45 

 TRANSISTORS 



In leaving transistors to the end of this book there is no suggestion that 

 the subject is only of marginal interest to the electrobiologist ; rather, this 

 chapter is in the nature of a 'stop press' item. Following the announcement 

 of the invention of the transistor from the Bell Telephone Laboratories^ of 

 America in 1948 there elapsed an interval of some 5 years for further 

 improvements and the development of the necessary manufacturing 

 techniques, after which transistors with predictable characteristics began 

 to appear upon the market. Once in the hands of the circuit engineers 

 there followed a further period during which the usefulness of the transistor 

 as an active circuit element — and hence as an alternative to the thermionic 

 valve — was investigated. In recent years the results of these endeavours 

 have led to the publication of a veritable spate of transistorized devices. 

 So rapid is the pace of development in this field that it is difficult to write 

 about the subject without one's efforts becoming almost at once out of date. 



It would be unwise to predict that the transistor will completely oust 

 the valve in electrophysiological or allied work; indeed, the transistor is 

 at present at a disadvantage compared with the valve in two particulars 

 of great interest to the electrophysiologist. These are: it is difficult to make 

 a transistor amplifier having a high input resistance, and it is also difficult 

 to make a direct-coupled transistor amplifier which is not seriously prone 

 to drift; but as a.c. coupled amplifiers to work from low-resistance prepara- 

 tions, certain pulse generators, high voltage power supplies and voltage 

 stabilizers may all be successfully made with transistors, the subject is 

 clearly deserving of his attention. Advantages of the transistor are small 

 size, absence of microphonicity, excellent power efficiency (no heaters to 

 supply) and ruggedness (transistors can withstand accelerations which 

 would destroy a valve). The question of transistor noise is at present 

 receiving much research effort; it seems probable that future types will 

 surpass valves in this respect but generally speaking those available at 

 present are not as good. The performance at radio frequencies in contem- 

 porary transistors is also unsatisfactory, but electrobiology is mainly 

 interested in the frequency band 0-20,000 cycles, so this aspect is of minor 

 importance. At the time of writing transistors are between two and three 

 times as expensive as valves. 



Transistors are of two types, 'point-contact' and 'junction'. The point- 

 contact type was announced first, and has certain properties not possessed 

 by the junction transistor which are important in some pulse circuits. 

 Nevertheless, it appears — at least in the field in which we are concerned — 

 that the junction transistor is emerging as by far the more important, and 

 the point-contact type will not therefore be discussed. 



In Part II, the use as diodes of junctions between P and N type semi- 

 conductors — germanium or silicon — was mentioned. The triode junction 



677 



