THE CALIBRATION OF INSTRUMENTS :,'.'.> 



Millivoltmeter Leads. External shunt ammeters must be 

 calibrated with the same set of leads connecting the shunt and 

 the millivoltmeter that is to be used in the subsequent test. 

 Frequently special leads 30 or 40 feet long must be used in order 

 to remove the millivoltmeter from stray fields, or to allow it 

 to be placed where it can be easily read. Such leads should be 

 of large diameter to reduce the resistance, and should be provided 

 with proper terminals. 



In all cases when connecting the millivoltmeter and the shunt, 

 care must be taken that all contacts are clean and firmly set up. 

 The leads should be carefully examined to see that they are not 

 broken inside the insulation or where they are soldered to the 

 terminals. 



Thermo -electromotive Forces. The material of the resistance 

 strips used in shunts should have a small thermo-electromotive 

 force when opposed to copper. Manganin is the best. The 

 existence of a thermo-electromotive force may be demonstrated 

 by sending full current through the shunted instrument for a 

 considerable time and then breaking the main circuit. The 

 millivoltmeter will not return at once to zero. This effect may 

 be differentiated from the zero-set by breaking the millivolt- 

 meter circuit. Switch-board shunts are likely to be defective 

 in this respect, and should not be used in very careful work until 

 they have been investigated. 



Effect of External Temperatures. Variations of room tem- 

 perature produce only small errors in soft-iron ammeters with a 

 spring control, for as the spring weakens, the permeability of the 

 iron decreases in such an amount as practically to compensate for 

 it. In the voltmeter there is an additional source of error in the 

 alteration of the resistance. The windings will be of copper and 

 the series resistance of a material with a zero temperature coef- 

 ficient. The net effect will depend on their relative magnitude; 

 it will be small in high-range instruments. 



The only effect on current dynamometers, with the coils in 

 series, is to alter the spring, 0.03 or 0.04 per cent per degree C., 

 causing the instrument to read too high. The effect on dyna- 

 mometer voltmeters is on the resistance as well as on the spring. 



With a rise of temperature, the magnets of a moving-coil 

 voltmeter decrease in strength, the springs weaken, and the 



