napoli

1
RICH status report Transversity Meeting A March
05 -Jlab

• Detector Performances
• Possible upgrade

2
The PID Challenge

• Very forward angle ---gt high background of p
  and p
• TOF and 2 aerogel in not sufficient for
  unambiguous K identification !

3
KAON Id Requirements
Signal Vs. Background
Process Rates
signal (e,eK) bound state 10-4 10-2
accidentals (e,e)(e,p) (e,e)(e,p) (e,e)(e,k) 100 100 0.1

• Very forward angle ? high background of p and p
• TOF and 2 threshold Cherenkov aerogel are NOT
  sufficient for unambiguous K identification
• RICH DETECTOR

4
JLAB RICH detector, similar to the ALICE and STAR
RICH
MIP
Cherenkov angle resolution
Separation power
Relevant quantities NP.E. N. of detected
photons AND sq angular uncertainty
5
(No Transcript)
6
JLAB Hall A RICH MonteCarlo Simulations
N. of detected photoelectrons
Separation power
p, K Separated by qp qK 30 mrad
p, K Separated by ? 6.8 s
7
JLAB Hall A RICH some components
8
Rich MWPC performance at lower HV
HV 2100 V
HV 1900 V
9
JLAB Hall A RICH OPERATING conditions
Gas Pure Methane (Minimize Photon Feedback, High
Q.E.) High Voltage 2100 Volts for a gain
of 8x104 Grid Voltage 250 - 450 Volts
Optimal trigger to read-out delay 400 ns
(peaking time of gassiplex response)
MIP charge and hit size
cluster charge and hit size
10
RICH Performances key parameters
p Cherenkov angle reconstruction
Npe for p and P
Nclusters
Cherenkov average angle (rad)
Angular resolution
Npe p/p ratio
11
RICH Performances PID
p/K population ratio
Aero Selected p
Angular resolution
Aero Selected K (!)
Aero Selected P
Separation power
Aero Selected K on a large sample of filtered
data
This would accept 10-4 pions x p/K
ratio ?1/100 pion contamination
. But NON GAUSSIAN TAILS GIVE AN IMPORTANT
CONTRIBUTION !
Kaon selection
12
Rich PID Kaon selection results
Time of coincidence for Aerogel Selected Kaons
w/o and w/ rich
AERO K
AERO K RICH K
p
P
K
13
RICH PID Pion rejection factor
Time of coincidence for Aerogel Selected Pions
effect of Rich Kaon selection
AERO p RICH K
AERO p
p
p
N.Evts in the peak Backgnd subtr. 64656
N.Evts in the peak Backgnd subtr. 63
Pion rejection 1000
14
JLAB Hall A exp e94-107 Preliminary Results on
12C Target
Missing Energy Spectra
Aerogel Kaon selection
12C(e,eK)12BL
RICH Kaon selection
Missing energy (MeV)
15
(No Transcript)
16
(No Transcript)
17
250 ps
18
p
p
Pion rejection 1000
K
K
Possible improvements
- MWPC stability for high rates For single
rates 60 KHz HV2100 V is OK In the range 60
KHz 100 KHz HV2075 V is OK Above 100 KHz HV
must be reduced further (running at reduced gain
with moderately good performance seems to be
feasable) - p/K separation for pgt2.5 GeV/c
Doable just replacing the radiator
DAQ rate bottleneck (1 KHz) can be ovecome
replacing part of the readout
19
RICH electronics upgrade
The HMPID ALICE RICH DAQ scheme
fbD27..0
fbD31..0
VME to Local Bus Interface
LOC_ADD11..0
LOC_ADD3..0
DILO 5 Boards (ADC and DILOGIC)
Column Controller (1 to 8)
Segment Controller
LOC_CS
GASSIPLEX
LOC_CS
LOC_R/Wn
LOC_R/Wn
RCB BOARD
SEGMENT
Front end digitization/ multiplexing On board 48
multiplexed channels (instead of 240) Clock rate
up to 10 MHz
20
HV 2100 V
HV 1900 V
21
(No Transcript)
22
Rich p/K separation for p gt 2.5 GeV/c
Radiator C5F12 n1.24 sCh 5mr
Radiator C6F14 n1.29 sCh 5mr
4 s separation at 2.5 GeV/c
4 s separation at 3.0 GeV/c
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
Danilyuk, Novosibirsk, RICH2004
aerogel
photodetector particles
L 4.4 cm at 400 nm Layer t
n n mm desired
measured 1 6.0 1.030 1.0297
2 6.3 1.027 1.0268 3
6.7 1.024 1.0234 4 7.0
1.022 1.0213
Produced in May 2004
27
RICH SOFTWARE PRESENT STATUS

• PID algorithm is based on
• recognition of the Clusters (that is continue
  spots of fired pads) on the cathode planes.
• 2. Identification of the baricenters of the
  clusters with the Cerenkov photon hits on the
  cathode planes
• 3. Calculation of the Cherenkov photon emission
  angle through Cerenkov photon hits on the cathode
  planes.
• 4. Calculation of the average of the Cherenkov
  photon angle distribution and check of the value
  obtained with the expected emission angle (three
  checks, one for each hypothesis on the particle
  to be identified Pion, Kaon and Proton).
• 5. Three c2 tests (one for each hypothesis on the
  kind of the particle) to check the obtained
  Cerenkov photon angle distribution with the
  expected one.
• 6. A procedure to cross out the signals from
  noise. This procedure is based on the c2 test
  values and is performed when none of the
  hypothesis on the particle hitting the Rich seems
  statistically significant.
• - The p rejection factor obtained with the
  algorithm described above is 999/1000

To be done

• -A better method to determine the Cherenkov
  photon hits on the pad plane (so far identified
  with the baricenters of the clusters) will be
  used. This method will employ an algorithm
  (Mathieuson formula) that takes in account more
  accurately of the charge distribution in the
  pads.
• -The systematic errors in the particle entrance
  angles in the RICH will be lowered using a new
  algorithm. The particle entrance angles will be
  assumed as those that make the Cherenkov photon
  angle distribution variance minimum (so far
  particle entrance angles in the RICH are given by
  the tracking chambers).
• An accurate check on RICH parameters (above all
  the radiator refraction index) will be performed.
• -The RICH analysis code is still slow. Some
  improvements has to be done to make it faster.
• - Correction for temperature variation (n(?))

28
Conclusions

• RICH detector excellent Kaon Identification
  and clean Kaon signal over a large p and p
  background for p2 GeV/c (5 ?)
• limited speed ( 1KHz) --gt electronic upgrade (
  2KHz) ongoing
• separation _at_2.4 GeV/c 3.9 sigma (extrapolated)
• possible improvement(s)
• Changing radiator and proximity gap
• 5sigma _at_2.4 geV/c
• 3 Sigma at 3 GeV/c
• Dual radiator for further improvement ??