DESIGN PROCEDURE 



From the specification, a provisional block diagram of the proposed 

 apparatus follows at once {Figure 44.1). The only point to notice at this stage 

 is that if time marks are required for direct visual inspection, it is essential 

 that they appear similarly in successive traces ; that is, the train of pips must 

 not march to right or left, otherwise setting up the trace length becomes 

 extremely difficult. Thus, if the time marks are to be derived from the mains, 

 the master oscillator must be synchronized with the mains also. 



We now consider the contents of each block in more concrete terms. 

 Initially we propose well-established and trustworthy circuitry, perhaps as 

 follows : 



Master oscillator 

 Time base 



Delay circuits 

 Pulse generator 



Main amplifier 



Pre-amplifier 



thyratron relaxation oscillator 



sanatron (2 pentodes) and paraphase amplifying 

 valve to derive push-pull output, say another 

 pentode 



double triode flip-flops 



If the shape of the output pulse is not critical, a 

 very simple possibility here which gives a 

 stimulus moderately free from earth is an elec- 

 tronic version of the traditional induction coil; 

 i.e. a power valve is used to interrupt the current 

 in the primary of a loose-coupled mutual induc- 

 tance, to the secondary of which the preparation 

 is connected. So for two channels 



It would be unwise not to use double-sided ampli- 

 fication here. The cathode ray tube which it is 

 proposed to use has a deflection sensitivity of 

 IjlOO/VAsmm/V, so with a final anode poten- 

 tial of 2,000 V, the deflecting voltage required to 

 sweep the full 6 in. is about 300. The gain 

 required is therefore 6,000. A good pentode can 

 give a gain of 150, but not under low-noise con- 

 ditions, when a figure of 30 is nearer the mark. 

 In addition, since the amplifier is direct-coupled, 

 gain may be lost in the interstage couplings. 

 6,000 is an awkward amount. It is just too much 

 for 2 stages, but rather too easy for 3. It will 

 be possible to use considerable negative feed- 

 back. A reasonable division of labour would be 

 to secure (including couplings) a gain of 1 5 from 

 the 1st, 20 from the 2nd and 30 from the 3rd. 



Total 



This too will be double-sided. A gain of 100 is 

 beyond the capability of one low-noise stage, 

 but may easily be achieved by two. 



Total 



1 valve 



3 valves 

 2 valves 



2 valves 



6 valves 



4 valves 



Grand Total 18 valves 

 + rectifiers if used 



At this point it is possible to produce a first estimate of material cost. As 

 a rough guide, which is about right at the time of writing, the author allows 

 £2 per valve, with its associated circuitry, + £5 for sheet metalwork + £5 

 for each power pack + cost of cathode ray tube (£20). At present, then, it 

 looks as if the unit will cost about £70. 



The next stage is to see if there are any catches in the scheme as so far 

 proposed, and if so to modify it as necessary. For example, the thyratron 



672 



