ARCING OF ELECTRICAL CONTACTS IN TELEPHONE SWITCHING CIRCUITS 551 



ia = 10^ amp/cm"^ and /„ = 0.5 amp one gets: T = 2400°K at t = 

 4 X 10~^ sec and T = 1600°K at / = 1.1 X 10"^ sec for Pd. 

 The corresponding thermionic emission is obtained from 



ith = ATe 



2- ~ kT 



_■> 



with .1 = 60 amp cm~" deg.~ and <p = 4.99 volts for Pd . At the 

 termination of the arc, t = o, ith = 0.048 amp/sec^ at ^ = 4 X 10"^ sec, 

 ith = 0.032 amp/cm" and at ^ = 1.1 X 10~*sec, ith = 6 X 10"^ amp/cm.^ 

 The respective rates of electron emission from the arc spot are 1.5 X 10^°, 

 1.0 X 10 and 1.9 X 10 electrons/sec. This indicates that the initiating 

 electrons may be furnished by thermionic emission if the charging time fol- 

 lowing the first arc is of the order of or less than about 5 X 10~^ sec. This 

 time is more than an order of magnitude too small compared to the 

 charging times involved in the data of this section. One may, therefore, 

 exclude the thermionic emission as an explanation for the low arc initia- 

 tion voltages obtained. 



The initiation of reversed arcs, however, may be enhanced by ther- 

 mionic emission from the previous arc spots since the recharging times 

 involved, tQcY ^, are usually very small. I and c are usually of the orders 

 of 10"' henry and 10" ^ farad and the charging time is of the order 

 10"^ sec. 



ESTABLISHMENT OF GLOW DISCHARGE AND TRANSITION INTO AN ARC 



For the circuit in Fig. 1(b), it was observed that on break of the 

 contact, glow discharge was observed under certain circuit and contact 

 surface conditions. An obvious requirement was that the voltage across 

 the contacts should exceed the glow discharge voltage of the contact in 

 the surrounding atmosphere. This requirement alone, however, was not 

 .sufficient as in some cases no glow could be detected, in others glow was 

 established and maintained and in other instances glow was followed by a 

 transition into an arc. In this section is presented an experimental study 

 of the conditions that determine the nature of the discharge. 



Cathode Current Density in Sialic Normal Glow 



First, measurements were made of the cathode current density in a 

 static normal glow. This was done for palladium and gold contacts in dry 

 atmospheric air at 25°C. In each case the cathode was the flat end of a 

 cylinder and the anode was a larger parallel flat surface of the same ma- 

 terial as the cathode. The circuit in Fig. 12 was used. The contacts were 



