Module Production ___________ Report on the Catania Status

1
Module Production___________ Report on
theCatania Status
Salvatore Costa Università di Catania and INFN
Sezione di Catania
2
Outline

1. Module accounting
2. Bonding
3. Production rate
4. Bonding experience
5. Pull Test results
6. Failure analysis
7. ARCS LT Tests
8. ARCS test analysis
9. LT test analysis
10. Usage of central DB
11. Logistics transportation of modules

3
Module Accounting

Module Type? Total IB34 IB12phi
Received from Bari Gantry 61 28 33
Bonded 58 (51 in DB) 27 30
ARCS tested 56 (47 in DB) 27 29
OK 53 26 27
bad 3 1 2
LT tested 49 (0 in DB) 27 22
OK 49 27 22
bad 0 0 0
4
Production Rate

• Initially slow due to bonding commissioning (we
  lacked long-standing experience with previous
  bonding jobs), but now on schedule. To illustrate
  our progress
• First box received, with 8 Modules, was completed
  (bonded ARCS tested) 23 days after arrival
• Last 2 boxes, with 9 Modules each, have both been
  completed within 2 working days after arrival
• Current bonding capability is 4-5 Modules/day
• Long Term keeps up with no difficulty with
  current incoming flux of 1 9-Module box / week

5
Bonding Experience

• Became full-blown Bonding Center on 5 Dec 2003
• Previously designated Bond Repair Center
• Had passed official CMS Qualification Tests
  (bond test 1 2)
• Bonding experience with previous experiments
  none
• Equipment
• 2 Hughes 2470-II bonders
• 1991, 1987
• automatic but no Pattern Recognition or
• bond foot deformation feedback
• 1 ATV GM-100 pull tester
• manual
• Machine downtimes each time a new machine fault
  occurred that we had not seen before.
• Time spent to learn to locate the source of
  the fault, what action to take and with what
  procedures.

Higher probability of bond failures or accidents
6
Bond Pull Tests Results
IB34
IB12phi
sampling
5g
7
Bond Failure Analysis
53 Modules 34048 bonds repaired repaired repaired repaired unrepairable unrepairable unrepairable unrepairable
53 Modules 34048 bonds modules affected strips affected modules affected channels affected
Totals 44 84 563 1.65 10 19 29 0.08
Machine failures 42 80 506 1.48 7 13.2 14 0.04
Bond foot misplacement 2 3.8 9 0.026
to stick 42 80 506 4 7.6 4 0.012
Pinholes (caused by bonding?) 1 1.9 1 0.003
Operator mistakes 2 3.8 57 0.17 3 5.7 15 0.04
Mishandling 1 1.9 7 0.02 2 3.8 7 0.02
Wrong ref 1 1.9 50 0.15
Wrong par (lifts, pinholes) (1) (1.9) (384) (1.11) 1 1.9 8 0.02
8
Bond PA-Sen in TIB L12
correct
misplaced
9
ARCS Test Analysis

• All tests performed following the official TIB
  procedure
• ARCS 7.0 used for long time (7.1 installed
  operational but not yet fully tested)
• Tonys macro used initially, starting mid March
  Riccardos update (now v3.1.0)
• Flagging efficiency with Riccardos macro
  100
• Out of 53 good modules, 51 are grade A, only 2
  are grade B
• No IV problems
• A total of 55 defective strips found over 34048
  bonded strips
• 29 caused during bonding
• 26 pre-existing Sen pinholes APV-PA bond opens
  Hyb faulty chan

open
chan
10
ARCS Analysis IV
We wrote a script which queries DB to compare
measured IV with the Sensor one. We also
implemented Temperature correction. NO IV
problems up to now
IV data not available for all the sensors
11
Long Term Test Analysis

• 49 modules tested with TIB Official Scenario
  (about 72 h)
• Some modules retested with a 1-night scenario due
  to failure on the last record
• First, cold last records always recorded
• Some failures of the sw before Marcos patch on
  I2C
• 1 failure due to DAQ problems with Lt_1.22
• No failures since Lt_1.24 was installed
• Riccardos macro used for the analysis (now
  v3.1.0)
• No new defective strips appeared after LT test
• From time to time some noisy strips appear and
  disappear

12
3 Bad Modules

• In 3 Modules, all from the very last group of
  Modules received, we have observed plain failure
  of the ARCS Test after bonding (they had passed
  with flying marks the test upon arrival)
• 2 Modules (30200020035289, 303) fail Fast Tests
  Low Voltage test I(V2.50) 850 mA (max 750 mA)
• Bonds are among the best weve made, both
  mechanically and aesthetically
• No visually evident damage from possible
  mishandling
• ESD?
• 1 Module (30200020035305) fails to power up the
  Hybrid
• Damage to connector?
• Has something changed in our working environment
  or procedures?
• All these modules Hybrids are marked with Alans
  blue dot
• Deep investigation needed to understand these
  problems

13
Usage of central DB

• Bonding data always timely uploaded to TrackerDB
  since the beginning
• for all ARCS LT data, xFLAG 1.4.3 has been used
  to produce XML files 53 ARCS 49 LT
• XML files checked for consistency with root files
  and macros output
• No serious problems reported
• Only few root files need to be redone due to old
  ARCS bug in the header
• 86 of ARCS XML files already uploaded
• LT XML files not yet uploaded

14
Logistics Transportation

• No major problems with modules received from Bari
• In first 2 boxes only the styrofoam blocks on the
  sides were too far apart, so Modules (on
  transport cradles) slipped out of their slots,
  but no damage to the modules resulted. Problem
  was quickly corrected in newer boxes
• Some dirt (mainly small fabric? fibers)
• We blow a 1.5 atm air jet on Module before
  starting visual inspection
• Our storage space in clean room can hold 8 more
  boxes (72 Modules),
• After, better to start shipping to Pisa
• or we will have to figure out how to carve out
  more storage space

15
ARCS LT Test
The Catania CMS Module Production Crew
Inspection, Test on arrival, Bonding
S. Costa, N.Giudice, F.Noto, F.Raimondo,
C.Sutera, C.Tuve
S. Albergo, M. Chiorboli, M. Galanti, G. Gatto,
A.Tricomi