970 



THE BELL SYSTEM TECHNICAL JOURNAL, SEPTEMBER 1952 



from 800 to 1300 cycles per second. A switching ai'rangement makes it 

 possible to apply either of the output voltages or their sum or difference 

 to a rectifier-meter combination. From the readings of the meter the 

 correlation can be computed according to the relation 



Pl2 



= (>S' - If)/4:VJ^2 



(6) 



Vi and V2 are rms values of the individual noise voltages, and S and D 

 are the rms values of their sum and difference. The equivalence to 

 expression (5) can be seen by noting that 



S' 



D' = (vi + ^2)^ — (^1 — V2)' = -4^11^2 



Results of correlation measurements on three bridges are shown in 

 Figs. 14-16. In each case the calculated curve is shown for reference. 

 The values of / were calculated from decay measurements on optically 

 injected holes, as described by J. R. Haynes," using a value for mobility 

 of 1700 cm /volt-sec. In Fig. 14 the agreement with the theoretical model 

 is very good. The scatter in the points is due to fluctuations in the 

 noise, which are quite large in the band used for these measurements. 

 In Fig. 15 the agreement could be made quite good with a lateral shift 



- 0.6 



2 0.5 

 o 



0.1 0.2 03 0.4 0.6 0.8 1.0 2 3 4 6 8 10 2C 



RATIO OF LIFE PATH TO SEGMENT LENGTH 



Fig. 14 — Noise correlation. The solid curve is calculated, the points experi- 

 mental. 



