ELECTRIC WAVE FILTERS 279 



As mentioned above, the purpose of the filter shielding was to 

 reduce the cross- talk between filters mounted side by side; the effect 

 of the shield upon the coil inductances was an unwelcome incidental. 

 When the coil adjustment was considered, however, the shielding 

 effect of the copper can was of no interest, but its effect on inductances 

 was of paramount importance. Efforts were made to devise a dummy 

 shield which would simulate the standard filter shield in its effect 

 upon the inductances of the coils mounted inside but in which the 

 coils would be readily accessible for adjustment. As it was expected 

 that the adjustable coil would enable resonant frequencies to be 

 located at their desired values within limits of ±0.1 per cent, it was 

 necessary that any dummy shield developed for use in the adjusting 

 process should reproduce on the coils the effect of the filter shield to 

 at least this figure. In the ideal case, it would have been desirable 

 that the difference between the dummy or adjusting shield and the 

 standard shield be no greater than the difference between two standard 

 shields. 



Several different designs of adjusting shields were tried in the 

 attempt to realize this ideal. The model finally adopted was con- 

 structed in the following manner: A box of the same height and width 

 but slightly shorter than the standard shield was built, open at one 

 end. Laboratory tests showed that for coils placed inside the shield 

 near the closed end, the reduction in inductance caused by the shield 

 was almost entirely unaffected by conditions at the other end of the 

 shield. The far end could be closed or left open with an effect which 

 was negligible in comparison with the total inductance change caused 

 by the shield. This condition, of course, held only for coils in the 

 closed end or at a distance of at least three coil diameters from the 

 open end of the shield. Since the maximum number of solenoidal 

 coils mounted on a panel in the type " C " system filters was four, it was 

 necessary only that the adjusting shield simulate the standard for 

 two coil positions, as a filter panel could be turned end for end and the 

 other two coils inserted in the shield after the first two coils were 

 adjusted. The method of adjustment employed was the same for all 

 the preliminary models as for the final design. Small holes were 

 drilled in the top of the shield corresponding to the standard coil 

 locations in the filters, and through these holes the coils were adjusted 

 by means of hard rubber rods with small buttons at their ends. For 

 each coil position a pair of these rods was provided, so arranged that 

 one could be rested on each of the long sides of the coil cradles. Then, 

 by pushing on one rod or the other, the operator could move the 

 cradle to the desired position. 



