A high-resolution TAC measures the time interval from the first accepted start pulse to the next stop pulse. It ignores all subsequent start pulses and any additional stop pulses until it has finished converting the first pair of start and stop pulses. If either input is receiving randomly distributed pulses at a very high counting rate, the TAC will prefer to analyze the pulses arriving earlier on that input and will suppress the pulses that arrive later. This will distort the measured time spectrum for correlated start and stop events. The distortion can be controlled by limiting the counting rates at the start and stop inputs. From Poisson statistics,¹ it can be shown that limiting the average random counting rate r at both start and stop inputs to

r <0.01 / T_range         (3)

will ensure that the number of suppressed pulses in the analyzed time range T_range will be less than 0.5% of the number of accepted pulses on the respective input. This condition is adequate to ensure less than a 1% distortion of the time spectrum.

For a short time range, T_range = 50 ns, the condition in Equation 3 limits the counting rate to 200,000 counts/s at both the start and stop inputs to the TAC. This counting rate is still high enough to require an MCA with a conversion time of 5 µs or less in order to keep up with the data from the TAC.

¹Ron Jenkins, R. W. Gould, Dale Gedcke, Quantitative X-Ray Spectrometry, Marcel Dekker, New York, 1981, First Edition, Chapter 4.