308 A High-Frequency Alternator. 



Parallel bearings working in two loose sleeves were tried, 

 and also combinations of conical and parallel bearing surfaces; 

 but the simple conical bearings shown in PI. V. fig. 1 were 

 found to give the best results, probably owing to their shorter 

 length allowing the inductor more freedom to spin on its 

 axis of gyration rather than forcing it to spin on its mecha- 

 nical axis. An ideal method would have been to allow the 

 inductor to spin like a top and choose its own axis of 

 rotation ; but this could not be done with the actual alternator, 

 as the clearance between the inductor and the pole-tips was 

 less than 0*1 mm. 



Experiments were next started to see whether the efficiency 

 of the alternator could not be increased so as to allow the 

 small amount of power required for the experiments to be 

 obtained with an inductor having more than 90 teeth. An 

 inductor having 204 teeth was therefore constructed. 



By placing search- coils on different parts of the field-ring 

 it was found that the E.M.F/s induced were much larger in 

 those coils placed near the poles than in those placed on any 

 other parts of the ring, from which it seemed that the changes 

 in the number of lines of force produced by the movement of 

 the inductor were confined to the neighbourhood of the pole- 

 tips, probably due to the magnetic leakage and to the eddy- 

 currents set up in the laminations. Coils were therefore 

 wound right on the pole-tips themselves. Using the coil on 

 one pole-tip only, which consisted of 10 turns of No. 30 

 double cotton-covered wire, it was found that the output of 

 the alternator was 10 times as great as could be obtained 

 with the original armature-winding on the ring itself. 



With this inductor and the coils on the pole-tips as shown in 

 PL V. fig. 1, it was finally possible to obtain a H.M.S. current 

 of 0*1 ampere at frequencies up to 120,000 ~ per second. The 

 maximum P.D. at 100,000 ~ per second with 0'1 ampere 

 flowing was 2 volts, and on open circuit 3*6 volts. As this out- 

 put, though small, was ample for the purposes of the research, 

 no attempt was made to further increase it. The wave-form 

 of the current, when this inductor was driven so slowly that 

 the frequency was only 2670 ~ per second, is given in PL VI. 



An illustration will perhaps convey some idea of how high 

 a frequency of 120,000 ~ per second really is. In plotting 

 curves for ordinary frequencies of 50 to lOO^per second, a 

 scale often adopted is 10 inches for 100~. If it were 

 attempted to plot a curve up to 120,000-^per second to this 

 scale the curve paper would require to be 12,000 inches, or 

 nearly one-fifth of a mile long. 



At the time of constructing the present alternator the 



