1120 THE BELL SYSTEM TECHNICAL JOURNAL, OCTOBER 1951 



frame grid. The L3 grids are a further development. The major difference 

 between the earlier frame grids and these is in the method of bonding the 

 0.0003'' lateral wire to the side rods. In the earlier grids a gold braze was 

 used to bond the laterals to the side rods. This necessitated heating the 

 unit to approximately 1070°C to flow the gold. The newer grids have the 

 lateral wires bonded by a glass glaze which allows the process to be carried 

 out at approximately 700°C. There is a differential expansion between 

 molybdenum and tungsten of about five to four. The net result of the 

 reduction of temperature in this process is that the tungsten wires are 

 stretched less at the lower temperature and thus when returned to room 

 temperature have higher residual tensions. This is important because the 

 higher the residual tension the higher the resonant frequency of the lateral 

 wires. This in turn means that the noise level of the tube due to vibration 

 or shock will be reduced since loose grid wires will give rise to microphonic 

 noises. Tighter wires also decrease the possibility of grid to cathode shorts. 

 Tests have shown that an increase of about 25% in the resonant frequencies 

 of the lateral wires can be expected as a result of using the glass glazing 

 technique as compared to the gold brazing technique. 



It is interesting to note that the residual stress in the lateral wires of 

 these grids is of the order of 200,000 pounds per square inch. This figure is 

 roughly ten times as great as the allowable working stress for steel beams 

 such as are used in the construction industry. 



When the glazing technique is used, the grid is gold plated after the 

 glazing operation has been completed. The gold is used to inhibit thermionic 

 emission from the grid wires. This is a necessity for tubes of this type when 

 used in the circuits for which they were designed. The need for the plating 

 exists because of the proximity of the grid wires to the hot cathode and 

 their unfortunately favorable position for receiving a deposition of active 

 material from the cathode during its processing and operation. The desired 

 amount of gold on the grid wires is that which will cause a diameter in- 

 crease of about 0,00002". This is an extremely difficult increase to measure 

 because the measurement must be non-destructive, since it is made on the 

 finished grid and is used as a production control. The method used to date 

 has been an optical measurement at a magnification of about 500 X. 



A very high degree of precision, compared to that previously available, 

 has been obtained for some of the parts whose dimensions are critical. 

 The cathode sleeve is now obtainable with minor axis limits of ±0.0003''. 

 The mica discs are now made with the critical holes to that same tolerance. 

 The frame grid side rod is made to ±0.0001". These are the basic elements 

 of the tube and, after inspection has shown them to be acceptable, their 

 assembly becomes close to that of a conventional tube. The inspection of 



