46 BELL SYSTEM TECHNICAL JOURNAL 



The voltage wave form of an entire " B " transient, covering the time 

 from the initial separation of the contacts to the final subsidence of 

 the voltage charging the line wire, is shown in Fig. 3, and the a-c. com- 

 ponents of the current in the range from 20 kilocycles to 20 megacycles 

 are shown in Fig. 4. The low-frequency components of the current are 

 comparatively weak. A line and load relay were chosen to give a 

 relatively simple transient with the important components at frequen- 

 cies which could be photographed. A 500-ohm Western Electric 

 U-type relay and a line of 300 ft. of No. 22 switchboard pair were used. 

 The mate wire of the pair was grounded at both ends. The currents 

 and voltages were not photographed simultaneously but the types of 

 the transients were correlated by repeated observations. A current 

 picture will not exactly correspond to a voltage picture, as the tran- 

 sients produced by successive operations of a contact are never 

 identical. 



The "B" transient may be explained as follows, using as a basis the 

 simple circuit of Fig. 1. The steady current is established and the 

 contacts start to separate, moving apart at a speed, which is at first 

 surprisingly slow (about an inch a second). The contacts have been 

 deformed by the pressure between them, and as this is relaxed the 

 current density and the temperature at the contacting areas rapidly 

 increase until at some light pressure the area becomes so small that 

 the current explodes it. There may be some necking out of the softened 

 contacts before this and under some conditions there are indications of 

 a metallic arc lasting a fraction of a microsecond, but at any rate an 

 initial rupture occurs between hot and soft metal areas. 



The wire has been at ground potential, but the battery plus the 

 collapsing magnetic field of the load relay commence to charge it at a 

 rate depending on the line and relay winding capacity and the relay 

 inductance and losses. In ten or twenty microseconds, it has reached 

 at the contacts a potential of from 50 to 200 volts. This is below the 

 voltage at which sparkover due to ionization of the air can occur, but 

 something usually happens which recloses the circuit. This is believed 

 to be caused in somewhat the same manner as the " preclosures " men- 

 tioned earlier. It is probable that a cold point discharge reheats the 

 contacts. This is followed by a collapse of the voltage to about 15 

 volts above zero in the direction of the previous voltage, indicating the 

 formation of a metallic arc. This lasts a fraction of a microsecond 

 and the voltage then drops to nearly zero, suggesting that the contact 

 areas heated by the field current and the arc have been drawn together 

 in solid metallic contact. The line is discharged with an oscillation of 

 comparatively low damping (which is characteristic of the line wire) 



