618 



BELL SYSTEM TECHNICAL JOURNAL 



<n 16 



'C 14 



10 12 14 16 18 20 22 24 26 28 30 32 34 36 



SECOND test: microphonic level in decibels below 1 VOLT 



Fig. 2 — Reproducibility of microphonic noise level measurements using a commercial 

 socket with a constant, artificial, mechanical stimulus (100 No. 102F Tubes). 



sockets, it has been observed to be as low as 3 db and in others as high 

 as 8 db. 



In order to show that this random variation is not due to the tube 

 itself, experiments have been made with two forms of suspension which 

 minimize the reaction of the mounting on the vibration of the tube and 

 so reduce as far as possible the effect of variation in coupling. In one 

 set-up, the tube is hung by a single thread of rubber, stretched to its 

 elastic limit, the electrical connections being made by very light, 

 flexible leads fastened with light clips directly to the prongs of the 

 base. In the other, the tube is clamped lightly between two large 

 blocks of very soft sponge rubber, and the electrical connections are 

 made through rhercury cups into which the base prongs dip. In both 

 cases, the agitator is a light pendulum striking the base or bulb of the 

 tube. The two mountings give very similar results, and are charac- 



