example, consider the case of a simplex lap winding having coil edges per slot, a brush width of 2^ segments, and A = 0. With the help of a diagram like Fig. 231, but with A
six

equal to zero, it is readily shown that the successive phases of the short-circuiting of neighboring coils will follow the order shown in parts a, &, c, etc., of Fig. 234, where the shaded coils
indicate short-circuit conditions.

During a

brief interval the
is,

condition

shown

in

diagram

c

will exist,

that

all

the

coil

edges in a slot will be simultaneously short-circuited; a little later, as in diagram e, six coil edges are again short-circuited, but
four are in one slot and two in the next
slot.

A

study

of Fig.

234 shows that when

coil

edges

1

and 2 leave

short-circuit they are subject to the effect of mutual induction from the simultaneously short-circuited coils 3, 4, 5 and 6, all of which occupy the same slot. When coils 3 and 4 leave shortcircuit they are subject to the mutual induction of coils 5 and 6, which are in the same slot, and of coils 7 and 8, which are in the

next

slot; obviously, because of this separation of the short-circuited group of coils, the inductive effect upon coils 3 and 4 will be smaller than in the case of coils 1 and 2. Similarly, when

COMMUTATION
coils

261

5 and 6 leave short-circuit they are acted

upon by the mutual

induction due to the simultaneously short-circuited coils 7, 8, 9 and 10, all of which are in the slot adjacent to that occupied by 5 and 6, hence the inductive effect upon these two coils is still
less

than in the case of

coils

3 and 4.

The commutating

condi-

tions are, therefore, not the same in all of the winding elements, and their short-circuit current curves will have different forms.

An additional disturbing feature arises from the fact that when the successive coils of a slot, as 1-2, 3-4, 5-6, of Fig. 234 break contact with the brush, they are not identically situated with respect to ,the adjacent pole tip, consequently the e.m.fs. generated in
(a)
1

(b)
5
1

3

3

5

135

(d)

IDD
IDD

II
II
246

Dll Dll

Dll

IDD IDD

FIG. 234.

Successive phases of short-circuit in adjacent

coils,

A =

0.

each of them during the final stage of the short-circuit (by cutting through the fringing field) will be different. This is due to the fact that the successive commutator segments are evenly spaced,
while the
234, coils 1

being grouped in slots, are not. Thus, in Fig. and 2 are ahead of 3 and 4, etc., with respect to the direction of rotation, and their short-circuit terminates when they are in a weaker field than that which acts upon 3 and 4 when the latter leave short-circuit. Similarly, coils 5 and 6 leave shortcoils,

circuit
field

when they are subjected to the action of a still stronger than that which acts upon 3 and 4. If, therefore, the com-

mutating e.m.f. acting upon coils 1 and 2 is just sufficient to overcome the e.m.f. of self- and mutual-induction therein, it will be more than sufficient to balance the smaller inductive e.m.f.

262
in 3

PRINCIPLES OF DIRECT-CURRENT MACHINES

and 4, and much too great in coils 5 and 6. In the latter coils there will be a condition of overcommutation, and under these
circumstances every third commutator segment may become blackened because of the possible excessive current density. In order that there may be no marked difference between the field
intensities acting

upon the various coils of a slot while they are undergoing commutation, the angle subtended by a slot should be small. For this reason the number of slots per pole should not be less than 12, and preferably greater than 12, and the angle subtended between the edges of a brush should not exceed one-twelfth of the angle from center to center of the poles. 1

oo
1

(O
5
1

3

1

3

3

5

II

III
II
(/) 79

an
ii
135

IQD
1

CO 3

5

79

11

135

11

00 79

11

II

IDD

II

III 246
FIG. 235.

DM
8
10
12

IDD
8
10
12

246

246
coils,

8

10

12

Successive phases of short-circuit in adjacent

A =

1.

The order of commutation illustrated in Fig. 234 can occur only in full-pitch windings, since it is in such windings that the back pitch, y if is made nearly equal to 2S/p. Chord windings (fractional pitch) are, therefore, characterized by the condition

A^O.
Case
occurs
2.

It follows

from the above analysis that the second
variation of this case

case arises

when A ^ 0. An interesting when A = 1, that is, when

->

(17)

Thus, if a winding has pitches that satisfy equation (17) and is arranged so that each slot contains six coil edges, the brush
1

Gray, Electrical Machine Design.

CO MM U TA TION

263

of adjacent coils will be as case, pairs of coils, like 2

width being 2j^ commutator segments, the order of commutation shown in Fig. 235. In this particular and 3, 4 and 5, etc., enter and leave

short-circuit simultaneously. 152. Selective Commutation in

Wave Windings.

A

study of

the simplex wave winding shown in Fig. 81 (p. 93) will show that the several brushes of one polarity are connected to each other

not only by an external conductor but also through the winding by way of the coils that they short circuit. The figure also shows
that the resistances of these internal paths are not equal because of the varying areas of brush contact; further, the short-circuited
coils are

fringing fields

not at any instant identically located with respect to the through which they are moving, hence the e.m.fs.
field,

generated in them by rotation through the are not the same in any two of

though small,

them.

Both

of these facts are

responsible for an unequal division of the total armature cur-

rent

between the several
The unequal comof

brushes.

ponents
,

the total

current
.

from brush to brush in cyclical order, in such a way that Kirchhoffs laws are continuously
shift

. 236. FIG. 00 Diagram showing duration of short-circuit.

satisfied.

This shifting of

the current values at the brushes in the case of
called selective

wave windings

is

In the case of a simple ring is simply the time required the duration of short-circuit winding for a given point on the commutator to move through an arc
equal to the width of a brush. But it will be seen from Figs. 231 and 233 that this simple relation does not hold in lap and wave windings, since a coil is short-circuited only when similarly numbered segments are simultaneously touched by the brush.

commutation. 153. Duration of Short-Circuit.

These segments being displaced with respect to each other, the time of short-circuit may be either greater or less than in a ring
winding. Consider the case of a multiplex lap winding, Fig. 236 (drawn to represent a duplex winding); the distance between corre= sponding edges of similarly numbered segments will be m0

Carter has shown 1 that the same condition. hence T = b/v c or the same T is less than this in multiplex lap windas in a ring winding. Effect less Wide Brushes. hence p'. that is. In simplex lap windings a/p = 1. on the other hand. March. Short-circuit of coil 1 will endure while the edge m of the is p until edge n reaches q. the term mft = ft in the for T must be multiplied by the number of double pole pitches in the region from which the brushes have been omitted. F. 1910. when the brush spans several segments. the coefficients 1 Electrical World. as in Fig. other things being equal. is the displacement corresponding to the distance between a given brush and the next brush of the same polarity. in somewhat modified form. If one or more pairs of brush sets are omitted. LV. Therefore. In wave windings. m will be at p'. Fig. Consider the case of a ring winding in which two coils are simultaneously short-circuited. is the width of a segment. Simultaneous of Commutation of Several is Coils.264 PRINCIPLES OF DIRECT-CURRENT MACHINES where ft ft. in a ring winding. the short-circuit exists over the distance from p to equals b which - (mft is ft) = b + ft (l -} The time of short-circuit then (18) where v c is the peripheral velocity of the commutator. as in the preceding sections. 154. exactly similar result is obtained in the case of wave windings where the number of brush sets equals the number of poles. When n brush moves from position at q. Vol. 804. 233. . When the brush width equal to or than that of a commutator segment. also holds when several coils are short-circuited simultaneously. above equation . if some of the brushes are removed. T is greater than ings. 237. sparkless commutation generally requires the satisfaction of the condition R b T/L > 1. the necessary An correction can be applied by remembering that mp. W. since a/p < 1. p.

Current paths. and the condition to be satisfied is that this shall not approach infinity as x approaches zero. currents in the coils di* J of the two coils being L and M. 237. FIG. such that xb is the brush overlap on the receding segment. R b Vo _x _ Xl Ec2 = (20) where Xi = (3/b and x is a linear function of the time. Evidently. at the end of . dt ^+E Rb f .COMMUTATION of self- 265 and mutual-induction respectively. Then. 1 - X = and since the above equations may be written L L di dii M di M di. ci = (19) di. 2 The expression jO - is a measure of the current density at the receding segment. necting leads. neglecting the ohmic drops in the conthe differential equations that determine the become (see Art. 143). several coils simultaneously short-circuited.

Let the lowest power be x a so remain finite if .266 PRINCIPLES OF DIRECT-CURRENT MACHINES i\ the commutation period = IQ and x = finite. Also. > -\- -fff M Aa) X A \ a \ T = / I^Rc r> i ^& r> 2 r^ ~ ** " ^ i ^ 2 1 #i_l L Xi 1 X EcZ = If a finite quantity (24) than unity. If there are any powers less than the first. -Tdx ' = Bx a ~ l for small values of x. if current iz is represented by the same short-circuit curve as i\ and therefore analytically by the same equation but with a y shift of the x coordinates. preserve the finite value of the lefthand terms we must have less a were a ~ l ^ Aa + and ^B = ARb (25) ^Aa + ^B Eliminating = Q (26) A and B between these two equations R T b . 0. both sides o equation (21) will become infinite. and (22) and transposing -}- E ci a finite quantity (23) and . x would approach infinity as x hence to approaches zero. terms involving higher powers become 1 a a~ l = = negligible and iQ + ii Ax and -r- Aax . Substituting these values in (21). so that the ratio becomes 0/0. then the ratio will there are no powers of x less than the first.-. that as x approaches zero. and the current (^ this i) must remain + can be expressed as Assume now that the sum of a series of x ascending powers of x (or of the time). (L T ff.

= (30) In other words.f. was pointed out above that if infinite current density a must be greater than unity. _ In this inequality the term j^f 2 takes the place of of a single coil. L in the discussion covering the case Proceeding in a similar manner. when a coil undergoes commutation in the presence of others that are simultaneously short-circuited.f. would be con- stant and equal to (L 2i /T). the effect is the same as though its self-inductance had been increased. hence + SM X) -r + ohmic drops + commutating e. the general differential equation takes the form L -jj + 2M X is -jj- + ohmic drops -f- commutating e. Carter has shown that in general >i where Li Miz (28) M Mi 2 i L2 M 3 23 M n3 Mi n 2n M M: (29) M 3n When more than two coils are simultaneously short-circuited.m. the commutation were linear in ( the coils. = where dt If the rate of change of current in one of the di all coils.m. -r. hence (27) is 2 the condition to be satisfied.COMMUTATION But is 267 it to be avoided. x. .

L.f. flux ^> 3 linking with the end connections beyond the edges of the core. SLOT LEAKAGE FLUX. acting upon an elementary tube dx will then be 4?r X . 238.f. the other side The magnitude and distribution of the flux is therefore not the same on the two sides of a coil. the self-excited may be separated into three parts: the slot from wall to wall of the teeth crossing and completing its path through the core. Practically without exception the windings of all direct-current machines are arranged in two layers. 1. The coil edge contains z = Z/2S conductors. Calculation of the Self -inductance. 155. in Slotted ArmaThe self-inductance of a coil has been shown to be equal . 2. In to the number of flux linkages per ampere. 239. The flux < Fig. Coil Edge Occupying the Bottom of a Slot. as indicated by <p 2 .268 PRINCIPLES OF DIRECT-CURRENT MACHINES tures. divided by 10 8 the case of an armature coil embedded in a slot. 238. 4ir current is ^. flux passing from tip to tip of the teeth within the space tips.m.m. The m. in the one side of each coil being in the top layer.z gilberts. as indicated by flux linking with the coil 1. The The between pole 3. whose total m. ^^ Paths of leakage flux (6) coil. Fig. ("> FIG. surrounding The number of linkages due to each of these fluxes will now be computed separately. per unit bottom layer. a.

or x'2 dx = 3b a 47T (31) Above the coil the m. This flux inks with j-z conductors.f. Within ie region h 2 the flux nitude uniformly distributed and has the mag47T ind since it links with all of the conductors. found by integrating in from to hi. all dimencentimeters.m.2. 269 The 47T^ bs Wh? I'dx where sions I' is being in the corrected length of the armature core. hence the number of linkages due to of linkages the entire /hrough depth of the coil is v The total number occupying bottom of slot. that of the iron part being negligible in comparison. The denominator of the above expression represents the reluctance of the air part of the path. has the constant value is yy. L " lh = To 9 zH ' 67 (32) .COMMUTATION assuming that the lines of force pass straight across the lux produced in this elementary path is then slot.

as in Fig. 2406 it be- FIG. The flux h 47T I'dx .270 PRINCIPLES OF DIRECT-CURRENT MACHINES In the same way the inductances due to the flux in the region A 3 and & 4 are X and 10. In the case of straight this reduces to slots. the resulting expressions for L\ t identical with those for Lib except that h% is replaced by h'z The total inductance of the coil due to slot leakage is then T T I T fQS or less than twice Lib. Assume from tip to tip of teeth are made up of that the lines o a straight portion 242.^rr + rJ (a) (35. force 6 . Coil Edge Occupying the Top of a Slot. 240. and with the shape comes (6) of slot Fig. Fig. N /Q7 tin arc 6. and of two quadrants of an elementary path dx is circles. (33 < * 7r 7' 4 60 (34 is The total inductance due to slot leakage L lb = L' 16 ?-+. Fig. Coil occupying straight slots. 2. Using same methods as in case a.8 = ~. 240a (SG. 241. TOOTH TIP LEAKAGE FLUX.

2 Arnold (Die Gleichstrommaschine) uses the entire pole pitch as the superior limit. Gray (Electric Machine Design) uses only one tooth Elektrische Maschinen Apparate u. width.4 Niethammer 2 3 L = IogM (-) \s / . 1 FIG. 1. Anlagen. and gives results that agree fairly well with actual meas- urements. END CONNECTION LEAKAGE FLUX. 242.O. 2386) is the total free length per element.ll X 10~ 8 (40) where s is the diagonal of the rectangular coil section (including insulation between turns) and I/ (see Fig. Stuttgart. 139. Values of L 2 calculated by the above equation will be somewhat too The large because no account has been taken of the effect of neigh- boring slot openings in reducing the flux.COMMUTATION hence the total inductance for both sides of the /"W(r- 271 coil is - 2 X dx bo ! 'J JQ + TTX l' 10 9 X T. Various approximate formulas have been developed for calculating the inductance due to the end connections. Coil occupying top of slot. 241.46 Iogio[l + ^^- (39) b) represents the distance between pole expression (r limit of the integral is taken to be half of and the superior tips. . FIG. this amount on the assumption that the coils undergoing commutation are approximately midway between pole tips. p. 3. Tooth-tip leakage. 1904. 1 The limit of integration used in the above equation has been checked by numerous tests. gives z 2 Z/[o.

p. i.g.. or L = Li +L +L 2 3 = i O. H a- ULi. di = nab) - . 243. Electric Generators.8 lines per ampere-conductor per cm.272 PRINCIPLES OF DIRECT-CURRENT MACHINES Arnold 1 gives L = 3 zHf zHf [o. is at the rate of on the average. the flux linked with 4 c. Die Gleichstrommaschine.s..092J X 10~ 8 = [0. lines per ampere-conductor I. 243). 1900. 2nd ed. particular case. Relative slot dimensions. FIG. . Cross section FIG.J (42) Parshall and Hobart have published 2 the results of measurements of the inductance of armature coils of commercial machines from which they deduce the coils 1 that.46 logio (j~) 0. 376. Vol.46 logic . 244. The total inductance of a winding element is the sum of the inductances due to the several parts of the leakage. of coil. i i Hobart calculates the end-connection leakage on the basis of a flux of 0. of free length.(^p) 0.e.8 (41) where ds is the diameter of a circle whose circumference equals the perimeter of the coil section including the insulation between turns (see Fig. 2 .23] X 10.

ratio of depth to width is 5 : 1 (Fig. Thus.7] X 10. Taking z = I and V = is equivalent to about five slots in the cm.8 1. The mutual inductance of two coils is equal to the number of flux linkages with one of them when a current of 1 ampere flows through the other. one above the other. is ^ X 0. the coil edges lie In this case two lie dis- tinct conditions are possible: (a) the coil edges (b) side by side. divided by 10 8 The previous discussion of the simultaneous short-circuiting of several coils indicates that there are two cases to be considered one in which the coils in question occupy the same slot. 1 Lu=or (M + M) X1010~ 8 z 8 . and 0.11 X 10~ 8 10~ 8 henries. COILS OCCUPYING THE SAME SLOT.8 of 273 "embedded" length lines per of wire (10 lines per inch). lie the other in the bottom layer. which space between poles. an average of ^ (% + 1%) X coil. results of the foregoing formulas when customary dimensions are inserted. but it is open to the objection that the designer must exercise great discretion in selecting the proper unit value of flux to fit the dimensions of his machine. 244) and in which T #0 is approximately 10. Calculation of the Mutual Inductance.s.16 + 15. (a) It is clear that if the coil edges 18 side by side in the slots. 156.16 lines per ampere-conductor per cm. consider a machine with straight open slots whose c.73 or a total of 4. of "free" These values check fairly well with the length (2 lines per inch).g.COMMUTATION per cm. The value of L 2 for one side of the lo glo [1 with and V = 1. of length. M. Parshall and Hobart's method is rapid and simple. : X in 1. which they lie in different slots. corresponding to 4. . or one in the top layer. = = = 1 3.. the other .05 X 10~ 8 . ampere-conductor per cm.

hence the mutual inductance is due only to the slot and tooth-tip fluxes.274 PRINCIPLES OF DIRECT-CURRENT MACHINES also throughout their entire lengths. and the interthe Same . mutual inductance. (a) Both Coils in In this case the sides of the coils will be parallel Layer. In Fig. therefore. the slot flux linkages are 10 same as on the other side of the coil. or the -} 2 2. to each other throughout their entire lengths. the end connections run in opposite directions. a coil. 245.m. The linkages due to the tooth-tip leakage flux are obviously the same as in the calculation of L.z gilberts per ampere the elementary tube dx] the slot flux linkages are FIG.f. Finally. there is and therefore no great error in writing M = L (6) Since the coil edges are in different layers. 245 the cross-hatched areas represent the two sides of On the left-hand side the inducing coil is at the bottom of the slot and 4ir exerts a m. On the right-hand side. of on ^. 6 J (43) COILS OCCUPYING DIFFERENT SLOTS. Flux leakage paths.

f. Fig.m. -b -I X) (r " (44) 10 9 is somewhat too large since in carrying out the integration continuously to the pole tips the effect of the slot openings This value is ignored. 246. coils FIG. Considering the tooth-tip leakage first. the self-inductance of a single On this basis . coil edge 2 with a m. cult to estimate.the free lengths. coils Mutual inductance. 246. Mutual inductance. in adjacent slots. coil edge 1. not in adjacent slots. The mutual inductance due to end-connection leakage is diffiArnold recommends taking it as one-half of coil.COMMUTATION 275 linked flux will consist of tooth-tip leakage flux along the embedded portion and end-connection flux along. of z gilberts In the tube dx the flux linkages are then 47T To* op I'dx and the mutual inductance due to coil is this flux on both sides of the FIG. acts 4?r Yp: upon paths surrounding per ampere. 247.

M.276 PRINCIPLES OF DIRECT-CURRENT MACHINES - M"vt = zHf fo. due to tooth-tip leakage. remain the as before. = M' = 13 ^ is 4?r zH' .. M ls . the above equations become but are M'i 3 = 2 X 7 L o zH r J 60 + 7r(2* + 1U zH' - X) ^ and 7T logio jr~~T~o~T~~~ OQ -f. 247.05] J X 10~ 8 (45) and If 12 = M' l2 + M"i2 slots. logio 2 b~~+2-jrt (51) These values are somewhat too large inasmuch as the the slot openings has been neglected. the end-connection leakage reduces to zero because the coils separate and run in opposite directions after Then leaving the slots. 7? It has been pointed out that a sufficient criterion of T7 the inequality ~rf~. 157..F. Coils not in the Same 3t ductances M' 12 and M'i In this case the mutual inLayer. effect of The Commutating E.6. (46) When the coils considered are not in adjoining placed as in Fig.2 L logio ( ) \ S / - 0.025] J X 10~ 8 (48) It is not necessary to carry the computation beyond the case shown in Fig. 6 same M arid - M' - + ^ (T " 6) zH' lo (50) . 247 for the reason that the numerical values become relatively small and the brushes are seldom so wide that coils are simultaneously short-circuited in more than three consecutive (6) slots.l logio ( ) L \ s I - 0.ZlTl Mu" = K^3 = z 2 Z/ [o. 4. >1 not in itself ..

in the short-circuited coils.m. therefore. among the relation emax < 2. termines the limiting conditions. An examination of the fundamental differential equation L^ + iR + -r : di RbT y . however. therefore. V in armature reaction. because it is bound up with the relation E T <io(R 2/4). would lead to the conclusion that large values of L and very small values of Rb and T are permissible. but this is naturally incorrect. This is the condition connection with while it is appears. (I Q j + . the latter condition holding when the brushes + In are in the neutral zone or only slightly displaced therefrom. Obviously. it is possible to have the + 7? T7 condition j L < 1. This last relation. If. some other relation must exist which de(5). It therefore.m.f. latter are generally small in larly and the ohmic drops at the transition surfaces.m.. particuthere is when considerable brush lead. L -r.m. mutation process the reactance voltage. the above relation is of importance in the case of resistance commutation.fs. x RbT i) p . this e. desirable to keep the average reactance voltage small. other words.f. will set up currents in . For instance. the commutating E c '. it is even more important to keep the commutating e. concerned in the com.m. within limits. that the brush must be so placed that at no-load the com- .COMMUTATION 277 sparkless commutation. consider a machine which is to operate with that has already been discussed in Chap. that brushes fixed at a definite angle of lead at no-load the commutating field will have a certain value and will generate an e. the brushes are given a decided displacement from the neutral so that E T > i Q (R 2Ri). hence the conclusion that the commutating e.f.f..m. these coils and sparking may result if the e.f. so generated is the sufficiently high to strike an arc between the brush and commutator segments as they break contact. The comparison with the others.m. (i - \ ^) +E = rn c t\ shows that there are in general three e.. e. It follows. taken alone.f. The same thing is r> m therefore true in regard to the relations which follow from them ^ >1. should be at all times nearly equal and opposite to the reactance voltage.

248.f. the commutating field decreases gradually with increasing load. may even reverse under load. 248. that there are no special commutating In constant-speed generators and motors of the separately excited or shunt type. When the operating under load conditions. the change being unfortunately in the wrong direction inasmuch as the commutating e. and in the commutating zone will have the forms indicated in Fig. which represents the case of a generator with a forward lead of the brushes. It will be seen distribution due to the field that the strength of the commutating field decreases with increasing load.. V. the commutating e.fs. of the which act in that region.m.f.m. and armature windings can be determined by the methods described in Chap.m. and the difference between mutating machine is them must again be within the sparking limit. the reactance voltage set up in the short-circuited coils will be opposite in direction to the commutating e.278 PRINCIPLES OF DIRECT-CURRENT MACHINES e. The field strength in the commutating zone resultant of the m. In actual calculations this last fact made use of because of its greater simplicity. The commutating e. 248. devices present. or if the brushes are not given a sufficient lead. de- . type of machine under con-The curves of flux Armatur Field FIG.m. If. should increase with the load. commutating field would also be the resultant of the individual -fields produced by the field winding and by the armature is is due to the and armature windings there were no saturation the field winding acting separately. and the disturbing effects of saturation must then be allowed for according to the judgment lar of his experience the designer and with the particu- sideration. if the field excitation remains constant. If the armature is magnetically too powerful. assuming.f. as in all the pre- ceding discussions. Commutating field distribution.m. in the manner indicated in Fig.f.f. will be within the sparking limit. because of armature reaction. therefore..m.

There may also be pulsations of the flux as a whole. Under load conditions the commutating e.m. thereby giving rise to peripheral oscillations of the armature flux in the interpolar space. /sec. In the case of fractional pitch windings it is necessary to add algebraically the unequal e.fs. pitch.8 windings. and v is the peripheral This assumes that the coils are of full velocity in cm. (53) If with similar is reservations concerning coils chorded now nac the by the brush. All of these effects are of high frequency and induce in the short-circuited coils e. of rapidly changing direction which are superposed upon the main e. the flux density in the commutating zone. 158.f.8 < 12 to 18 volts (54) It is evident from these considerations that the value of BO. To strength in the neighborhood of the commutating pole tip should shade off gradually instead of abruptly.fs. During the period of commutation the rotation of the armature periodically changes the positions of the teeth and slots with respect to the pole shoes. must be sufficiently under control to allow the brushes to be placed in a position which will insure good commutation under all loads. per element becomes E = jg cl I'v (B - Bo) X 10. due to periodic changes in the reluctance of the magnetic circuit if the surface of the teeth presented to the poles does not remain constant. number of turns per element. The changing current in the short-circuited coils produces a further pulsation of the flux in this region.fs. so that both sides are at the same time cutting fields of equal intensity. They give .m. generated in each side of the coil.m. Pulsations of Commutating Field.280 PRINCIPLES OF DIRECT-CURRENT MACHINES where B is Q is Z/2S the the average strength of field in the commutating zone. considered in the preceding sections. a condition which can be realized fairly well by making the airthis end the field gap at the pole tips longer than it is under the central part of the pole shoes. number of we have the 17 simultaneously short-circuited relation 2E CO = n -^ Zo sc I'v B X 10.m.

Sparking Constants.COMMUTATION rise 281 to saw-tooth notches in the short-circuit current curves. 159. Thus. and may therefore be used as a check upon the magnitudes of the constants selected before the design has proceeded too far. Also.9 + /m - . Z^ We = and where v d dcom = = = = peripheral velocity of armature diameter of armature diameter of commutator. Substituting these values in the expression for er we have = C | I'vq where 27T/3F &. X 10. of periphery. er . it has been shown that b rp _ + Vc and L where F is a function have also =^~ zH f XF of the dimensions of the machine. ~ j 2 = ampere-conductors per cm. Obviously these pulsations can be reduced by using numerous small slots with few coil edges per slot. er The expression = -~ L which has been shown to be of considerable importance can be put into another form involving the principal design constants of the machine.

and if the short-circuit current curve is symmetrical with respect to the point O (Fig. undergoing commutation in a generator. Obviously.-Magnetizing action the short-circuit current curve is of short-circuited elements. q ampere-conductors per inch of periphery) it becomes approximately 40 X 10 6 160. In the a posicoil clearly exerts a demagnetizing magnetomotive force upon the main magnetic circuit. Reaction of Short-circuit Current upon Main Field. currents will be by the brushes. these statements are to be reversed in the case of a motor. But the mid-point of the commutation period occurs when the coil is in the geometrical neutral. and will then exert a magnetizing effect. When set a generator is running without load. If commutation takes place. Fig. in the case of overcommutation. With a forward lead up in the coils short-circuited of the brushes the current in the short-circuited coils will have the direction in the b position. . not symmetrical. Let a and b. In case of undercommutation the resultant action will be demag- netizing. 250. 218). and FIG. hence there will be a resultant magnetizing action. it has an average value. on a coil tion the the average. 250. a backward lead of the brushes will evidently result in a demagnetizing action. Fig. The direction of the current flow will depend upon the direction of the field in which the coils are moving. while in the b position the action is a magnetizing one. upon the direction of displacement of the brushes.282 PRINCIPLES OF DIRECT-CURRENT MACHINES r7 The quantity -~ Vvq may be considered as characteristic of corn- mutating conditions. one or the other of the two effects will preponderate. v in feet per and in minute. 250. represent the initial and final positions of of . the current will have the direction of the b position during the greater part of the time of short-circuit. in the geometrical neutral. about 20 X 10 6 In English units (7 in inches. using metric units. Thus. and. The magnitude of shown these no-load short-circuit currents may be sufficiently great . therefore. the demagnetizing and magnetizif ing effects annul each other.

E. being stationin ary in space.f.fs. The distorted air-gap armature current can action of the flux due to the magnetizing of the fluxes which would be of as be thought compounded m. the brushes takes set so that commutation are place in the geometrical will not be acted upon by the coils the short-circuited neutral. The Armature Flux Theory. then. 251.m.f. ture flux in the short-circuited coil is in the same direction as the original.f. IX).f.f. must be neutralized by an e. G. wiped out by the opposing m. Production of armature flux.fs. 251.m.I. p. to effect the reversal. of opposite direction and somewhat sufficient sur- greater magnitude plus e.. of an interpole or its equivalent (see Chap.m. Lamme.m.m. 1911.E. Its Application to Interpole A Theory Commutation and Machines. vantage that in the 1 it commutation of emphasizes the physical phenomena involved It does not involve the explicit process. This figure brings out the fact that the e. in order that there may be a This reversing e. them in exactly the same way that the main flux acts upon the conductors under the poles.f. but the armature flux. will generate e. or the brushes may be kept in the neutral axis and the armature m. If. current flow and therefore tends to prevent the desired reversal of current. Strictly speaking. may be obtained by shifting the brushes until commutation takes place in a sufficiently intense part of the field flux. this is not exactly true when saturation of the iron part of the magnetic circuits exists. armature the field and acting separately. A full treatment of the subject of commutation based on this 1 It has the adanalysis has been worked out by B.m. generated by the arma- FIG. produced by 161. Evidently this e. 2359. Part 3. but it will serve as a first approximation to the truth. A. . XXX. The state of affairs in a generator will be as shown in Fig. field component of the flux.f. Trans. hence also the experimentally determined open-circuit characteristic. Vol.COMMUTATION 283 to materially influence the field flux.m.m.m.

2. 218? NOTE. This is a consequence of the motion of the coil. space as will be clear from Fig. certain of the magnetic fluxes may remain practically constant in value and direction during the entire period of commutation. 250.fs. The armature of a 6-pole machine has a simplex lap winding of 546 conductors arranged in 91 slots. If. of self-induction. the commutator has 273 segments and has a diameter of 42 cm. This problem may be solved most simply by plotting a curve showing the variation of contact loss throughout the period of commutation. in short-circuit current curves are sinusoidal. which shows a short-circuited coil in two successive positions. of which there are several. in the paper rethe stationary armature flux. Construct dia- . for while this flux does reverse with respect to the coil.fs.284 PRINCIPLES OF DIRECT-CURRENT MACHINES self- consideration of the e..m.f. PROBLEMS What is the ratio of the energy losses due to brush contact resistance two machines. for the case of a machine which has a brush covering 2^ commutator segments and in which the 1. ferred to there occurs the following statement: "According to the usual theory. and curves showing the variation of current density at five points equally spaced along the brush arc.2 cm. however. Construct a curve showing the variation of current density at a commutator segment. of and mutual-induction. before and after the reversal of the current. However. this self-excited flux e. and the brush contact arc is 1. though these are implicitly involved in the e. during the commutation of the coil the local magnetic flux due to the coil is assumed to be reversed. This is but one instance. generated by cutting For example. as indicated in curves c and d of Fig. is considered as producing the be eliminated in the computain other words. to show where there is apparent contradiction of fact in the usual mathematical assumptions made in treating this problem." reversal of the self-excited flux which is responsible for the a mere mathematical however. 3. the latter is then to be taken as the resultant of the field flux armature flux and that part due to the armature turns outside of the of the commutating zone. in the zone in which the commutation occurs. of self-induction is not. The energy loss will then be proportional to the area under the curve.m. it does not reverse its direction in The e.m.m. which are identical in all respects except that in one of them the commutation is linear and in the other sinusoidal. it should tion of the commutating which is flux. abstraction in the inductance theory.f.

each case. and that of Problem 4 runs at 94 r. and the thickness of the insulation between the top and bottom layers is 0. assuming that the lines of force pass Straight across the slot (horizontally in Fig. The armature of a 10-pole machine has a quintuplex wave winding of 1230 conductors arranged in 205 slots. (a) assuming full-pitch windings.2 cm.p. 64. + tributed uniformly over the cross-section of the circle..m. below the periphery. each conductor has a cross-section measuring 0. 52. The opening at the top of the cm.COMMUTATION 285 grams showing the sequence of commutation in the coils occupying two adjacent slots. the center of the circular slot being If the slot contains z conductors. taking into account the mutual induction due to simultaneously short-circuited elements.m. Construct diagrams showing the sequence of commutation in the coils occupying two adjacent slots.. 52)? . 4. of the pole pitch. and that the pole arc is 70 per cent.1 cm.9 cm. An armature has semi-closed circular slots of radius r cm. as shown in of slot is ri The diameter the right-hand diagram of Fig.18 by 1. (6) assuming a pitch that is less than full pitch by one slot.. of the armature of Problem 3 is 59. what is the inductance due to slot leakage flux. Assuming that the total length of the conductors of an element is twice the length of core plus three times the pole-pitch. The machine of Problem 3 runs at a speed of 430 r. Find the duration of the period of short- circuit in 6.1 cm. wide by 2. the commutator has 615 segments and a diameter of 131 cm. dis(r r-i) cm. 5.12 cm. the brush contact arc is 2 cm.p. and the slots are 1 cm. 7. deep. the total length armature core is 32 cm. deep. -The thickness of the insulation at the bottom of the slot is 0. find the total inductance of an element.. wide and r 2 cm.. p.

whether the above ratio is unity or greater than unity. possibly. consequently the compensating winding is connected in series with the armature. Clearly. is the primary cause armature current of the action magnetizing of the field distortion which in turn necessitates the shifting and thereby brings into existence the demagnetizthe armature. one. if the transverse of action ing magnetomotive force of the armature were balanced by an equal of the brushes in space. then. In either case. to assist in commutation. the compensating winding must be traversed by the main armature current or a fractional part thereof. nearly complete neutralization of the armature flux will result. there will exist in the commutating zone a reversing flux which increases proportionally with the armature current (unless saturation of the iron of the magnetic circuit sets in) which is precisely the the current in the short-circuited condition necessary to secure good commutation at all loads. If the ratio of compensating ampere-turns to armature ampereturns is unity. if the armature magnetomotive force is overcompensated. there will exist in the neutral zone a component of flux having the proper direction to reverse coils. Principle of Compensation. and may or may not be provided with a diverting shunt. and the brushes could then be permanently fixed in the geometrical neutral axis. The problem of compensating armature reaction then consists of two parts. if the ratio is slightly greater than unity. as in the case of the series winding of a compound machine. the prevention of the field distortion and the consequent elimination of the demagnetizing action of the arma286 . the distortion of the field and opposite magnetomotive force having the same distribution would be completely eliminated and brush displacement would be unnecessary except. Moreover.CHAPTER IX COMPENSATION OF ARMATURE REACTION AND IMPROVEMENT OF COMMUTATION The cross or transverse 162.

the main pole pieces that carry the coils of the shunt winding are bolted to the yoke. reaction is practically Since in the case shown FIG. Compensating Devices. These wedges also serve to reduce the cross- . embedded in slots in the pole The space distribution of the compensating winding being very nearly the same as that of the armature winding. In larger multipolar machines the number of compensating turns per pole should be I/a times the armature turns per pole. J. the Diagram of com- pensating winding. the production of a commutating e. In a German patent granted to 163. the main ring or yoke being clamped between cast- machine of this iron frames. 252. where a is the number of current paths through the armature. or a part of it. and the cores that carry the com- pensating winding are held by the bolts which pass through the main poles and by the wedges which hold the commutating lugs in position. Fig. the other. E. 252 the current in the armature conductors is half of the total current. the more important. in Fig. 253 illustrates the construction of the field frame of a Company. Later.m. H. Of these two. is ing faces. first is the there in 1884 exposition of the principle of Menges cuited coils is compensating armature reaction. Ryan and M. The compensating windgenerator. type as built by the Ridgway Dynamo and Engine The entire magnetic circuit is built up of sheet steel stampings.COMPENSATION OF ARMATURE REACTION 287 ture. the neutralization of N tl __ - Jrp^ armature complete. The patent specifications call for the use of a stationary compensating winding wound around the armature at the sides of the poles and traversed by the arma- ture current. number of turns in the compensating winding should be half of the effective number of cross-magnetizing armature turns. Thompson experimented extensively along similar lines and developed the method shown diagrammatically in 252 for the case of a bipolar Fig. for the purpose of neutralizing the reactance voltage of the short-cir- the latter and reversing the current in them. in 1892.f. in such a direction as to oppose the magnetizing action of the armature.

Under load FIG. conditions the magnetomotive force of the compensating winding acts in the directions shown by the dotted lines. section of the magnetic path and so keep down magnetic leakage. Frame of Ridgway generator.288 PRINCIPLES OF DIRECT-CURRENT MACHINES from pole to pole. showing slotted pole face. 253. Flux paths in poles of compensated machine. Section q of the bridge . FIG. assuming generator action and clockwise rotation. 254. 254 illustrates diagrammatically a portion of the magnetic machine embodying the above device. circuit of a Fig.

m. 255.m. acted upon by a resultant m. 255. Closely akin to the Thompson-Ryan device is an arrangement 19 .fs.COMPENSATION OF ARMATURE REACTION is 289 acted upon by two m. 253) and winding in them more than the normal number of conductors. and section p is acted upon by two m. hence it produces a local field of FIG. coils in place. The arrangement of the compensating winding and main field winding is shown in Fig. Frame of Ridgway generator.m. but the not materially increased on account of the initial saturation of the iron. This effect is accentuated by making slots m and n somewhat larger than the others (see also Fig.fs. having the same direction. the proper direction to assist the reversal of the current in the short-circuited coils. flux is The result is that the central tooth t is of opposite direction.f. which makes it a north pole under the assumed conditions.

and in opposition to the armature m. The field structure closely resembles the stator of an induction motor.f. M.m. The entire prevention of field is distortion not necessary for successful commutation. the compensation can be made complete.m. Commutating Devices. Instead of using a field frame of the salient pole type with the addition of a slotted ring. in the Sayers winding. the field structure consists of a slotted cylinder wound with two independent sets of coils in Fig.f large enough to neutralize the reactance voltage. principal consideration ing coil field of sufficient . there is no compensation of armature reaction. is introduced into the short-circuited coil by an auxiliary winding which is so placed as to cut a part of the main flux during the commuta- . 257.m. The FIG. Thus. but the reversing e.290 PRINCIPLES OF DIRECT-CURRENT MACHINES due to Deri. and since the reluctance all of the magnetic path is the same along diametral paths.f. C. and concentrically surrounding the armature. up a magnetomotive force acting along axes midway between the poles. 257. Deri's arrangement of main and compensating windings. sets FIG. Fig. to insure the presence of a commutatintensity to generate in the short-circuited is an e. 256. as indicated The main winding. Bayers' winding. 256. produces poles whose axes are indicated by the arrows. the compensating winding. 164.

also that of Swinburne.m. interval. therefore. 258. 1890. l magnets. 258. and since the field intensity at the latter increases with increasing current. excited In the latter. the energy that The simplest method at the brush contact. this arrangement motors. are placed in the FIG. that suggests itself for this purpose is the use of commutator leads of high specific resistance. is device of historical rather than as is Sayers practical interest. 259. the brushes may be so constructed as to interpose more resistance in the path of the short-circuit current than in that of 1 Journal Inst. The main coils are connected to auxiliary coils which lie behind them with respect to the direction of rotacoil is tion so that commutation is not dependent upon the flux density at the leading pole tip. small U-shaped electroby the main current. as in the ordinary machine. 259. Instead of inserting resistance in the circuit in the above is is. the commutating e. A limit is set to this automatic adjustment of commu- The tating conditions by the saturation of the trailing pole tip. High resistance leads to commutator. Swinburne's com- mutating device. during that undergoing commutation. must be handled which serves to limit the current and.COMPENSATION OF ARMATURE REACTION tion period coil 291 and which. Fig. Another method that has been used for the prevention of commutation difficulties consists of the insertion in the circuit of the short-circuited coil of an auxiliary resistance neutral zone. manner. as indicated in Fig. open to the serious objection that the main current must flow through a set of these extra resistors with consequent heating and loss of efficiency. p. It used in certain types of alternating-current however. At not in circuit. increases with it. of Electrical Engineers.f. but upon that at the trailing pole tip. . FIG. is in series with the all other times the auxiliary The auxiliary coils are in reality merely extensions of the commutator leads which have been wound around the armature. 106.

\ structural arrangement of the interpole machine is shown in The magnetomotive force of the interpoles Fig. present the intervale or . in the manner indicated in Fig. It goes is of by means Commutation of a shunt around the interpole winding. The final adjustment for sparkless operation made by varying the air-gap under the interpoles by means shims. the principal features of which are discussed later. however. and the mica insulation between segments should have a rate of wear as nearly as possible the 1 same as that of the copper. Chap. 481. without saying that the mechanical construction of the commutator and brush rigging must be such as to insure perfect contact and absence of vibration.f. 260. . while the short-circuit current resistance of the carbon must pass transversely through the higher and copper in series.292 PRINCIPLES OF DIRECT-CURRENT MACHINES by making them of alternate longitudinal layers the main current. commutating pole. The copper provides a path of high conductance so far as the main current is concerned. better design. Special brush with special difficulties. in Machines having no Auxiliary Devices. is so adjusted that the m. the is device that exclusion | of all used to the practical others. 260. of the armature is slightly over- The compensated. 1905. 65. FIG. The copper of the commutator should be of uniform quality so that a true cylindrical surface may be preserved. in order that mechanical resonance may not occur. This has the effect of considerably increasing the resistance of the circuit at the last stage of the commutation process. Electrical World. however. due to Young and Dunn. The natural period of vibration of the brushes and brush holders should differ from that of any vibration which may possibly be impressed upon them by the motion of the commutator. It is possible. wherever conditions is commutating ITT T. At the present time. of carbon and copper. auxiliary insulated section. or 165. II.m. 1 provides for the final rupture of the short-circuit at an auxiliary carbon brush insulated from the main brush. p. that brushes of this type may fail to operate satisfactorily for the reason that the short-circuit current may pass from one copper layer to the next by way of the commutator A surface instead of through the intermediate layer of carbon.

fs. but a distinct limit is set by the fact that in such windings at least one side of the shortcircuited coil will lie close to a pole tip where the field intensity changes sharply. based upon the assumption of full-pitch windings. resistance Consequently hard carbon should be used where commutation is necessary. tendency toward constant drop with increasing current density is obvious. set by the voltage drop at The the brush contact. 1. harder the grade of carbon used in the brushes the greater will be the drop. determined by the relation is fjl -j LI > 1. and is incompatible with the requirement of a fringing field of gradual slope. windings the two sides of a coil in fields of unequal strength. The average reactance voltage er = ~7p L< as 2i Q Rb should T~) be less than 1 volt.fs. so that the e. The summation of the commutating e. or S#c0 = This equation is n sc -^rvB Q X 10~ 8 < 12 to 18 volts. 261 (taken from Gray's Electrical Machine Design) shows the variation of brush The contact drop with current density for an average carbon. generated in the coils under a brush must not be greater than 12 to 18 volts. somecommutation are undergoing times of the same polarity. This results in a decrease of the short-circuit e. Fig. The limiting value of e r . Both effects indicate that chording the windings is advantageous. At the same time there is a decrease in the in the case of fractional pitch demagnetizing action of the armature. generated in them may tend to annul each other. this feature greatly restricts the zone through which the brushes may be rocked. to which due regard must be paid to For the sake of completeness they insure sparkless operation. The contact drop in the direction from commutator .COMPENSATION OF ARMATURE REACTION Reference has already been the 293 made in the preceding chapter to more important magnetic and electrical factors concerned in the commutation process.m. are here recapitulated and in addition there is given a summary of the methods adopted to secure them. 2i Q Rb which is of the order of 1 volt.m.m.f. 2.

To secure sufficient contact area to carry the current the axial length of the brushes is adjusted so that the average current density is in the neighbor- hood of 30 amperes per sq. in order to cut down the mutual induction of the simultaneously short-circuited coils. 3. cm. in. Relation between brush drop and current density.2 10 20 30 40 50 60 70 80 90 100 Amperes per Sq. in.) for hard carbons and as high as 65 amperes per sq.294 PRINCIPLES OF DIRECT-CURRENT MACHINES to brush is generally somewhat higher than that in the direction from brush to commutator.5. remedy is hence if SEco turns out to be too large the a series-parallel or full-parallel winding. 261.o S 0-8 0. The p/a a maximum special difficulties are in series (two- circuit) windings. An sq. is sq. . Considerably higher densities are permissible with metal (copper) brushes.5 times the width of a commutator segment. up to about 160 amperes per sq. if and since the ratioit is fixed by usual is follows that p/a must be kept small ratio encountered. in. cm. the brush width seldom exceeds 3. average of about 40 amperes per customary. Since the number of simultaneously short-circuited coils is given by sc _b p @' a practice at from 2 to 3. (10 amperes per i.Inch FIG. The brush width should not be too great. in. (5 amperes per sq.6 o 0.) for soft carbons. average carbon. In practice.

the load current will have little effect in producing distortion. the main field should be powerful. if possible." in comparison with the armature field.1 field ampere-turns per pole. the saturation acts.8 to 0. It was shown it is reversal of the Chap. otherwise there is the possibility that the machine will "flash over." exceed 20 volts and should be maximum difference of potential selection of a large diameter carries with it the possibility number of slots each containing but few coil sides. Since the armature magnetomotive force acts upon a different path from that subjected to the main field excitation. . though in most cases the factor is 0. in simplex lap windings. as an increase of reluctance. The presence of using a large The numerous armature slots means further that there will be a number of them between pole tips to cut down oscillations of the commutating flux and minimize pulsations of the of sufficient flux as a whole. of course. involves a moderate average value of volts per segment. should not less than 15 volts. by increasing the number The value of L can also be made of commutator segments. reducing the Both of these considerations point to the desirability of a design which the ratio of diameter to length of armature core is A large number of commutator segments relatively large.COMPENSATION OF ARMATURE REACTION 4.1. small by selecting a relatively short axial length of armature. or stiff.9 instead of 1. 5. in Q 77! E -T- -- = ^~ which t value. 295 The inductance of the coils can be kept within limits by number of turns per element. The between adjacent segments in machines of the series types should never exceed 40 volts. V that in order to prevent the commutating field by the distorting action of the in armature necessary to observe the relation armature ampere-turns per pole < 1. If at the same time those parts of the magnetic circuit which are influenced by both the armature and field excitation are saturated at no-load. thereby reducing the mutually inductive action. for a given number of armature conductors. " In other words. it is clear that the disturbing effect of armature reaction can be reduced by introducing additional reluctance into the transverse magnetic circuit of the armature. or conversely.

265. 263. namely the air-gap and the iron parts adjacent The the above methods at first sight. designs are those which include both the features of additional reluctance and saturation of the iron. in Figs. 267. 266. therefore adds addition of reluctance to the path of the armamore or less to the reluctance of the main the circuit. Their effectiveness is open to question behind the inasmuch as the armature flux will tend to pass around The most effective slots rather than across them. FIGS. FIG. FIG. or by making the bore of the pole faces eccentric. 264. Longitudinal slotting of pole cores. as in Fig. 262. 262.G. FIG. the FIG. Chamfered and eccentric pole faces.296 PRINCIPLES OF DIRECT-CURRENT MACHINES first of is not as simple as it appears armature magnetomotive force acts upon a path which includes the most important part of the main magnetic circuit. FIGS. . 265 and 266. Construction of laminated pole. the extra reluctance is 265.m. alone involve a longitudinal slotting of the pole cores. 268. cost of the hence requires more field excitation and increases machine. The ture m. FI. Designs embodying this principle FIG. 262. 267 and 268. FIG. FIGS. as shown 263 and 264. 263. usually obtained by chamfering the pole tip. 264. thereto.f.

266.COMPENSATION OF ARMATURE REACTION as in Fig.000 per sq. V. however. 5' = (1. FIG.000 lines per sq. the desired saturation is obtained by using a long thin tip. or.25 to 2) |_360a 1. Commutating Poles. This is particularly true in the cm. It is. The length of the air-gap at the pole tips can be com- puted from the equation r. as the pole shoe alone may be built of stampings and bolted to a solid pole core. in. case 6. often pole tips. or inter- have been extensively adopted for the improvement of commutation in generators and motors in which sparkless operation would be difficult or impossible of attainment under ordinary conditions. only adapted to machines that run in one direction. The requirement of a fringing commutating field of gradually changing intensity is met by properly shaping the tips of the pole shoes. not necessary that the entire pole be laminated to secure this construction. poles. Fig. Pole of Lundell generator.). Commutating poles. Examples of such machines are turbo-generators and shunt motors which are required to have a wide . of course. Saturation of the armature teeth acts in the same way as saturation is of the The tooth density purposely made high. in which case the laminaup to the required thickness in such a manner that the projecting tips are alternately on opposite sides. 297 at The saturation feature is most important the trailing tip in the case of generators and at the leading tip in the case of motors. as in railway motors. it is. 269. in the case of laminated poles. 268. by using tions are built a stamping of the form of Fig.2AT t as 166. 267. shown in Chap.6 E . of motors which are required to operate with the brushes fixed at the geometrical neutral. (21. 269 illustrates the pole construction of a Lundell generator which embodies a number of these features. therefore as high as 140. a plan view of the pole face as seen from the armature surface is then like Fig.

270. the intensity of the field under it must be greater than under a pole of full length in the ratio of full length to actual length. InterInterpoles are also extensively used in series railway motors.60 and 0. obviate the for the various poles necessity expedients commonly employed of in ordinary machines. This is is not objectionable.65. instead of be equal to.m. since the essential feature the production. The flux distribution outside of this zone is of minor importance. if the machine were a motor revolv- .m. the calculation of the leakage factor being made in the manner outlined in Chap. s. of a reversing flux of proper intensity.m. required to drive the commutating flux through the transverse path n.f.f. of ampere-turns to balance those of the armature and to supply the m. taking into account the m.f.fs. as discussed in Chaps.m.298 PRINCIPLES OF DIRECT-CURRENT MACHINES range of speed. supplied by the main poles. the ratio of pole arc to pole pitch 0. of armature and interpoles are different. both the breadth and length of the interpoles The presence should be kept as small as possible. Complete neutralization of the armature flux is not possible by the use of interpoles for the reason that the space distributions of the m. the distance moved over by a slot while the coils it are undergoing commutation. The calculation of the winding to be placed on the commutating poles presents no It is necessary to provide a sufficient number special difficulties. The figure represents a bipolar generator revolving in the clockwise direction. Fig. They are superior to forms construction involving compensating windings in the pole faces because of their greater simplicity.70. or slightly in greater than. of interpoles increases the magnetic leakage of the main poles. VI and VII. however. Winding of Commutating Poles. but if the interpole is shortened. To reduce the leakage to a minimum. 167. is generated. in those parts of the path of the lines of induction which are common to both magnetic circuits. with due regard to fringing of the flux. and the span of the main poles made smaller than in ordinary machines. in the commutating zone only. The axial length of the of can be made than that the main poles. is The span of the interpole should usually between 0. IV. S. for it is less interpole immaterial in what portion of the coil the neutralizing e. N.

a condition that would be met automatically the upon reversing the current through the armature. to be generated. Fig. . and in opposite If there were no saturadirections in the other two quadrants.------&u --~ an(^ _ o ^> . 270.fs.COMPENSATION OF ARMATURE REACTION 299 ing in the same direction. flux. d> Magnetic 5i> * circuits in interpole machine. here of the \$ = v& where t v is the coefficient is of dispersion main poles. in the other two. . of the main and commutating poles act in the same direction in two of the four quadrants of the armature core and of the yoke.m. 270 reveals the fact that the m.f.m. in which of the case the flux in the armature core would be + \$t in two FIG.f. would produce a proportional < tion each m. and ing flux ~ m the other two. since two windings are in series. quadrants. \$t 4. and \$ it = v&i where Vi The magnitude the of the commutating flux <{ is given by where B ig the flux density in the gap under the interpole. the polarity of the interpoles would have to be reversed.m. where <. is the work- produced by the Similarly. interpole. coefficient of dispersion of the interpoles. is determined by the value of the commutating e. . in two ' of the quadrants of the yoke the flux would be ^S .

m. f to a armature armature core. the term coils In the L 2 due . the corrected breadth and length . The above discussion applies directly to the bipolar machine of Fig. of the interpole. 5. (Chap. .300 PRINCIPLES OF DIRECT-CURRENT MACHINES b'i and and l' t are respectively. ampere-turns for the two sets of teeth opposite the interpoles = = = = ampere-turns ampere-turns ampere-turns ampere-turns for the for the core. in accordance with KirchhofTs law. the algebraic of all the m. d { than the under the Corresponding to this value of <f>.fs. that due to the two interpole windings. VIII). let AT = ic ampere-turns required by the two interpole cores and shoes AT = ig it AT = A T' a A T" a A T' v A T" y sum ampere-turns for the two interpole air-gaps. Effect of The presence expression of Commutating Poles upon Coil Inductance. Thus. in addition to the m. these corrected lengths are greater actual lengths interpole. AT and that due to the armature. main flux < by from 3 to 4 times the air-gap. and with obvious modifications applies also to multi- polar machines. must be zero. and to each flux density there will correspond a definite number of ampere-turns which may be determined from the appropriate B-H curves. provided the interpole coils are A larger number of turns may be used if a diverting shunt is placed across the interpole winding. 168. in the closed magnetic circuit. b to c Then. abcdefa. e to f for the yoke.fs. It follows. listed above. ird ' _= '' P ^ be wound on each not shunted. In this circuit there act. larm ~ 2 Z p 2 _Zia ^ == Trd . and to that of the there will be definite flux densities in each part of the closed magnetic circuit abcdefa. 270. that the 1 number of turns to interpole is ^ ~ ATi r 2-a *. then. A Tarm) where t. commutating poles causes an increase in the inductance of the short-circuited L = LI +L +L 2 3 under them. c to d for the yoke.m.

from the center of the commutating pole of corThe tooth-tip flux within the rected breadth b i} Fig.f while a backward lead causes a com. 271. and the average inductance.46 logic [l + Compounding Effect of Commutating Poles. it (*' - I'd X 1. a generator produces a demagnetizing effect and consequently reduces the generated e. is fT^2 d% f .m. was shown that a forward lead of the brushes in the case of V. dimensions being expressed in centimeters. 271. . In Chap.COMPENSATION OF ARMATURE REACTION to tooth-tip leakage. r limits' of the pole is 10 10 Z 45'i &'< - FIG. Tooth-tip flux under interpole. considering both Yi Pi) sides of the coil. as follows: is 301 affected. 5i r^_ If l'<l'i. (Vi - rO 2 10 t In the above equations all b'i is the corrected gap length.. Its value may be computed Suppose the center of the slot containing a coil edge to be a distance x cm. there must be added to the above expression a term z 2 jgi 169.

Fig. the generator poles. the effect of the oppositely directed fluxes due to the interpoles (shown by the hatched areas) being to annul each other. In the same manner a backward displacement of the . these effects of Generator (O FIG. 272. 272a represents the conditions when the brushes are in the geometrical neutral axis. If the brushes are displaced in the direction of rotation. the total flux is reduced by the difference between the hatched areas e f h and bed and in addition by the sum of areas a b c and e f g. Thus. Compounding effect of interpoles.302 PRINCIPLES OF DIRECT-CURRENT MACHINES If is provided with commutating brush displacement are accentuated. pounding action. as may be seen from Fig. Fig. the armature winding between adjacent brushes of opposite polarity is then acted upon by the flux due to a main pole only. 272. 2726.

go on indefinitely. because the interpoles eventually become saturated. poles.f..f. The speed under these conditions would rapidly fall. 272c.m.m. that the armature current again rises. . is proportional to the active and to the speed. flux time that saturation of the interpoles sets in. the speed has been continuously increasing. Similar considerations show that in the case of commutating pole motors a forward lead of the brushes will increase the flux and therefore reduce the speed. especially will this be if the rotating parts have large moment of inertia. The result will be a rapid increase of counter e. and an increase of is armature current to produce the necessary acceleration of the armature.m. The and so still increased current further strengthens the interfurther weakens the field and accelerates the But the counter e. due to reduced flux is offset by the increase due to greater speed. there will be a corresponding decrease of counter e. and so repeating the above-described cycle of changes. VII it was shown that a considerable backward displacement of the brushes of a com- mutating pole motor reversals of rotation. in which case the machine would true become a generator drawing upon its kinetic energy of rotation to send current back to the supply line.m. may If result in a continuous succession of actual reversal of direction does not occur. results in an increase of the flux linked with the armature winding. Fig. The tendency to accelerate the armature will then continue until the decrease of counter e. The decrease of field strength cannot.f. while a backward lead will decrease the flux and raise the speed.COMPENSATION OF ARMATURE REACTION 303 brushes. for since the effect of a backward displacement weaken the active field. In Chap.f.m. and the momentum of the armature will cause the speed to continue to increase even after the flux has reached a practically constant value. pulsation of speed may to result. possibly to a value greater than the line voltage.f. but up to the armature. however. thereby producing increased speed. causing such a reduction of counter e.

due (a) (6) (c) to the heating effect of the current flowing through the resistances of the armature winding. the field winding. 1." (c) brush friction. The losses may be grouped under the following heads:.CHAPTER X EFFICIENCY. . due to hysteresis in the armature core and teeth. faces. amperes flows through a resistance of r 304 The power consumed when a current ohms is i 2 r watts. is the net power output less In every dynamo-electric machine than the gross power input. THE IRON OR CORE (a) LOSSES. up by the current in the (d) eddy currents in the end plates of the armature. of i The Ohmic Losses. the short-circuit currents in the commutated pulsations of the flux set short-circuited coils. 2. the brushes and brush contacts. or "wind- age. etc. due to bearing friction. 3. and pole THE MECHANICAL (a) (b) LOSSES. the difference being consumed in the internal losses of the machine. coils. teeth. 4. 171. ADDITIONAL MISCELLANEOUS LOSSES caused by (a) (6) (c) eddy currents in the armature conductors. THE OHMIC LOSSES. friction between the moving parts and air. Sources of AND HEATING Loss. in non-insulated bolts through laminated core. (6) eddy currents in the armature core. RATING 170.

In long-shunt compound machines.59 ohms. Ta = Po sa (1 + ~ 0. the loss (total) is (6) Pf = C z i a rf + 2 is r s watts is where ia ia = is i + is in case the machine a generator. and = i in case it is a motor. po the length is expressed in feet and the cross-section in circular = 9. RATING AND HEATING armature and " field 305 wind- The combined ohmic or ings are z i r losses in the commonly referred to as the " (a) The armature copper motors is loss in all copper losses. In general.TS = E is t watts if (5) r a includes the resistance of the regulating rheostat field one used.016 ohm. In shunt machines the field loss is winding Pcf = where is 2 i* r8 = #. degrees a If = number of armature circuits in parallel. mils. if these dimensions are expressed in centi- meters and square millimeters. p (6) = 0.00420 a" ohms at (2) where p la sa t = specific resistance of the armature copper = total length of wire on the armature = cross-section of the armature conductors = working temperature of the armature in Centigrade C. respectively. is The field copper loss in separately excited machines if r f 2 P cf = and watts is (3) in plain series generators and motors 2 PC/ = ia ?/ watts field (4) where r/ represents the combined resistance of the and its regulating shunt if the latter is used. 20 . types of generators and (1) Pea = *aV a Watts at the where the armature resistance is to be computed or measured normal working temperature.EFFICIENCY.2 .

351. The Core Losses. 0.. 172. especially at light loads. I. Soft carbon brushes. i a and i 8 (c) The ohmic loss at the commutator depends upon the drop of potential at the transition surface between commutator and brushes. to 1. as well as upon the amount of current flowing.55 to 0. Copper brushes (65 to 160 amperes per sq. thereby giving rise to a molecular friction in the mass of the through power armature core.017 to 0. common.1 volts Very hard carbon brushes. Vol.7 volts 0. in. r a being increased to include the average brush contact resistance. This hysteresis loss can be represented by an empirical equation due to Steinmetz of Pha = where rjfVB a ^ watts (9) !Die Gleichstrommaschine. drop of potential at each brush is Ae. 261).6 volts 0. the loss is Pec If the = 2i a Ae watts (8) volt With the usual type of carbon brushes. 1906. In the Armature Core.2 Medium carbon brushes. ed. to include the This method may lead to inaccuracy.306 PRINCIPLES OF DIRECT-CURRENT MACHINES compound machines the total field loss is In short-shunt P cf = i rf z z -f i s r s watts .03 volts.) It is to 1. of the mutator in that sufficiently ohmic loss at the comarmature winding. p. (7) the above relations again holding between i.9 1.4 to 0. Ae is approximately 1 when the brush current density has values common in ordinary practice (see Fig. . Values of Ae as determined 1 by Arnold are as follows: Very soft carbon brushes.5 volts 0. The relative motion between the armature core and the magnetic field of force produces a periodic reversal of the loss magnetism of the core. (a) Hysteresis Loss. however. 2nd.

also portional to its volume. 273.06 0.05 0.0062 X 10~ if British units are used (weight 77 pounds. expressed in watts per pound per cycle per second. if volume is ex- 77 pressed in cubic inches and flux density in lines per sq. in. RATING 77 AND HEATING 307 of the core = a constant depending T)Yl upon the material f = ~T = the the number of magnetic cycles per second V = Ba = If the volume of the core maximum and value of the flux density in the core.. of the core is pro.13 0.. 150000 50000 0.11 0. = 0.1-' 0.- 1 Cycle per Second FIG.0021 X 10~ 7 density in lines per sq. metric units are used in the above equation flux (volume in cubic centimeters rj= 0.02 0.08 0.03 0. using the above .14 0.09 0. in.).0017 X 10~ 7 Since the weight.15 Watts per Lb. as a function of which case = in the flux density expressed in lines per sq.07 0. flux density in lines per sq.EFFICIENCY. value of r. in. 273 shows the variation of the hysteresis loss. W. The curve of Fig. the equation for the hysteresis loss can be written Pn a in = rifWB a 7 watts (10) 0.10 0.04 0.). Curve of hysteresis loss.01 0. cm. for ordinary sheet steel.

308

PRINCIPLES OF DIRECT-CURRENT MACHINES

In the Armature Teeth. The flux density varies from section to section because of the taper of the teeth, and it is not correct to compute the hysteresis loss by substituting an average value of flux density in the above equation.
Consider an element dx, Fig. 274, at a distance x below the
of the tooth; its
tip

volume

is

dV =
where k
is

blkdx

=

(b t

-

bi

~

b '*

Ikdx

that the total flux
flux density will

the lamination factor (from 0.85 to 0.90). Assuming is the same at all sections of the tooth, the

vary inversely as the width of the section, or

ri
tr dx
1

FIG. 274.

Computation

of hysteresis loss in teeth.

where
tooth.

B

t

is

The

the actual, or corrected, flux density at the top of the hysteresis loss in the element is then

dPht

.
-

dx

arid the total loss per tooth is

(11)

Since the volume of a tooth

is

the above expression can be written

EFFICIENCY, RATING

AND HEATING
1 -

309

'*

X

o

(12)

is

In other words, the expression for the hysteresis loss in the teeth similar to the general expression, but with the addition of the

factor

The

ordinates of Fig. 275 give the value of this factor for various
b' t /b t
.

values of

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

FIG. 275.

Correction factor, hysteresis loss in teeth.

(b)

Eddy Current

Losses.

That part

of the core loss

due to

eddy or Foucault currents can be approximately calculated by the formula derived below, but it is more usual to determine the
loss

under known experimental conditions for reasons that will appear later. Consider a radial element, Q, Fig. 276, of one of the armature core stampings. Let the thickness of the stamping be t, and let ct be the radial depth of the core, where c is a numeric. When the element is in the vertical position OA the flux passing through
,

its lateral

walls

is

a

maximum, and when
zero.

it is

in the horizontal

This change of flux occurs four times revolution in a per bipolar machine, or, in general, four times per magnetic cycle. The changing flux induces an alternating e.m.f.
axis

OB

the flux

is

310

PRINCIPLES OF DIRECT-CURRENT MACHINES

and sets up a corresponding alternating current which may be assumed to flow in paths like those indicated in the lower part of the figure; an elementary current path is then bounded by similar rectangles of widths 2x and 2(x -j- dx), and lengths 2cx and
dx), respectively. elementary circuit will be

2c(x+

The change of flux through such an 4B a X 4cz 2 lines per magnetic cycle,
is

where

Ba

is

the

maximum
where /
volts.

lines per second,

flux density in the core, or 16B a cx 2f the number of magnetic cycles per

second.

The average

e.m.f. in the

elementary circuit
is

will

be

lQB a cfx 2

X

10~ 8

The

resistance of the path

r4cx 7 1 hdx

+

4x "[ hcdxl

FIG. 276.

Elementary paths

of

eddy currents.

where 7

is

the specific resistance of the material of the core.

The

loss in the

elementary path

is

1\

hdx

c)

and the

total loss

is

Pea =

~~

7x~io 16

c2

+

X * dX

==

ijo

7X10 16

c2

+

1

EFFICIENCY, RATING

AND HEATING

311

But

hct 2 is the

volume
7

of the element, hence the loss is
,.2

R. 2/2/2

X

10 16

c2

+

1

(volume of tooth) watts

(13)

This equation shows that the eddy current loss varies as the square of the flux density, the square of the frequency of the magnetic reversals, and the square of the thickness of the laminations;

and inversely

The equation cannot, however, be
sults,

as the specific resistance of the core material. relied upon for accurate re-

because the actual distribution of the current

may

differ

considerably from the assumed distribution, and the laminations

1

Watts per Lb.

2

FIG. 277.

Curves

of

eddy current

loss.

are not perfectly insulated from each other, as has been tacitly assumed. Due to these causes the actual measured loss will be

from 50 to 100 per cent, greater than that computed from the formula. Thus, assuming

Da = / = t = r =
c2
2

10,000 gausses

60 cycles
14 mils
12
1

=

0.0356 cm.

X

10~ 6

ohms per cm. cube

=

(nearly)

312

PRINCIPLES OF DIRECT-CURRENT MACHINES

the loss in watts per pound by the above formula is 0.22, while the observed value for these data in the case of annealed sheet steel is 0.44 watts per pound. Fig. 277 shows the variation of eddy
current loss with flux density at frequencies of 25 and 60 cycles per second and for laminations 14 mils thick. The loss at other frequencies and thicknesses can then be computed by observing

that the loss varies as the squares of these quantities. Eddy Current Loss in the Teeth. Referring to Fig. 274, the eddy current loss in an elementary section of a tooth is

dP et =

X

volume
fir

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

FIG. 278.

Correction factor, eddy current loss in teeth.

where
T-J

e is

the eddy current constant.
A.rt.n-I
rt.

Integrating,

let.

~

-.

*x-I

1

i

tv

*/

efH

2

B X volume
2
t

of tooth

X

2

J-^

(14)
(13), in that it

This equation differs from the original equation, contains the additional factor
log e bt/b't

the value of which

is

shown

as a function of

b' t /b t

in Fig. 278.

EFFICIENCY, RATING

AND HEATING

313

Eddy Current Loss made in Chap. II to
pole faces.

in

the Pole Faces.

Reference has been

the cause of the eddy current loss in. the This loss is confined to a relatively thin layer at

the face of the pole because the direction of the induced eddy currents is always such as to damp out the flux pulsations that

produce them (Lenz's law).

The flux pulsation at any given point in the pole face will pass through a complete cycle of changes in the time required for a point on the armature to move over a distance equal to
the tooth pitch, that
gives
is,

in a time

f

t

=

,

,.

seconds.
of teeth

This

a frequency of

ft

=

p

=

^i

number

X

rev.

per sec.
Fig. 279 represents the variation of flux density at the pole face on the assumption that the curve of distribution is sinus-

FIG. 279.

Variation of flux density opposite teeth and
T>

slots.
T>

oidal.

The amplitude
if

of

the

pulsation

is

B =
r

-

^~

.

Then

v

= = =

peripheral velocity of the armature in centimeters per

second
fj,

p

permeability of the material of pole face specific resistance of material of pole face in absolute
electromagnetic units
1

the pole face loss in watts per sq. cm.

is

pp =
1

*ir STT \ HP

v

x

10

~7

=

fc2

x f STT \/~ \ HP
7

10

~7
(

15 )

Potier, L'Industrie Electrique, 1905, p. 35.

Rudenberg, Elektrotechnische

Zeitschrift, Vol.

XXVI,

p. 181, 1905.

314

PRINCIPLES OF DIRECT-CURRENT MACHINES

where k 2

=

/B'\
(

2

)

is

a function of the relative dimensions of
slots.

the air-gap, teeth and

1

has worked out the curve

of Fig. 280 as giving fairly satisfactory values of k 2 in terms of the ratio b a /5. If British units are used (B g in lines per
sq. in., v in feet per second,

the loss in watts per sq.

in inches, and ^ and p as above) in. of pole face is given as
t

Pv =
Total Core Loss.

1.65

X

10- 7 k 2 B a 2 ^l~ MP

(16)

more convenient

to

Except in the case of special designs, it is have access to the combined values of hysteresis and eddy current losses than to compute each of these losses Test methods separately.
lend themselves readily to the determination of the total core loss,

and from the

results of such tests

curves like Fig. 281 2 can be prepared. 173. Mechanical Losses.

FIG. 280. Constant for calculation of pole-face loss.
ble.

Bearing Friction and it is possible to While Windage. the loss due to bearing friccompute the loss due to tion, windage involves
(a)

and

(b).

so

many complex
its

variables that cal-

of

magnitude is impossiAs it is likewise impossible to separate the combined value the two losses as obtained by test measurements, they

culation of

are always grouped as friction and windage loss. This loss varies from 1 to 3 per cent, of the rated capacity in high-speed machines

moderate capacity, and from 0.8 to 2 per cent, in low-speed machines of moderate size. In large direct-connected machines the loss will be from J to 1 per cent. In very high-speed machines, such as turbo-generators, the loss due to windage will be increased.
of

Friction loss in the bearings varies with the
1

%
Vol.

power

of

Adams, Lanier, Pope and Schooley, Trans. A.I.E.E.,

XXVIII,

p. 1133, 1909.
2

From Gray's

Electrical

Machine Design,

p. 102.

EFFICIENCY, RATING

AND HEATING

315

the peripheral velocity of the shaft in the bearings, up to velocities of about 1800ft. per minute; at higher velocities it varies The windage loss, as in the case directly with the velocity.

power of the speed. But in both cases these losses are independent of the load on the machine. (c) Commutator Friction Loss.
of fans, varies as the third

Let d com

Ab

= p = / =
c

diameter of the commutator, inches total area of brush contact, sq. in.
brush pressure
(Ib.

per sq.

in.)

coefficient of friction.
Cycles per Second 50 60

30

100

10

15

20

25 30 Watts per Lb.

35

40

45

50

FIG. 281.

Total core
is

loss.

Then the brush

friction loss, in watts,

Pbf =

X

746

=

0.0059 d com nA b p cf watts

(17)
is

Ordinarily the value of p c is from 1.5 to 2 Ib. per sq. in., and / about 0.3 for carbon brushes and 0.2 for metal brushes.
(a)

solid
of

Eddy Currents in the Armature Conductors. When large, armature conductors are used in open slots, different portions the same section of the conductor may be simultaneously in
Under these conditions
e.m.fs. of

fields of different strength.

.................. long shunt ...... smaller conductors in parallel...............316 PRINCIPLES OF DIRECT-CURRENT MACHINES magnitudes will be generated from point to point of the and eddy currents will result... rjfVB a - 6 4 armature teeth .. of the loss due to the ohmic resistance.f MISCELLANEOUS LOSSES: Eddy currents in armature conductors ........ ^ to 3 per cent.. ........ (b)... Field : i a *r a 2 separately excited .. i r/ + is 2 rs Commutator .... etc..... i V/ + is zr9 2 compound.... efH 2 B a 2 V armature teeth ................................... The loss due to these eddy currents may be from 5 to 1 5 per cent... tfWB^Vt 2 z pole faces .................... (0................... depending upon Brush friction size > t and speed.......... greater than the true resistance... This is equivalent to saying that... O.... and cannot be computed..... This loss may be minimized by stranding the conductors or using different cross-section. if r/ . COPPER Loss: Armature ... the effective armature resistance is from 5 to 15 per cent. they are absorbed in the amount attributed to friction.... windage............. losses..... These are minor losses...... is z rs compound. k Bg ge& X 2 ^j X constant MECHANICAL LOSSES: Bearing friction and windage ............... Etc.....05 to 0....OQ59dcom nA.. 2 i a ra . i a 2 r/ shunt . yfVtB l t - G *-(rT X 5 -X 2 /p\Z~ Eddy currents: armature core .. ........ short shunt .. (c) and (d) Miscellaneous Losses Due to Short-circuited Currents... so far as the armature copper loss is concerned...... Summary of Losses.......................15) Losses due to currents in short-circuited coils.................... In testing.. and core 175. p ...... 2i a Ae series CORE Loss: 1 Hysteresis: armature core ..............

The true or commercial efficiency of a machine. . generally referred to simply as the efficiency. Then input = output + ohmic losses + stray power loss (21) In the case of generators. RATING AND HEATING 317 176. the output In either case. Efficiency of Conversion. input are the terminal voltage and the line current. the total losses (1) may be divided the pure ohmic or copper losses. is the power output to the gross power input. Electrical and Mechanical Efficiency. in motors. efficiency If = output T-T input = output " ri~~ losses output - + = input " ~~^ lossses " (18) ^ . respectthe ively.EFFICIENCY. That of a motor becomes For the sake into of convenience. . two groups. both In the quantities being expressed in terms of the same unit. is. and the ratio of this total electrical power to the total mechanical power input that 77 C is defined as the efficiency of conversion. and (2) the stray power loss. _ output + : copper losses electrical input input The ratio of the net electrical output to the total power . the expression input stray power loss = output + ohmic losses (22) represents the total electrical power actually developed in the armature. . ~ = efficiency input ot conversion = electrical - power developed -r r mechanical power input . True Efficiency. = stray power losses -. case of a generator the output is electrical and the input mechanratio of the net ical. which include all the remaining losses. . is mechanical and the input electrical. due to the flow of current through the resistance of the armature and fieldwindings. expression for the efficiency of a generator becomes t E and i Ei+P t (19) where P is the summation of all the losses.

input copper losses input _ output + stray power loss input the ratio of the net mechanical output or The mechanical efficiency is to the mechanical power developed.318 PRINCIPLES OF DIRECT-CURRENT MACHINES is developed in the armature rj e called the electrical efficiency. electrical efficiency = . generators and motors of the separately excited and shunt 177. and is. . therefore. rj m = mechanical efficiency = output + stray power loss . but this types. = output input . ~ . the stray tendency or less compensated. except in compound The stray the decrease of the flux as a whole. The Stray Power Loss. . the expression input copper losses = output + stray power loss (26) represents the total mechanical power developed by the machine. = (24) electrical power developed out P ut output + copper losses Evidently. . The distortion of the flux due to armature reaction tends to increase the core loss. f] = f\cf]e (25) In the case of motors. brush contact load losses. on the other hand. The and windage. by a simple loss in of a shunt be determined machine can power is more manner by running it as a motor without load at its rated voltage . friction loss and the miscellaneous In the case of machines which operate at approximately constant speed and constant flux. and the ratio of this power to the input is again the efficiency of conversion.copper losses It follows that rj = Ticrim (29) loss includes all stray power the losses except those due to the pure ohmic resistance of the various windings. power loss remains nearly constant at all loads within the working range. as. that _ is. or electrical output - = . generators. made up of the core loss. . for example.

P. and Mfg. July. data taken from these curves can then be used to calculate the efficiency at any load. Fio.320 PRINCIPLES OF DIRECT-CURRENT MACHINES loss and the friction and windage loss In this manner curves like Figs. .M. 282 and 283 tween the total stray power at the same speed. 8 Motor and Generator Testing. 1 8. Core loss of series motor. Elec. 1913. 100 200 300 400 500 COO 700 800 900 1000 1100 K. Westinghouse and 11. Friction and windage as a function of speed. 1200 1000 800 GOO 400 200 100 300 Volta 400 500 FIG. pp. 282. Sec.. Railway motor test. 283. Co. may 1 be determined.

average 1000 200 250 350 400 450 Amperes Output FIG. remain constant. Let it with Load. including rheostat STRAY POWER Loss: Core loss Friction and windage 5500 watts. 2000 watts. 3500 watts. RATING 178. 1250 watts.EFFICIENCY. flat-compounded long-shunt generator has the following COPPER Loss: Armature and series field at full load Shunt field loss. 1000 watts. 284.volt. Losses and efficiency. there will be a constant loss of 7750 watts at all loads and a variable loss increasing according to a parabolic law from zero at no-load to 5500 watts at full-load. 21 . Commutator loss. except those in the armature and series field. Condition for be assumed that a certain 250-kw. Assuming that all these losses. 550.. Variation of AND HEATING 321 Efficiency Maximum losses : Efficiency.

the efficiency 77 ~ 550* __ of the efficiency are is shown graphic- computed from the equation 55(H' + 7750 + (r + r )i f a 2 where i + 2. . E being assumed constant. (31) . necessary to use greater refinement than in the preceding example. A second differentiation will verify the statement that the condition thus determined is for minimum it value of the denominator of the fraction.27 The above example consist of a constant is typical of those cases in which the losses term and a term that varies as the square of the load.322 These PRINCIPLES OF DIRECT-CURRENT MACHINES losses and the variation ally in Fig. Pconst . friction will The bearing friction. or Ph+e = is (Ph+e ) Q +d c is (33) the core loss at no-load and a constant. 284. windage and commutator remain constant as before. t -\ . the denominator with respect to i and equating to Differentiating zero. + r. h (r a clear that 77 will have is its maximum t value when the denomi- nator of the fraction a minimum.) = i\ra + = Pconst = i a *(r a + r/) (32) which agrees with the statement above. Writing t E t i E E it is . In such cases the efficiency has its maximum value when the variable loss equals the constant loss. the following corrections may be applied: 1. or ^ 77) + (r. The core loss (due to hysteresis and eddy currents) may be is When taken to increase linearly with the load current because of the compounding where (Ph+e )o 2. effect.

and (E t irj)* ra n short-shunt machines. is due to brush contact drop Pb 4. and a . = and C 2 ^ 2 + 1 Motor and Generator Testing. and <2 . r/ in short-shunt machines. the negative sign in the case of motors. RATING 3. a line current.erm that varies as the square of the current. E field loss is 2 t in plain shunt and in long-shunt compound machines.erm. hence the efficiency is given by an expression of the form term that varies directly with the 77 ~ ____ E i E i+~Pconst + Cli + C E t t t _ 2Z 2 E + i-^i t + Ci + C* (36) Differentiating the iition for maximum denominator and equating to efficiency is found to be zero. 6. the positive sign being used in case of generators. is nearly equal to i a 2 (7a ?"/) 0). = ia \20X total brush area 2 +V (35) The armature The shunt loss is i a r a as before.EFFICIENCY. Westinghouse Electric and Manufactur- ng Co. . AND HEATING be taken as 1 323 The total brush contact drop may 20 so that the loss X total brush area in sq. the con- or Pconst = C 2i 2 (37) n this expression lo-load (i Pcon st is . 5. evidently equal to the total loss at or that part of the load loss which aries as the square of the current. The series field loss is i a 2 rf in long-shunt machines. of all the losses therefore includes a constant The summation . in.

All-day Efficiency. If this load is carried by a motor whose fixed losses are 5 per cent. The all-day efficiency of a machine 180. hours). curves of power input. Fig. Load curve. 285. and is also affected by the ratio of its fixed and variable losses. so that the area under the curve is proportional to the energy output.EFFICIENCY. is the ratio of the net energy output to the total energy input during a working day. B and C. The ordinates of the load curve represent power output and the abscissas represent time. . and whose variable losses are 10 per cent. The all-day efficiency of a machine is dependent to a large extent upon the shape of its load curve A. A. it should be so designed have a relatively high armature resistance in order to make the efficiency a maximum at or near the point of average load. it is important that the all-day efficiency of a motor that runs continuously should be as high as possible.. 285. RATING AND HEATING 325 as to occasionally at full-load or overloads. Inasmuch as charges for electrical service are based largely on energy consumption (kilowatt- 012345678910 Time in Hours FIG.

it was specified that the allowable rise of temperature of the parts of a machine (excepting railway motors) should be as follows: armature and field windings. and subtracted if it was higher. but it is particularly specified that the observed rise of temperature must never exceed the limits given in the following table: of TABLE OF HOTTEST-SPOT TEMPERATURES AND OF CORRESPONDING PERMISSIBLE TEMPERATURE RISES . bearings... emphasis is placed upon the highest permissible temperature of the hottest The rise spot as well as upon the maximum rise of temperature.328 PRINCIPLES OF DIRECT-CURRENT MACHINES (1914) rules.. These rises of temperature were based upon standard conditions of a room temperature of 25 C. temperature in the case of air-cooled machines (excluding railway motors) is based upon an ambient temperature of 40 C. the correction to be added to the observed rise if the room temperature was below 25 C. for each degree difference between room temperature and 25 C. should be corrected by % In the new rules (effective December 1. a barometric pressure of 760 mm. and normal conditions of ventilation. 40 C. 50 C... 55 C. It perature differed was further provided that if the room temfrom 25 C. commutator. 1914).. the observed rise of temperature per cent...

ducted at ambient temperatures not lower than 25 C. effect of types of In the case of machines intended for operation at an altitude of 1 Water-cooled oil transformers are exempt from this reduction. or of any insulation whose temperature would be affected by the heat of the commutator. AND HEATING 329 may be the ambient temperature at the time of the It will be noted that the above temperature limits recognize the advances in the art of constructing insulating materials that have been made since the adoption of the superseded rules. The temperature limits of commutators difficulties Current per brush arm so constructed that no from expansion can occur are as follows: Max. The temperature of insulation used in the commutator. must be subject to special guarantees by the manufacturer. 1 for each 100 meters by which the altitude exceeds 1000 meters. This is due to the fact that numerous tests have shown that the effect of variations of the ambient temperature is small. 95 C. pending the accumulation of more extensive data at higher temperatures. RATING whatever test. The new rules abolish the requirement of a correction of the rise of temperature due to a difference between the ambient temperature at the time of the test and the standard reference temperature (except in the case of air-blast transformers which are not considered in this text). the limit has for the present been set at 125 C. recommended that tests be condirection. however..EFFICIENCY. must in no case exceed the limits prescribed in the table of hottest-spot temperatures. . machines designed for maximum operating temperatures in excess of 125 C. These temperatures hold for the metallic parts only. 200 amperes or less 200 to 900 amperes 900 amperes and over 130 C. permissible temperature 130 C. and even higher. obscure and of doubtful It is. less 5 for each 100 amperes increase above 200. The observed high altitude in increasing the temperature rise of some machinery is recognized by specifying a reduction of the normal permissible temperature rise to the extent of 1 per cent. While it is known that insulating materials coming under the head of Class B can be supplied to withstand maximum temperatures of 150 C.

the correction C. to the highest temperature observed. instead of The ambient temperature to be measured by means several thermometers placed at different points around and halfway up the machine at a distance of 1 to 2 meters. the temperature of the armature is referred to a weighted mean of the pit and room temperatures. The temperature pit. the weight assigned to each being based on the relative proportions of the machine in and above the pit. Where machines are partly below the floor line in pits. or by thermo-couples. but the size of the oil cup must be increased with that of the machine under test. THERMOMETER METHOD. is is 5 C. including measurements by mer- cury or alcohol thermometers. To this end the thermometers are to be immersed in oil in a suitable heavy metal cup. (2 inches) high. When this method is used. (1 inch) in diameter and 50 mm. The smallest size of oil cup permitted by the rules consists of a metal cylinder 25 mm. for example. Three methods of determining temperatures of the various parts of a machine are specified. except that when the thermometer is applied directly to the surface of a bare winding. one or the other of these methods 1000 meters or satisfactory being adequate for commercial 1. and must be sufficiently deep to. tests. a massive metal cylinder with a hole drilled partly through it. by resistance thermometers. stantly in the pit must be of the portion of the field structure conreferred to the ambient temperature in the . such as an edgewise strip of 15 conductor or copper casting. insure complete immersion of the bulb of the thermometer. This hole is filled with oil. a test at any altitude less than 1000 meters and no temperature correction is necessary. the hottest-spot temperature is estimated by adding a hottest-spot correction of 15 C.330 PRINCIPLES OF DIRECT-CURRENT MACHINES is less. and protected from drafts and abnormal heat radiation. this time lag being greater the greater the size of the machine. The object of thus increasing the size of the oil cup is to avoid errors in the calculations of temperature rise due to the time lag between changes of temperature of the machine and the surrounding air. any of these instruments being applied to the hottest accessible part of the completed machine.

Detectors of this kind will assume a temperature practically equal to that of the adjacent coils. when this method is used. one and one between the top and bottom two-layer windings. . or between the coil in at least core. and wedge in single-layer windings. In the case of resistance measurements the temperature be computed from the formula temperature in degrees CenFrom this it follows that the rise of temperature of a tigrade. that temperature shall be taken as the highest observable temperature and a hottestspot correction of 10 C. but is to be used only with coils placed in slots. and a sufficient number shall be employed to insure locating the hottest spot. A correction of 5 C.5 + R t) ( ^(i -ft l) (40) resistance of winding at 6) degrees R = t resistance of winding at degrees. where t is the initial = where Rt+e (234. and a 10 correc- C. This method consists in the determination of the temperature of windings by measurement of their increase of resistance.EFFICIENCY. This method is not permitted in the case of low resistance field coils where the joints and connections form a considerable part of the total resistance. is to be added to the highest reading in the case of two-layer windings with detectors between coils and between coils and slots. in order to increase the probability of method revealing the highest observable temperature.5 -f- t). 3. added thereto. RESISTANCE METHOD. Whichever yields the highest temperature. RATING AND HEATING 331 2. careful check readings must be taken by means of thermometers. plus 1 1000 volts above 5000 volts terminal voltage (single-layer windings are commonly used in alternators. They should be placed between the coils coils and in the case of two sets of locations. method involves the use of thermocouples or resistance coils located as nearly as possible at the estimated hottest spot. for each tion in the case of single-layer windings. winding is given by the formula coefficient of copper is to I/ (234. This IMBEDDED TEMPERATURE -DETECTOR METHOD. seldom or never in direct-current machines). The thermocouples or resistance coils are built into the machine. but without disassembling the machine.

The statement of the resistance. by thermometer. measured in kilowatts. with current and voltage adjusted to give losses equal to those in service will . ventilation must be defined. shall not exceed 100 after one hour's continuous nominal rating must also include the corresponding voltage and armature speed. motor rounding C. The nominal rating of a railway therefore. arbitrarily defined as the mechanical output at the car or locomotive axle. TABLE OF MAXIMUM TEMPERATURES AND TEMPERATURE RISES The temperatures obtained on stand test. The continuous ratings of a railway motor are defined as the inputs in amperes at which it may be operated continuously at one-half.332 PRINCIPLES OF DIRECT-CURRENT MACHINES Operating conditions in 183. when operated on stand test with the motor covers and The system of cooling system. at any other normally accessible part. and if cooling is by means of forced draft the volume of air on which the rating is based must be given. run at its rated voltage on a stand with the covers arranged to secure maximum ventilation without external blower. ing temperatures for short periods than in other types of machines. at the air. if any. without exceeding the specified temperature rises tabulated below. not exceeding 90 commutator and 75 C. which causes a rise of temperature above the suris. the variable nature of the load makes it more difficult to give a definite rating to railway motors. as measured by C. Heating of Railway Motors. the case of railway motors are much more severe than in ordinary motors because of restricted space and the nature of the It is therefore good practice to permit higher workservice. arranged as in service. The rise in temperature. Further. three-fourths and full voltage respectively.

EFFICIENCY. RATING AND HEATING 333 .

the temperature rise on the winding should not exceed 55 C. of the temperature rise on stand test in the case of enclosed motors. rated current and voltage for a time sufficient to give approximately stationary temperatures. is commonly referred to as the output equation. when expressed in algebraic form. the armature.334 PRINCIPLES OF DIRECT-CURRENT MACHINES ventilation due to the motion of the car or train. and from 90 to 100 per cent. speed and dimensions of exists by Kapp. originally derived G. the temperature rise in actual service will be from 75 to 90 per cent. between the rating. successor to the American Association of Motor Manufacturers. 184. in ventilated motors. "The temperature two hours immediately following the normal load After a 25 per cent. This relation. Thus. on commutator after a run of normal windings. The machine should also be capable of standing a momentary overload of 50 per cent. Output Equation. 185. Temperature Specifications of Electric Power Club. A definite relation. by thermometer. let E = ^a = t \l/ rated terminal voltage rated armature current ratio of pole arc to pole pitch q = = ampere-conductors per unit length of armature periphery. on all and 45 C. and on the commutator 60 C. protected and enclosed types. be higher than in actual service because of the absence of the In general. overload in current at rated voltage for run. are summarized in the table on page 333. composed of representative manufacturers." Specifications for direct-current motors of the open. Since E = > I eT^To* ^ (nearly) and 3> = BgU = Bg l (nearly) . the losses being the same in both cases. and for various classes of service. The Electric Power Club. has adopted : the following specification for direct-current generators rise shall not be more than 40 C.

E. " depends upon the design constants" of the machine. 653.. FIG.I. Vol. for it is closely related to the thermal characteristics. conductors per inch of armature periphery. similarly. in. but principally upon B g and q since the range of values of ^ is limited. resistance per inch of length r = p length =s = 1 1 x Y~ = h a Y~h a 1 Trans. Each conductor ' will carry a current of its - Qi amperes. . 286. Calculation of copper loss per sq. and its cross-section will be is h circular mils . 286 represent a portion of the armature " surface (shown as a smooth-core type for convenience) of 1 inch square. of armature surface. 1905. RATING AND HEATING is 335 the power output of the machine in kilowatts KW where " = Eia * ^ 60 X 10" (42) The numerical value of this coefficient. Bg is clearly a measure of the degree is called the output coefficient of utilization of the q magnetic material of the machine. in part a measure of the specific utilization of armature copper. XXIV.E. and let h be the current density in the armature conductors expressed in circular mils per ampere. A. 1 Thus let q be expressed in ampere- a cir.. as is has been shown by Adams. tj Bg and q. Let Fig.mils.EFFICIENCY. p.

in. capacity. of is dli r X 1a = fl -jr watts per sq. The ratio q/h (watts per sq. (43) The value or less.000 lines per sq. The energy losses in any machine are converted into heat and cause a rise of temperature whose final value depends upon the heat capacity of the materials of the structure and upon the facility with which the heat may be radiated or otherwise dissipated.000 lines per The value of generally lies between sq. in. more complex structure Let Q = s = heat generated per second.336 PRINCIPLES OF DIRECT-CURRENT MACHINES working tempera- since the specific resistance of copper at the ture of the armature The i r 2 loss very nearly per conductor is then is 1 ohm per circular mil-inch. cm. weight of the body in kg. in large machines. The temperature will become stationary when the rate of heat generation becomes equal to the rate of dissipation.--r = ijfi T watts ah ia But the number -.5 in large machines running at high peripheral speeds (6000 feet per minute) where the ventilation is more effective. is = -. 0. in. but may be as high as 2. to about 800 to 850 in machines of 1000 kw.000056 (large machines). /%A \a I 2 a -. ia hence the loss per sq. in kg-calories = amount of heat specific heat of the substance required to raise 1 kg. Heating and Cooling Curves.000015 (small machines) and 0. in small machines up to 60. coefficient of W = A = a = cooling = amount of heat in kg-cal. in. radiating surface in sq. .iia/ of conductors per inch of i r z armature periphery armature surface in. 186. Values of B g range from about 40. up of q varies from about 400 in machines of 20 kw. It is of interest to derive the homogeneous body for the reason that law of heating and cooling of a it throws light on the conof a generator ditions obtaining in the or motor. due to copper loss) is generally in the neighborhood of unity for ordinary peripheral velocities of 2500 feet per minute. 1 C.

In this case no heat is being developed. RATING AND HEATING 337 dissipated per second per sq. J> which gives . Cooling of the body. consequently and the fundamental equation becomes 2. so that Qdt Transposing. () Q = (48) = sWdB is + Aa(0 t 0i)dt If the temperature 22 O degrees when = 0. 1. (0 -0i)-> = ^ rise of (46) and this is the limiting equation (46) may temperature be written the body. The last Q = aA(B - 0i)* = co (47) which expresses the fact that when the temperature becomes stationary the rates of heat production and dissipation are equal. cm. . Heating of the body. of radiating surface per degree difference of temperature between body and 0i = = surrounding medium temperature of body in degrees Centigrade temperature of surrounding medium in degrees Centigrade. The remainder will be dissipated. = sWdB ~ + Aa(B - Oi)dt (44) sWdB Q.Aa(et e.EFFICIENCY. to the amount Aa(0 Q\)dt kg-calories. In a time dt the temperature will increase by dO degrees.sw d6 o Aa(0-0i) When t = . and the body absorbs sWdd kg-calories.) Assuming that = 61 when = 0. During this interval the heat liberated amounts to Qdt kg-calories.

as is turned upside down with respect to the shown in Fig. the slope of the curve at the . but with a change of sign. Heating and cooling curves. 287. \ 20 10 12345G789 Time in Hours FIG. Differentiating the heating equation _aA - * ~sW and substituting t = origin is found to be in the result. Hence. in this case. . is. = (0 - sW If 0i (49) as the equation of the cooling curve. the heating and cooling curves are of the same logarith- mic shape. j. the equation of the cooling curve is the same as the variable part of the heating equation.sW ( JQ Aa(9 l = that 0.338 PRINCIPLES OF DIRECT-CURRENT MACHINES dt Jo which gives \ dd = . if = (B 0j)* = 00 is the temperature at the beginning of cooling equal to the limiting temperature at the end of the heating period. 287. but one other.

805. per (51) where a v is given in meters per second.seconds. cal. From this it follows that of surface. The heating equation can then be written To substitute numerical values in the above equations. would be reached in a time sW T = . 1907. the slope of the curve at the origin gives the rate at which the temperature would rise if all the heat were If the temperature continued to rise at this rate. per (52) where .0038 (1 0. but not upon its cooling area or the nature of the radiating surface.09 for copper.) kg-cal.11 for iron 0. T cxA. the following relations obtain Q = = s = s = 1 0. r . limiting temperature is rise.00127?. 0.250*0 sq.89 sq. + a = 0. X (loss in kw. RATING that is.00127*0 C. AND HEATING 339 dependent upon the mass and material of the body. + 1 per 0. hence. in. cm. of Ott also showed that London Electrician. The value of a may be found from the experimentally determined fact that when air is blown across the bare (or thinly varnished) surface of an iron core its surface temperature will rise 1 C. cm.0245 (1 0. or (1 = 12. called the time constant of the body. all of the heat is absorbed and none of it is radiated. the absorbed.EFFICIENCY. if the surface of the coated with a thick double layer of varnish the radiation expressed in feet per minute.) watts per sq. per sec.25*. per sec. + the velocity of the air in meters per second this is equivalent to 0.) watts per sq. .527 X (loss in kw. (1 + X 10~ 6 kg. if v is in feet per minute. when the radiation is 0. where v is .core is 1 v is The experiments Ludwig Ott. per sec. per sec. at the first instant. p.906 0. in. X 10~ 6 Ib-cal.2386 0. per 1 C.) Ib-cal. In fact.

v being in feet per minute. cm. the rise of temperature for a radiation of per sq. (53) ^-j agree in general with the . . meters per second while for a heavily varnished surface 333 a I + 0. .0030 0. cm. ybeingin per second. 0i = constant X . the irregular distribution of heat evolution and the thermal capacity of the insulation. in.0194(1 + 0. in. 0... per 1 C.UUuo4v for a heavily varnished surface Other writers give the value of this constant as follows: .00127z> _ JL rr for a bare surface 52 . 187. peripheral velocity of armature. per 1 C. the temperature rise is given by 61 _ w ~ 460 1 a + O.00054y) watts per sq. Heating of the Armature. The temperature 1 (in degrees Centigrade) for a radiation of expressing v in feet per minute. likewise.107 v) + Temperature rise computed from the above formula will not observed rise in actual machines because it neglects the transfer of heat from the winding to the core. Taking the value of the radiation for bare or thinly varnished surfaces. cm. radiating surface where the constant is in each case to be determined by experiment.3 1 7v + 0. or vice versa. But in general terms it will be true that the ultimate rise of temperature can be expressed by the equation watts dissipated B . is watt rise found by putting a w = 1. rise of surface meters temperature.or 0. The experimental results of Ott referred to above may be changed to a form applicable to the rotating part of the machine.107v 1 (55) Accordingly.340 is PRINCIPLES OF DIRECT-CURRENT MACHINES (1 watts per sq. 71.. ~f~ U. sq. is watt per sq.Mv where w = a = v = total watts dissipated total radiating surface.

Arnold 1 300 '+0. . 1 + O. RATING Metric units AND HEATING English units 85 _ 1 341 55C Kapp .EFFICIENCY. . 288. 1 + 0. tween the curves of Fig. curve represented by Arnold's formula lies nearly midway between the two corresponding to Ott's researches.3 \/v embodied sq. It will be observed that the ipheral velocity in feet per second. 288 A is large part of the differences bedue to the fact that there is no .10 354 1 + 46. which shows the temperature per watt per a function of the per- 1000 2000 3000 Feet per Minute 4000 5000 6000 FIG.5_ O.00051*. 288. 1 + 0.0213 \/v The above expressions rise of are in Fig. 99 0. as + 0. at least for values of v within the usual limits of practice. in.OOOGy Wilson 640 . Relation between temperature rise and peripheral velocity of armature.lv + 0. . 1 + O. Esson .180 645 Thompson 100 1 . rise of be noted that all of the above formulas for computing temperature are more or less uncertain. unless applied to a machine of the same type as that from which the constants were It should experimentally determined.OOOSlz.00091z.

outer cylindrical surface. nor does a include their surface. The value of c will be smaller the greater the ratio of heat conductivity along the laminations to that across them. is v are expressed in square inches and feet per minute The numerical coefficient in the numerator is t( be taken near the lower limit of its range when the ventilatior good. In using this formula. The radiating heat. to estimate the rise* of temperature of the core.00051 v where a and respectively. p. 805. Consequently. London Electrician. . in other words. the exterior of a mass will depend upon the heat conductivity and since the conductivity along the laminations is much greater than across them (Ott found it to be from 50 to 100 times greater) it follows that unless the core is very deep the greater part of the heat will be dissipated from the cylindrical surface. 1907. of the flow of heat from the interior to and direction magnitude useful in in different directions. for the rise of has been pointed out 1 that a rational equation temperature of an armature should be of the form It where l = j^-. the value of w to be inserted in th( formula 1 is Ott. 2di -f c2a 2 1 + bv (56) 2i = sum Sa2 c of cylindrical cooling surfaces = sum of end surfaces = a variable coefficient less than unity. however. The Arnold formula for rise of armature temperature (40 to 70) is Ol ~ w a 1 + 0. but including the surface of the end connections as well as of the core: others include the exposed sides of the core in addition to the Evidently all exposed surfaces are but not to the same extent.342 PRINCIPLES OF DIRECT-CURRENT MACHINES Some absolute agreement as to what constitutes the radiating surface. it should be notec that w does not include the watts dissipated in the end connec tions. the rise* of temperature of the armature core is to be distinguished fron that of the end connections. writers specify the outer cylindrical surface only.

Electrical World. Dimensions of radiating surfaces. by conduction through the The temperature pole cores and yoke. p. London Havill. 289. <JM . 805. for the insulating compound is a better inside the coil varies upon the depth of the 1 For the results of elaborate studies of field coil by Neu. plus half the lateral area of the walls of the ventilating ducts. heating refer to articles XXXVIII. 1901. being highest near the middle of the cross-section of the coil and lowest on the exposed surface. and by direct radiation. or (Fig. RATING AND HEATING of winding ^~r- 343 w = total core loss of a -\ embedded length r~n ~~c total of length X winding <58) The value recommended by Arnold is the cylindrical surface of the core. Impregnated coils run cooler than ordinary coils. by eddy is The heat from point to point in a manner depending winding and upon the nature of the insulation. and by radiation dissipated in three ways: by convection in the surrounding air. 289) a = irdl + ird aver h(2 + X nv ) (59) In the case of the end connections. 188. plus the two end surfaces. currents. 1 The field coils are heated not only by the i z r losses that occur in them but also by the losses in the pole faces caused from the armature.EFFICIENCY. Vol. . Levine and and by Ott. Electrician. greater than that given by' the above formula. and in enclosed machines it is about twice as great as given by the formula. 56. 1907. p. In semi-enclosed machines the temperature rise is about 50 per cent. - free length of . Heating of the Field Coils. total length of winding winding (60) a = 2irdL FIG.

344 PRINCIPLES OF DIRECT-CURRENT MACHINES it heat conductor than the air rise of replaces. per 1000 feet per minute. greater than the average rise. the value of C for open type machines. The value of C decreases with increasing peripheral velocity of the armature. is from 70 to 80. C means the rise of temperature due to a radiation of 1 watt per unit area. the maximum rise at the hottest point being from 12 to 20 per cent. The average rise of temperature of the entire winding is from 40 to 60 per cent. Measurement of the temperature by the increase of resistance gives the average rise of temperature of the winding as a whole. the internal surfaces of the ducts are not as effective as an equal area on the outside.00005?. the latter of by taking the mean is determined thermometer readings at the middle and ends of the exposed cylindrical surface. and machines from two to three times greater than given (62). This is an average the decrease of C being somewhat greater if the coils are value. less if the coils are long. The greater surface presented to the air by reason of this construction permits of increased radiation. or C = where v is 75(1 - 0. of the Formulas for computing the form 9 rise of temperature of coil field coils are _ watts lost in radiating surface of coil Different writers assign various values to the constant C. because of the cooling effect of the yoke. greater than the average rise of temperature of the exposed surface. and under standstill conditions. In machines of the protected type C is in enclosed approximately 50 per cent. however. greater than the above value. and somewhat short. by approximately 5 per cent. depending upon the selection of what constitutes the radiating surface. due to fanning action. If the radiating surface is expressed in square inches and is taken to mean the outer cylindrical surface exclusive of the exposed end. by equation Field coils of the ventilated type of construction are made of concentric parts with an open space of about J^ inch between them.) (62) peripheral velocity in feet per min. Obviously. For a given temperature rise the ventilated coil will . with an average of 75.

thereby lowering the excitation and the core loss. and by the flow of the by current across the contact resistance between commutator and The rise of temperature can be computed from the brushes.44 shunt field resistance of . AND HEATING 345 dinary coil.p. 190. is Experience shows that a reduction in horse-power rating of about 30 per cent. and the increase of speed permits a will loss reduction of the flux per pole. 281). 189. PROBLEMS ohm and a 169 ohms. accompanied by an increase of speed of 20 per cent. it is clear that its rating must be reduced to avoid excessive temperature rise.5 amp. 1 where = = total loss at the TTUcomvcom commutator v peripheral velocity of commutator in feet per minute. assuming that the stray power loss remains constant at its no-load value. 1. RATING radiate about 50 per cent. The commutator is heated to brush the due losses friction. Rating of Enclosed Motors.m. taken as equal to 50. reduction in horse-power rating decreases the current and con2 sequently the i r losses.. C may' be formula W W 2\. the efficiency of conversion and the electrical efficiency when the armature current has its full-load value of 25 amp. When running without load the armaFind (a) the stray ture current is 1. than an orwhat amounts to the same thing. The core loss decreases notwithstanding the increase of speed. Heating of the Commutator. A 220-volt shunt motor has an armature resistance of 0. and approximately as the first power of the speed since the hysteresis loss is always greater than the eddy current loss.EFFICIENCY. If a motor of the open type converted into one of the enclosed type. (6) the true efficiency. the core loss varies nearly as the square than outweighs the of the flux density. or. more watts per sq. for the effect of reduced flux density more effect of increased frequency of the magnetic reversals (see Fig. P&/. power loss and the loss in the shunt winding. and the speed is 997 r.. The give a temperature rise within standard limits. in.

Further. Boosters may be 191. impossible to exactly realize this form of characteristic without auxiliary devices. the booster should have an external characteristic It is consisting of a straight line through the origin. but it may be approximated sufficiently closely for practical purposes by designing the booster as a series-wound generator with flux densities well within the point of magnetic saturation. the magnetic circuit must be laminated throughout in order that eddy currents set up by a change in the flux may not be of sufficient magnitude to retard the change of flux and so make the machine sluggish in its action.CHAPTER XI BOOSTERS AND BALANCERS. the voltage of the booster should also be proportional to the current.f. the line drop is directly proportional to the current.m. The hysteresis effect illustrated in Fig. of the The compensation. An obvious use for a booster is to raise the voltage of a generator. 104. depending upon the polarity of its excitation. or of a section of the bus-bars of a central station. its generated e. The Series Booster. Boosters. in case the load is of such character as to require the same voltage Since as receiving devices at or near the source of supply. if the excitation is of such character that the main flux is subject to wide variations. and should be reduced to a minimum. is especially objectionable in boosters. but are generally directconnected to a shunt motor taking current from constant potential mains. 116. p. by an amount sufficient to compensate the ohmic drop in a feeder supplying a distant load. by means drop 347 . being added to or subtracted from that of the circuit. is armature driven by any form of prime mover. of a series booster. 192. in other words. TRAIN LIGHTING SYSTEMS A booster is a dynamo-electric machine whose connected in series with a circuit.

. or if the demand A lead storage battery when discharged to the permissible limit gives 1. the booster The booster armature is charging period. The field winding of is connected across the main bus-bars.f. when nearly exhausted.f.S to provide for the contingency that the If. Connections of shunt booster. with consequent economy in cost of fuel: In a system of this kind a so-called shuntbooster is used to force charging current into the battery against the latter's counter e. thereby is assist- ing the battery to discharge if its voltage for current is unusually heavy. adjustment being made when the readings of the am- FIG. meters indicate that it is necessary. load.65 X (E/l. thus insuring a fairly uniform load on the generator during its working period. the total number of cells required is E/l.f. hence the machine is really separately excited. the voltage of the system is E.m. may be used to carry the At the end of a prolonged charge the battery voltage will then have risen to 2. never across its own armature. battery alone.m.. therefore. of 2. hence the booster must be .65 volts per cell to give it the overcharge" that is periodically required to keep it in good condition. and when fully charged requires " an impressed e. the connections being shown in Fig.8 volts per cell. may be added to that of the battery. low. 290. hence the above connection booster voltage is in series with the battery during the to supply a relatively small of the field winding. the booster e. The means of the field by manually controlled rheostat.m.m. and is called upon e.BOOSTERS AND BALANCERS 349 other times the battery takes charging current from the generator. 290.S). By inserting a reversing switch in the field circuit.f. or by using a reversing rheostat.

in a 110-volt system.15 volts per cell. the maximum booster voltage will be 52 volts. will reduce the battery voltage to the minimum value of 1. and it is permissible or even desirable to allow the voltage of the power circuit to fall when there is a heavy rush of current. as in hotels and office buildings.8 61 when fully discharged. the current must be considlow. Fig. /. of the motor that drives the booster need be only from two-thirds to three-fourths of the volt-ampere capacity of The capacity the booster for the reason that mum current the voltage is during the periods of erably reduced. in a 110- volt system. overcharge. flowing uniformly for eight hours. The shunt field winding of the booster. The normal.m. lo .350 PRINCIPLES OF DIRECT-CURRENT MACHINES capable of generating E (^rir - l) = OA7E volts. Their purpose is to adjust the battery voltage to the line requirements to compensate for the changes in voltage due to varying conditions of charge and discharge. for in that case the booster e. Thus. 291 represents a type of installation frequently used in such a case. as on starting an elevator. but when a fully charged battery begins to discharge its terminal voltage is 2.8 volts per cell. it is necessary to maintain a constant lamp voltage. the number of cells required will be 110/1. In 194. be not greater than one-half of the eight-hour rate. therefore requiring =-r 2. is connected across the constant potential lighting bus and its magnetizing effect is opposed by that of . 10 of them as end-cells. The design of the booster will then be completely determined when the maximum discharge rate of the battery is known. (eight-hour) discharge rate of a lead battery is defined as that current which. can be made to oppose that of the battery to a sufficient extent to bring the terminal voltage to the proper value. Consequently in such a system 61 cells would be installed.f. The Constant-current or Non-reversible Booster. thus. 290 the cells shown at the right-hand end of the battery are the end-cells which are cut in and out of circuit by means of an end-cell switch. field is The pro- number of end-cells may be reduced if the booster vided with a reversing switch. isolated plants supplying a lamp load and a fluctuating motor load. = = 51 cells. the current during overcharge should when the latter delivers its maxiand when the voltage is highest. In Fig.

by an amount very nearly proportional to the current. but this results in an load is increased excitation of the series winding of the booster and a reduction of its generated e. being no more than is excess of the normal amount. if the current is equal to the eight-hour rate.volts under At normal load the adjustments are such that the battery neither charges nor discharges. the sum of the voltages of generator and booster equals the open-circuit voltage of the battery. there is an initial tendency to draw the increased current from the generator. Constant-current or non-reversible booster. words. its terminal voltage rises or falls. and the resultant differ- produces a booster voltage that acts in the same and which is from 10 to 15 .BOOSTERS AND BALANCERS 351 the series winding. . 350). S. voltage at the battery terminals is therefore disturbed. in other ential excitation direction as the generator. rate (see p. normal load conditions. when discharged at the is greatly decreased if it is discharged at . and at the one-hour rate 1 If 8-hr. The current through the armature and series field of the booster is therefore not constant. the total variation of a few per cent. When a storage battery charges or discharges. The entire lighting and power load is then carried by the main generator. a decrease of the a momentary weakening of the series excitation of the booster and a charging current therefore flows into the battery. 291. If the motor suddenly increased. as indicated by the arrows.f the original condition of balanced . its capacity the capacity of a storage battery is C 8 amp-hr. but it is substantially so. respectively. Conversely. motor load results in sufficient to cause the battery to take up the fluctuations of current above and below the average value. thus. as the term constant-current booster might imply. the 1 change of voltage is 0. The excitation due to/ is normally the greater of the two.05 volts per cell.relieves the generator of the current in FIG.m. and the battery discharges and .

352 PRINCIPLES OF DIRECT-CURRENT MACHINES (equivalent to four times the current at the eight-hour rate). cells when charged to a normal voltage of 2 the load calls for a supply of current equivalent to the eight-hour discharge rate of the battery. provided the battery is initially fully charged. p. but in the type of service to which the non-reversible (and other automatic) boosters are adapted. by Lamar Lyndon. thereby preventing the electrolyte from The relation between discharge rate (ri) and penetrating to fresh material. The linear variation of booster voltage of course requires that the magnetic circuit be worked on the straight part (see Fig. over and above the normal supply amperes.8 volts in 1 hr. The reduced capacity is due to the fact that the high ^ rates of discharge produce chemical changes of great velocity in a thin surface film of the active material. p. if the current is such that the voltage per cell falls to 1.. it is clear that C 8 = 8i 8 and C n = ni n whence. p.) If is is the current corresponding to the 8-hr. and Foster's Electrical Engineers' Pocketbook.. This can be accomplished by so proportioning the series winding that an increase of the current through it from I aver to I avcr (1 -|. (Cn} is given approximately by the formula ' (See data in Storage Battery Engineering.21 volts per cell. 1913.. it follows .p) will produce the necesof laver 115 X 0. 875. 113). 100. the booster voltage must be lowered by = 5. If sume that the voltage of the motor circuit is 230. the current is said to be the 1-hr. the magnetization curve Change of battery voltage with varying of conditions of charge can be compensated by manual regulation of a rheostat in the shunt field of the booster. 115 volts per cell. 98. rate. say. where pX 100 is the prescribed percentage variation of booster current. rate. 7th ed.75 volts. 3d ed.05 sary change in field excitation. substituting the above approximate relation between Cn and Cs. asrequiring. rate. the fluctuations of load causing alternate charge and discharge are so rapid that the general condition of the greater rates. that .2 to 0. Thus. The function of the booster is then to produce a change of voltage at the battery terminals corresponding to the charge or discharge rate demanded by the load. and i n the current corresponding to the w-hr. and the capacity falls to Cs amp-hr. the variation is 0. For example. the corresponding capacity in amp-hr.

acts in the same a in such direction as the battery. the reversible booster shown diagrammatically 292 is frequently used. though in both types the shunt and series field windThe object of the booster is ings are differentially connected. 195. The non-reversible booster is systems in which the average motor load is small Booster. and the fluctuations are considerable. e. and the current through the shunt field of the booster must be adjusted so that its magnetizing effect exactly neutralizes that of the series winding With increased demand on the line there is a slight increase current of through S. plus the ohmic drop in the ar- The charge mature 23 of the booster itself. consequently.m. large average load.m.f. to hold the load on the generator at a constant value equal to the average load on the system. the battery must neither charge nor discharge. Referring to Fig. 292.f. the battery is so designed that its open circuit voltage is equal to that of the system. FIG. Differential or reversible booster.BOOSTERS AND BALANCERS battery suited to 353 changes very little. 292. S. as in a railway system having a in Fig. On the other hand. and a resultant magnetization of the booster direction that the generated e. if the load . Reversible In systems in which it is not desirable that battery discharge shall be accompanied by a drop of voltage of the power circuit. It differs from the non-reversible booster in that the current through its armature is not unidirectional. generated in the booster armature must therefore be equal to the drop of battery voltage that corresponds to the disrate demanded by the load. when the load on the system has its normal (average) value. leaving the battery to take up the fluctuations. discharge of the battery then takes place. It is adapted to systems in which the average load is large compared with the range of the fluctuations.

Moreover.M. which has a the load has normal voltage equal to that of the line. or its fractional part thereof.f. To compensate these changes The capacity the excitation of the shunt field must be adjusted by hand regulation of a rheostat in series with the shunt winding.F. its actual voltage may vary over a considerable range. no current through the booster field winding. and in which the booster has a simple shunt winding. ing. that is to say.) is shown diagrammatically in The field coil / of the booster B is in series with Fig. therefore.f. Auxiliary Control of Boosters. System (Controlled by the Gould Storage Battery Co. The adjustments are so made that when average value the exciter E produces an e. 196.354 falls PRINCIPLES OF DIRECT-CURRENT MACHINES below normal the shunt winding overpowers the series windand the booster voltage is added to that of the generator.m. Both the reversible and the non-reversible boosters described in Articles 194 and 195 have the disadvantage that a given change of current in the series coil of the differential winding produces a definite voltage. 293. is generated in the booster armature. no e. There is. a given change of current in the series winding of the booster will not always automatically result in the desired rate of charge or discharge. without re- gard to the fact that the change of battery voltage corresponding to each rate of charge or discharge varies with the condition of the battery. Although the open-circuit voltage of the battery is nominally equal to that of the generator and of the system. of the booster is determined by the fact that maximum current and maximum e. neither charges nor . The Hubbard Counter E. These systems have practically superseded the types of differential boosters described above. depending upon the state of the battery charge.m. and the battery. 197. occur simultaneously. latter being in turn excited the armature of a small motor-driven exciter E. the heavy current that must be handled by the series winding requires a conductor of large cross-section and a machine frame of excessive dimensions and weight per kilowatt of capacity. To obviate these difficulties there have been developed several automatic systems that regulate the battery by external means. with the result that a charging current flows into the battery.m.f. the field of the by the main generator current. equal and opposite to that of the line.

293. Hubbard counter e. When the station output has its average value the pull in such a manner of the electromagnet is balanced by a spring that the lever presses upon the piles of carbon plates.m. Automatic booster regulation. 294. The main output of the station passes through a coil S consisting of a few turns of heavy strap copper. The carbon piles are connected to each other at . results in 355 An increase of load an increase of the current of the exciter. 198. Entz system. and the battery then takes a charging current. electromagnetic pull on a core attached to one end of a pivoted lever L. and a flow of current is established through the booster field winding in the proper direction to generate in the booster armature an e. the polarity of the booster reverses. Shunt FIG. the generated e.f. R i and R%.BOOSTERS AND BALANCERS discharges. The Entz System (Electric Storage Battery Co. system of booster regulation.m. with forces that make the resistances of the two piles equal to M each other.f. that assists the battery to discharge. a decrease of load weakens the field of the exciter.f. and produces an FIG. above the average value through the series coil of the latter then exceeds the line voltage.) of external control for installations of large capacity is illustrated in Fig. Conversely.m. 294.

thereby generating an e. at their lower ends the carbon piles are connected to the terminals of a small auxiliary storage battery. in the exciter armature and energizing the field of the booster.f.356 PRINCIPLES OF DIRECT-CURRENT MACHINES the top and to one terminal of the field winding /' of a small motor-driven exciter E. The combination of the resistances RI and R%.m. the auxiliary battery and the exciter field winding /' is entirely similar to the a Wheatstone bridge. for the main battery. the spring overpowers the pull of S and the resistance of R 2 becomes less than that of Ri producing M t a reversed flow of current through /' and. the use of the auxiliary battery is not absolutely necessary in any case. the middle point of which is connected to the other terminal of The armature of the exciter E supplies current to the field /'. and the pressure on RI will be increased. equal to that of the line. In case the load falls below its average value. The polarity of /' is by the same causes that make the galvanometer in the bridge circuit deflect one bridge arms are varied. will neither charge nor discharge. the field winding/' circuits of is affected the equivalent of the galvanometer. so that the booster voltage acts in the same direction as the line voltage and forces charging current into the main battery.m. of the booster B. therefore. If the load current increases there will be a tendency to increase the generator current through S. which has a normal voltage winding. through / also. So long as the resistances of RI /. this causes a reduction of the resistance of RI and the auxiliary battery will send a current through /'. The two equal halves of the battery correspond to the ratio arms of the bridge. The field windings of E and B are connected in such order that the booster voltage adds to that of the main battery and a discharge results. The purpose of the auxiliary battery is to avoid imposing unequal loads upon individual cells . generated in either the exciter or the booster. there will be no difference of potential between the terminals of /' consequently there will be no e. or a part of it. Under these conditions the main battery.f. and RI and R 2 to the variable and unknown resistances. . way or the other as the resistances of the In installations of small capacity the exciter and the auxiliary battery can be dispensed with. and R% are equal to each other and all the cells of the auxiliary battery are equally charged. may be used directly.

Hg. this disturbs the balanced condition of the circuit through/ and a current will flow through it in such a direction that the generated e. If the exciter is omitted.BOOSTERS AND BALANCERS of the 357 main battery. connected in series across the line. 265. range pressure only hence the unbalancing of the bridge circuit can produce only moderate current through circuit/' (or/). the lever turns in the A . but this can be done only when the capacity of the booster and the magnitude of its field current are small.f. illustrated (General Storage Battery Co. Bijur system.) of in Fig. the size of the carbon piles is limited by the fact that the practically constant current through S can produce of variation the a narrow on carbon plates. designed to have a normal S or by the The battery FIG. with the result that the booster remains unexcited. are each provided with a series of taps connected to a set of contact points of graduated Spring lengths. which dip in or out of the small troughs mercury. The Bijur System external control. 199. then connected to points which have no difference of potential between them. the connections shown leading to /' are transferred to /. 295. as at PI and Pz. also utilizes the The equal ratio arms Ri and #2. An increase of load causes a slightly increased flow of current through S and the resulting counter-clockwise movement of the lever short-circuits more and more of the resistance R\ as the movement proceeds. of the booster causes the decrease of load will make the pull of battery to discharge. the spring overpower that of the coil S. open-circuit voltage equal to that of the line. consequently it Automatic booster regu- lation.m. Wheatstone principle of the or bridge. as the lever L is tipped one way or the other by of the control magnet IS restraining spring. potentiometer circuit. will neither charge nor discharge when the system is carrying its average load if the control apparatus is adjusted so that under these conditions the lever is horizontal for the booster field / is .

between the voltages on the two sides of a three-wire system (see Art. Balancers. The result of this condition of neutral equilibrium is that a change of current that unbalances the forces by an amount just sufficient to overcome the friction of the moving The parts will produce a continuous movement of the lever. b and c) represents three methods of possible connecting a balancer set for the purpose of maintaining equality. while the other unit will act as a motor and so limit the rise of voltage on that side. the voltage on the more heavily loaded side will fall while that on the more lightly loaded side will rise. 296 (a. short-circuiting more or less of R z and the current through / reverses. VI). A partial compensation of this shift of the neutral may be effected by the system of field connections in voltage shown on the heavily loaded side in Fig. 123. whereas the former fixed variation of generator current. thereby tending to still further balance the voltage on the two sides of the system. the load becomes unbalanced. Fig. thereby checking the extent of the voltage drop. in this case the drop will weaken the field of the motor. clockwise direction.358 PRINCIPLES OF DIRECT-CURRENT MACHINES . Chap. differs from other systems of external control in that the latter involve a variation of generator current proportional to the battery charge or discharge. but the drop in speed of the balancer. If. while at the same time the rise in voltage on the lightly loaded side will strengthen the field of the generator element.m. 2966. causing a reversal of the booster e. thus tending to increase its speed. or approximate equality. causes a response from the battery to the desired extent with a This follows from the fact that the magnet S and the restraining spring are so proportioned that with a given current through S the pulls due to them balance each other at all points within the range of motion of the lever. 200. with the connections shown in Fig. excitation of the booster will then go on increasing in the proper direction to relieve the generator of all but the initial variation. but the . and a flow The Bijur system of charging current into the battery.f. 296a. will prevent the generator element from assuming a sufficient part of the unbalanced load to maintain the potential of the neutral as nearly constant as would be the case were the speed to remain constant. Under these conditions the unit on the heavily loaded side will act as a generator. due to the load on the motor element.

mounted its own storage battery.E. be satisfied by Lighting.E. 1 The condition to. 129-227. 296c. b c set. The axle-lighting system. any system of train lighting is that the lamp voltage shall be maintained at a constant value independently of the number 201. 3. this automatic action The was dependent upon an unbalanced voltage. in which a small generator. Vol. in the system of 296c it depends upon the unbalanced current. is driven directly from the axle. where the series windings are connected in such a manner that the current in the neutral excites the gensible erator cumulatively. Perfect regulation is however posif the units comprising the balancer are compound-wound as in Fig. Connections of balancer three-wire system. XXI. The head-end system. while in the motor it acts differentially. . voltage on the heavily loaded side is therefore kept up by the combined effect of increased excitation and increased speed.and it is therefore possible to adjust the compounding to maintain perfect equality of voltage on both sides of the neutral. but whereas in the system of Fig. in which a under each car. iSee Trans A. 1903. Train lamps in use and independently of the speed and direction motion of the train. pp. Fig. . single constant voltage generator placed in the baggage car or on the locomotive supplies current to the entire train.BOOSTERS AND BALANCERS 359 balance cannot be perfect for the reason that the automatic response of the balancer depends for its inception upon an actual unbalancing of the voltage. 296.. The straight storage system. in which each car is equipped of of with 2. In the case of steam railroads three methods of electric lighting are in use: 1.I. a FIG. 2966.

Generators of the ordinary types do not possess inherent operating characteristics suitable for such service. in others it is mounted on the locomotive. 202. Voltage Regulation in Train of prevent objectionable lamps. as during switching. The complete equip- storage batteries. automatic regulation must be provided to compensate variation Lighting Systems. and at low train speed. In the axle-lighting system the maintenance of constant voltage is complicated by the fact that the speed of a generator from the car axle will not only vary through wide the machine but must be capable of operating in either limits. generally one for each car. or else be reinferior illumination placed by fully charged batteries. The variation of battery voltage between the extremes of full discharge and full charge requires the use of an automatic regulator in order to maintain constant voltage at the lamps. direction. in some cases the turbo-generator unit is installed in the baggage car. This follows from the fact that in the former case the generator is driven from the engine. It has the disadvantage that the gradual exhaustion of the battery results in toward the end of long runs. The standard (lead) for Pullman consists of 16 cells. corbattery equipment sleepers in order that the lights ment must include responding to a nominal lamp voltage of 30 volts. The batteries must be charged at terminal or division points. the candle-power of To the . which always runs in the positively driven same direction. axle-driven generators must be used in connection with storage batteries in order that the lights may not go out when the train is stationary or when the speed is so low that the generator voltage is less than the normal lamp voltage. and to make a machine of ordinary type conform to the requirements. Naturally.360 PRINCIPLES OF DIRECT-CURRENT MACHINES The straight storage system was used in the earliest installations of electric lighting on steam railroads.wound generator is driven by a turbine taking steam from the locomotive. may be operated when the train is parted. The design of generators for automobile lighting is similar to that of axle-driven machines for train lighting except that there is no need to provide for reversal of the direction of rotation. In the head-end system a single compound. more or less elaborate regulating devices must be used.

centrifugally operated switch arranged to establish the connection between the generatoV and the battery when the speed and generator voltage have reached predetermined pick-up values. full discharge. hence the latter. lamp circuit or by the utilization Examples of these methods are given in the following 203. to overcome voltage variations due to change of speed. as the battery ap- proaches the fully charged condition. It is possible to arrange the regulatory devices in such a manner that the voltage and current output of the generator are controlled by the battery voltage. The various methods of regulation may be classified In some as either mechanical or electrical (or electromagnetic). charge and systems the maintenance of constant voltage also involves regulation for constant current output from the generator. for it takes no account of the fact that the charging current of a lead battery should " taper. for reversing the polarity of the generator terminals when the direction of rotation is reversed. Regulation. or by a slipping clutch. 297 illustrates diagram- . delivers constant power. consisting of a rocker arm on the shaft.BOOSTERS AND BALANCERS for the variation of battery voltage full 361 between the extremes of and. in the case of axle-driven generators. Under the heading of mechanical methods of regulation may be included those axle-lighting systems in which the generator voltage is controlled by the slipping of the driving belt." that is. become gradually less. when in use. such regulation is not entirely satisfactory. In these systems the speed of the generator is maintained constant when the load increases above a definite predetermined load which causes slipping to occur. and to break the connection when the speed is below the assigned limit. In the Stone system the generator is provided with an automatic device. Under electrical all or electromagnetic methods of regulation is may (a) be grouped which voltage control of resistance in the variation by the automatic systems in obtained in the exciting circuit of the generator. there is also an automatic. or else to make the generator control the line voltage and therefore also that of the battery. Fig. as in the Stone generator. or (6) of the armature reaction of the generator to secure the desired characteristics. Resistance articles.

As the charging current increases. and circuit. Since the response of the and \$ 2 is dependent upon a variation of EL the . Voltage regulation by means of resistance in main line (Gould system). To prevent excessive overcharge of the battery of solenoid due to high generator speed. pile . the plunger V is ar- .362 PRINCIPLES OF DIRECT-CURRENT MACHINES matically a type of automatic regulator which operates by varying the resistance of a pile of carbon disks connected in the main lamp the battery voltage E B rises above normal. the upward pull of solenoid B relieves the pressure normally exerted upon R by the weight of the plunger of B. increasing its resistance to a sufficient extent to absorb the greater part of the the movement r. as during the charging. This . 297. Generator method Field Regulation. If of its plunger increases the pressure on the carbon thereby reducing its resistance and permitting an increased flow of current through solenoid \$ 2 The motion of the plunger of \$2 then releases the pressure on the carbon pile R. increase of EB solenoids Si as an ohmic drop in R. lamp voltage cannot be held absolutely constant. by the two solenoids The carbon pile R is acted upon V and B. the former responding to changes of the generator voltage due to change of speed. the contact is closed and that the generator is charging the K any increase of speed will tend to increase both the generator voltage and the charging current. lamp voltage EL tends to increase also. and the latter to For example. thus increasing the field resistance of the generator and lowerbattery. . 204. 298 illustrates a of regulating the generator voltage by the variation of a R FIG. ing its voltage. resistance R in its field circuit. Fig. assuming that variations of the battery current.causes an increased flow of current through the solenoid Si. but the variation will be small. E 204. the use of the solenoid Si and pile r increases the sensitiveness of the response of >S 2 to a change in L The lamp regulator is used in conjunction with a generator regulator described in Art.

When the speed falls below this pick-up speed. is field rheostat generator and battery erator is of the gensufficiently high to generate a voltage capable of actuat- established when the speed ing the solenoid series the generator current. 299. reinforces the pull of the shunt windS. thereby increasing the field resistance and lowering the generator voltage. is closed. voltage overpowers that of the generator and a reverse current flows through the series winding of K. The entire load is . 298. Voltage regulation by means of (Gould system). Voltage regulation using vibrating contact. with the result that the net force acting upon the plunger is not sufficient to hold the contacts closed against the gravitational pull. and the connection between The automatic switch K FIG. flowing through the winding of switch K. FIG. the battery ing S'.BOOSTERS AND BALANCERS 363 ranged so that the increased line voltage causes it to relieve the pressure on the right-hand side of R.

respectively. Combined field and line regulation (Safety Car Heating Lighting Co. is illustrated in Fig. carries the entire generator current Coil its full rated value. Fig. is controlled by the pressure exupon it by levers operated by the plungers of coils S and A. The resistance of the carbon pile C. Field and Line Regulation. 300 is a diagram of connections of a system of train lighting which. 298. sented diagrammatically at PC. 300. 298. but involving a vibrating contact analogous to that of the Tirrill The automatic switch regulator. 205. somewhat similar to that of Fig.). thus reducing the generator closes contact B Coil V operates similarly if the generator voltage rises voltage. too high because of high rotative speed. the field Thus. is A. & described separately in Articles 203 and 204. includes independent regulation of generator voltage and of lamp voltage. shunted across the set to hold the generator voltage at 39 volts on equipments having 16-cell lead batteries (2. 299. XY .45 volts per cell). if the battery charging current exceeds the safe limit. but the solenoids B and V act on the same magnetic circuit and open and close the contact C. by opening the short-circuit The object of the resistance in series with the voltage coil A. in series with the shunt field erted Coil winding of the generator. and at 78 volts on "60-volt" equipments having 32 cells of lead batteries. repre- Another method. S and is adjusted to hold the current at line. K operates in the same manner as in Fig. If Edison batteries are used.364 PRINCIPLES OF DIRECT-CURRENT MACHINES The pole-changer is then carried by the battery alone. coil C and momentarily short-circuits winding of the generator. these voltages are set at 43 volts and on part of 86 volts. like the system FIG.

the friction of the brushes against the commutator holds the rocker ring against one of the stops and the brushes are then in the proper position for sparkless commutation. thus preserving the original polarity of the generator. 300a shows cludes a diagrammatic form another system which inlamp regulator and a generator field regulator. which assists the shunt coil in holding the switch tightly closed. Fig. E. the two piles being connected in parallel with each other. The effect of variation of temperature upon the voltage coils of the generator and lamp regulators is compensated by means of resistors. In this system the armatures of the magnets controlling the generator and the lamp circuit are provided with air dash-pots having graphite plungers. When the machine is running in one direction. and which accel- erates the opening of the switch when the generator voltage falls below battery voltage. the charging current will then automatThe limitation of generator ically taper down to a safe value. having zero temperature coefficients. the ring being free to rotate through 90 degrees between a pair of stops. pivoted armature when the generator voltage equals the battery voltage. The lamp voltage is controlled by the pair of carbon piles. The pull of this electromagnet is opposed by a spring. The in . placed in series with these coils. thus bringing into action the series coil.BOOSTERS AND BALANCERS 365 limiting the generator voltage to 2. F. Reversal of the direction of rotation causes the rocker ring to be turned through 90 degrees against the other stop.45 volts per cell of lead battery is to prevent excessive overcharging of the battery when the . fully charged. current imposed by coil S prevents overloading of the generator battery is due to lamp load or to charging an exhausted battery. the design being such that the armature of the electromagnet will remain in any position within the limits of its travel is normal. in series with the lamps. The pressure upon these piles is due to a system of levers and a toggle joint actuated by a coil connected across the lamp mains. The automatic switch for establishing the connection between when the lamp voltage K generator and battery at train speeds above the pick-up speed is The shunt coil lifts the similar to others already described. The four brush arms of the generator are mounted on a rocker ring carried on ball bearings.

Company system of train adds to that of the series coil. the resistance R is short-circuited. thereby connecting the battery to the generator. 207) and in the automobile lighting generator lation of generator voltage charges.366 PRINCIPLES OF DIRECT-CURRENT MACHINES automatic switch coil K is closed in response to the pull of the voltage their when the generator speed and voltage have attained proper values. the Wagner Electric Manufacturing Company (Art. running in the direction of charge. The AHM FIG. At the same time the ampere-hour meter. 300a. Lighting and Heating lighting. Regulation made by 210). since it is dependent upon the inherent characteristics of the generator. S. Reguby making use of armature reaction under load conditions is exemplified in the Rosenberg train lighting generator (Art. contact C is closed. thereupon the switch pull of the shunt coil S is energized. tends to be maintained at constant value by the action of the carbon pile rheostat in circuit with the generator field. is AHM. This type of regulation. and when the battery is has reached its point of charged and the contact needle N contact. . The charging current. 206. that is. the generator supplies current directly to the load and the battery neither charges nor dis- by Means of Armature Reaction. may be classed as electromagnetic. so that there results a sudden reduction of generator voltage and the battery is thereby "floated" on the line. U. and current flows through the shunt coil of the regulator F. flowing through the series coil of the regulator F.

but it will be understood that with suitable modifications the principle is applicable to multipolar machines.f. 393. instead of being at right angles thereto as in the ordi- nary type of machine. the direction of and (2) that independent produces a current which. supplies current to the lamps train is at rest and also to the shunt field producing the polarity indicated by nn. 1905. The axis of commutation of the brushes BB is in line with the axis of when the winding //. which must be used in connection with the generator if the latter is to function properly. properties are which it develops an e. 301. p. The FIG. 367 Lighting Generator. beyond a certain speed. Its distinctive The Rosenberg Train first berg generator.BOOSTERS AND BALANCERS 207. battery E. The Rosendescribed 1 in 1905.m. the poles. remains (1) that it is of the direction of rotation. 301. ss. shown in Fig. embodies a number of interesting structural features and has operating characteristics that make it suitable for train-lighting service. Diagram of Rosenberg train-lighting generator. . The brushes 1 BB are connected to the battery terminals through Elektrotechnische Zeitschrift. The diagram practically constant no matter how much the speed of connections of a bipolar machine is is increased.

B neutralizes the field excitation due to //. placed at right angles to the in the same position as the main brushes of an ordinary generator.f. definite limit beyond which the main current delivered by the BB cannot increase. such diffifield is of culty as might still exist is further overcome by notching the It is clear that there is a pole faces opposite the main brushes. M the lines of force of this M the motion of the armature conductors through this cross-field then generates an e. becomes counterthe of the cross-field Mb also reverses.. and direction clockwise) the effect of this double reversal is to preserve the original polarity of the brushes BB. Rotation of the armature through the magnetic field set up by // will produce a flow of current through the short-circuited armature along axis bb. As is clear from the figure.f. represented by the arrow B opposes the excitation due to the field winding //.m.. hence no e.m.m. and this in turn prevents the direction of rotation M . thereby creating a powerful cross-field.368 PRINCIPLES OF DIRECT-CURRENT MACHINES an aluminum cell C which offers a very high resistance to the flow from the battery to the armature and only a very small resistance in the direction from the armature to the battery.m. In case . clockwise through pole rotation will result in a cross-field directed from left to right. In addition to the main brushes there auxiliary is.f. brushes M . circuited polar axis. this limit being reached when the armature m. of current property of the aluminum cell prevents the discharge of the battery through the armature when the train is at rest or when running at a speed below that at which the generator picks up this its load. field finding a path of low reluctance the shoes. The fact that the armature m. and current along the BB axis in such a direction that the armature m. B the due to the field winding means that the excitation opposes field flux parallel to the BB axis is small. for in that case there would be no e.f. bb..f. The weak field in the BB axis obviates commutation difficulties that would otherwise arise due to the short-circuiting of winding elements under the middle of a pole face. excessive current through the short-circuit bb. is reversed (that is. .m. thereaxis. b. that is a pair of shortbrushes.f. in the main brush axis. differs widely from the ordinary generator in that The machine what is usually the main secondary importance with respect to the cross-field. and current in the bb It follows.m.

The generator may be driven either by a belt from the car axle or directly by mounting the armature on the axle itself. (6) Elektrotechnisdie Zeitschrift. may be set by adjusting the rheostat in the field circuit //. assumed constant Eb = e. so that the be considered to be proportional to the m.BOOSTERS AND BALANCERS fore. (5) Substituting (2) and (3) in \$& 1 = CiC 2 c 3 i a (n/if n a i a )n 1005. Messrs. Vol. = constant exciting current in field winding // = main current output ia = short-circuit current in axis bb it n fif = ampere-turns due to // n ai a = effective ampere-turns of armature in axis BB n b ib = effective ampere-turns of armature in axis bb E = terminal voltage of line. saturation of the magnetic circuit is neglected. . Kuhlman and On Hahnemann have developed 1 the quantitative relations between the speed and current output of the machine operating as a generator.p.m.m. p.f. 525. generated in short-circuit bb <# = field flux in axis BB = field flux in axis bb 3>& = ra armature resistance (including brushes) t If flux may it.f.m. the basis of the foregoing qualitative study of the physical phenomena occurring in the machine. 369 that beyond a certain speed the machine will deliver a Any desired limit to the current practically constant current. that pro- duces the following relations will hold Eb = c^ B n \$B = \$6 = E = t (1) (2) c z (n/if C 3 n6 na ia ) ia r a (3) i there results (4) Ci\$b n (4). 24 XXVI. Thus let n = speed if of the armature in r.

- ~^ is the ratio of the maximum ohmic drop in . the field circuit controls (ia)max by fixing the value of if (c) If the machine is to act as a generator. (ia)max .--fl>a This result also shows how the rheostat in . ( E ^ = \ ~ c^ria o t I c4n n2 - a Et n2 But the term a K TO. ia must be positive. (M B ) would exactly neutralize the field excitation.m. would be a dead short-circuit on the line (or battery). at standstill. If ia indicating a flow of current (6) n = oo } ia = -Tl a or n a ia = nf if which means that at infinite speed the armature m. i a = ~. i a which means.f. hence _ or . simply. that were it not for the aluminum cell C the armature. From (a) If equation (8) the following conclusions TT - may be drawn: n = 0. or l (i ) \ 'ajmax . i i a ra (7) a E ra t -E ra -f c^n a n 2 _1 2 t JL^ (> where C4 = Ci C 2 C 3 9 2 I II. This condition therefore determines the limiting current output of the machine running as a generator.370 PRINCIPLES OF DIRECT-CURRENT MACHINES this value of 2 and substituting t 3?b in (5) E = and Ci c 2 c 3 -i (nf if n a i a )n 2 n f if ~ na J. the negative sign of into the armature from the line.

suppose that the generator is to supply a maxi50 amperes at a terminal voltage of 50 volts and that it is to pick up its load at a speed of 300 r. resulting in the expression (il) which represents an equilateral hyperbola. 302. still it follows that 5 is speed smaller in comparison with unity and that with increasing it rapidly approaches zero. it HQ = / \l -.5 X 10 6 This is is the equation of the curve shown in Fig. with only slight error. 302 . and since this ratio must be small from considerations of efficiency. la = Mmax ~ ~ pi -i '~ 2 is (9) Equations (8) and (9) show that the current zero when (10) ftf\\ H = and that increases.p. From (10). the current in the short-circuited determined by combining equations (2). or 4 = - The dashed portions of the curves of Fig. It is seen that 4 is a function of c 3 and this is dependent upon the reluctance in the path of the cross-field <&.-I^t r-r-r = I J \c 4 na ^ a ) m ax (ia } max */ Et - \c4nt - rapidly approaches as a limit as the speed For example. of variation of 4. mum current of ' 300 50 c 4 na 1 X 50 and from (9) .m. 4. (3) and (9).BOOSTERS AND BALANCERS 371 the armature to the line voltage. The manner path 66. The curve showing 4 in Fig. 302 is based on the assumption that 4 = 60 amperes when n = 300. Therefore the denominator of ia (8) may be considered equal to unity and the expression for becomes.

An examination of Fig. and while not entirely accurate because of the neglect of terms involving ra in the pres100 90 80 70 60 8 & * 50 40 30 20 10 2 4 000 800 1000 1200 1400 Uevoluti ns per 77 Min -20 -40 -GO -80 -100 FIG. Rosenberg generator. 302. ence of small values of n. correspond to negative values of ia (indicating motor action).372 PRINCIPLES OF DIRECT-CURRENT MACHINES computed from equations (9) and (11). 301 will show that in two of the quadrants of the armature winding the currents i a and 4 flow in the same direction in the conductors. they depart only slightly from the correct curves within the range shown in the diagram. Relation between current and speed. and in the other two quadrants .

Operation of Rosenberg Machine as a Motor. and let the starting impulse be in the clockwise direction.BOOSTERS AND BALANCERS they flow in opposite directions. let Fig. but it has no torque at standstill. direction it will continue to run in that direction. and there can be no cross-field to react upon the armature current until rotation through \$ B produces current and But if the armature is given a start in either flux in the bb axis. 301. and the currents themselves are i a = 32 and ib = 36.m. 303. The figure is drawn in . ^ \^a)max I _ and reaches a maximum d(4 value when 1 + i b) 2E t E t -1 = o or when n = - ~^* Substituting the values used above. . Thus. . react upon each other. Rosenberg machine when supplied with current from an external O) FIG. case is . for the armature current and the field flux due to // have their axes in the same direction and cannot.p. Relation between currents and fluxes in Rosenberg machine operating as motor. therefore. 373 former The Et ' total current in the . The reason for this absence of starting torque is clear from Fig. 303 represent the same machine shown in Fig. source will operate as a motor. the line. Et _ " 1. instead of supplying it to. the speed corresponding to this maximum current in the conductors is 500 r. The 208. . v ^ _!_ 2 .. 301 but taking current from.

consist of ordinary resistors. reaction the for armature the of field. the standstill current value from (8) is whereas the true computed from (9) is i a = ia = The range of speed through which 500. motor action occurs is from n = to ^ n = \ \ C^Uf I~~P Without going into further particulars it will be clear Fig. had the initial rotation been counter-clockwise. to the torque due to the reaction All that is necessary is to interpret negative values of i a in those equations as current input to the motor. as in Fig. developed by Inthe Electric Storage Battery Co. serves. that the speed characteristic is similar to that of a cumulative compound motor. is illustrated in Fig. Analytically. The resultant torque is the difference between these two opposing torques. therefore. when the constants are those used in the foregoing discussion. would have been opposite to that shown. marked 304). The direction of 4 is determined by applying Fleming's right-hand rule for generator action. but some care should be used in applying equations (9) and (11). the arrows on the armature conductors of part (a) show the direction of flow of ia and those in (6) serve . together with a special method of voltage control. Two of the bridge arms. similarly for ib . 302).374 PRINCIPLES OF DIRECT-CURRENT MACHINES two parts in order to show with greater clearness the effect of the two pairs of brushes.10 ohm. and between \$B and 4 always opposes the rotation. " series There is also a compensating winding. the characteristics of this motor when supplied from constant potential mains are involved in the equations derived in the preceding article for the case of the generator. especially at low speeds. rotation and. of a . In either case the reaction between <& and initial ia produces a torque in the same direction as the accelerate the armature. it is connected between opposite points Wheatstone bridge circuit (marked " control bridge" in Fig. marked R. and therefore also of \$6. corresponding to a maximum armature drop of 10 per cent. Thus. 304. . modification of the Rosenberg type of generator. A terminals. because the term involving r a in equation (8) is not then negligible as has been assumed." neutralizing purpose due to the main generator current. if r a = 0.. the direction of 4. stead of connecting the shunt field winding across the machine 209. while the other two. 301.

Electric Storage hence the readjustment of bridge currents increases the main The extent of field excitation and raises the terminal voltage. If it is desired to give the battery an overcharge. that overcharging of the batteries must be accomplished during daylight runs. is determined by the value of the resistance R' be observed that the closing of the lamp circuit through the triple'. has a third coil connected between the generator . therefore. the rise of voltage and. and the field current will change to a sufextent to readjust the generator voltage. circuit effect of Auxiliary Switch GOAmp. the overcharge switch. 304. the magnitude of the charging It will current.Fuse FIG. in addition to the usual shunt and series and the coils. Rosenberg generator and control Battery Co. It follows. circuits. the ficient lamp having been previously disconnected. thereby reducing the generator voltage to the normal lamp voltage and preventing damage to the lamps because of high voltage during charging. therefore. have negative temperature tion points of the bridge not coefficients. is opened. pole switch short-circuits R'.BOOSTERS AND BALANCERS marked IR. This has the reducing the voltage impressed on the bridge. 375 The junc- connected to the control field are connected directly across the line. in the same manner as though the generator voltage had itself decreased. A variation of generator voltage will then alter the difference of potential between the control field terminals. which normally short-circuits the resistance R'. The automatic switch.

and the switch will open when the current in the series coil drops to zero. Diagram of connections of Wagner automobile lighting generator. B. 305 (except for the omitted in recent models) are identical with those disclosed in an English patent (No. The Sayers generator. No. In this machine there is an auxiliary brush b connections of the Wagner machine. 1914. I. series field winding. the main brushes. though in a quite different manner. therefore. . 305. determines the closing of the switch by the difference be- tween the voltages of generator and battery. The Wagner Automobile Lighting Generator.376 PRINCIPLES OF DIRECT-CURRENT MACHINES The pull due to the main shunt coil is insufficient to battery. Fig. Washington University Studies. and the shunt field connected between this auxiliary brush and one of the main brushes. Pt. The auxiliary coil. Sayers in 1896. 9364) issued to W. close the switch. was driven in such a direction that the portion of the armature winding included between the terminals of the shunt placed midway between is winding field 1 winding occupied the trailing half of the pole faces. or to keep it closed without the pull due either to the auxiliary coil or the series coil. FIG. Vol. intended for constant speed operation. July. 210. 1 The principle of the utilization of armature reaction embodied in the Rosenberg generator is also used in the Wagner automobile The lighting generator. 1. Langsdorf. in conII.

there is given below a discussion of the theory of its operation and a derivation of its characteristics. it is driven Wagner ma- chine presents a number of features of considerable technical interest. the shunt winding tapped tion of rotation. the two windings being connected in series. the machine was also used as a for cranking the engine. 305. 305) and an auxiliary brush b placed midway between the main brushes and ahead of brush BI with respect to the direcIn other words. is respondingly. Fig. accordingly. worked out on the assumption that the series winding is present. the total flux under the trailing half of the pole face increased. In the form here described. but the shunt winding. but it is interesting to note that the form of the equations indi- .f.BOOSTERS AND BALANCERS 377 sequence of armature reaction under load conditions. and the generator design has been modified by the omission of the series field winding. current for this purpose being taken from the battery. In later models. instead of being connected across the main brushes was connected between one of the main brushes (Bi. the magnetizing effect of The series the shunt winding being the greater of the two. thereby giving rise to a compounding But in the Wagner machine the direction of rotation such that the shunt winding is connected across that part of the armature winding lying under the leading half of the pole face.m. winding was connected in the main circuit in the usual manner. as The of the automobile. and the e. the in a practice of combining both motor and generator functions series motor single unit has been discontinued. each provided with its own commutator. reduced to an equivalent two-pole model for the sake of simplicity. and moreover the machine is intended for variable instead directly of from the engine constant-speed operation. generated in 'that part of the armature winding included between the terminals of the shunt field winding increased coraction. As originally constructed. as shown in Fig. the machine had four poles and an armature wound with two distinct two-circuit windings. The field winding consisted of shunt and series coils connected differentially. that part of the armature winding lying under the leading halves The diagram of connections shows the machine of the poles. experience having shown that its effects contributed little or nothing The analytical theory presented below has been of value. however.

m. as is the case when lead batteries are For if the machine is delivering current at some given value of speed.f.m.f. and by tion of the armature. objectionable in the case of automobile lightif the machine were to be used would be necessary to add to the equipment ing for reasons stated above. provided the battery voltage remains substantially constant.f. prevent a further increase of current.m.m. cates that this winding exerts only with the results of tests. and the direction of this current is such that it sets up an additional demagnetizing action in the air-gap under the leading half of the pole face (see Fig. thus checking tend to Referring to Fig. away from the lead- ing pole tip. produces a considerable current in the short- circuited element. generated in the armature between brushes FIG. thereby reducing the e. this e. two ways: (1) by increasing the magforce of (2) netomotive the decompounding shifting the flux. 305 will show that the operation of this machine is not independent of the direction of rotation. but the increased current used.. because of increased cross-magnetizing acseries winding. therefore. 306). an increase of speed will tend to increase both the generated e. There is a third and most important effect that arises from the fact that the auxiliary brush b shortan element of the armature winding that lies opposite tha middle of the pole face and in which there is generated an active circuits e. it will be clear that the connections will make the machine regulate for constant current without regard to change of speed. 306. An examination of the diagram of connections given in Fig. for a reversal of the direction of rotation will cause the shunt winding to subtend the trailing half of the poles and radically is This feature not it alter the operating characteristics. and the current.m.378 PRINCIPLES OF DIRECT-CURRENT MACHINES minor effects. netizmg effect -Demagdue to BI and 6 and consequently weakening the shunt excitation. reduce the generated e.f. and is actually found that beyond a certain speed the effect of rising speed is to cause the current to fall off from a maximum value.f. but for train lighting . 305. It is evident that this double demagnetizing effect will ultimately it short-circuited coil. will weaken the field in and.

are.f. ANALYTICAL THEORY The sis following symbols recur frequently throughout the analyand are tabulated below for convenient reference: t E = Ea = = i = n = r = = r/ ra = i 8 a constant line voltage total e. then. It follows. currents ii.m. generated in the line current armature shunt field current speed in r. respectively. that i shown entering brush B 2 can be thought of as dividing equally between the two paths leading to brush BI. these two currents being inversely proportional to the resistances of the paths through which they flow.m. 305. 305. i z and i 3) indicated in Fig. Yi s and %>' s respectively. and that the shunt field current entering brush b may be considered It will to divide into two parts. that the . n.BOOSTERS AND BALANCERS 379 an automatic switch so arranged that a reversal of rotation would simultaneously reverse the terminals of the series winding (if present) and change one of the shunt terminals from brush B\ to brush BZ. resistance of shunt field winding resistance of series field winding armature resistance measured between main brushes and = = series field turns per pair of poles Z = number of armature conductors \$> = flux per pole d = diameter of armature = length of core T = pole-pitch = air-gap = ratio of pole arc to pole-pitch = number of poles p a = number of parallel paths through Bg = flux density in air-gap. ii l = Ai + l Ais (12) .p. including brush contact resistance shunt field turns per pair of poles nf I' <5 \l/ armature the line current be clear from the diagram of connections. Fig.

acting upon this circuit will be 'o > na rf of which.m.. or 17.f. due to current i.f.f.380 PRINCIPLES OF DIRECT-CURRENT MACHINES i flowing through the armature winding a cross-magnetizing magnetomotive force distributed produces over the armature periphery in the manner shown by linearly The main current the sloping line aba' of Fig. the m.m. If the length of the air-gap.m. 307. the field inten- sity at the pole tip will be given and the demagnetby ^7: oo 1U ^ ^-r. sidering the closed magnetic circuit indicated by the dashed which is drawn so as to include all of the armature con- ductors under a pole face. will produce a transverse field whose intensity at any be taken to be proportional to the m. izing flux represented by the hatched area A is 2' 10 10 326 (15) . The resulting demagnetizing effect in the leading half of the pole face. because of the high reluctance of but point.m. will be consumed in the air-gap at each pole tip (assuming that the reluctance of the iron part of the circuit is negligible in comparison with that of the air-gap).m. Distribution of m. Con- a' FIG. the flux will be much less than proportional to the m. at that between the poles.. corrected to take account of the effect of the slots. 307.f. between brushes b and BI.-^-^-. the magnetic circuit in that region. Under the pole faces this m. will then be represented to a sufficient degree of approximation by the cross-hatched area A. line P. Fig. and will have a distribution point may represented by the saddle-shaped curve. u 5 cm.f. 307.

due to current it . the effect X'2 N FIG. it will be clear that the currents in the two layers of the first and third quadrants neu6 to %Z tralize so far as magnetic effects are concerned. current at a due to 1 47T o 77. 308 then given by v ~ Ir47r ' 3Zis ' r_ ~ 4ir3Zi s 2L10 3~26~ 4 10 ^26^ 47r 3Zi s rl (16) 10 2566 Consider now the effect of the current Yi8 flowing from brush brush BI by way of the long path through conductors. This fact will be made clear when it is remembered that each conductor in this particu- lar belt of the armature winding has a return conductor in the opposite quadrant. Distribution of m. this being a consequence of the fact that the winding has two layers. 309. shown by the trapezoidal-shaped figure of Fig. 300.m. T is > anc* ^ ne corresponding intensity at that point is Tyr qnrjj. with that of Fig. leaving for con- sideration only those in the Comparing this distribution second and fourth quadrants. 310.m. Again bearing in mind that the actual space distribution of this belt of current is as shown in Fig.BOOSTERS AND BALANCERS The current 381 %i 3 flowing in the j conductors between brushes 6 and BI results in a peripheral distribution of m. Under the leading half of the pole face between brushes b and BI. 308.f.f. brush b is Z 3 A . this belt of point field opposite The m. on the whole. as indicated in Fig. The magnetizing is flux represented by the hatched area in Fig.f. 308. a magnetizing one. it becomes evident that the distribution of transverse flux due to identical' with that Yi & is due to %i s) but that the total effect is only . of this current is.m.

Taking ^ = becomes (nearly) (19) In addition to the demagnetizing effect caused by the trans 1 verse armature reaction. that .7. 309 FIG. FIGS. 309. This brush short-circuits elements that are moving through the strongest part of the field. 310.382 PRINCIPLES OF DIRECT-CURRENT MACHINES one-third as great as that due to the latter. Since the is opposite the center of the pole face. It may be assumed. or 10 325 (18) FIG. therefore. so far as armature reaction is concerned. there remains to be considered the effect of the short-circuit current in the armature element under brush b. and 310. Consequently the total magnetization contributed by the currents %ia and %i t is <t> m = 3'TO 4 4?r [1 - (2^ - 2 I) ] = 4rr ZisTl^ To" -l)*] (17) The face is net reduction of the flux in the leading half of the pole then. so that the current set up in them may have considerable magnitude. (18) Distribution of components of exciting current. 0. given by it is the difference between (16) and (17). the field intensity at that point is not appreciably affected by the cross-magnetizing action of the main current i and not to any great extent by brush b t the currents %4 and }i a.

(20) the net flux in the leading half pi. Q^\$ - ai + Pis - yn) Z'n - i s ra ~^~ Substituting in (21) the value of there is i3 from (14) and solving for itt obtained (noting that (22) . generated in the armature between brushes BI b is then (%< where Z' + Pi. is directly proportional the speed. - yn)Z'n = d OU Xr^n \/ ins' /\ JLU P being the number of poles in the of parallel paths through the machine and a the number arma- This is the e. it becomes necessary to divide the right-hand member of (19) by the number of pairs of poles in the machine in order that the expression may be perfectly general. therethe short-circuit current as well.BOOSTERS AND BALANCERS 383 fore. Since this short-circuit current is always so directed that it produces a demagnetizing effect in the leading to half of the pole face.m. responsible for the production ture winding.^ * ^ 8p (*** p = Since the total flux per pole of the pole face is is <l>.yn The and e. whereas the actual machine is multipolar. The values of a and p follow immediately from equation (19). Y^ ai ai + . and upon the resistance of the element and the + brush contact. its value is derived in equation (40).f.f. of the shunt field current z a hence . but since (19) was derived on the assumption that the machine was a bipolar one.m.m.f. generated in the short-circuited element and. it follows that the total demagnetization in this half of the pole face is given by an expression of the form Pis (ai yn). where 7 is a constant depending upon the average flux density in the air-gap. the e. therefore. the number of turns per armature winding element.

Substituting (22) first in (24) and then in (26). (26) the final form of (26) being obtained by substituting the value of ii from equation (12). may magnetic saturation is be taken as proportional to the net if and \$ = and the Ci(n a i8 is n fi) (23) total generated e. Ea = where 3>Z'n = cn(n 8 i s - n f i) (24) (25) Since the total generated e.curve which shows the relation between the current output i and the speed n.f. there result.^ (30) Equation (30) represents a . that is. there is the further relation that Ea = E + t irf + i. neglected.f.m.)j8Z' 7 + - ^ aZ'] + (ra+r^ + y^}. must be equal to the sum of the terminal voltage and the ohmic drops in the series field winding and in the armature.m. the flux excitation. and E4 l -o) = E>+ i r [ '+ r '-MaZ>n+ %)].384 PRINCIPLES OF DIRECT-CURRENT MACHINES useful flux 3> is The produced by the differential action of the shunt and series field windings. (29) Dividing (27) by (28) and solving for it is found that *+yZ'n* + E a t + (ra +r. respectively. and reducing.(2r a } = E t + i(r a + rf ) +^ i.jnr + %r a z/nl (28) where R = ra pZ'n i. The form of .

and the FIG. as shown by the curves y d and y d Fig. 311. represented by a curve of the form y n or y^. 311 are not drawn to any particular scale. The" shape of the curve corresponding to (31) will be that of .BOOSTERS AND BALANCERS 385 this curve may be found by noting that i is the quotient of two functions of n. and will be that curve y n if the three roots of y n = The curves of curve y n if two of the three roots are imaginary. but are merely But whatever the roots of y n = 0. are all real. . of Fig. one of which (the numerator) is (31) and the other (the denominator) yd is = n 2 (an s - &nf )cZ' +n\cn f (r s T (ra + rf )(3Z' - a Z' + (r a rf ) (r s - (32) The function y nj since it is a cubic equation. Curves representing equations (31) arid (32). function yd will be represented by a parabola with vertical axis. will in general be Fig. illustrative of typical forms. 311. 311.

and the roots will both be 2 if # <4ws. that the latter condition is the only one that can arise in practice. shown.386 PRINCIPLES OF DIRECT-CURRENT MACHINES the curves will cross the axis of ordinates at a distance below the origin given . . o it intersects the axis of ordinates at a distance r 2 ~| + r. 311) must be as close Inas possible to the origin and point B as remote as possible. is no demagnetizing effect due to the winding element shortcircuited by brush 6. easily In any case. For which would mean physically that there 0. 311. that there is no current also the points in and suppose that 7 = . the curves representing y n and y d must be so related that their ordinates corresponding is machine will both be positive for a large range of speed (in the positive sense). in which case the parabola two points both of which lie to the left of the origin. Similarly. to. The curve showing the relation between i and n can then be obtained by dividing the ordinates of the cubic curve by the ordinates of the parabola. or in other words. in which where m= q s (an s cnf (ra = = (r s + |) -^ ) - (r + r. if the to operate satisfactorily as a generator. will cut the axis of n in .) (r. + %r if r^2 -j- The two roots will be real q 2 > 4ms. spection of equation (31) will suffice to show that the constant n 7 is the controlling factor in determining the shape of curve y' which it intersects the axis of n. by E t (rs + %r a ). as in curve y" dj Fig. curve y' It is very corresponding d imaginary. the intersections with the axis of n of the parabola represented by (32) can be found case by putting yd = 0. which means that the parabola representing the function yd does not intersect the axis of abscissas.) (ra -f fr a ) -^J above the origin. on inserting the actual values of the constants comprising m. Obviously. q and s. which means that the function y n must have three real roots as in curve y' n and that point A (Fig.

312. and in particular to determine the condition that i will give rise to will readily be seen that three real roots of the function y n = 0. the numerator of (30) reduces to y = ' . the curve showing the relation between and n will have the form of Fig. Fig. it is important to investigate this case further. then. In this case the curve obtained by dividing the ordinates of the straight line L by the corresponding ordinates of the parabola would have the form shown in Fig. the current through the machine will be negative for positive values of speed. other hand. 311. to the analysis of equation (31). FIG. it is easily dn = Z'ri2 - - ra yZ'n - ^1 If = -[ScnsJ Z'n - r aJ Z'} dy. curve y'n. 311. FIGS. Proceeding. ' is put equal to zero. 312 and 313. for it if two of the roots are imaginary. Since in practice 7 will always be greater than zero.E t (r8 (34) which represents a straight line shown as L in Fig. Relation between current output and speed. . 313. it is asymptotic to the axis of n at infinity. when 7 > 0.BOOSTERS AND BALANCERS in 387 then this short-circuited element. it is found that the condition which determines the position of the points of of curves y n or y n is maximum and minimum n = Gen. seen that dy. 312 within the range of generator On the action. which would mean that the machine would run as a motor and not as a generator. as in 7 = FIG. 313.

388

PRINCIPLES OF DIRECT-CURRENT MACHINES

which means that one of these points lies to the right of the There is a point origin and the other to the left of the origin.
of inflection in the curve

where -~%

=

0,

that

is,

when n =

points of maximum and minimum. that the three roots of y n = shall be real, the function can be written
or

midway between the To find the condition

which, on substituting
.
.'

-x +

becomes

^
--} cnj
1 J

.:..-,,;

(36)

~

+ !2cV\ x - -( + cn yl l^^J^ + r Z' 6\2Z'
L

7 VIZ'
s

+

en)

-

~^ 108c n 8 3
3

3

=0

(37)

This equation

is

of the

form
X3
%X
-frj

=
3

which

will

have three
2

real roots

if

the absolute value of |= 27

is

greater than

j-.

In this particular case

and

E

(38)
t

cn s yl

Z'

6\2Z'

'

cn s /J

108c 3 n s 3

To reduce the values of and 77 to simpler and more manageable form, it is necessary to evaluate the characteristic constants that enter into their expressions. Thus, from (23)
flux per pole
Bg^/rl
\fsrl

amp-turns per pair of poles

k

X

1.65^5

1.6&5

where k

is the ratio of the excitation required for the complete magnetic circuit to that required by the double air-gap. Unless extremely high flux densities are used, k will be somewhere be-

BOOSTERS AND BALANCERS
tween 1.2 and 1.4, with an average value of about from (25) that
c

389
1.25.
It

follows

=

CiZ

'

=

'

~
8

(39)

The magnitude

of the constant

7

is

eration that the demagnetizing flux

cf>

determined by the considcaused by the current in

the element short-circuited
If

by brush

b is given

by yn,
.

or 7

=

the e.m.f. in the short-circuited element
is
,

is e s

c .,

the average

current in the element will be
resistance of the short-circuit,
47T

c>

=

,

where

r is the

apparent

and
'

_Z_

.

*

~

10 2S
25

ls c
-

4ir

Z trle

s.c.

10 2S 45

r

where S

is

the

number

of

commutator segments.

In practice,

^
o^
1>

as a general rule.
es

But

e,. e .

is

given by

1

.

c.

--

X X

1C 8

S

"

2

hence

7

=

=
n
2400
c

10 8

W V
in (38), it will

Substituting these values of
2

and 7

be found that

terms involving r a and negligibly small in comparison r in with the others. Dropping a comparison with r s the values
r a 3 are
,

of

and

17

become
12
dr /S\* /I 10 10 E Etdr (\* / ir,*Z'BT*l* \Z) \

X

t

+
I

1.09 kZ\
)

<n

=

38.4/c

X

10 10

E

t

rs 8 2 r

(41)

/S\"
I

Substituting these values in the criterion

l
1

3

27
The
factor

^
4

2

|

arises from the fact that in a series winding a single brush
series.

short-circuits

elements in

390

PRINCIPLES OF DIRECT-CURRENT MACHINES
it

and reducing,

follows that
k'
2

10.85

rs 2 5B g ap

1.09/cZ\

3

\8

(42)

The

physical meaning of the last result

is

that unless the resist-

ance of the short-circuited element, including brush contact resistance, is greater than a definite limit, the demagnetizing effect of the short-circuit current will so far weaken the field
at the leading pole tip as to prevent the machine from building up; that is, an extremely small value of r would result in the

machine taking current from the
all

line, or running as a motor at In practice, the contact resistance between the carbon brush b and the commutator will be sufficiently high to insure that r will exceed the critical value, which

positive values of speed.

means that the characteristic curve of Fig. 313 will always cut the axis of n in two points lying to the right of the origin and in one
point to the left of the origin; the latter point has no practical significance, since it relates to the condition of running back-

ward. The three points of intersection with the axis of n can be located by finding the three roots of (37) and then substitutSince the coefficients of (37) are ing for x its value from (36).
3

2

necessarily so related that

^>

~T,

equation (37)

is

the irreducible

case of Cardan's rule, and the solution must be found onometric methods. The three roots of (37) are 1

by

trig-

sin

(60(43)

sin (60

where
sin

=

~

le

V27
1

Chauvenet's Plane and Spherical Trigonometry, pp. 99-100.

BOOSTERS AND BALANCERS

391

of Fig.

hence the abscissas of the three points of intersection of the curve 313 (only two of them are shown) are

m

= -^Qcn s
r,.

+

2

x 3 \/f

sin

ye
3

(60

- HO)

(44)

Substituting the values of

and

17

from

(41) in the expression

for sin 6 given in equation (43),
3.28/b

Z

sin 8

=

rs 't!-'*i

r

a5B g p
i

no^y\

a

/^\

It

has been previously pointed out that satisfactory operation

and of Fig. 311 shall of the generator requires that the points be separated as widely as possible; and it is also important that
point

A

B

A

(corresponding to the speed n\) shall be as close to the
It follows, therefore, that sin 1^0,

origin as possible in order to secure the

pick-up -speed.
also sin

6, shall be small. Inspection of that the conditions to be satisfied to meet this requirement are

advantage of a low and therefore equation (45) shows

as follows:
1. Make the resistance r as large as possible, that is, make the brush b of hard carbon with the smallest practicable dimenThe desirability of this design is involved in the discussions.

sion of Fig. 312.
2.

Design the armature winding with one turn per element

-)
3.

Design the shunt
(

field

winding so as to have a low

resist-

ance per turn
4.

\n s

small

/

)

.

Use a simple two-circuit armature winding

(a

=

2).

5. Make the air-gap as small as possible. 6. Select a moderate value of flux density in the air-gap and make Z correspondingly large. In other words, the armature

should be magnetically powerful.

392

PRINCIPLES OF DIRECT-CURRENT MACHINES
note that for moderate speeds the cubic 311 does not deviate greatly from the straight For the slope of the line L, as determined by (34), has the

It is interesting to

curve
line L.

y' n of Fig.

constant value

E

t

(-^

+

&Z

r

]

,

and that

of the cubic curve

is

given by
yZ'(2cn.n*
hence,

-

r.n),

when the speed n is low, as in the vicinity of the pick-up speed, these two slopes will not differ greatly, and will approach This fact may be used to deterequality as 7 approaches zero.
will give rise to the maxicurrent output of the machine. It is possible, of course, to determine this condition accurately by differentiating equation (30) with respect to n and equating the resulting expression

mine approximately the condition that

mum

to zero; but this procedure develops excessive complications for the reason that the differentiation gives an equation of the fourth

degree in n.

If, therefore, equation (30) is simplified by dropping terms containing 7, the approximate expression for i be-

comes

ran 2

+
=

qn

+

s

Putting in the condition that

0, it is

found that

n

=
CUs --

r__ ~
R7> + pZ
_L_

/.

a

^

^

(

CUs

V-g-

+
_1_

R7> / } pz

m

^

/.

m ( CU
\-g

Inspection of (46) will show that

is

the approximate value of the pick-up speed of the machine. Furthermore, the constants q and s are very small compared
ra,

with

so that the terms
ra

and
ra

in

(47) are

negligible.

It

follows, therefore, that the

maximum

current output will occur

BOLSTERS AND BALANCERS
at a speed

393

greater than twice the pick-up speed; this checked by test results. It has been previously pointed out that experience has shown that the series field winding can be omitted without materially The analytical reason affecting the operation of the machine. It will be for this fact is clear from the form of equation (30).

somewhat

conclusion

is

observed that the factors rf and n/, respectively the resistance and the number of turns per pair of poles of the series winding, appear only in the denominator, that is, in equation (32), which
is

represented graphically by the parabola of Fig. 311; since

= 2n(n. - W)cZ>

+

cn,r

it is

clear that within the
,

working range of speed

(to the right

the principal effect of r/ and n f is to alter the slope of the parabola, but only to a small extent since in any case
of the origin)

their numerical values are relatively small.

INDEX
(Numbers
refer to pages)

Armature, characteristic
circuits,

number

of

167 98

Abampere, Abcoulomb,

definition

of.... 11,

17
12

coils

105, 268, 295,

definition of

Abvolt, definition of Acceleration of trains
Acyclic generator Air ducts, armature
-gap, area

12

conductors, number of .... core construction
excitation

300 96 62
128

227 72

disk....;

63, 89

63 120 ampere-turns 119, 120 142 chamfering required by flux density 65
.

distribution
145, 147, 152, 155 121 fringing correction

All-day efficiency Alternating-current machines..
e.m.f. rectification

Alternator

generated e.m.f
Altitude, effect
rise

325 44 52 44 46, 50
.

88 62 e.m.f. generated in 56 145 field, shape of 283 flux 340 heating 133 magnetizing action of reaction 133, 137, 205, 286 361, 366 regulation by resistance 57, 305 \ 88 ring

drum

eddy currents

.

.

:

.

winding see Winding Automobile lighting generator
366, 376
Axis, of

on temperature
329 328
15

commutation

137, 241

Ambient temperature Ampere, definition of
Ampere-conductors Ampere-turns
air-gap

geometrical neutral neutral

armature core
armature, per pole cross-magnetizing demagnetizing field, per pole
interpole pole cores

146 25 119, 120 128 142
138, 141 138, 141

Axle lighting system

133 134 359

B
Back ampere-turns
e.m.f
138, 141

Balancer

191,

199 358

142 300
128 124 129 44 142

Battery ampere-hour capacity
350, 351

and shoes

discharge rate

350, 351

teeth

end

cells

yoke Armature
amp-turns per pole

train

equipment

350 360
349, 351

voltage range Bearing friction 395

314

no-load. 293. booster control. 259 self -inductance. 243. 276. 237 256. motor Chord windings demagnetizing 208 101. Coefficient of coupling 37 dispersion or leakage. 293 field 142. 115. mutual inductance of 273 100 numbering of order of commutation. 292 262 Commutator blackening 19. 295. contact resistance 240. 373 162 Bipolar machines 59 separately excited genera- 138 demagnetizing amp-turns. tor . 243. 354 motor capacity for shunt 350 348 negative non-reversible 350 reversible 353 . 241 criterion 246. . 301 Commutation 135.. commutation 267 161 Characteristics armature compound generator motor load . 263 period of 240 resistance selective. constant current 350 differential 353 Entz control system 355 Hubbard control system. 297 poles. 102 . 262. 243. 323 friction loss holders lead 315 66 137 pressure sets required 'wide. 138 347 Booster 354 auxiliary control 357 Bijur system of control. 263 sinusoidal 238. . law of Rosenberg machine. 239. 76. magnetic potential due to . . in wave windings. 306. 300. 300 105 Coil. 248 259 successive phases under-. 33.f 242.. 244.396 Bijur system. current density 67. 249. 254. . 171 Capacity of machines Carter correction factor criterion for 326 122 . 347 348 348 231 292 dummy .m. 280. 246. 113. . 71 . . 273 335 self-induction. 143 . cross-magnetizing turns amp- motor series 204 168 209 173 generator motor shunt generator .. 279. 295. 267 286 improvement of linear 238.300 79 343 . INDEX 357 10 Characteristics. 246.. . 247 on axis of 16 drop of potential at 163. 249. . 268. 297... . armature 36.369. 66. 116 Biot-Savart. . 168 181 211 20 290 Commutating devices e. 256 240 voltage without auxiliary devices. field 293 heating of field intensity 295. mutual-induction output series shunt track return Bridge control Brushes'.. 262 237. 248 over238. 293 effect of.166 construction 60.117. 298. 264 adjacent coils axis of 137. 129 . effect of 67 93 264 width Building up of generators 294 75.. 71. .. 83. 268.

129 147. Current. coefficient of. 296 pole 128.117. 295 287 Compensating devices Compensation of armature reaction 286 77 Compound excitation . number of ture arma96 350 D Demagnetizing action. 382 301 Conductors. at commutator segment 250 due to.89. 141 corrected Constant current booster 171 regulation Contact resistance of brushes 240. 345 62 peripheral velocity 97 pitch segments. . 282. absolute unit of continuous density in brushes 211 17 11. turns . 263 43 Dynamo 58 construction of 59 bipolar and multipolar . 181 eddy or Foucault force due to heating due to induced 62 10 15 motor characteristics effect 188 211 inter- 8 Compounding poles of 15 practical unit of. direct direction of force 12 generator characteristics. 329. compo. Drum 380 148 demagnetizing nent. 152. . . 214 37 Coupling. starting of Direct current ampere-turns for corrected length heating of machine Discharge rate of battery Disk armature . 137. 247 143 of Contactors Continuous current rating of motors Control of boosters bridge series-parallel shunt motor speed.. law of Counter e. current density 250 number of 96 voltage betweeen 61. 382 ampere-turns 138.INDEX Commutator diameter friction loss 397 magnetizing ampere: . windings coils Dummy . 89 . 314.. 315 'heating 61. Controllers Cooling curves Core. 378. 332 354 231 227 217 229 336 62 128 123 component Deri winding cross-mag- netization 148 Diamagnetic substances Difference of 289 1. 334.. 254.f 199. generators . 188 motors 235 . 304. 378. trical 12 magnetic Differential booster compound motor motors. 141 . 155 Division of load.m. Cumulative compound motor.138. generators in parallel . coefficient of 246. armature 233 45 326. 107 102 Duration of short-circuit 237.. 61 Cross-field. 249. 306. 315. 267 Criterion. 339 losses. Distribution of flux in air-gap 7 353 211 215 44 44 350 63. sparking 368 Cross-field magnetizing action of 137. 25 elec- potential. . 45 293 44 67.. short-circuit 238. 300 ampere-turns for 3 Coulomb. . Dispersion. . 345 65.

295 81 ' rheostats step of windings Element of winding 90 100 . 214 9 magnetic pulsations of reaction short-circuit 280 generated and induced . 56 solenoid straight wire at pole tips ".. shape of coil heating 343 commutating. magnitude of induced of mutual-induction rectification of alternating of self-induction 48 12 34 115. tractive effort of 348 8 13 Electromotive force of alternator 38 8 46. 12 regulation of voltage. number of Dynainotor End cells 96 350 105 connections. 244. 155 Electric Storage Battery Co.. 297 control of speed 219 cross368 displacement of windings. 16 17 11 141 1 counter direction of induced 199. system of train 374 lighting Electrical potential 9. quantity of 361 11 Electrolysis Electromagnetic induction Faraday's law of Electromagnets. train lighting currents upon pulsations of unit of 52 33 52. 13 Feeders.. 50 142 ampere-turns per pole 145 armature. 142.. 295. 15 362. armature core pole faces . 311 313 pole faces teeth 312 Efficiency 304. armature conductors core 309. field excitation INDEX 73 68 Elements. form of E 62 66 351 loss. train lighting Electricity.. 293 of direct current generator. negative Field. 249. 276. 246.398 Dynamo. 149 297 fringing 343 heating of coils intensity due to circular coil 3 commutating 242. 98 44 dynamo 37 energy of magnetic excitation 73. 324 mechanical 318 321 variation with load Electric Power Club 334 Eddy currents. 317 325 all-day commercial or true 317 of conversion 317 electrical 318 maximum 321. Faraday's law of electromagnetic induction 8. air-gap distribution 348 of 147.. 55 12. 37 Entz system of booster control. 271 Energy stored in magnetic field. 343 leakage flux 268. 152. 364 reversal of commutating 142. 282 regulation. 355 187 Equalizer 107 Equipotential connections heating lines and surfaces of 8 see Excitation dynamos Field Excitation. 280. .

mechanical equivalent of. classes of temperature limits Intensity. 8. 181 division of load 8 24 2 magnetization by Insulation. homopolar or uni72 76. 362 Gravitational potential II 8 145. 320 121. 123 Fringing of flux Froelich's equation 179 . 293 Heating of armature 328. characteristics. 328. Inductance of armature see Self-induction coils polar building up. . 10 direction of. 300 304 . 147.f 3 48 43 Imbedded-detectors Induced current e.m.. 374 162 75. 171 Induction. 71. 329. 354 116 306 Gauss. electromagnetic lines of 185 combined output compound.. 143 Hubbard Hysteresis loss. 79. 262. 271 268 268. magnetic. 331 8 12. due to current. field axis of coil Lundell parallel operation polarity Rosenberg separately excited series shunt stability of operation three-wire 188 297 186 83 366. 15 of end-connections 343 of field coils 343 of railway motors 332 326 Heat run 33 Henry. . train lighting. system control of booster 314. 152. 376 ing 21 Gilbert. 48 acyclic. 83. turbo- Wagner automobile density air-gap pole cores teeth distribution... .INDEX Field windings 75. 332 3 16 17 '. 359 Heat. 270 tooth-tip 20 5 from magnet pole Force.f. 155. demagnetizing Friction loss effect of. definition Generated Generator e. 4 Flux 399 70 light- Generator. 129 16 268. due to current. 168 173 170 191 328 328. . 340 commutator. 220 279.in. 32 right-hand rule. 283 Head-end system. 343 12 Fleming's left-hand rule 9. leakage end-connection slot 117. 333. 345 curves 336 due to current. definition 72 Homopolar generator 328 Hottest-spot temperatures . air-gap 24 65 65 297 366. solenoid of magnetization 39 71 Interpole machines motors Interpoles Iron loss.75. 12 linkages Foucault currents see Eddy Currents Fractional pitch windings 101.. 297. definition Gould system of train lighting.. 76.

. 314. 306. armature 199 16 losses 304. armature commutator field mechanical miscellaneous stray power 305 306 305 314. 129 Laplace.m.. reluctance 6 20 26 13 130 Magnetism. enclosed. Lenz'slaw 12 32 Magnetomotive force 359 Lighting of trains Lincoln adjustable speed motor. 309. characteristics. INDEX eddy currents. 221 Linear commutation. 91. air-gap. 129 flux. uniform flux force. 271 pole cores shoes slots 19.308. M Magnetic circuit. magnetic 137 117 coefficient of. 30. 248 24 Lines of induction 8 equipotential force Magnet flux pole. law of Lead of brushes leakage potential of coils Leakage. . Joule.113. Loss line Lundell generator. 199 306. 211 counter e. .312. law of examples of application. and tubes of . 214 copper. 324 3 Linkages. 83 24 Magnetization curves of dynamos. 199. 308 314 . 108. 263 distribution. 257. 320 361 regulation of voltage Miscellaneous losses. 304. number of flux 20 166 325 304 Load characteristic curve Losses core or iron. 320 pole face teeth total 350 capacity. lines 145 4 4 3 L Lap winding.. definition of 15 26 27 1 K Kirchhoff's laws.. 345 43. Mechanical characteristics. 143. shunt booster compound. 315. 315. 115.400 Iron loss. 238. . 100. motors equivalent of heat 4 Motor 313 308. . 312 314. 117. . 320 315 317. 316 Maxwell. unit 2 39 133 21 2 5 from efficiency definition of Maximum 322. by induction intensity of Magnetizing action of armature Left-hand rule. tooth-tip 268 268. rating of interpole power of summary of 318 317 railway heating of reversing 85 235 345 220 199 224 332 214 .. 116 . 270 Fleming's 14. due to current 263 10 direction of 10 12 117.. field . .268.313 306. end-connections..f direction of rotation division of load. 243. 165 297 hysteresis total. 244. residual 75.

19. unit. 297. shunt. 238.. . .f. 12 6 energy gravitational 8 6. ampere-turns for . construction of flux density in 65. 364 71. 373 204 209 208 217 303 202 215 Peripheral velocity of armature. 218 273 34 36 leakage shoes. coil . Mutual inductance.. ..INDEX Motor. armature induction coefficient of . effect on temperature rise . 20 199 O Oersted. 2. 243. interpole starting of differential . .. 300 128 296 65 131 Multi-voltage speed control. 59 154 9. . .and demagnetizing effect 139 . . 262 losses Q Quantity of electricity. corrected length 341 23 27 68 97 83 121 199 torque 68 Motor-generator 94 Multiplex windings Multipolar machines 45. compound. . . construction eddy current loss X Negative booster feeder leakage flux saturation unit magnet Poles. . 332 rating 332 stand-test 326 Rating continuous and short-time 326 345 enclosed motors Railway controllers motors Parallel distribution . . 11 R 229 224 225 cycle of operations 332 heating 226. Rosenberg separately excited series 401 shunt speed regulation. commutating magnitude of field.. .159 . . commutating cores. choice of number of Potential curve. . 279. 25 237. interpole motors. changer 361. . 186 windings Paramagnetic substances Period of commutation 92.. 313 130 296 . pulsation. . .m. 263 . operation of generators. discovery of Ohmic 304 armature 305 commutator 306 field 305 335 Output coefficient 334 equation Overcommutation. 246. ampere-turns for. . . 59 cross. . electrical. . 298. . . 5 Neutral axis geometrical Non-reversible booster 348 348 134 133 . definition Pulsations. 350 Power magnetic of motors e. Permeability Permeance Pig-tail connectors Pitch of windings Polarity of generators Pole arc. 128 65 66. 280 55 303 26 10 speed. 100 1.

103 Reentrancy. generators. armature 133. 91. 300 e.. . lamp Tirrill 362 217 360 362. motors speed regulation windings Simplex windings Simpson's rule Sinusoidal commutation. train lighting motors. 282 348 76 76. . 256 Rocker ring 66 Rosenberg generator 366. 295. railway motors.226. 1 15.238. compensation of components of Saturation of pole tips teeth Sayers' winding generator 296 297 290 376 Segments of commutator. applications of. 217 9 Right-hand rule. of 33 Separate excitation Separately excited generator.m. 83 203 Rheostatic control of speed. motor Series booster 204 347 : Reluctance Resistance. Ring winding 89. characteristics . . characteristics in parallel 305 brush contact 240. num- Rectification of alternating e.. 238 382 237. . 268. 300 32 Self-induction coefficient of. effect on speed.402 Rating. 247 commutation 240 field. 92.. 195 26 distribution excitation 159 74 75.. 247. 374 machine as motor 373 Rotation of motors.. 282 . 178 measurement of temperature rise by 331 regulation. 241.. 295.263 115. 128 248 . S Safety Car Heating 208 217 76 94 . Fleming's . .361. train lighting Regulator. . definition 161 by armature curve field reaction. 366 for constant current 171 168 362. field windings Short-circuit of adjacent coils. Reaction. shunt motors voltage. 279.f 52 94.. 295 current density 250 Selective commutation 263 Self-excitation 74 Self -inductance 33 of armature coils. direction of 85 discharge motor starting current curves demagnetizing effect duration of reaction of Shunt booster excitation of. 364 227 159 103 259 Reversible booster 353 81 distribution Rheostat.. 101 362. degree of Regulation. core loss shunt windings Series-parallel control 186 221 209 320 184 75..33. & Lighting system of train lighting 364 . 366 of short-circuit currents. speed.f.. Reactance voltage. 364 96 average voltage between.. 268. . 73 162 ber resistance. . INDEX 332 293 205 286 137 regulation of voltage by..m. .. 61. . 361. . 364 . 248. train lighting.. 173 generators in parallel 187 . 168 armature winding 57.

403 . . 328. 361 . 339 Tirrill 217 227 Speed-time curves Stability of operation of generators 170 332 Stand-test. 333. . 235 Time constant.5 70 see Storage battery see Battery system. interpole motors regulation. 331 328 328. 256 Swinburne's commutating device.. definition of 63. field intensity on axis. system of train lighting 4 72 366 Teeth. 243. 330. compound Thermometer. 291 Unipolar generator U. Torque compound motor motor shunt motor Track return booster series separately excited motor . 339. 329 330. 170 177 303 regulator Tooth-tip leakage flux 195 268. leakage flux pitch shape of Slotted armatures Smooth core armatures Sparking constants criterion. 268 97 64 120 119 17 281 246. 270 64 328 Ventilating ducts Volt. 207 211 208 348 38 359 Tractive effort of electromag- nets Train lighting .. 295 . 215 rheostats 203 Stone train lighting generator. 270 199. . 329. straight Stow multi-speed motor Straight storage system Stray power loss . ambient 124 312 297 308 268. S. 341 Electric Power Club time-constant-. . 334 339 of Speed characteristics. 214 . 204 Thompson-Ryan winding Three-wire generators 330 287 191 motor shunt motor effect on of. Solenoid. 202 212. Lighting & Heating Co. shunt motor.INDEX Slot. Uniform magnetic field . 359 221 359 318 Turbo-generators Two-circuit windings winding -layer windings Wave 104 Surfaces. ampere-turns for eddy current loss in flux density in hysteresis loss in leakage flux shape of Temperature. Undercommutation 239. measurement resistance motor series by separately excited motor. characteristic see U 8 Characteristics. 214 Temperature correction hottest-spot rise . 332. . temperature rise. railway motors Starting of motors 202. on series generator on shunt generator pulsations. 344 15 Voltage commutation 240 between commutator segments 61.317. 209 208 separately excited generator 165 Thury system 185. specifications. equipotential Susceptibility 24 . Tubes of force 3. 267 212.

armature chord or fractional pitch closed-coil 258 263 314. 91. 101. 144. 94 simplex 104 two-layer wave. 376 of speed 305 92 Ward Leonard system control ring 89. 144. .INDEX control of speed limits of batteries 218 349. 264 field 75. . series Watt. 343 298 commutating poles 289 Deri 287 Thompson-Ryan Yoke.91. 79. 366. multiplex 263 94 98 9(: reactance regulation 247. ampere-turns for construction 129 Young and Dunn brush 65 292 . 293 161. 76. 91. number of circuits 360 conductors open-coil parallel pitch of W Wagner automobile generator lighting 89 92 97. 143. 293 89 52 89 drum 102 dummy 90 element equipotential connections 107 98 field displacement 100 step distributed.. selective commutation Windage Winding. . 262. 258. definition of series-parallel Wave windings.256 108 218 15 in . 320 88 104. 279. 351 lap 91. reentrancy resistance 94.241. . 263.. . 103 57. 101. effect of. 257.

.

LIBRARY UNIVERSITY OF CALIFORNIA IE on the last date stamped below^ LD 21-100m-l2.'46(A2012sl6)4l20 .

UNIVERSITY OF CALIFORNIA LIBRARY .