SOME CIRCllT ASl'ECrS ()!■ THE TRAXSISTOH -VI 



the voltages on grid and plate are usually taken as independent variables; 

 the grid and j^late currents are taken as functions of the voltages. It be- 

 comes natural, then, to measure tubes with regulated power supplies having 

 low impedances to keep the voltages constant, and one is then naturally led 

 to describe tubes in terms of admittances. Now the trouble with this scheme 

 for transistors is that many of them oscillate when connected to low im- 

 pedances, that is, many transistors are short-circuit unstable. To avoid this 

 dilBculty it is convenient to measure with high impedances in the leads; 

 the analytical counterpart is to regard the currents as independent variables, 

 leading naturally to a description of the transistor in terms of impedances, 

 as shown in the figure. 



This description by open-circuit impedances happens to be a good one 

 for many purposes, but there is nothing final or unique about it. In fact at 

 high frequencies one of the other descriptions becomes more convenient. 



By interpreting the '^ equations as circuit equations, one is led directly 

 to the first equivalent circuit of Fig. 4. A little consideration shows why the 

 •g's are called open-circuit impedances. For example, if the second mesh is 

 open-circuited, then the equation say that %i is the ratio of input voltage 

 to input current, that is, the input open-circuit impedance; while '^21 is the 

 ratio of output voltage to input current, that is, the open-circuit forward 

 transimpedance. Sunilarly '^12 is the open-circuit feedback transimpedance 

 and "^22 is the open-circuit output impedance. Most of the subsequent dis- 

 cussion is concerned with low frequencies, where the unpedances reduce to 

 resistances. 



This equivalent circuit for small signals is only one of many possibilities. 

 Another, which is in fact more frequently used, is shown on Fig. 4. It con- 

 sists of a T of resistors, each of which is associated with one of the transistor 

 leads, and a voltage generator in series with the collector lead whose ratio 

 to the emitter current is also of the dimensions of a resistance. The elements 

 of this equivalent circuit are related to the former one by a simple sub- 

 traction. The other equivalent circuit on Fig. 4 is obtained by converting 

 the series voltage generator to the equivalent shunt current generator, 

 whose ratio to the emitter current is now a dimensionless constant which 

 we shall call a . 



These circuits, as well as all the other numerous possibilities, are equiva- 

 lent in the sense that they all give exactly the same performance for any 

 external connection of the unit. These three, however, are particularly well- 

 behaved in that usually none of the circuit elements is negative; they are 

 readily accessible to measurement; the association of the various circuit 

 elements with corresponding regions withiii the transistor appears to have 

 some physical significance; and, finally, the parameters are not too dread- 

 fully dependent on the exact operating point used. 



