Transcript Diapositive 1 - indico in2p3

Optical components: PMT Blocks
Tile Upgrade Workshop (CERN- February 2008, 8 and 9)
François Vazeille
 Two main possible effects must be considered: particle rates and/or radiations.
Iron
µ metal
3 in 1
Plastic
support
PMT
and
Divider
Mixer
Small pieces
 3 components are considered here:
- Light mixers.
- PMTs + Dividers (since they were optimized with PMTs).
and not the 3in1 cards to be reviewed in front-end electronics.
 Passive components,
such as the Plastic molded supports (NORYL) are not affected by radiations.
1
Light Mixers
 Present LHC design
• Produced (10 000) and tested in the Optical laboratory
of the Faculty of Mathematics and Physics, Charles University in Prague.
• Shape and dimensions + drawing optimized by Clermont-Ferrand:
- Square cross section + length
 The best light uniformity on PMT photocathode.
- Length
 The best PMT location inside the magnetic shielding.
- Final drawing
 Special corners for the assembly inside the Plastic NORYL support.
• Material: Extruded acrylic PMMA (Polymethyl Metacrylate)
PLEXIGLASS XT 2270 clear-transparent (92%)
with highly UV-absorbing material
 To kill Cerenkov effect of direct particles.
2
 Do we need a new SLHC design?
• Drawing: No modification … since same magnetic field and same PMTs.
• Material
1. Clear-transparency level before starting SLHC?
- Natural ageing?
- Ageing due to the light effect (Natural light, Fiber light)?
- Ageing due to radiation?
 Should be measured on some Light mixers after some LHC years.
- Light mixers taken in the worst positions.
- Needs access to the corresponding Drawers.
2. UV- dopant amount before starting LHC?
as above.
3. SLHC effects: increase of above effects
- Transparency decreased
 Light yield decrease expected.
- UV dopant decreased + 10 times more direct particles
 More Cerenkov effect expected.
BUT HOW MUCH?
3
• Next actions before taking decisions
- Measures after some LHC years: transparency and UV-dopant efficiency.
- Bibliography works about:
- Natural ageing.
- Radiation effects on PLEXIGLASS.
- UV-dopant loss with time and radiation + direct particles.
- To complete the bibliography study if there are doubts or missing information
by experimental tests:
• Intense light bath.
• Radiation tests.
Comment:
Remaking the Light mixers and changing them should a very big job!
4
PMTs
 Present LHC design
• Produced (10 140) and tested in 7 Institutes
(Clermont-Ferrand, Dubna, Lisbon, Pisa, Urbana, UTA, Valencia)
• Design optimized by Clermont-Ferrand + other Institutes (Pisa, Valencia…)
 7 years collaboration with Hamamatsu: PMT R7877  8 stages
Metal dynode structure with a metal package.
- Compactness and stray magnetic field effects.
- Gain (~105) and linearity (over 16 bits, non-linearity < 2% at the upper range).
- Dark current…
 Do we need a new SLHC design?
• Very likely no.
• Information from long term studies
- TJNAF (USA): same PMTs working and tested before/after at Clermont-Fd
and used again after several working years in true conditions.
- Pisa results (see next talk.
• A maintenance must be foreseen: replacement of bad PMTs.
5
Dividers
 Present LHC design
• Produced (11 000) and tested by Clermont-Fd
 45 mm
• Design optimized by Clermont-Ferrand + Chicago (2 ground levels):
- Optimization of the PMT working:
• Gain and linearity.
• Impact of additional continuous current
(MB events)
 See the performances (next slide).
- Minimum of cables:
direct connections with PMT and 3in1 card.
- No sensitivity to radiations: only passive components.
6
• More on design and performances
- Choice of resistors of voltage divider
 driven by the current flow.
- Applied rules for a pure resistor Divider:
• Rule 1: Pile-up events
100 times the mean anode current
Worst case of pile-up events: ~2 µA
 voltage Divider current of 200 µA.
Warning: Current simulations gave 576 nA as maximum current for Layer 1
 There would be an additional safety factor of about 4.
• Rule 2: Intense pulses
Storage capacitors in the last stages
to balance the voltage drop due to large current flow.
- Performances:
from the study of a set of 250 PMTs over the 10 140 requested PMTs.
Anode current
100 nA
2 µA
5 µA
Gain variation
~ 0%
> 1%
> 2%
7
• Do we need a new SLHC design?
- The maximum current will go from 2 to 20 µA
 The Rule 1 would be no longer satisfied
only a ratio equal to 10 between the current flow and the anode current
or 40 if we take the simulation value of about 500 nA.
 Bigger sensitivity to MB events and worst linearity behavior,
in particular for very high energy events (the best expected events!).
- That should affect only a part of the cells: close to the beam.
- Using the Laser should be a way to recover the linearity
- HOW MANY CELLS WOULD BE AFFECTED?
- SHOULD WE ACCEPT SOME NON-LINEARITY?
- SHOULD WE CHANGE EVERY DIVIDER OR ONLY A FRACTION OF THEM?
• Next actions before taking decisions
- New simulations (or new analyses of existing simulations)
in order to know better the MB currents in every cell.
- To look at deeper the R&D results of tested PMT/Dividers
+ Test Bench results.
- To decide what is acceptable from the Physics goals, knowing that the Laser
8
is a useful tool to solve the problem.
- If requested, is it possible to make an upgrade?
YES from the design point of view
Already made at Clermont-Ferrand by Michel Crouau for LHCb,
with several solutions considered to increase the gain on the last dynodes:
- Booster Divider (Additional supplies on last stages).
- Cockcroft-Walton Divider (Oscillator supplied by a LV).
- Transistor Divider (Transistors on last stages).
Chosen solutions: less components are added.
Divider with
2 Transistors
transistor base
gain DC stability versus anode current
DC gain versus anode current
Resistor
(resistor Divider
base)
4
gain variation (%)
35
30
gain variation (%)
25
20
15
10
5
3
2
1
0
0,1
-5 1
10
Ian(microA)
100
1000
anode
0
1
10
Ian(microA)
100
1000
anode
 500 µA flowing in the Divider is sufficient to deal with a 100 µA anode current.
9
 Chosen Divider in the LHCb experiment
for the 100 Hamamatsu multi-anode PMTs.
Comment:
Remaking the Dividers and changing them should be a very big job!
10
First conclusions
Component
Light
Mixer
PMT
Divider
Is this a
component which
MUST be
upgraded?
If yes, why?
Likely no
Perhaps the
material
but not the drawing
No
No if a
compromise is
accepted,
otherwise yes
- Increased MB
current
- Working/Linearity
affected for:
• First cells
• High energy jets
Program?
If no, is this a
component
that, with high
probability, will
not need
upgrading?
- Bibliography
- Light tests
- Tests after
some LHC years
The proposed
program will
lead to a
certitude
- See long term
tests
- Detect bad
ones
Maintenance
- Deep analysis
of Tilecal R&D
- New look of MB
simulations
- Physical goals
- New design
possible
A compromise in
between
decreased
performances
on some cells
and possible
corrections
(Laser)
11