ELECTRICAL WAVE FILTERS 



425 



ones described, and hence a very sharp selectivity can be obtained with 

 this circuit. It is possible to save elements in this filter by using two 

 primaries for each coil, putting one primary in one series or lattice arm 

 and the other in the corresponding series or lattice arms as shown by 

 Fig. \6D. Only half the number of elements per section are required. 



By replacing the series condenser of the series arm of Fig. \6A by a 

 parallel condenser, it is possible to change the filter fi-om a low-pass 

 to a high-pass filter. Condensers in series, or in parallel with both 

 arms result in wide band-pass filters. It is possible to obtain a wider 

 pass band with this type of filter than with the single coil type since the 

 resonances will be spread over a wider range of frequencies. 



In a good many cases it is desirable to have unbalanced filter sec- 

 tions rather than the balanced type which results from the use of a 

 lattice network. This is particularly true for high impedance circuits 

 for use with vacuum tubes. Since the lattice type section is the most 

 general type, it gives the most general characteristics obtainable. 

 The filter sections described here can in some cases be reduced to un- 

 balanced bridge T sections by well known network transformations, 

 with, however, more restrictions on the type of attenuation character- 

 istics physically obtainable. 



A very simple bridge T network, which is equivalent to a lattice 

 network of the kind shown on Fig. 13, with two crystals replaced by 

 condensers, is shown on Fig. 17. This section employs mutual induct- 



Fig. 17 — Single crystal bridge T band-pass filter. 



ance, and the resistance ^^ shown is necessary in order to balance the 

 arms of the equivalent lattice. This type of network is able to repro- 

 duce some of the characteristics of the lattice filter, but is not so general 

 and is, moreover, affected by the dissipation of the coil to a larger extent 

 than the equivalent lattice. 



'" The use of this resistance was suggested by Mr. S. Darlington and practically 

 all the work of developing this filter has been done by Mr. R. A. Sykes. 



