426 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1952 



than unity the emitter to ground voltage will begin to fall because of the 

 larger collector current increments driving the voltage of the node .¥ 

 negative more rapidly than the emitter current drop through Ve would 

 normally carry it. This transition point is called the peak point. If 

 then a{ri) + /?&) is sufficiently large, in this sense, the input resistance 

 may be negative in this region II. When the internal node voltage has 

 fallen to a value near that of the collector terminal the "valley point" 

 has been reached. At this point, the emitted hole current has reduced the 

 collector impedance to a minimum value beyond which a is essentially 

 zero; the transistor is said to be saturated. From this point on the in- 

 put impedance again becomes positive and is determined almost entirely 

 by the base and emitter impedances. By terminating the emitter- 

 ground terminals in various ways with resistor-capacitor-bias com- 

 binations, such a network can be made to perform monostable, astable 

 or bistable functions. Under such conditions, the emitter current and 

 correspondingly the collector current switch back and forth between 

 cutoff and saturation values. For example, in Fig. 16 is shown a value of 

 emitter bias and load resistance such that there are three possible 

 eciuilibrium values of emitter current and voltage. It may be shown that 

 the two intersections in regions I and III are stable whereas that in region 

 II is unstable. Hence, if the stable equilibrium is originally in I, a small 

 positive pulse Ap applied to the emitter will be enough to switch from 

 stable point I to stable point II and conversely, — A^. will carry it from 

 the high current point to the low current point. The circuit designer is 

 interested in reproducing in a given circuit (with different transistors 

 of the same type) the follo^vdng points of the characteristic: 



a — The off impedance of the emitter — he desires that this be greater 

 than a certain minimum. 



b — The peak point Vep — he desires that this be smaller than a certain 

 maximum. 



c — The value of the negative resistance — he desires that tliis be greater 

 than a certain minimum. 



d — The valley point Ves , Ls — he desires that these be greater than 

 certain minima, and 



e — The slope in region III — he desires that this be smaller than a 

 certain maximum so that he ma}^ control it by external means. 



It may be shown that these conditions can be satisfied for useful 

 circuits by specifying certain maximum and minimum boundaries on the 

 static characteristics. Fig. 17 is an idealized set of input or emitter 

 characteristics. By specifying a minimum value for the reverse resistance 



