PMT Calibration Results and the MC simulation

1
PMT Calibration Results and the MC simulation

• Shigeru Yoshida, Chiba University
• http//www.ppl.phys.chiba-u.jp/

2
Outline

• Overview of the IceCube PMT calibration for the
  detector MC
• Absolute QE/CE
• Uniformity of CE
• Charge Resolution/WaveForm
• The Response Modeling
• A DOM-analog MC prototype

3
The DOM/PMT calibration
Gain/Waveform (by S.Yoshida Mons2003)
QE/CE Uniformity (by H.Miyamoto Mons2003/this
meeting)

• DOM 2D Scan
• (Scheduled this year)

4
The Absolute QE/CE measurement
Reflectors/Attenuators
Energy Meter
5
Attachment for the IceCube PMT

• Photon beam with 13 mm diameter scans the cathode

6
CE Uniformity2D scan
PMT
UV LED
R-guide
Rotation-bed
Motor
7
The Collection Efficiency Lego plot
8
Example The relative Collection Efficiency
SF0030 vs SF0050
SF0016 vs SF0050
9
(No Transcript)
10
(No Transcript)
11
Charge Resolution/Waveform
In the Freezer
Temp 32degree
12
PMT Setting in the Freezer
Base circuit
Diffuser attached to UV LED
13
SPE Charge Spectrum
14
SPE Waveform
15
The calibration data for the MC simulation

• The absolute QE/CE data of 4 PMTs now under
  measurement
• The CE uniformity data of 16 PMTs available
• The SPE spectrum and the waveform _at_ -30C At HV
  from 1600V to 2000 V for 4 PMTs
• The Gain uniformity data of 1 PMT (possible to
  add 3 more PMTs)
• We are ready to add more PMT data this summer in
  the pre-shipping measurement.

16
The DOM response Modeling
Charge Response/Waveform
17
SPE Charge Response

• Sspe(q) (1-Pe)exp-q/qt Pe NG(q-q0/qs)

18
SPE Charge Response ContdSome GOOD News

• qt/q0, qs/q0 universally scaling

19
SPE Charge Response ContdSome GOOD News

• Global qt/q0, qs/q0 can reconstruct the response
  of ALL PMTs at every gain

20
And some BAD News

• GAIN seems to have some position dependences.

21
(No Transcript)
22
We see it in the CE scan data

• .

CE X Gain

• We may be able to ignore this because
• so small in terms of AREA
• Simulating MPEs response would be OK
• ONLY SPE simulation might be problematic

23
SPE Waveform Model

• Gaussian fits well!!

24
SPE Waveform Model

• FWHM of a SPE pulse depends on PMT gain.
• But the change is small 200 psec or so

25
Some Issues and Proposal

• Uniformity exhibits significant tube to tube
  variance. We do not want to implement 4800 tables
  in the DOM MC.
• SIMULATE the response for the broader light beam
  as a function of angle and compare it with the
  measurement at UW.

26
Some Issues and Proposal (Contd)

• Localized gain variance has been found.
• Use the AVERAGE behavior ?

27
Some Issues and Proposal (Contd)

• Localized gain variance has been found.
• Just ignore it!

28
DOM Analog MC prototype
DOMVPmtWaveform.cc
DOMVPmtCharge.cc

• (1-Pe)exp-q/qt Pe NG(q-q0/qs)

GEANT4
DOMVPmtQe.cc
DOMVGlass.cc DOMVGel.cc
29
Summary

• The PMT calibration data is ready for the
  detector MC
• Modeling the PMT response was done.
• A 1st prototype PMT MC module is being built
  under the root framework
• Running this MC would address some concerns like
  the tube-tube variance
• Please bookmark
• http//www.phys.chiba-u.jp/research/IceCube/PMT
  /