| |
Counting
In counting applications, the analog input pulses that cross the
discriminator threshold are converted to standard logic pulses at
the output of the timing discriminator. These logic pulses can be
counted in a counter/timer or in a multichannel scaler. Usually, the
discriminator threshold is set just above the noise level, so that
all the real events are counted without counting noise pulses. If a
narrow band of pulse-heights must be selected for counting, the
Model 583B can provide that function with its dual discriminator
thresholds.
The maximum permissible counting rate can be restricted by the
pulse-pair resolving time of the system. When the inherent width of
the detector pulse is not the limiting factor, timing discriminators
can offer pulse-pair resolving times in the range of 5 to 65 ns. If
dead time losses must be restricted to <10%, a 5-ns pulse-pair
resolving time permits average counting rates up to 20 MHz. This
same pulse-pair resolving time will handle burst rates up to 200
MHz, provided there is a 5-ns spacing between pulses.
The fast detectors commonly used in single-photon and single-ion
counting (photomultiplier tubes, microchannel plate detectors,
microchannel plate PMTs, and channeltrons) usually have pulse widths
short enough to attain the 5-ns pulse-pair resolution of the fastest
discriminators. Other types of detectors, such as scintillation
detectors, germanium detectors, and silicon charged-particle
detectors, deliver pulse widths that limit the pulse-pair resolving
time to much longer values. In most applications the maximum
counting rate is limited by the detector and the electronics that
precede the timing discriminator because of factors other than
pulse-pair resolving time. In such cases, although the pulse-pair
resolving time of the timing discriminator does not determine the
maximum permissible counting rate for the system, it still affects
how closely randomly arriving pulses can occur and yet be recognized
as two separate events. This, of course, determines the dead-time
losses in a counting experiment. |
