24 



BELL SYSTEM TECHNICAL JOURNAL 



greater precautions are necessary for accurate results. There must 

 be no initial direct capacity in arm CS), or a correction will be re- 

 quired. Possibly variable capacity ratio arms would be preferable 

 to resistances. 



Null-Impedance Bridge Method for Direct Capacity, Fig. 4 



Assuming that the electron tube supplies the means of obtaining 

 an invariable true negative resistance, Fig. 4 shows a method which 

 determines any individual direct capacity from a single bridge setting. 

 The bridge arms are replaced by a Y network made up of two resist- 



Fig. 4 — Null-Impedance Bridge Method for Direct Capacity 



ances R, R and a negative resistance — R/2; the Y has then a null- 

 impedance between corner ^ and corners ^, (? connected together^. 

 The three terminals 1, 2, 3 of the network to be measured are con- 

 nected to corners 2), 6^, 6B and a balance obtained by adjusting the 

 variable standard condenser C . Then di = C regardless of the 

 direct capacities associated with Cn and C\ since these capacities 

 either are short-circuited between corners SB, (2 ox ^,(? or are between 

 corners 58, S) and thus outside of the bridge. 



Correct adjustment of the negative resistance may be checked by 

 observing whether there is silence in telephone Ti after the balance 

 has been obtained. Assuming invariable negative resistance, this 

 test need be made only when the bridge is set up, or there is a change 

 in frequency. The bridge may be given any ratio Z1/Z2 by employing 

 a Y made up of impedances Zi, Z2, and — Zi Z2/(Zi + Z2). 



'See appendix, section 4, which also describes a transformer substitute for theY. 



