SHIELDING IN HIGH-FREQUENCY MEASUREMENTS 575 



place the completed equipment in an electrostatic shield to avoid all 

 coupling to the measuring circuit or from one to the other. This is 

 usually done by mounting the apparatus on a metal panel and placing 

 it in a wood box with a sheet metal lining. By this means it is possible 

 to place the auxiliary apparatus as close to the measuring circuit as 

 desired, without introducing errors. 



As far as the internal shielding of the auxiliary apparatus is con- 

 cerned, the same rules hold as for other apparatus described. How- 

 ever, in the case of vacuum tube equipment, wherever any gain is 

 introduced, coupling between parts of the circuit must be reduced in 

 proportion to the gain introduced between these parts. This is 

 usually accomplished quite readily by suitable mechanical design and 

 may be insured by placing each stage in a separate grounded shield, 

 although this is seldom necessary. 



Conclusion 



It has been impossible to go into very great detail in this brief paper 

 on the subject of shielding. The attempt has been made, therefore, 

 to outline a few general rules and to give representative examples of 

 typical measuring circuits. It will be noted that the examples have 

 been limited largely to the bridge circuit. This is because our ex- 

 perience has shown that this circuit is the most flexible and accurate 

 over the whole of the frequency range over which precise impedance 

 measurements have been made, and because the problems of shielding 

 it are sufficiently difficult and varied to give satisfactory examples of 

 the solution of rather complicated problems. The principles of 

 shielding given have been found to apply equally well at all frequencies 

 and it has been found that up to the maximum frequency at which 

 precision measurements have been made, the shielding methods de- 

 veloped for use with moderate frequencies require practically no 

 modification as the frequency is increased. Experience with measure- 

 ments and measuring circuits up to 2,000 kilocycles makes it appear 

 probable that when precision measurements are made at still higher 

 frequencies, the shielded bridge circuit will continue to remain the 

 most satisfactory measuring circuit. 



