272 BELL SYSTEM TECHNICAL JOURNAL 



and condensers were the precision of the bridge standards and the 

 bridge operating technique in the shop it seemed unlikely that more 

 precise location of the filter characteristics could be obtained until 

 these factors were improved. However, another way was found. It 

 has been pointed out in preceding paragraphs that critical points, 

 such as cut-off points and attenuation "peaks" are usually dependent 

 upon the resonant frequency of a pair of elements, or perhaps of several 

 pairs. It is much more important that these critical frequencies be 

 correctly located; that is, that the L-C product of the elements be of 

 the correct value, than that either element alone be of the correct 

 value. Therefore, if some way could be found of adjusting a combina- 

 tion of elements associated together in a filter so that the L~C product 

 would be of the correct value, a considerable improvement in the 

 precision of the attenuation characteristic location might be expected. 

 If the elements could be adjusted so that a negative deviation in 

 condenser capacity, for example, could be compensated for by a 

 positive deviation in coil inductance, the L-C product, and therefore 

 the resonant frequency, would be correct, and the effect of the indi- 

 vidual element deviation would be greatly minimized, though not 

 entirely eliminated. Either an adjustable coil or an adjustable con- 

 denser would make this procedure possible; and, as such a resonant 

 frequency adjustment could be made to the precision of an oscillator 

 calibration, or about 0.05 per cent, if necessary, the errors of in- 

 ductance and capacity bridge standards, as well as initial adjustment 

 error and variation due to potting would be eliminated at the resonant 

 frequency, and their effects minimized greatly at other frequencies. 



Development was therefore started on an adjustable type of in- 

 ductance. As the simplest and most inexpensive coil was, of course, 

 the most desirable, efforts were made to modify the solenoidal air 

 core coil used in previous carrier filters. One scheme considered was 

 to mount a small copper vane in such a way that it could be moved 

 into or out of the field of the coil; the eddy-currents in the vane set 

 up a small counter field that neutralized a part of the field of the main 

 coil and thus reduced its inductance. Another method made use of a 

 small "pancake" wound coil mounted vertically on the side of the 

 main coil so that it could be moved up or down, into or out of the field 

 of the main coil. This small coil was short-circuited, and therefore 

 produced an effect similar to that of a copper vane. The model 

 finally adopted, however, although using the same principle of adjust- 

 ment, was made in a somewhat different form. A rectangular rotor 

 was constructed to fit around the main coil. This rotor w^as pivoted 

 to the main coil on its short sides and could be rotated through an 



