532 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1954 



straight-forward at least to the design objectives of No. 1 and No. 2 of 

 Table I. 



The best unit measured to date showed ao > 0.96, fa ^^^ 25 mcps, 

 Tb c^ 60 ohms, and Cc ^^^ 1-8 mmf. These values agree quite well with 

 those expected from the resistivities and layer thicknesses employed. The 

 unit oscillated at 95 mcps with Vc = —30, Ig = 1.0 ma. Connected in a 

 common emitter \ddeo amplifier working from a 75-ohm generator im- 

 pedance into a load resistance of 2,150 ohms shunted by 5 mmf of ca- 

 pacitance, this unit produced a power gain of 23 db at 500 kc, falling to 

 20 db at 3 mcps and 15 db at 10 mcps.* In an uncompensated common 

 emitter tuned circuit, this unit gave 20.5 db at 10 mcps with 3 mcps band- 

 width between the three db points.* It has been operated with a collec- 

 tor voltage of —90 volts. 



SUMMARY 



The designed elimination of donors and acceptors from a thick col- 

 lector depletion layer introduces a new design variable in junction tran- 

 sistor triodes. The new structure (p-n-i-p or n-p-i-n) is believed capable 

 of development into the microwave frequency range. Several factors 

 which were of second order importance in p-n-p and n-p-n units such as 

 emitter depletion layer capacitance and collector transit times become 

 significant in limiting ultimate performance. The thick depletion layer 

 permits operation at higher voltages than were previously possible in 

 any but low frec^uency units. 



Moderately good results have been obtained already. Units having 10 

 mil emitter diameter, 15 mil collector diameter have produced stable 

 gains without compensation of 20.5 db at 10 mcps and have oscillated 

 at 95 mcps. 



The junction transistor now promises to be a serious competitor to 

 high vacuum triodes over a much larger range of frequencies and power 

 levels than before. 



ACKNOWLEDGMENTS 



J. A. Morton and R. M. Rj^der have strongly supported and en- 

 couraged this work. J. W. Peterson and W. C. Hittinger have collaborated 

 in and contributed to the experimental studies. The models constructed 

 and tested are the products of the persistent efforts and many useful 

 suggestions of J. A. Wenger, J. McGlasson, and L. P. Meola. Many 

 others, particularly those engaged in semiconductor materials research 



* These measurements were made by L. G. Schimpf. 



