HEYLAND ALTERNATOR 267 



It is evident that we may replace the actual commutator of a 

 multipolar field of 2 P poles by a commutator consisting of eighteen 

 icnts only (which would correspond to a two-pole construction), 

 but running at P times the speed of the field-magnet, and having its 

 brushes spaced 120 apart. Such an equivalent two-pole commutator 

 (running at a correspondingly higher speed) is shown in Fig. 161, 

 and on account of its greater simplicity it is more convenient to 

 consider than the actual multipolar commutator. 



The entire field winding is divided into six groups of coils. The 

 field coils may be wound in the ordinary way if the number of pole- 

 pairs is a multiple of six ; otherwise, each pole would be provided 

 with two or more windings the coil being wound with two or more 

 wires laid side by side (in a two-pole magnet, six wires would be 

 wound side by side) so as, in each case, to subdivide the entire field 

 winding into six groups of equal resistance. These six groups are 

 then connected to the twelve active sets of segments as shown in 

 Fig. 161, and across the ends of the groups are connected non-induc- 

 tive resistances as shown. 



Let us suppose that the brushes are in connection with the 

 secondary of a transformer whose primary is across the generator 

 terminals. By shifting the brushes the commutator may be arranged 

 to occupy any desired position relatively to a brash at the instant 

 when the current through that brush is at its maximum value. Let 

 the brushes be so adjusted that when any brush is receiving maximum 

 current it rests in the position of symmetry on the active segments 

 being in contact with the four middle active segments, as shown in 

 Fig. 161. The field winding is then in the best position for receiving 

 current, while most of the non-inductive resistances are not traversed 

 by currents at all. In order to make this point quite clear, we may 

 imagine the brushes displaced through 90 (\ period), so that the 

 relative position of brushes and commutator at the particular instant 

 considered is as shown in Fig. 162. It is at once evident that now 

 the current flows mainly through the non-inductive resistances. 



Assuming the brushes to have been placed in the most favourable 

 position, corresponding to Fig. 161, as the current through brush I is 

 decreasing, the active segments gradually recede from brush I and 

 approach brush II. After one-third of a revolution, which corre- 

 sponds to ^ period, the active segments will be under cover of 

 brush II, which is now receiving maximum current and delivering 

 it to the field winding in the same direction as was done by brush I. 

 After another one-third of a revolution, the active segments will come 

 under cover of brush III, which in its turn is now receiving maximum 

 current, and so on. Thus, although the current through the field 

 winding will vary slightly (the period of its fluctuation being clearly 

 one-third of that of the generator period), the variations are kept 



