MAXIMALL Y-FLA T FILTERS IN WA VEGUIDE 709 



The theoretical treatment of maximally-flat filters presented here has ig- 

 nored the dissipation in the elements. Better agreement between expected 

 and observed characteristics would be obtained if this had been taken into 

 account. The observation of .3-db loss and 2-db loss in the four-cavity and 

 the fifteen-cavity maximally-flat filters is indicative of the amounts of added 

 insertion loss to be expected because of dissipation in the elements. In 

 addition to the increased loss at midband, we should expect a rounding of the 

 insertion loss characteristic near the cutoff frequencies, and a broadening of 

 the standing wave characteristic at frequencies well beyond cutoff. In 

 many applications, however, these effects can be ignored. 



Concluding Remarks 



In the foregoing, the design of maximally-flat band-pass filters has been 

 treated in detail. The treatment of other types of band-pass and band- 

 rejection filters is beyond the scope of the present paper, although much of 

 the material presented here may be of use in designing such filters. In 

 fact, almost any filter consisting of a ladder network of inductive and capaci- 

 tive elements in series and in shunt can be simulated in waveguides by fol- 

 lowing these principles. Emphasis on the maximally-flat filter has been 

 deUberate for two reasons: it gives a type of transmission characteristic that 

 is useful in microwave work; it is simple to design. 



Acknowledgement 



Many members of the Holmdel Radio Research Laboratories have influ- 

 enced the evolution of the technique of building waveguide filters. A firm 

 foundation was laid by the pioneering work of G. C. South worth^^ and A.G. 

 Fox. Intimate association with W. D. Lewis and L. C. Tillotson fostered 

 many stimulating discussions which clarified doubtful issues. The com- 

 ments of M. D. Brill and S. Darhngton are deeply appreciated. The un- 

 tiring effort of R. H. Brandt in making skillfully the accurate measurements 

 which were necessary is gratefully acknowledged as is also the invaluable 

 support of all the many other people without whose cooperation and con- 

 fidence the work would have been impossible. 



References 



1. E. L. Norton, Constant Resistance Networks with Applications to Filter Groups, 



B. S. T. J., Vol. XVI, pp. 178-193, April 1937. 



2. W. D. Lewis and L. C. Tillotson, B. S. T. /., Vol. XXVII, #1, pp. 83-95, January 



1948. 



3. E. L. Norton, U. S. Patent #1,788,538, January 13, 1931. 



4. W. R. Bennett, U. S. Patent #1,849,656, March 15, 1932. 



5. S. Butterworth, Wireless Engineer, Vol. VII, #85, pp. 536-541, October 1930. 



6. V. D. Landon, R.C.A. Review, Vol. V, pp. 347-362 and pp. 481-497, 1940-41. 



7. Henry Wallman, M.I.T. Radiation Laboratory Report #524, February 23, 1944. 



8. S. Darlington, Journal of Math, and Pliys., Vol. XVIII, pp. 257-353, September 1939. 



