464 ELECTRICAL MEASUREMENTS 



The commutator is very likely to be troublesome, especially if 

 any sparking occurs due to the presence of dust or oil. 



The fact that a meter has a high torque is advantageous only 

 when it is associated with a light moving element. The ratio 



Full-load torque 

 Full-load speed X friction 



should be large if the accuracy of the meter is to be only 

 slightly affected by the wear of the jewel, pivot and commutator. 



Of late years the weight of the movable element has been 

 much decreased by the use of a spherical self-supporting armature 

 and an aluminum brake disc. 



The effect of changes of room temperature on the accuracy 

 of the meter should be reduced to a minimum. It is evident 

 that the net effect of temperature on the resistance of the disc 

 and on the drag magnets should be balanced by the change in 

 the resistance of the armature circuit. The average temperature 

 coefficient of the meter between 20 and 40 C. should not be 

 more than 0.2 per cent, per degree at either 10 per cent or 100 

 per cent of full-load current. Tests show 8 that the temperature 

 coefficients of representative direct-current watt-hour meters of 

 American manufacture vary between +0.26 and -fO.07 per 

 cent per degree C., most of them being about -fO.l per cent. 



As the armature circuit carries considerable current and is 

 in part made of copper, its resistance (R in formula 1) will 

 rise and decrease the driving torque when the meter is first 

 connected in circuit; consequently, tests should not be made 

 until the permanent state of temperature has been reached, 

 which may require about 20 minutes. 



When a load is thrown on the meter, the heat liberated in 

 the current coils also raises the temperature of the copper- 

 wound armature and increases the resistance of the potential 

 circuit, thus decreasing the registration. 8 This effect increases 

 with the load current and, up to the time of attainment of tem- 

 perature equilibrium, with the length of time the current is left on. 



From equation (2) it will be seen that the effect of 'he self- 

 heating of the meter on the magnetic brake is to decrease the re- 

 tarding torque and cause the meter to speed up. However, as 



