COLD CATHODE DIODE 



resistance, which is 70 V/1 mA = 70 kQ. Referring to Figure 7.4, it is clear 

 that the goodness of the stabihzation depends on the attenuation of incre- 

 ments in V^, that is, on R being much larger than 200 ohms. 



There is, however, a catch here. Suppose we choose a tube current of 

 1 mA as lying nicely in the centre of the makers' recommended range, and 

 suppose Fj = 100 V. A 250-fold reduction in fluctuations of K,^ sounds 

 attractive, so let us aim for this. Then R must be 50,000 ohms and the 



Figure 7.4 



voltage drop across it will be 50,000 ohms X 1 mA = 50 V, so K„ = 150 V. 

 Now let us try to do four times better. For a 1,000-fold reduction we want 

 R to be 200,000 ohms, hence the drop is 200 V and so F„ = 300 V. Suppose 

 now K„ undergoes a 10 per cent upward change in value. In the first case 

 it alters from 150 to 165 V, so the increment is 15 V and the proportion that 

 is added to V^ is 1/250 x 15 = 0-06 V. In the second case the change in 

 F„ is 30 V and the proportion of the change added to K,, is 1/1,000 X 30 = 

 0-03 V, i.e. only twice as good. If we aim for a 4,000-fold reduction we get 

 R = 200 ohms X 4,000 = 800,000 ohms, and the drop across it would be 

 800 V and F„ would have to be 900 V. A 10 per cent increase would then 

 be a change of 90 V, and the fraction appearing at the output is 1/4,000 X 

 90 = 0-0225 V, an improvement hardly commensurate with requiring 900 V 

 to produce 100. 



In point of fact there is little to be gained by making F„ more than two 

 or three times F„ and the proper thing to do is to get F„ from some con- 

 venient point in the circuit and then to decide on the glow current, rather 

 than the other way round. The performance of the circuit is improved if 

 R is increased, and R can be increased until the tube current approaches the 

 lower limit advised by the makers, below which the tube performance will 

 become unsatisfactory. It is a good general rule in any case with soft valves 

 to have low currents if possible, because the hfe falls off inversely as a high 

 power of the current — about the fifth. 



Voltage stabilizer tubes 



As with the voltage reference tube, one or more of these are arranged in 

 series to produce the required steady output voltage K, from a supply voltage 

 K„ subject to fluctuation, but in this case it is intended that they be used to 

 stabilize the voltage supplied to some sort of load, represented by Rj^ in 

 Figure 7.5. If the load resistance is also liable to fluctuations, such that the 

 load current varies by an amount 61 j^, and if when the load is maximal a 

 'keep alive' current I^min flows through the tubes, then clearly when the load 

 is minimal the tube current must rise to /,min + II- It emerges that voltage 

 stabilizer tubes may have to pass appreciable currents, typically 40-100 



121 



