77(5 



TII1<: BELL SYSTEM TECHNICAL JOUUNAL, MAY 1!)57 



26 



24 



< 



K 



_) 



O 20 



> 



o 

 a. 

 < t 8 



16 



14 



20 



10 20 30 40 50 60 



X 103 

 NUMBER OF OPERATIONS 



40 



60 



80 100 



120 140 

 X 103 



Fig. 3 — Measurements of arc voltage at cathode arcs between a carbon-plati- 

 num pair of electrodes, and between a carbon tungsten pair — at successive rever- 

 sals of striking potential. 



carbon is much higher and quite \'ariable in the range from 20 to 30. One 

 is tempted to conckide from this that the vapor in an arc between active 

 palladium contacts is predominantly the metal of the electrodes, not 

 carbon vapor. A more sound conclusion, however, as will be pointed out 

 later, is that the source of electrons on the cathode of an active arc is 

 palladium metal rather than carbon.* 



When contact surfaces are very heavily carbonized, an arc voltage 

 substantially higher than that characteristic of the metal of the elec- 

 trodes is sometimes observed for a short time at the beginning and at 

 the end of an active arc occurring at the discharge of a capacitor into an 

 inductive circuit. An example of this is shown in the oscilloscope trace 

 of Fig. 4. The higher arc voltage at the beginning of this arc, when the 

 current was extremely small, is interpreted as the initiation of the arc 

 between carbon surfaces, and the enhanced voltage at the end is evidence 



* A very simple experiment has been carried out which proves conclusively 

 that the character of an arc of the type which we call a cathode arc (see below, 

 and Ref . 3 and 4) is determined by the properties of the cathode, and not by those 

 of the anode. This is perhaps self evident, l)ut a direct test is reassuring. Tlie test 

 is simply the observation that, for an arc of the cathode type Itetween electrodes 

 of different materials, the arc voltage is sul)stantially the same as it would be if 

 both electrodes were of the cathode material. The test is made by reversing the 

 potential between the electrodes rejieatedly, and after each reversal observing 

 that the arc voltage changes gradually from that characteristic of the anode to 

 that characteristic of the cathode. After each reversal the arc begins to clean from 

 the cathode the anode material that was deposited there before the reversal, when 

 wluit is now the cathode was the anode. Accompanying this cleaning, the arc 

 voltage goes uj) or down until it reaches the value characteristic of the cathode 

 itself. Measurements obtained in this way are reproduced in Fig. 3 for a carbon- 

 platinum pnir of electrodes and for a carbon-tungsten pair. 



