1476 THE BELL SYSTEM TECHNICAL JOURNAL, NOVEMBER 1953 



The data shown in Fig. 3 and subsequent data are plotted for basic 

 4-1-lB stacks. Tests were made with 60-cycle sinusoidal supply voltage. 

 The * 'forward voltage drop" is expressed as the rms volts required to 

 produce a specified current in a moving coil dc ammeter connected di- 

 rectly to (short circuiting) the output terminals of the rectifier stack. 

 For stacks w4th more than one cell per element, the voltage drop is ob- 

 tained by multiplying the observed drop in Fig. 3 by the number of cells 

 in series per rectifying element. 



The reverse current is measured by applying a specified rms voltage 

 to the ac terminals with the dc terminals open circuited and noting the 

 rms input current after the current has stabihzed. When selenium stacks 

 have been ''off voltage" for some time, a relatively high reverse current 

 is obtained for the first few seconds. The current then decays approxi- 

 mately exponentially. Usually, the current will stabiUze after voltage 

 has been applied for 5 to 10 minutes. 



It should be emphasized that these curves are only typical characteris- 

 tics. Selenium cell manufacture requires individual testing of each cell 

 before assembling into a stack. Cells are graded by their electrical char- 

 acteristics. Large variations exist between the lowest and highest grade. 

 Each manufacturer sets up his own standards regarding the variations 



Fig. 2 



10 20 30 40 50 60 70 80 90 



AMBIENT TEMPERATURE IN DEGREES CENTIGRADE 



Temperature de-rating curves for long-life expectancy. 



