BIOLOGICAL AMPLIFIERS 



If no special attention is paid to the input cathode-follower grid current, 

 this figure may well be 10~^ amps. The eifect of introducing a current of 

 this magnitude into the hypothetical cell is to produce a spurious potential 

 difference across the membrane of (8 X 10') X 10~^ V, or 80 mV. Clearly 

 the circumstances of recording are not very physiological. 



One often hears or reads statements such as 'the cathode follower valve 

 was selected for low grid current'. Reference to Figure 39.8 — which shows 

 the form of the grid-current characteristic for a thermionic valve — makes it 

 clear that any valve may be made to pass zero grid current by choice of 



Into 

 valve 



I Out of 

 I valve 



HT+ 



Input 



Figure 39.8 



HT- 

 Figure 39.9 



appropriate bias (point A), but that at this point the input resistance, 

 equal {dVgj^)l{dIg) is rather low. A more usual working point in capacitor- 

 coupled circuits, where the grid current is diverted down the grid leak, 

 would be point B, where the input resistance of the valve is high. 



Experience shows that if a suitable cathode follower valve is operated 

 under 'low-noise' conditions, reduced heater voltage and low anode current, 

 the order of input resistance corresponding to point A working is neverthe- 

 less sufficiently high to be large compared with a microelectrode of resistance 

 50 megohms or so. It appears that for this particular application point A 

 rather than the more usual point B working is satisfactory. 



Consider the direct-coupled triode cathode follower circuit in Figure 39.9. 

 The current through the valve is determined by the ratio of the negative 

 supply voltage to i?^, since the cathode and grid are at approximately earth 

 potential. The grid-cathode bias depends on the positive anode supply, 

 since the anode-cathode and grid-cathode voltages are related by the rather 

 constant parameter [i. It emerges, then, that the grid current can be set to 

 zero by appropriate choice of HT+, 



This makes it a good deal easier to choose a suitable specimen for this 

 important position. Instead of rejecting valves out of hand because their 

 zero grid current point does not occur at the particular value of HT+ 

 (arbitrarily chosen) which is in use, one can concentrate on selecting for low 

 noise and microphonicity. If a number are found which are satisfactory in 

 these respects, efforts can be directed towards choosing one whose grid 

 current characteristic cuts the /^ = axis at the minimum slope; this speci- 

 men will then hold its low grid current setting best. 



624 



