42 



LAYOUT AND THE CONTROL OF INTERFERENCE 



The interference to be discussed in this chapter is that due to the action of 

 stray fields on the input circuit of high-gain amplifiers. Spurious responses 

 generated by the apparatus itself, e.g. due to intermittent faults, inadequately 

 smoothed or stabilized power supplies, etc., are dealt with in Chapter 43. 

 Interference minimization will be considered with special reference to 

 electrophysiology, where the problem is particularly acute, but the principles 

 of interference control are quite general and applicable in other work : these 

 principles should be understood because they have to be applied whenever 

 a new experiment is set up. The apparatus designer can, by appropriate 

 design, do something towards producing gear which is insensitive to certain 



Figure 42.1 



Figure 42.2 



kinds of interference, e.g. by using differential amplification, but it is chiefly 

 with the user that the onus of obtaining unadulterated signals at the amplifier 

 input rests. 



Interference is borne in three ways : 



Electric fields — Most interfering electric fields arise in the manner shown 

 in Figure 42.1; the point P, driven by the interference generator Gj, alter- 

 nates in potential with respect to earth in the neighbourhood of the amplifier 

 input terminals. The equivalent circuit is shown in Figure 42.2. Q is the 

 coupling capacitance between P and the 'live' input lead, Qn the amplifier 

 input capacitance, R^ the input resistance, and Rq is the signal generator 

 resistance — in electrophysiology an electrode. 



Clearly the effect produced by P is minimal when C<. is small (which is 

 simply saying that P should be removed as far as possible) or when Qn is 

 large, or when Rq or R\^ are small. Since the efficiency of signal transfer of 

 the arrangement is poor unless R^^ is much greater than Rq, it follows that 

 in practice it is Rq rather than R^^ which will be the determining factor. 



It follows that experiments using silver wire or wick electrodes are easier 

 to carry out than those employing microelectrodes ; for the former we have 

 Rq of the order of ohms X 10^, whereas for the latter the figure is about a 

 thousand times greater. Furthermore, with the low resistance electrode, C\^ 

 may be allowed to be quite large, perhaps 100 pF, but the same value in 



656 



