322 ISOTOPIC TRACERS AND NUCLEAR RADIATIONS [Chap. 10 



10.16. Counter Resolving Time. The shortest interval in which two suc- 

 cessive events can produce in a counter separate discharges of sufficient 

 amplitude to register as counts is referred to as the resolving time t. Since 

 the actual insensitive time of a counter following a discharge varies somewhat, 

 the measured value of t is the average of these intervals. For most counters 

 the resolving time is in the order of 10~ 4 sec, which is approximately the 

 collection time of the positive ions formed in the discharge. 



The value of r for any counter can be ascertained by two general methods 

 in use at the present time. The first is an indirect method in which t is 

 calculated from the observed counting loss, and the second is a direct method 

 where the length of the interval is measured with electronic devices such as 

 the synchroscope. 



The first of these methods is most often used since it requires no special 

 equipment and can be carried out relatively quickly. Two samples, I and II, 

 with nearly equal activities are prepared and measured as follows: Sample I 

 is counted alone giving a counting rate n\\ without disturbing it sample II 

 is placed in counting position and the total counting rate nn of I + II 

 observed; sample I is then removed and the counting rate n 2 of sample II 

 obtained. From the three counting rates and that of the background tib, the 

 resolving time is calculated [35] with the formula 



n\ + «2 — W12 — fib 

 T = z 2 2 mm 



The total number of counts in each measurement should be sufficiently 

 large and of roughly the same magnitude to ensure a small probable error in 

 the counting rate since r is estimated from small differences between large 

 numbers. 



The formula given here is one of several that have been suggested for 

 computing the resolving time from a single set of measurements with a paired 

 source. Although the various formulas differ considerably at high counting 

 rates, they lead to nearly the same value for r at low rates where the formula 

 above is valid. It is essential therefore to use test samples whose activities 

 provide counting rates at levels where 100 {ri\ + n 2 — n^/n^ is no greater 

 than several per cent. At the same time, however, the activities should be 

 high compared to the background count to ensure accuracy in a reasonable 

 length of time. Normally, samples with counting rates in the order of 

 1,000 cpm are found most useful. 



The direct method for evaluating r requires a linear pulse amplifier and an 

 oscilloscope with accurately known sweep frequencies. The width of the 

 pulse in microseconds can be estimated directly from the oscilloscope trace 

 and taken as a first approximation for r. A more useful device is an oscillo- 

 scope equipped with a single sweep circuit and a persistent screen. The 



