D

1
DØ Silicon Hookup and Commissioning
Directors Review on Installation and
Commissioning
Marcel Demarteau Fermilab
Outline

• Goals and timeline
• Overview of Silicon Read out
• Stand alone commissioning
• Cosmic Ray Stand
• Barrel / Disk read-out
• Commissioning in Collision Hall

2
Goals

• Ensure data integrity at the level of 1014 -
  1015, i.e allow for loss of 1 bit in 1014 - 1015
  bits transferred
• Provide adequate operating margin for device
• Ensure adequate diagnostics tools and controls
• Exercise control logic
• Exercise Data Acquisition and filtering scheme
• Exercise offline software

Timeline

• Commission read out electronics in phases on the
  bench without detectors attached
• Integrate final components as soon as available
• Read out small scale setup in final configuration
• Enlarge scope of read out

3
Current Status

• Complete read out system installed in cleanroom
  in LabC
• Final version of all read out components, except
  Interface Board
• System run in checksum mode in Level 3 node
• SVX chips downloaded to read out set data value
• All channels read out data constant
• Calculate checksum (sum of all bits) in VBD
• Data transferred to L3 node checksum compared
• If checksum different read in and analyze even

4
A) Ran with multiple VME Read out Buffer (VRB)
boards, without low mass cable and long
transmission line, with direct regeneration of
signals at hdi, 3-chip hdis only System
equivalent of 22,000 channels Data transfer on
VME at 12 Mbytes/s

• Very low noise on data transfer backplane
• Proper handshaking between read out elements
• Data integrity gt 1013 ( 1
  non-fatal error every 30 minutes)

B) Ran with single chain (hdi) in close to final
configuration, including low mass cable, 30
3M/50 conductor cable. Optimized impedance
matching for data and control lines on sending
and receiving ends

• Successfully operated all hdi types
• Mapped out operating margin for all hdi types
• Data integrity gt 1013 for all hdi types with
  comfortable operating margin

C) Written raw data to disk, unpacked and
analyzed offline within L3 framework and online
monitoring program (Examine)
5
DØ - D7
Data bus lines
Clock strobe Read out at 53 MHz
DValid
18.8 ns
Strobe channel id
Strobe data
Bus lines D0/D7 WCLK and DVALID signal risetime
of 2-3 ns
Upper limits
7-8 ns Operating Margin
6
Schedule and Resources

• 12/05/99 Expand to 6 chains with low mass
  configuration
• 12/15/99 Implement new interface board
• 50 ? 80 conductor cable with SVX power provided
  through cable
• monitoring capabilities added

Verification of data acquisition and data
integrity will be ongoing effort
7
Cosmic Ray Setup

• Install up to 12 ladders in spare Be support
  structure and exercise full read out system, DAQ,
  L3 and on/offline software, controls and
  monitoring
• First step in trying to operate full barrel disk

Array of 4 scintillator counters form external
trigger Momentum analyzing steel, pT? gt 2 GeV/c
Data Rate 1 cosmic/min
Ladder installation in progress
8
Setup

• 6 ladders read out through low mass cable in
  final configuration
• 6 ladders read out through high mass cable
  Total of 9216 channels
• Barrels will be cooled and interlocked
• First cosmic ray recorded 12/20/99

Ingredients

• Detector Control and Monitoring
• Secondary Data Path
• Data Acquisition
• Online data monitoring (Examine)

9
Detector Controls and Monitoring
All slow controls and monitoring uses 1553
protocol 1553 interface to Si detector only
available with new Int.Brd.
1553
Detector
Controls
Readout Crate
Controls Crate
Ethernet
EPICS Clients SVX Download Low Voltage
High Voltage 1553 Devices Rack Monitor
Control Room PCs
UNIX Hosts
Available

• Download of Sequencer and SVX chips, not
  optimized
• SVX and bias Voltage control done, need
  implementation

In progress

• 1553 Monitoring devices, currents, temperatures
• Cooling monitoring and interlock
• Oracle database design population and interface

10
Secondary Data Path
Intended for calibration processes and
monitoring For Silicon calculation of pedestals
and gains
L3 Node
Available

• Front-End software finished as stand alone

In Progress

• Porting to cosmic ray setup Jan 00
• Exercise system Feb 00
• Database interface first pass March 00
• Database retrieval April 00

11
Run Configuration

• Module addressing and read out
• chip id, pedestal, gain, ladder barrel, Temp
  address, I address

Readout Configuration
Monitoring Devices
Geometry parameters
Control Parameters
Calibration Constants
12
Data Acquisition

• All pieces to exercise full data acquisition
  chain are in place
• Read out to L3 system extensively tested
• Online Examine SMT analysis code tested on raw
  data
• First pass at full read out with 12 ladder test
  Dec 99/Jan 00

• Will slowly ramp up to enlarge read out and
  complexity of tools
• L3 Tools data unpacking, hit finding, track
  reconstruction

13
Timeline for development

• Dec 99
• 6-12 ladder setup and test controls and
  monitoring
• Dec 99 - Jan 00
• Cosmic Ray data taking online monitoring with
  Examine
• Feb 00
• First pass at L3 filtering and alignment
• First pass at secondary data path
• Mar 00
• Database integration

14
Timeline
So far discussed development on small scale
system (12 ladders) and commissioning of read out
system. Move to full subsystem in March 00.

• March 15/00
• Hook up of full H-disk, 48 detectors, 36864
  channels exercise all components
• April 1/00
• Hook up barrel, 72 detectors
• April 15/00
• Hook up barrel and F disk, 96 detectors (10)
• May 1/00
• 15 more detector assemblies to go (5 barrel/disk,
  6 F disks, 13 H disks)
• one week for read out and debugging per assembly
• Done Aug 15/00
• Note, doubling the time per assembly (23
  assemblies) would give completion date of Oct
  7/00. Still adequate (7 H disk assemblies can go
  last)

No assembly (barrel or disk) leaves the silicon
facility without having been read out.
15
Meanwhile in the Pit
SVX EMULATOR
Monitoring Control
3/6/8/9 Chips
3M
Low Mass
IB
Optical Link 1Gb/s
Sensor
NRZ / CLK
1 5 5 3
VBD
VRB
VRB
VRB
VRB
68k
VME
VRB Controller
L3
HOST
Secondary Datapath
Examine
Concurrent with work at the silicon facility, a
complete shake down of the read out system will
take place at the assembly hall of final read out
configuration downstream from the interface
board, including online data checking (Examine).
See M. Johnsons presentation
16
Barrel and disk hookup at DAB

• Hookup of cooling system immediately
• Run coolant above dewpoint, so no damage to
  detectors. Procedure verified at silicon facility
  
• Starting 9/26/00, 5 weeks allocated for check out
  of silicon detector
• 1.5 weeks allocated for cabling each detector
  end. Note, read out has HDI-enable line,
  therefore, system can be fully powered at hookup
  time, thus instant hdi checkout. Procedure should
  be developed at silicon facility during full
  barrel/disk tests
• Complete Barrel and F disk hookup in 3 weeks.
• Assuming 7days/week, 12hrs/day, this gives 20
  mins of debugging time per hdi
• Two additional weeks for debugging complete
  barrel and F disk assembly. Hookup of H disks in
  parallel.

17
Resources
18
Summary

• Read out system will be verified continuously.
  Current level of data integrity for single hdi
  would be adequate.
• Partial barrel with cosmic ray data taking first
  step to final system configuration
• All system components will be exercised at the
  silicon facility and procedures developed for DAB
  hookup and commissioning
• Many DAQ components exist already in stand alone
  form, but are not integrated.
• All full assemblies will have been read out at
  the silicon facility before installation.
  Development of procedures at facility crucial.
• Schedule is tight but feasible