GENERATION, CONTROL, AND MEASUREMENT 245 



ardizing laboratories, such as the U.S. National Bureau of Standards, 

 and are used only for the calibration of secondary-standard lamps 

 (Coblentz and Stair. 1933). 



Secondary standards, in the form of incandescent lamps that have been 

 carefully selected, aged, and standardized against a primary standard, 

 can be obtained with a certificate containing the precise operating con- 

 ditions under which the lamp was standardized (ibid.; National Bureau 

 of Standards, 1949; Teele, 1953). It is necessary that these conditions 

 be reproduced accurately because of the critical manner in which the 

 radiant flux emitted by incandescent lamps is dependent upon voltage, 

 current, and ambient temperature. The most important precaution to 

 be observed is to avoid subjecting the lamp to mechanical shock and 

 overvoltage. With careful use a standard lamp should serve the recjuire- 

 ments of the average research laboratory for many years without 

 restandardization. 



Since the incandescent lamp is a pure resistance at power frequencies, 

 it can be operated on either alternating or direct current with sufficient 

 precision for a secondary standard. The only limiting factor is the avail- 

 ability of meters that are sufficiently accurate for a-c measurements. The 

 less expensive moving iron vane, rectifier, and other types of a-c meter 

 cannot be calibrated directly with a d-c potentiometer. For precise meas- 

 urements of alternating current, power meters capable of accurate reading 

 on both alternating and direct current must be used. The two common 

 types are the dynamometer and the thermocouple meters, but they both 

 have nonlinear scales and are relatively expensive. In addition, the 

 thermocouple meters are easily damaged by overload. It is often more 

 practical to operate the standard lamp on batteries or a regulated rectifier- 

 type d-c power supply (Fig. 3-40), so that d-c meters can be used. 



For any incandescent lamp the current and voltage are dependent 

 variables. Therefore the lamp can be adjusted to the calibrated work- 

 ing conditions with either an ammeter or a voltmeter. In general, the 

 current measurement is preferable because, as the lamp ages, evaporation 

 from the filament causes its diameter to decrease. At constant current, 

 the current density of the filament increases and partially compensates 

 for this loss. The voltage measurement may be used to indicate whether 

 the lamp characteristics have changed significantly. 



Although the U.S. National Bureau of Standards recommends oper- 

 ating the standard lamps in a room with a curtain containing a suitable 

 aperture between the lamp and the detector, it is usually more convenient 

 in a small laboratory to operate the lamp in a housing such as that given 

 in detail in Fig. 3-40. 



The housing is a plywood box with a hinged cover and two sets of 

 controls. One is for the d-c power supply for the radiometric standard 

 lamp, as already described, and the other is for the 500-w standard of 



