EQUIPMENT FAULTS 



time in which to make the necessary checks. The correct approach here is 

 by a kind of graduated amputation. With a good supply of spare fuses 

 available, find the positive terminal of the HT smoothing capacitor and 

 unsolder the HT+ busbar supplying the whole of the rest of the apparatus. 

 This isolates the power pack. If the fuses again blow upon switching on, 

 suspect the smoothing capacitor or the rectifier input capacitor of internal 

 short-circuit: if they do not, the power pack is in order and the trouble lies 

 elsewhere. Switch off, remake the HT+ connection and pull out all the 

 valves (not the rectifier). Try switching on again. If the fuses blow again, 

 shake the apparatus vigorously to see if any pieces of metal are lodged 

 among the wiring. Suspect composition resistors comprising HT potential 

 dividers — perhaps one which normally constitutes most of the divider resis- 

 tance has gone 'low'. Any divider resistor which is discoloured, or has 

 cracked paint, has been subjected to overheating and may be the offender. 



If the dividers seem to be in order, the trouble is probably a valve taking 

 a greatly excessive current, most likely a power valve. Replace the valves 

 one by one until the culprit is found. With the offending valve pulled out 

 again, check the associated circuitry, looking particularly for excessive 

 screen potentials or insufficient bias. 



(2) Partial failures 



The method here is to check the operation of the unit stage by stage, start- 

 ing at one 'end' and working towards the other. With apparatus containing 

 pulse circuits it is best to monitor the waveforms with an oscilloscope, 

 beginning at the 'front' : thus if the defective piece of gear is a stimulator, 

 begin by checking the output of the relaxation oscillator, then verify that the 

 delay circuits are being triggered properly by the oscillator, then that the 

 shock generator is being triggered by the delay circuits, and so on towards 

 the output terminal until the defective stage is found. This done, the oscillo- 

 scope may be exchanged for a voltmeter, and the relevant valve potentials 

 checked. Potential checking is made much easier if one knows what the 

 reading ought to be, and to this end it is a valuable, if monotonous, practice 

 to measure and note down in a special log book the potential readings to 

 earth from all pins of all valves of all newly completed apparatus, together 

 with the ohms-per-volt of the meter used. In general an instrument of at least 

 10,000 O/V is advisable. Any dial settings which affect these potentials 

 should, of course, also be noted. 



If in the defective stage one of the voltages is clearly much other than it 

 should be, the faulty component — anode load, bias resistor, etc. — should be 

 deducible. If all seems to be correct, look for open-circuited coupling com- 

 ponents, e.g. capacitors or germanium diodes. 



If the faulty piece of apparatus is an amplifier it is often better to work in 

 the opposite direction, i.e. from the output back towards the input. This is 

 because an oscilloscope is already present in the form of the cathode ray 

 tube which is fed by most biological amplifiers. Using a sine-wave oscillator 

 provided with a calibrated attenuator, inject, via a large capacitor (say 4 /^F 

 paper), a signal direct on to one of the Y deflector plates of the tube. If the 

 trace produced is of steady amplitude and brightness, the tube supply cir- 

 cuits are in order and the tube can be calibrated in terms of mm/V. Armed 



667 



