DESIGN FACTORS OF THE 1553 TRIODE 517' 



2. Even rough calculations, such as the coaxial line approximations 

 used in Appendix 2 are close enough to the facts to indicate whether 

 the design is close to an optimum with respect to such parameters 

 as output spacing, anode diameter, grid diameter, and the like.' 

 More accurate calculations and experiments can give more precise 

 answers to these questions. 



3. Some considerations such as cathode activity, tube life, heater power 

 and so on have not yet been included in the analysis. However, 

 systematic optimization for such parameters as are treated quanti- 

 tatively is greatly facilitated. In general, each different figure of 

 merit leads to a somewhat different optimum and hence a different 

 version of the tube. 



The design of tubes by the method of figure of merit has been outlined. 

 The method is very general, but in essence has just three steps: 



1. Formulate the system performance of the projected device with the 

 aid of a figure of merit. 



2. Find how the figure of merit depends upon the parameters of the 

 tube, such as spacings, current, etc. 



3. Adjust the tube parameters, subject to physical limitations, to op- 

 timize the figure of merit. 



Acknowledgments 



The development of this microwave triode has required not only the ex- 

 pert and highly cooperative services of a large team of electrical, mechanical, 

 and chemical engineers but also the indispensable assistance of skilled tech- 

 nicians, all of whom worked smoothly together to develop these new ma- 

 terials and techniques to a point where they are specifiable and amenable 

 to quantity production. It is not practical to mention all those who have made 

 significant contributions to this development. The contributions of A. J. 

 Chick, R. L. Vance, H. E. Kern and L. J. Speck, however, are of such out- 

 standing nature that mention of them cannot be omitted. 



APPENDIX 1 

 Derivation of the Figures of Merit 



Gain-Band Figure of Merit 



Let the problem be stated as the design of an amplifier tube to operate 

 with as large gain over as wide a frequency band as practicable. As a 

 standard environment, we use a single-stage amplifier working between 

 equal resistive impedances. For three reasons this standard is suitable: it 

 is simple; it corresponds closely to practicality in many cases especially 



