Status ATLAS SCT 2002

1
Status ATLAS SCT 2002

• NIKHEF Contributions to ATLAS
• Magnets discussed by Frank
• Trigger
• DCS
• Muon MDT and alignment Covered by next talk
• SCT
• Talk Outline
• SCT Progress Past Year
• NIKHEF SCT Work Past Year
• Future NIKHEF Work

2
ATLAS Detector
3
SCT Progress Barrel

• 1st Carbon Fibre Cylinder Delivered
• Mostly in spec, some difficulties achieving the
  very high precision machining requirements they
  had
• Ready to start mounting brackets
• Some delay because of last minute change in
  layout for clearances
• Modules
• UK and Japan sites qualified for production
• Nordic Group and USA have made modules but still
  not perfect
• Japan in routine production since summer 2002 and
  now speeding up
• UK started routine production
• System Test
• Many modules run simultaneously
• With good grounding and shielding, no extra noise
  and good rejection of external noise

4
Barrel System Test
5
SCT Progress - Endcap

• Modules
• Hybrids K4/K5/KB/Filters story
• Mid 2001 realised K4 had serious problems
• Started program to be sure of solving it
• 1. Investigate/understand the problem difficult
  with binary (hit-or-miss) readout
• At NIKHEF we realised it is a feedback from
  Digital Power Supply to Analogue side
• 2. Fix K4 (NIKHEF)
• Filters in digital power supply
• Worked well on NIKHEF modules, but could not
  rescue a Liverpool module
• 3. New version K5 (Freiburg)
• 4 layer ? 6 layer better AGND plane
• Now working well
• 4. Fit Barrel Hybrid on Endcap Module KB
  (Geneva)
• Worked as well as Barrel Modules (still some
  feedback)
• Cooling finally worked well, but more material
  and ?cost?
• Would have been a lot of extra engineering work
• Decided early 2002 to go ahead with K5

6
Feedback Model (S. Peeters simulation)
7
Feedback Reduction with Filter Glued on Chip

• S-curves Before Filter

After Filter
Filter Circuit glued to ABCD Chip Joop
Roverkamp/Henk Peek
8
Radiation Hardness of Endcap Modules

• June 2002
• Two Endcap Modules Irradiated
• Very noisy, out of spec
• Optimists Maybe they got overcooked?
• August 2002 Two more irradiatedbehaviour much
  better, well inside spec
• Pessimists Maybe the August ones were
  undercooked
• The Pessimists were right
• Status now
• Some chips out of spec on irradiation most OK/on
  the edge
• Are Endcap modules worse than Barrel? (I.e., is
  K5 bad?)
• Dont know any mechanism why they should be
• Hybrid power supply lines all OK after
  irradiation
• Could be statistics? Barrel were lucky?
• Fix We can raise threshold from 1 fC to 1.2 fC
• Fixes noise rate little effect on efficiency or
  tracking
• Raise bias to 450 or 500 V
• Regains full charge collectione efficiency
  Modules hold this HV

9
NIKHEF SCT Progress
Module Noise Model Simon Peeters

• In addition to feedback also has calculated the
  fundamental noise limit
• Many significant contributions
• Many input parameters, some were unknown and had
  to be measured
• End result
• Modules are 10-20 noisier than predicted
• Best quality chips are closer to prediction
• Conclusion we have a good hybrid, close to ideal
• Butstill some things not understood, e.g.
  anti-noise at low signals

Chip-0 Output Noise
10
Module Manufacture 5 Inner K5 Modules Made

• Aim was to test K5 and develop assembly
  techniques and QA
• All are in specs most important parameter within
  1 micron
• Sent to test centres
• 3 to system test
• 1 for thermal tests
• 1 for irradiation

11
Discs

• Order placed with PCI in July 2002
• First 2 Discs Delivered Last Week
• DU01 In Spec
• DU02 Not flat enough
• Needs work with PCI to see what can be done
• Cylinder and Wings contract imminent
• NIKHEF Cylinder due July 2003
• Cheap-O Prototype
• Torlon Inserts Succesful
• 300 per disc low X0
• Now glued to disc
• Next machine flat
• Send to UK for Services tests

Flatness of DU01, side 1 0.4 mm
12
Cheap-O with inserts, ready for measurement
13
Tooling to machine discs flat

• Discs 0.5 mm flat Modules need better
• For placement For Services Clearances
• We machine inserts to 0.1 mm flatness

• Stiff-back rohacell sandwich made
• Most fittings made
• Ready Feb 2003

14
Future NIKHEF Work

• NIKHEF Site Qualification
• We make five more modules in Jan/Feb
• If all OK, we are ready to receive components and
  start manufacture
• These 5 modules can go into ATLAS
• Endcap Module Production
• Officially starts April
• Hybrid delayed thoughlooks optimistic
• Discs
• Sort out flatness
• Finalise disc layout details
• Test services, inserts etc. incorporate minor
  improvements
• Start machining discs

15
Disc Assmbly

• H026
• Glue inserts
• add services

16
H023/24/25 Assembly

• Module assembly
• Modules-to-disc
• Disc-to-Cylinder
• Testing
• C3F8 Evaporative Cooling
• Whole disc readout

17
The End 2005 Ship One Endcap to CERN