Transcript DIRC Background Status Part I Updates and Background Scaling J. Va’vra and T. Hadig 11/7/2015

DIRC Background Status
Part I
Updates and
Background Scaling
J. Va’vra and T. Hadig
11/7/2015
0
DIRC main source of background: Radiative
Bhabhas striking Q4 magnet inboard flange
• Source discovered empirically
with a Geiger counter.
• Shielding erected step-by-step
slowly building empirically a
proof that things are improving.
• A final shield erected in 2001, and
the whole job completed in 2002.
• Carsten proved later, with a
Turtle program, that a major
source of DIRC background is
the radiative Bhabhas striking
Q4 inboard flange, which is too
small.
11/7/2015
1
Background Estimate and Reality
Thomas Hadig
Carsten estimated in Feb 2002:
R=13kHz/A I(HER) + 18 kHz/A I(LER)
+ 10 kHz/(10^33/cm^2/s) L
Comparison to average scaler rates in
March and April 2003:
All in all: good agreement
but some dependency on running conditions
and position in DIRC SOB.
11/7/2015
2
DIRC background projections and PMT aging
J. Va’vra, 2003 •
•
•
OUTPUT RELATIVE TO TIME ZERO
PHOTOMULTIPLIER LIFETIME - SbCs dynodes (9125) - ETL curves
•
1
0.8
•
0.6
0.4
min
DIRC PMTs may have reached
~30-40 Coulombs in some areas
up to this point
0.2
typ
After ~10 years of running ?
0
0
10
20
30
40
50
60
70
80
90 100 110 120 130 140 150 160 170 180
COULOMBS DRAWN
11/7/2015
•
Input: HER, LER and LUMI
projections of John Seeman (2003)
Independent analysis of G. Wormser
(1999), J. Va’vra (2000) and C. Hast
(2000 and 2002).
C. Hast’s 2002 analysis was done after a
massive DIRC shielding was put in
place, and therefore it predicts lower
scaler rates. Based on this analysis,
DIRC should be OK until 2006, and
beyond. (due to shielding and TDC
upgrade)
However, all this does not include the
trickle injection running conditions, and
any possible difficulty related to
running with large beam currents.
Without any shielding we would have
~1 MHz per PMT by 2007.
However, it looks like, if ETL Co.
data are correct, that the total dynode
gain will not drop by more than 40%
after ~10 years of running, given the
present shielding and the Seeman
2003 prediction.
3
•
•
•
•
DIRC Background Limitations
Before TDC upgrade (till summer 2002):
TDC deadtime of about 5% at 250kHz
After TDC upgrade (since summer 2002):
TDC deadtime estimated to be 5% at 2.5MHz
ADC deadtime estimated to be 5% at 1MHz
Small number of Out-of-order errors at high rates
fixed by changing DFB firmware
(Dominique, Chris, Philippe in summer 2003)
Tiny number of errors remain in stress tests (Chris)
Other limits:
•
Maximal number of hits shipped
currently we stop after 5000 to 6000 hits, mark event as
corrupted.
•
Readout time and volume (see Chris’ talk)
Offline: Tight timing cut, bkg has small effect on final data quality.
11/7/2015
4
However, there are still hot spots around BaBar
J. Va’vra, June 2003
• Radioactive hot spots correlate with the
regions where the beam is large, according to
Witold.
•
•
•
•
•
11/7/2015
Forward LER side (LER going in):
a) ~12m from IP near downstream collimator
b) ~25m from IP near upstream collimator
Forward HER side (HER going out):
a) ~7m from IP near Q5 exit
b) ~3-4m from IP near coupling of backward
raft to the central tube
Backward LER side (LER going out):
a) ~14m from IP near QDCX2 magnet
b) ~27m from IP near QFCX1 magnet
c) ~7m from IP at the exit side of Q5
d) ~3-4m from IP near coupling of backward
raft to the central tube
Backward HER side (HER going in): quiet
A suggestion:
Move both LER collimators from the
forward side to the downstream side of
Babar.
5
Energy spectrum of a general background
J. Va’vra, November 2000
Na22 source calibration:
CsI detector with a PMT:
• A typical background around
PEP-II consists of a few MeV
photons.
•
PEP-II background next to DIRC:
•
•
11/7/2015
The spectrum shape is independent
of how the CsI detectors is oriented
relative to DIRC, or if there is a 1”
thick lead shutter in or out. This
means the whole cave is filled with
this type of background.
Hard to shield - one needs a lot of
lead. This was successfully done
over past few years, which enabled
DIRC to operate without a glitch as
far as the background.
The detector was calibrated with a
Na22 source (0.511 and 1.27 MeV
gammas).
6
New updates in BaBar background display
• Increase readout
frequency in Run 4:
CsI: 0.5 sec (was every 2 sec)
Quartz: 0.5 sec (was every 2 sec)
DIRC scalers: 2 sec (was 10 sec)
DIRC HV PS current: 0.5 sec
PMT outside DIRC: 0.5 sec
RPC: 2 sec (was 10 sec)
EMC diodes: 0.5 sec
Diamond detectors: 1 sec (was 2.5 sec)
SVT diodes: 2.5 sec (was 2.5 sec)
•
•
Run 4: Based on a large radioactivity
measurements with a hand-held Geiger
counter, two CsI detectors (CsI 29 & 30) were
placed on the backward side on the outgoing
LER beam.
CsI #29 is only ~10 meters from BaBar. We
may want to put some lead shielding to block
this source.
11/7/2015
•
•
•
Two CsI crystals replaced on LER
in Forward direction.
Three crystals recalibrated.
Two new CsI crystals placed on
LER exiting from Babar near the
very radioactive hot spots.
7
Geometry of some detectors used in this study
(some changes after 2003 run)
Amp.18,VSAM 27
(next to Q2)
Q5
Q4
HER
20
LER
Amp.21,VSAM 30
(outboard Q5)
Amp.22,VSAM 31, 3043
(Downstream collimator)
2
19
3
HER
Amp.17,VSAM 26
(downstream of Q1)
17
1
Q1
Q2
Amp.19,VSAM 28
(inboard Q4)
Amp.20,VSAM 29
{outboard Q4)
14
IP
4
LER
Amp.24,2-nd VSAM 1, 3075
(Upstream collimator)
Amp.23,2-nd VSAM 0, 3054
(downstream of a bend)
Amp.12,VSAM 21
Amp.11,VSAM 20 (Luminosit y chamber
(outboard Q5), 7044 downstream), 7043
Q1
5
Q2
6
Q5
8
9
7
IP
LER
Amp.9,VSAM 18
(inboard Q4)
Amp.8,VSAM 17
(inboard Q4)
10
11
Amp.7,VSAM 16
(inboard Q4)
Amp.6,VSAM 15
(near Q2)
11/7/2015
14
13
HER
Q4
Amp.5,VSAM 14
(upsream of bellows)
12
Amp.13,VSAM 22
(HER collimator
downstream), 7042
Amp.10,VSAM 19
(between Q4&Q5), 7045
(East)
Amp.14,VSAM 23
(HER collimator
upstream), 7042
8
Run 4 - how we are doing ?
• A huge background in
Quartz_HER and new CsI
#29 and #30 in the outgoing
LER from BaBar. A real effect
! May want to put some
shielding next to entrance to
QDCX2 magnet on LER.
11/7/2015
9
Conclusion
• Background estimate still applicable.
• … will this be true for trickle injection ?
• If projection is valid,
no problem for DIRC till 2010.
• More hot spots found, need to add shielding.
• Improved monitoring with new sensors and
faster readout.
• … more in second talk … stay tuned !
11/7/2015
10