Inner PMT system design group report

1
DC Collaboration meeting at Chooz, June 9-11,
2005
Inner PMT system design group report
Yuri Kamyshkov / UT
2
Other presentations of PMT design group 1. PMT
mechanical supports David Reyna and Vic
Gaurino/ANL 2. Detector background rate Yuri
Efremenko/UT 3. Should we tilt PMTs?
John LoSecco/ND 4. Uniformity measurements Bob
Svoboda/LSU 5. PMT glass radioassay
Jerry Busenitz/UA
PMT design group is working on ANL
engineering and interfacing UA radio assay,
PMT supports LSU PMT uniformity in mag. field
tests, simulations ND and KSU simulations
UT PMT procurement, testing, HV system
see also website http//hepd5s.phys.utk.edu/dcho
oz/internal/PMT_Design/PMT_design.html
3
Baseline Configuration discussed in February
534 pc 8 PMT per detector with coverage
12.3 Light yield 190 p.e. / MeV from MC
156 p.e./MeV scaling from KamLAND With 3
spares, total required pmt counts is 1,100
pc 13 PMTs are not a backup option
anymore (too expensive) Tilting PMT??? (see
John LoSecco talk) Proposal to change the PMT
support (See Dave Reyna talk) from PMT
panels to individual mounts ? Length of PMT
teflon cable RG-303 will be 20 m
4
PMT radio assay for glass only (see Jerry
Busenitz talk)
compare
compare
5
What we are learning from PMT radio assay 1.
Background activities quoted by ETL and Hamamatsu
are approx. consistent with those measured
in Alabama 2. Expected activity for ETL 9354
(ultra-low glass) is 1.02 ? 0.52 Bq per PMT 3.
Expected activity for Hamamatsu R5912 (low glass)
is 5.2 ? 0.9 Bq per PMT 4. These data as well
as other sources of background (PMT bases,
supports, walls, liquid scintillator, etc.)
are included in the simulations for PMT
comparison (See Bob Svoboda and Yuri Efremenkos
talks)
6
Gamma Singles Rates from Radioactivity of
Detector Components
Scintillaltor
Buffer
Veto
Sand
Rock
Chooz
Vessel
PMTs
Acrylic
LS
Double Chooz
Steel
? - Catcher
Rock
Veto
Buffer
Scintillaltor
Y.E., May, 2005
7
Single rates (Hz, pe/pmt/pulse, av. anode current
in nA)
(Scintillator 2.1 MeV/g, density 0.8 g/cm3,
yield 190 pe/MeV PMT gain 107)
Average anode current per pmt will be 300 nA
(in near)
and 10 nA (in far
detector) ? PMT voltage divider current can be as
low as 0.1 mA (lt 0.2 W/PMT) Low current dividers
are being tested at UT.
8
PMT base circuitry used in Borexino
9
KamLAND base circuitry
10
Hamamatsu R5912 base circuitry used in tests at
UT
RG-59 cable 0.5m with SHV connector
11
Front-end input circuitry in Borexino
12
HV-signal splitter
Following Borexino NIM A430 (1999) 435
13
SPE signal at front-end input (preliminary)
at gain 107 50 ? at anode (takes half of
the signal) with 20-m RG303 (pulse FWHM 7
ns ? 10-15 ns) 1 spe on 50 ? load ? 4 mV
(at input) for threshold ¼ spe
discrimination level gt 1 mV
14
Magnetic shielding of PMTs
PMT sensitivity to mag. field as measured by
Hamamatsu for three 8 R5912 PMTs with DC light
source
? Effect of signal reduction is up to ?25 at
0.4G in DC mode, and probably larger in the
pulse mode up to ?50 at 0.4G (ETL). The effect
will be re-checked in PMT tests at LSU. ?
Reduction is not only due to multiplication
factor but (more dangerous) due to local loss of
photocathode efficiency for given orientation of
PMT in mag. field ? Most essential screening is
needed for spherical photocathode area.
Acceptable field level is lt0.05 G in the
direction transversal to the PMT axis. ?
Conventional ?-metal individual PMT shield is not
useful ? ?-metal mesh over photocathode is
being studied (LSU) but will reduce photocathode
area by 10-20
15
Magnetic shielding of PMTs
Earth mag. field at Chooz (from
http//www.ngdc.noaa.gov/seg/geomag/jsp/struts/cal
cPointIGRF) Bx 2.004E?5 T ? 0.20 G (North)
By 0.004E?5 T ? 0.00 G (West) Bz
4.306E?5 T ? 0.43 G (Down)
Acceptable field level lt0.05 G
Magnetic field of 15-cm Fe-shield around the
detector can be several times higher
Possible solutions (a) demagnetize 15-cm Fe
slabs before installation (was not successful in
Chooz-I) (b) use 5-cm (?) Pb shield instead of
Fe together with active Earth field compensation
coils. Used in Super-K, in KamLAND, ... We
have a quote from Nuclear Lead Co., Inc. _at_
Oak Ridge, TN to provide brick construction from
5-cm thick low-background Pb for 382K per
detector (expensive, but saves 10-cm radial
space) (c) global ?-metal shield between 15-cm
Fe and Inner Veto Vessel with thickness of
3 mm insulation 1-2 mm on each side
(probably expensive)
So far there is no satisfactory baseline
solution
16
Our 8 PMT purchase options
a. ETL9354KB, England (quotes, contacts) b.
Hamamatsu R5912, Japan (quotes, contacts) c.
Photonis XP1806, France (contacts) d. Burle, USA
?
17
Comparison of 8 PMT specs
DC required anode current linear range is up to
50 mA
18

Comparison of 8 PMT specs (contd)
19
Hamamatsu PMT used in
8 inch R1408 SNO / CANADA, LSND / LANL
R5912 MILAGRO / LANL, mini-Boone /
FNAL R5912-02 AMANDA /
South-Pole 10 inch R7081-02 ICECUBE /
South-Pole R7081-20 ANTARES
(0.1km2) / Marseille 13 inch R8055 NESTOR /
Pilos/Greece 20 inch R7250 KamLAND / Japan
R3600-02 SUPER-KAMIOKANDE /
Japan
20
Proposed Oil-Proof Assembly for Hamamatsu R5912
21
Proposed Oil-Proof Assembly for Hamamatsu R5912
Same base seal is used for Hamamatsu 17 and 20
pmts in KamLAND
22
Borexino-like potting offered by ETL
ETL is aggressively working on a new epoxi-based
seal (a la Hamamatsu) that will be ready for
demonstration this summer
23
P/V for R5912 Hamamatsu
24
Hamamatsu is further improving P/V (Example of
10 Hamamatsu PMT)
25
PMT pre-test at the production factory (Example
of data that can be provided by Hamamatsu,
included in cost)
26
Hamamatsu PMT prices R5912 ASSY
- 1,100 pc including low radiation glass,
potted base, 20-m teflon cable, all shipping
cost 4 mo aro, 100-150 pcs/mo (tax can be
exempted)
Total
1.568 M
27
ETL PMT prices in GBP present
conversion rate is 1GBP1.89 US ETL9354KB
1,100 pc 8" PMTs (u-low) 1.351 M or
ETL9353KB 1,100 pc 8" PMTs (low)
1.164 M voltage dividers 1,100 pc
0.104 M encapsulation
by ETL 0.270 M
RG303 cable 1,100 pc 20-m long
0.119 M possible other
encapsulation components shopping cost (typ
4)
Total
1.918 M
The only real advantage of ETL tubes is lower
background rate ( 23 Hz for ETL vs 58-65 HZ
for Hamamatsu)
28
Far PMT System Construction Installation Plan
(assumptions and constraints)
PMT manufacturers need 4 month aro before
starting production (Hamamatsu) PMT production
rate is 100-150 per mo (Hamamatsu) PMT will be
produced with sealed bases and 20-m teflon single
cable attached Cables will have no end
connector to facilitate cables pulling PMT
will be delivered to US for acceptance tests and
then delivered to Chooz PMT supports will be
manufactured in US and delivered to Chooz PMT
and supports will be cleaned before sending to
Chooz PMT-support assembly will be made at
Chooz followed by functionality test before
the PMT installation in the Far detector For
receiving PMTs and supports, for PMT-support
assembly and funct. tests we will need a clean
room in vicinity of Chooz with total area ? 100
m2 and transportation access. PMT final
cleaning before installation 534 PMT in the
Far detector will be installed by a group of 10
US people within 1 month SHV connectors will
be installed to the cable ends Cables will be
connected to grounded HV-signal splitting boxes
Functionality tests will be made after daily
installation of 20 tubes Data base
including all testing results location and
installation records
29
June 7, 2005
30
Questions that need Collaboration consensus
Do we agree on individual PMT supports? Do
we agree on no PMT tilting? Is Hamamatsu tube
radiopurity acceptable? Clean PMT assembly
room logistic at Chooz? US PMT assembly
group (10 people for 1 month) More ideas
needed on magnetic screening Need fast
feedback on assembly schedule/logistic
31
PMT measurements stand at UT