THE RELATION BETWEEN VOLTAGE AND CANDLE- 

 POWER IN MODERN INCANDESCENT LAMPS 



WILLIAM KUNERTH 



Among the operating characteristics of incandescent lamps, there 

 are certain features which are often overlooked by the man using 

 or dealing with lamps. The voltage is one of the chief factors upon 

 which the behavior of a lamp depends. It is well known that under 

 an excessive voltage, a lamp will not last long, but will last much 

 longer than its normal life if the voltage is below normal. Mathe- 

 matical formulje have been deduced from experimental results to 

 cover this relation. 



It is frequently found desirable to tell what the candle-power of a 

 lamp is when its voltage and the rated candle-power at some other 

 voltage are known. The accompanying curves show this relation as 

 determined experimentally, from a considerable number of G. E. 

 lamps. The Mazda B or Mazda vacuum lamps tested ranged from 

 20 watts to 100 watts ; the Mazda C2 or daylight lamps from 75 

 watts to 200 watts ; the Mazda C or gas filled lamps from 75 to 100 

 watts, and the carbon lamps as indicated on the curve. It may be 

 noticed from the curve that the 20-watt carbon lamp gives only 10 

 per cent as much light at 80 volts as it does at 110 volts. 



Everyone knows that there is a great deal of fluctuation of line 

 voltage on some commercial circuits. The candle-power of the lamp 

 will rise or fall with the change in voltage as shown by the curve 

 and this will have a corresponding effect upon the person using the 

 light for reading or working.^ 



Curves of this kind have the advantage that the candle-power de- 

 sired may be obtained with very little calculation. This is much 

 simpler and quicker and more accurate than the mathematical rela- 

 tion which is also given herewith for the different lamps tested. In 



the expression J- ^ ( JLY , I is the candle-power sought, V is 



1st \ Vsty 

 the voltage at which the lamp is operated, 1st is the candle-power at 



^See Iowa Acadamy of Science Proceedings. Vol. 22, p. 333. 



