Next: An optimized trigger for
Up: The model
Previous: the level 2 trigger
Figure 12:
Comparison between ideal and actual rates of recording signal
events on tape as a function of beam intensity, evaluated by the model
under the conditions at the end of the June 1998 run period.
|
As has been shown in the previous sections, the model gives a good
description of the full set of counting rates of relevance to our
trigger from very low intensities up to the beam current at which
Radphi has been designed to operate. The model now permits a decomposition
of the event sample collected at any given beam intensity into signal
and background components, and to compare the efficiency for signal
collection at various beam currents. The rate at which signal events
are recorded on tape, denoted , is given by
|
(15) |
which is to be compared with the corresponding rate for an experiment
without any dead-time effects, ,
|
(16) |
which is directly proportional to the beam intensity. The situation at
the end of the June period is summarized in Fig. 12. In
particular, it shows that increasing the beam current above 100nA does
very little to increase the rate at which signal is being written to tape;
in fact above 175nA it begins to decrease. Given that the proposal assumed
no dead-time effects and an intensity that corresponds to 250nA on this
figure, we are presently a factor of 4 short of our design goal in terms
of rate capability.
Figure 13:
Comparison between ideal and actual rates of recording signal
events on tape as a function of beam intensity, evaluated by the model
under the conditions at the end of the June 1998 run period.
|
Figure 14:
Comparison between ideal and actual rates of recording signal
events on tape as a function of beam intensity, evaluated by the model
under the conditions at the end of the June 1998 run period.
|
Given that no single stage of the electronics or data acquisition chain
was operating near its maximum rate (the phototubes and bases on the CPV
were running near their limit but they do not contribute dead-time within
this model) there is good reason to be skeptical that the overall losses
are so high. As a check on this result, Scott Teige analyzed the data
from the intensity scan taken at the end of the June period, and extracted
the yield of and mesons from their and
decays respectively. The yields are plotted as a real-time rate in
Figs. 13-14. The solid curves are the shape of the signal
(red) curve from Fig. 12. Fig. 13 in particular gives
striking confirmation of the model prediction.
Next: An optimized trigger for
Up: The model
Previous: the level 2 trigger
Richard T. Jones
2003-02-12