ICD Energy Calibration

1
ICD Energy Calibration Andy White January 16,
2002
2
? ICD in Run I -gt Expectations for Run II ? ICD
coverage in Run II 1.1 lt ??? lt 1.4
?ieta ? 12, 13, 14 ? Ingredients of the
ICD energy calibration
3
ICD Status
? All ICD components are fully installed and
ICD is being read out ? Timing of ICD relative
to Calorimeter understood ? 5/378 channels
give low/dead response - will need access to
fix ? Learning to use LED Pulser as monitor
4
South ICD - before fiber cables installed
5
(No Transcript)
6
(No Transcript)
7

• INGREDIENTS
• ADC to GeV Conversion factor
• Sampling fractions
• Tile to tile variation

8
The ADC to GeV conversion depends on the
following factors 1) The specific energy loss
(dE/dx) in the Bicron BC-400 scintillator. This
is a PVT scintillator so dE/dx (min) 1.956
MeV/(g/cm2) x 1.032 g/cm3 2.02 MeV/cm 2)
Mean MIP peak in teststand ADC counts. From
Marks distribution of June 2001 the mean for 368
channels was 135.7. (We were aiming for 140).
9
3) The relative gain factor between calorimeter
preamps (used on the teststand) and the ICD
preamps. This factor was measured at the
teststand to be 3.8. 4) The extra amplification
of x8.7 that we used to boost the signal on the
teststand. 5) The factor of 10 between the least
count of the teststand ADCs and the calorimeter
ADCs. The least count for the teststand ADC is 1
mV and for the calorimeter is 0.1 mV.
10
6) The cosine factor that accounts for the
angle of a normal to an ICD tile relative to a
straight line drawn from the IP through the
center of a subtile. There are three numbers, one
for each eta bin covered by the ICD Eta
cosine factor 12
0.592 13 0.633 14
0.671 However, NOTE that the sampling
fractions ALSO include this angular factor so
we must be careful not to use it TWICE! 7) The
thickness of an ICD tile. All tiles are 0.5 inch
thick 1.27 cm.
11
Therefore the average MIP peak position in
calorimeter ADC counts is given by
(135.7 x 10) / (3.8 x 8.7) 41.0
counts Using this with the other factors give
the energy deposition in an ICD tile as (
Cal. ADC count / 41.0) x ( 2.02 MeV/cm x 1.27
cm) Finally this gives the energy deposition
as Cal. ADC count x 0.06257
(MeV) Cal. ADC count/15982
(GeV)
12
Sampling fractions

? Present values are from a Monte Carlo
simulation by Vishnu using 20 GeV/c single
pions ? From caltables float Weight_icd1
72.6591 //ilayer 9, ieta12 float
Weight_icd2 69.5046 //ilayer 9,
ieta13 float Weight_icd3 63.4282
//ilayer 9, ieta14
13
Tile to tile variations
? Each tile fiber cable PMT combination was
calibrated using cosmics on the ICD teststand ?
These combinations stayed together on the
detector (apart from changes due to failures) ?
The variation from the mean response is
available and will be used as an additional
correction factor (see talk by Andre) ?
These data should be in DB as some run
dependence is expected
14
ICD tiles - teststand results
15
CONCLUSIONS
? All factors are in hand to achieve initial ICD
calibration ? Need to study effects of ICD
contributions on response across ICR ? Need
high statistics di-jet and jet-? samples -gt
float sampling fractions to optimize
correction -gt do we need phi dependent fractions?