In order to measure the spatial light output nonuniformity,
Thorn EMI 9822QKB photomultipliers are coupled to the CsI crystals
with optically transparent Dow Corning SYLGARD glue, and are
placed into a designated light-tight aluminum `Cosmic Muon Tomography'
(CMT) box.
Two pairs of flat, multi wire drift chambers (MWDC) are mounted
above the box, and
one pair is fixed below it. By having a total of three wire chamber pairs,
the muon trajectories can be both reconstructed and checked for scattering.
The wire chambers have an effective coverage of about 60
60 cm.
Under the box and chambers there are two 1 cm thick plastic scintillators
(`SB1' and `SB2'), and just above the scintillators there is a 5 cm
thick layer of lead bricks. Figure 3.2 shows a diagram of the
apparatus.
Figure: Monte Carlo simulation of penetrating cosmic muons interacting
with the CMT experimental apparatus. Two of a possible six CsI
crystals positioned in the apparatus are shown. The lower three black
rectangular objects are multi wire drift chambers (MWDC); two are
above the box and one is below the box. Also shown is a layer of 5 cm thick
lead brick below the lowest MWDC, and two flat plastic scintillators
at the bottom.
The lead bricks serve to help stop cosmic muons with kinetic
energies lower than 
160 MeV, as shown with the Monte Carlo simulation.
This ensures that the accepted muons are minimum ionizing particles that
travel straight through the crystals and the entire apparatus.
Six crystals can be positioned in the box at one time, and are connected to signal and high voltage cables which feed through from the electronics outside the box. The trigger for an acceptable cosmic muon event consists of at least one crystal being hit, in coincidence with a signal from both scintillators. The trigger rate is approximately 1 Hz, which means that each crystal requires one week to be reliably tomographed. Although the wire chamber information is not used to define the trigger, it is used offline to reconstruct the trajectories of the accepted muons and to reject particles with scattered trajectories. The electronics logic is shown in figure 3.3
Figure 3.3: Schematic diagram of the electronics logic for the tomography
apparatus. The trigger for a cosmic muon event is a signal from at least
one CsI detector, in coincidence with the two scintillators
SB1 and SB2.
The number of photoelectrons per MeV is measured by attaching a light
emitting diode (LED) to the CsI crystal. It is then clocked to flash at
10 kHz. The trigger for an `LED' event is simply a signal from at least
one of the six crystals in the box. The light intensity from the LED is
varied to produce peaks equivalent to
10 MeV to 100 MeV in five or six steps, which produces five
or six discrete peaks in the detector's recorded energy spectrum.
By plotting the square of the standard deviations against
the means of these peaks, one can extract the number of photoelectrons
per MeV from the slope, as shown in Fig. 3.4.
Figure 3.4: Graph of 
against Energy for pentagonal crystal
``S003''. From the inverse of the slope of the graph, one finds that
the number of photoelectrons per MeV for this crystal is
96/MeV.