SSD Update

1
SSD Update
Howard Matis
2
Outline

• What is the SSD?
• Why we need it?
• Some tracking results
• Current Status
• How to upgrade?
• Manpower needed to run the detector at its
  potential

3
The SSD exists!!

• The SSD is thin
• 1 radiation length - double sided Si
• The SSD lies at an ideal radius
• 23 cm - midway between IP and IFC
• The SSD has excellent resolution
• The SSD is too large to be replaced
• More economical to build other detectors

4
Two Layers of Tracking

• 95 mm ? 4.2 cm strips
• 300 mm thick
• Double sided
• Crossed at 35 mrad
• 20 ladders
• 67 cm long
• Air cooled
• ??? lt 1.2

5
SSD Facilitates Tracking from the TPC to IST

• ?xy 30 µm
• ?z 700 µm
• TPC pointing resolution at the SSD is 1 mm
• SSD pointing at the IST is 300 mm

Inside the IFC
6
SSD has been used in STAR analysis

• Needed for SVT tracking
• SSD is essential for SVT analysis

7
Data - Cu Cu 62 GeV Impact Parameter
TPC
TPC
SSD
SSD
SVT 1,2,3
SVT 1,2,3
1/p
1/p

• Resolution here includes vertex and hit
  resolutions. Real values 20 smaller.
• Huge difference in pointing between SVT and
  everything else
• At infinite momentum limit is 150 µm in XY and
  80 µm in Z (vertex resolution)
• At 1 GeV/c, it is 220 µm in xy and 150 µm in z
• z is our good (not shifting) coordinate!

8
Situation for Upcoming Run

• Ran detector in December
• System ran and stable
• Obtained pedestals and noise
• 70 of the detector good for physics - NOW
• Should be able to recover rest to 95
• L2 aborts acceptable for rate lt30 Hz
• SSD not useful for p-p running
• Jerome Baudot will visit BNL and instruct new
  people when ready to run
• Need beam and full detector

9
SSD Electronics and the Future
Module level - 768 strips per side ( P
and N side ) - Analog signal of 1.5 ?s -
50 mV/MIP - Readout speed 3 MHz
SSD level - 4 Readout Boards (RB)
located on the TPC wheels - Each RB
interfaces 10 ladders with trigger / daq /
slowcontrol - Readout speed 30 MHz
Ladder level - 2 Connexion boards ( 1
per side ) - 2 ADC boards ( 1
per side )
10
Current Readout Speed

• Front end Digitization
• FEE runs at 3 MHz
• 96 chips/ladder side with 128 channels/chip
• 4.1 ms to digitize a ladder side
• Transfer to DAQ via RDO
• Readout at 30 MHz
• 1 RDO/10 ladder sides
• 4.1 ms for readout
• 3 ms for 0s to match TPC DAQ format
• 7 ms total

RDO Board
11
How to Upgrade

• For 1.2 ms readout time
• Digitize at 10 MHz
• Have a RDO for 3 ladder sides - ?3
• No zeros to fill the ASICS ?2
• Opportunity to upgrade firmware
• Solve L2 abort problem
• Estimated cost (mostly to produce new RDOs )
  100 k
• Must verify power distribution and cooling
• Need to test concept

12
Lessons Learned from STAR Operations

• Detector must be tested with full system before
  each run at BNL
• Experts must be available to assist ASAP
• Come to BNL if necessary
• Train shift crews
• Testing slots are limited
• Schedule access is not fixed
• Scheduled accesses should stick to the schedule
• SSD must have a mother at BNL making sure
  detector is being used
• Must be regular maintenance of the detector

13
Need a team to succeed

• Physicists
• Subsystem leader (1/4 time)
• Software coordinator (1/2 time)
• Calibration and detector expert (1/2 time and
  resident _at_ BNL)
• Subsystem Expert (need at least 2 _at_ 1/4 time)
• Slow controls (1/8 time)
• Advice from present experts
• People who want to use it for physics
• Engineers
• French Support (need shops and some travel to
  BNL)
• Mechanical and electrical tech support (several
  weeks during shutdowns)
• Electrical Engineer (when problems occur)

14
Summary

• SSD provides valuable tracking for current
  detector and the Inner Tracker
• Low maintenance calibration
• Should be supported by STAR
• Can be maintained
• Should work for this Au-Au run
• STAR must provide sufficient manpower for the SSD
  to succeed

15
The End
16
Backup
17
Pedestal
Noise
Ladders 1 - 10
18
Pedestal
Noise
Ladders 11 - 20
19
Maintenance Needed Next Summer Shutdown

• Solve the problem with one of the LV power supply
  boards (ladder 5)
• Check all cooling pipes. Understand HV trips
  (cooling problem?)
• Solved if SVT removed
• One ladder side (7P) is silent
• Fix ADC Board
• Might need to replace ladder
• Check optical fibers
• Need specialized equipment. At BNL?
• Fix L2 abort problem
• Need 1 month of a SSD Engineer and Tonko

20
Runtime hardware support

• Requires less effort than the TPC or EMC
• Needs person at BNL watching SSD
• Advocating and assuring for its inclusion in the
  run
• Basic knowledge of system
• Answer simple questions from shift crews
• At least two people who can provide
• Backup
• Advise
• Help in access (need fixed access times from
  RHIC)
• Real expert who can consult
• FPGA Programming support
• Should only need this help during shutdowns

21
Shutdown hardware support

• Fix known problems
• Tend to routine maintenance
• Assist if changes needed to SVT cone area
• Routine work on infrastructure
• People
• Team of physicists to work on the detector
• Electrical and hardware techs to do
  infrastructure maintenance

22
Software support

• One person doing daily calibration and checks of
  the performance
• Jerome Lauret requires a software detector leader
• responsible for making sure SSD is ready when
  production starts
• People who need SSD for their physics analysis

23
(No Transcript)
24
(No Transcript)