M. Bonesini
INFN Milano- Bicocca
Some Comments on MICE TOF I design
The design of TOF1 is similar to the one foreseen for TOF0 (see note MICE-145 for further details). The only differences are due to :
· The operation in the fringe field of the first analysis solenoid
· The lower particle rate, as respect to TOF0 (depending on estimates
on average 700 KHz against 1.5 MHz)
Assuming that the residual component of B along the PMT axis (B//) can be reduced by a global iron shielding (1 cm thick , 10 cm gap), conventional R4998 PMTs may be used (see H. Witte, J. Cobb presentations or G. Gregoire last one). The only open problem is that all computations were 2-D and the perturbation from the magnetic PMTs individual shielding on the B field has not yet been studied (it requires a 3-D computation).
As additional remark fine-mesh PMTs may have production problems at Hamamatsu and their price has increased of about 50% in the last year.
The solution with conventional PMTs allows savings in the TOF1 readout,
as respect to the preliminary design
Figure 1 shows the behaviour of the Br and Bz components of the field for intergap distances of 100,120,200 mm. R4998 PMTs in the present delivered H6533MOD assembly (1 mm of m-metal shielding) have been tested to tolerate longitudinal fields up to 50 Gauss and orthogonal fields up to 100-120 Gauss. This points to an gap distance of ~100 mm, that is well suited to accommodate TOF1 with its support structure, and probably an additional soft iron shielding for each PMT. We can easily devise a structure that integrates the support of TOF1, the annular additional 1cm iron global shielding on a single stall.
Figure 2. Br and Bz components of the field with a 1 cm global iron shield for different gap thickness.
The main difference with TOF0 will be the dimensions: two X,Y planes
covering a 48x48 cm2 active area or smaller and the counter lateral segmentation: 6 cm instead of 4 cm.
At the BTF testbeam, 6 cm large BC420 counters have been tested in addition to 4 cm large ones, with no apparent deterioration of timing
due to bigger pathlengths fluctuations (see figure 2 for an MCA plot, giving st~45ps). As low-cost Ukrainian scintillators UPS-95F give similar results as the more expensive Bicron BC420 scintillators (albeit with a test on 4 cm large counters only), this last solution may be used in case of severe shortage of funds.
The same baseline-1 electronic chain foreseen for TOF0, based on CAEN V1290 TDC, may be used for TOF1 channels and seemS to be adequate.
Timing calibration is still based on impinging beam muons, using the detector X/Y redundancy. For the monitoring system a downsized
version is under study, based on a low-cost laser diode and a novel light injection system into the scintillator bars that may afford reduction in the required laser power up to a factor 103-104 (in collaboration with the MEG experiment).
Figure 1. (t_L-t_R ) for 6 cm wide BC420 counters from reanalysis of BTF MCA data (ORTEC 566 TAC resolution
was found 12.2 ps/count).
In table I a cost estimate for the new revised version of TOF1 is compared to what originally foreseen.
TOF1 TRD design / TOF1 revised designPMTs fine mesh / 87.5 KE / Conv. PMTs / 52KE
mechanics / 5KE / Mechanics / 8KE
Scintillators+light guides / 10KE / Scintillators+light guides / 5KE
cables / 3KE / Cables / 3KE
Patch panels … / 5KE / Patch panels … / 5KE
Laser syst (1/3) / 22KE / Laser syst (1/3) / 10KE
FE electronics / 18KE / FE electronics / 18KE
HV voltage (1/3) / 12 KE / HV voltage (1/30 / 9 KE
Total / 162 KE / 110 KE
Table 1. Cost estimates for TOF1 revised design (common items such as laser monitoring system, HV, … are assumed 1/3 for each TOF station)
1