THE CALIBRATION OF INSTRUMENTS 603 



source of error in alternating-current work. These effects may 

 be pronounced in wattmeters when working at low power fac- 

 tors; of course they are absent when direct currents are used. 



Current Distribution. Distribution errors may be met with 

 in alternating-current instruments with massive coils, the cur- 

 rent not distributing itself uniformly over the cross-section of the 

 conductor as it does when direct current is used. This also 

 may cause the alternating- and direct-current calibrations to 

 differ. 



Frequency and Wave Form. There is a possibility of error 

 in dynamometer voltmeters and wattmeters and in soft-iron 

 voltmeters, due to frequency, if the reactance of the instrument 

 becomes unduly high in proportion to the resistance. Espe- 

 cially should one be on his guard in investigation work where 

 abnormal frequencies are sometimes employed. An instrument 

 which is commercially correct at 60 cycles may be much in error 

 at 500 cycles. Eddy-current effects are much accentuated at 

 high frequencies. 



Soft-iron instruments may be subject to wave-form errors 

 arising from saturation effects, but with good modern instruments 

 no trouble is likely to be experienced. 



Induction instruments have errors all their own, due to the 

 fact that when the compensation has been adjusted to suit the 

 fundamental frequency, it will in general be incorrect for the 

 various harmonics. These instruments are designed for use 

 under definite conditions as to voltage, frequency, and wave 

 form, and though they are serviceable on distribution systems 

 where these things are fixed, it is unsafe to apply them indis- 

 criminately in general testing. The wattmeters and watt-hour 

 meters may have very serious frequency and wave-form errors, 

 especially at low power factors. 



Use of Transformers. The use of instrument transformers 

 introduces errors, due to ratio and phase angle, which vary 

 with the load. These are discussed on pages 577, 578. 



METHODS OF CALIBRATION 



The dates of all calibrations should be recorded and inserted in 

 the legends of the calibration plots, together with the numbers of t 

 instruments. 



