RESISTANCES 



constant-current type, control may be had by a parallel variable resistance 

 to 'bleed' some of the current {Figure 2.23). (Notice that no control is 

 achieved by attempting to bleed a constant-voltage generator, or by con- 

 necting a variable dropping resistance in series with a constant-current 

 generator.) If the generator is approximately matched to the load either 

 method may be used, but the variable series resistance is to be preferred 

 since it reduces the power to the load by reducing the power supplied by 

 the generator, whereas a variable bleeder works the other way round, and is 

 therefore rather uneconomical. 



A difficulty with the variable series resistance is that, to reduce progressively 

 the power supplied to the load to zero, the resistance must increase smoothly 

 to infinity, which is usually rather difficult to arrange. The problem can be 

 overcome by arranging the resistance as a potentiometer (Figure 2.24) when 

 obviously the load power is zero when the slider is at the bottom. When 

 the sUder is at the top some power is delivered to the load and some is 

 dissipated in the potentiometer. The latter fraction is small if the potentiome ter 

 resistance is high. 



G ~^^ I L 



Figure 2.24 



Figure 2.25 



We saw earlier that if no load is connected to a potentiometer the output 

 voltage is proportional to i?2- I^ the potentiometer is of the 'linear' variety, 

 i?2 is proportional to shaft rotation, so 



output voltage oc shaft rotation 



When a load is connected this is no longer true. In Figure 2.25, if we dis- 

 connect the load and apply Thevenin's theorem to the output terminals of 

 the potentiometer we see an open-circuit e.m.f., E', of 



R9 



in series with a resistance 



Ri + R^ + r 



,_ R,(R, + r) 

 R^ + R^ + r 



The equivalent circuit is therefore Figure 2.26, and on reconnecting the 

 load — of resistance 7?^ — we find the output current is 



i?l + i?2 + A- 



14 



