Layer0 Status Andrei Nomerotski 2/3/2005

1
Layer0 Status Andrei Nomerotski 2/3/2005

• Outline
• Introduction
• Module Production
• Assembly of Layer0
• Electronics
• System tests
• Schedule
• Mechanical and Installation issues will be
  covered in Bill Coopers talk

2
Overview

• Layer0 will improve impact parameter resolution
  and mitigate degradation of performance caused by
  radiation damage to the SMT inner layers after
  3-4 fb-1
• Important for tracking b-tagging at high
  luminosity and for Bs mixing measurement
• Layer0 uses RD and experience of the DZero Run2B
  Silicon Upgrade
• Many components prototyped in advance
• Contributing institutions Brown,
  CINVESTAV,Fermilab, Fresno, UIC, U.Indiana,
  U.Kansas, Kansas State, Louisiana Tech,
  U.Mississippi, MSU, Moscow State, NorthWestern,
  Rice, Stony Brook, U.Washington, Zurich

3
Layout of Layer0

• 6-fold symmetry, total 48 modules, 96 SVX4
  readout chips
• 8 module types different in sensor and analog
  cable length

Inside SMT
4
Run2B Silicon Closeout

• Run2B closeout is complete
• Results on L2-5 stave and grounding RD on Layer0
  have been presented at several instrumentation
  conferences and published in proceedings.
• NIM paper on the grounding RD submitted
• Though present Layer0 has a different design, the
  original Run2B Layer 0 allowed to prototype many
  aspects of the new design
• CF support structure
• Grounding/noise issues
• Assembly issues

Original Run2B support structure
5
Layer 0 Module Concept

• Sensor connected to hybrid with double-deck
  analog cable

analog cable
sensor
hybrid
6
Module Production

• We are in production since mid January 2005
• Had PRR in Nov 2004 very useful, implemented
  many recommendations of the committee
• 8 types of modules, 6 of each type
• Total 48 needed
• Have parts for 100 modules, plan to build 64
  modules to select best 48
• During Sept-Dec 2004 built 14 preproduction
  modules
• Most of them used for system tests
• Will discuss below
• Production components
• Module assembly
• Status and Production rate
• Module QC

7
Sensors

• Pitch 71/81 micron, intermediate strip, 70/120 mm
  long four types
• Ordered from HM 120 sensors (4 x 30) in March
  2004
• Received 112 in July and 8 in November 2004
• All IV, CV tested at Fermilab
• 3 out of spec due to large DI/DV - used for
  prototypes
• Only one(!) bad strip pinhole on one of last 8
  sensors
• Out of 120 x 256 30720 strips
• spec 1
• Four sensors fully probed at Stony Brook
• Results consistent with HM
• Four sensors and test structures irradiated at
  KSU
• Measured at Stony Brook
• Built two modules, see later

8
Pitch Adapter

• Two types of Pitch Adapter
• Ordered in June 2004, received in Oct 2004
  (Seigert)
• Satisfactory bond strength (6.8/- 0.8 g) and
  etching quality
• Before unsuccessfully tried two other vendors
• As backup ordered pitch adapters also from
  Advanced Thin Film Products
• Provided samples with very good pull strengths
• Received and evaluating first batches
• Building a module with this PA

9
Analog Cable

• 8 types of cables required (4 pairs different in
  length)
• 128 signal lines per cable
• Max length 34 cm
• Ordered in March 2004, delivered in July 2004
  (Dyconex)
• Total 200 cables
• Testing visual inspection no bad channels

10
Hybrid

• 48 hybrids required
• Had 267 tested SVX4.2B chips
• Received 119 hybrids in Sept 2004 (Amitron)
• 70 good, 49 out of spec
• out of spec cutoff edge off by 200 micron,
  can be used with modified fixtures
• Mechanical tests at Fermilab and KU
• Flatness lt 80 micron, thickness lt 800 micron
  (both better than spec)

• Assembled 101 hybrids at NXGen and 10 at
  Fermilab
• Yield (before burn-in) 90
• 7 hybrids left unstuffed, 1 broken
• gt50 of hybrids are burnt-in (KU)

11
Module Assembly Overview

1. Sensor and spacer are glued to kapton wrap-around
   circuit Spacers are glued to top cable
2. Bottom cable is glued to hybrid and pitch adapter
3. SVX chip is wire bonded to the bottom cable
4. Top cable is glued to bottom cable. Wrapping of
   sensor is completed
5. Pitch adapter is glued to the sensor
6. Cables are wire bonded to the pitch adapter and
   SVX chip
7. Sensor is bonded to pitch adapter

Day 1
Day 2
Day 3
Day 4
12
Module Assembly Fixtures
Stage 4

• A number of fixtures is required for all steps
• Typically have double quantity of fixtures
• Example below Gluing of analog cable

pitch adapter
hybrid
Stage 4
analog cable
Bond profiles for analog cable
13
Assembled Layer0 Module
sensor
pitch adapter
SVX4
analog cable
hybrid
14
Production Rate and Module QC

• Current status 10 production modules assembled
  since Jan 13
• Current rate 4 modules per week
• Have all fixtures to boost production to 8
  modules per week
• Module QC
• Mechanical inspections
• I-V curve
• Burn-in 72 hours with 200V bias
• Encapsulation of HV and GND bonds
• Another short burn-in before storage
• Results analyzed and put into the module database
• So far three modules with 0 defects, two with 1
  defect, one with many, others still need to be
  tested
• Additional module testing (done for a few
  modules)
• Several modules underwent long-term tests in
  April-July 2004
• Built and tested two modules with irradiated
  sensors (10 fb-1 equivalent)
• Thermocycling
• Laser tests

15
V-I plots

• Range 0-300V, typically current 50-150 nA,
  correlates well with sensor current

16
Examples of plots from hybrid burn-ins pedestals
L0 - 202
17
Tests of Irradiated Modules Shot Noise

• Good opportunity to check that we understand
  irradiated modules
• Normally stored in the freezer - warmed up to
  increase current
• Current vs. Temperature dependence is in
  agreement with expectations
• Minimal increase of noise wrt non-irradiated
  module at operational T
• Observed rising contribution from shot noise with
  rising T
• Shot noise scales as sqrt(Integration time x
  Current)
• 1400 e for 396 ns integration time at T10oC
  agrees with calculations

18
Support Structure Assembly Fixture
Collar North
North Modules Holders

CF Support Structure
South Modules Holders
Collar South
Support Mount North
Base Plate 90x20
Middle Support
CF Support Structure
Support Mount South
Rotation Control
The fixture will be mounted on CMM Table Lab C
at SiDET
19
Sensor Hybrid Holders

• Each module after mounting is tested electrically

20
Layer0 Electronics

• Layer0 will use 48 readout channels now used by
  Outer H-disks
• Readout chain needs to be interfaced to Run2A SMT
  electronics
• New wrt Run2A Hybrid-Jumper Cable-Junction
  Card-Twisted Pair Cable-Adapter Card
• Adapter Card isolates Layer0 ground and detector
  ground
• Needed because South is shorted to North by the
  support structure
• SVX4 needs to coexist with SVX2
• Status of Cables
• Digital Jumper Cables (KSU KU)
• Production cables ordered from Compunetics in Jan
  2005
• Have reduced thickness to 0.15 mm
• Twisted Pair Cable (KU) ready

21
Junction Card

• Two channels per card
• Cable routing and dressing prototyped with mockup
• Design finalized, will be ordered in Feb

22
Adapter Card

• Four channel per card
• Ground isolation tests done in Sept-Oct 2004
• Had good results with and without isolation
• Plan to reproduce them with the prototype support
  structure
• Design finalized, will be ordered in Feb

23
System Tests

• Effects of possible interference with support
  structure and between modules addressed with
  setups at Sidet
• Readout of modules installed on the support
  structure
• Readout of large number of channels, in the limit
  full Layer0
• Testing at conditions similar to operations
  addressed with setups at DZero
• Use real DAQ with Trigger Framework
• Preparation and exercising of online software
• Cooling system / Thermocycling

24
Prototype Support Structure

• CF support structure for Layer0 implements new
  grounding approach laminated ground mesh
  covering all surface crucial design feature
  ensuring low inductance path for GND and hence
  low noise
• Prototype is electrically and mechanically the
  same as final structure
• Have 8 L0 modules mounted on the prototype
  support structure since Nov 2004
• Simultaneous readout through simplified chain

modules installed on the support structure
25
Results with Prototype Structure

• Low total and random noise (no Faraday cage)
• the grounding scheme works well
• Observe pedestal shapes at fast SVX4 settings
• Caused by digital-to-analog couplings between
  hybrid and analog cable below the hybrid

Noise 1500 e S(MIP) / N 151 Coherent
noise very small
Noise for 4 modules
hybrid
cable
sensor
26
Pedestal Shapes

• Pedestal peak-to-peak difference
• At fast preamp settings 6-7 ADC counts
• At slow settings 2 ADC counts - acceptable
• Other remedies
• It looks that most of pickup comes from clock
  signals will have provisions to regulate clock
  amplitude at AC
• Have a spacer between hybrid and analog cable
• 0.2 mm spacer significantly reduces effect
• More testing in progress

Pedestals for 4 modules
fast preamp settings
10 ADC counts
slow preamp settings
10 ADC counts
27
Plans for system tests at Sidet

• Feb 2005 Read out 10 modules on the support
  structure through Junction Card-Twisted Pair
  Cable-Adapter Card
• Use prototype JumperCable-JC-TPC-AC
• Mar 2005 Read out the 10 modules though full
  chain (above Sequencer VRB)
• May 2005 Expand full readout to 48 channels
• June 2005 Read out full Layer0
• Use production JC-JC-TPC-AC
• Also thermocycle prototype support structure
  with 10 modules (Feb 2005)

28
System tests at DZero

• Installed two L0 modules, two hybrids, WIENER LV
  PS and power distribution during the 2004
  shutdown
• Successfully read out in special runs
• Data is being analyzed, noise looks reasonable
• However these SVX4 strings are not yet in
  operations, waiting for acceptable SEQC firmware
• Have a stand at DZero with IB, SEQ and VRB crates
  connected to Trigger Framework to debug firmware

• These SVX4 strings will be used
• To develop all slow control and monitoring online
  software
• Exercise offline software
• Measure S/N from real particles in the modules

4 ch. Adapter Card at HorseShoe
29
Other Integration Issues

• Some remapping choreography of Layer0/Inner
  H-disks/Outer H-disk readout and HV channels will
  be required during 2005 shutdown
• Isolated LV Supply (WIENER)
• Provides SVX4 power
• PS and heavy power cabling installed during 2004
  shutdown
• Power distribution to Adapter Cards will be
  installed during 2005 shutdown
• High Voltage Supplies
• Located outside of Collision Hall (in MCH2)
• Install more channels of Bira supplies
• Install new fanout box rated to 700 V
• Integrate new components to
• Slow Controls
• Interlock
• Online GUIs

30
Schedule

• Hybrid production essentially done (1/25/2005)
• Junction/Adapter Cards production ends on
  3/18/2005
• Module production ends on 4/22/2005
• Start module installation 4/1/2005
• Layer0 ready for system tests at Sidet on
  5/11/2005
• Layer0 ready to go to DZero on 7/19/2005
• Main schedule risks are in
• Design and preparation of assembly fixturing
• Full system tests

31
Summary

• Good progress on all fronts
• All parts in hand - started module production in
  January 2005
• Layer0 assembly to start in April 2005
• Detector should be ready to go to DZero in July
  2005
• Paying attention to system tests so far all
  looks under control