US Tracker Project – CERN - July 17, 2003 - Incandela 1 US Production Readiness CMS Tracker Week 17/06/03 CERN J. Incandela University of California Santa

US Tracker Project – CERN - July 17, 2003 - Incandela 1

US Production ReadinessUS Production Readiness

CMS Tracker Week 17/06/03 CERN

J. Incandela

University of California

Santa Barbara


• Fermilab (FNAL)• L. Spiegel, S. Tkaczyk +4 technicians+guest scientist

• Kansas State University (KSU)• T.Bolton, W.Kahl, R.Sidwell, N.Stanton, E.V.Toerne+grad student

• University of California, Riverside (UCR)• Gail Hanson, Gabriella Pasztor, Patrick Gartung, + post-doc

• University of California, Santa Barbara (UCSB)• A. Affolder, S. Burke, C.Campagnari, A. Gupta, D. Hale, C. Hill,

J.Incandela, S. Kyre, J. Lamb, S. Levy, C. Mills, B.Patterson, S.Stromberg, R.Taylor, D. White +3 technicians + 3 undergrads

• University of Illinois, Chicago (UIC)• E. Chabalina, C. Gerber, T.Ten

• University of Kansas (KU)• P. Baringer, A. Bean, L. Christofek, D.Coppage, X. Zhao

• University of Rochester (UR)• R. Demina, R. Eusebi, E. Halkiadakis, A. Hocker, S. Korjenevski, P. Tipton

Active GroupActive Group


Manpower (Non-faculty)Manpower (Non-faculty)

Non-Faculty FTE

0

10

20

30

40

2001 2002 2003


OutlineOutline

• Production model and parts flow

• Status of each production stage Gantries and curing stations Wire bonding Module test stands Database Rod burn-in with interlocks Rod shipment

• Current Capacity

• Parts inventory

• Problems and concerns


OverviewOverview

Layers Quantity APV/mod. Pitch phi Pitch stereo Microbond Wires

SS6 TOB5,6 1,800 6 122 - 4,147,200

SS4 TOB3,4 1,450 4 183 - 2,227,200

DS - rphi TOB1,2 1,150 4 183 - 1,766,400

DS - stereo TOB1,2 1,150 4 183 183 1,766,400

5,550 9,907,200

Type

• FNAL and UCSB will have equal capabilities and capacities FNAL production line

• Gantry operational – made ~24 operational TOB modules.• 1 of 4 DAQ and 4 of 4 ARCS

UCSB Production line• Gantry operational –made ~19+8 operational TOB modules• 2 of 3 DAQ and 5 of 5 ARCS

UCR module diagnostics and repair • 0 of 1 DAQ and 0 of 1 ARC

TOB Module Summary Table


Module Production PlanModule Production Plan

• Hybrids New UCSB Task: Wirebond

&Test:• Quick Test then Wirebond pa’s

• Thermal cycle with continuous ARC test and pitch adapter pulsing

Adds 3.3 Million bond wires

Ship half to FNAL

• Frames and sensors Received by FNAL Sample checks at Rochester Ship half to UCSB

• Module production: FNAL & UCSB Gantry fabrication of 12+ modules

per day per site (average=9 at peak)• Overnight cure • Cross-check on OGP

Wirebond 12+ modules/day/site• Recent significant improvements

Fast test with ARC/LED• Simple repairs

Overnight temp cycling with readout in “Vienna box”

Full characterization with ARC/LED Store for installation in rods

• Have storage for 1400 modules!

• Diagnostics and Repairs• Extremely Pathological modules will

be pulled out of production and sent to UC Riverside.


Rod ProductionRod Production

•Rods with optohybrids Ship to both FNAL and UCSB

•Rods at US Sites Modules mounted Capacity: 2-3 SS6 rods/day

per site Long Term Testing

• 8 rods simultaneously per site for 72 hours with 3 thermal cycles per day with data taking

Final test at US site

•Rods shipped to CERN

• Currently Received first rod with prototype

optohybrids at UCSB All tooling for handling the rods

• Rotation and handling fixtures, Module handling fixture, Torque wrench and insertion tool (no extraction)

• Cable forming tools, and other specialized tools

UCSB redesigned module insertion tools to use vacuum pickups.

Expect low voltage power supplies and cables in early August.

Plan to build and test first rod in September.

1st Rod burn-in stand ready at FNAL in October


GantriesGantries

•FNAL and UCSB Production to date FNAL+UCSB 24+27=51 functional

modules• Many have been used in test beam

•Currently can handle1 tray of 3 modules per day •Peak rate per location

Up to 4 trays per day (12 modules).

• Some inspected on OGP

Identical OGP AVANT optical inspection machines at FNAL and UCSB

Identical OGP AVANT optical inspection machines at FNAL and UCSB


FNAL StatusFNAL Status

• Recently commissioned 3 of 6 new assembly plates

• Corrections vary plate to plate

• Overall sensor placements are all well within specs.

5 m is not always met.

10 m spec is not a problem.

Final rotations are all under 6 mdeg


UCSB Gantry in New Clean RoomUCSB Gantry in New Clean Room

~30 minutes to run assembly program for 3 modules Including preparation (glue, parts,

DB), <1 hour to produce 3 modules


UCSB Data on ModulesUCSB Data on Modules• 10+6 more modules built in last month (19+8 in total now)

number

-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 160.0

0.5

1.0

1.5

2.0

2.5

3.0

entries : 21.000 mean : 1.3333 rms : 7.6116 min : -13.000 max : 16.000

number

-10 -8 -6 -4 -2 0 2 4 6 8 10 12 140.00.20.40.60.81.01.21.41.61.82.0



Data on Modules (3)Data on Modules (3)

number

-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 90.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0


•Completion of 3 new assembly plates imminent

•Stereo assembly plate under design –to be ready in September

•FNAL and UCSB need low grade parts for dummy modules – sensors & frames – to qualify new plates, and to recover from down periods as they occur in production.

• Silicon to frame angles all under 11 mdeg


WirebondingWirebonding

•Installed 3rd K&S 8090 at FNAL Now operational and will be used

for subsequent FNAL modules.

•UCSB wirebonder bond head problem fixed. Meanwhile learned to fully

automate bonding of entire TOB module using K&S pattern recognition: Approaching 20’ per module including setup time.

•UCSB considering K&S 8060 Adequate for hybrids Many used ones available: 20k$

to 50k$.


The TOB Electronic Testing CycleThe TOB Electronic Testing Cycle

Quick test hybrids on ARC Gantry makes modules.

Modules test on ARC

Assemble rods from modules Rod burn-in Rods shipped to CERN

Thermal cycled module

Wir

e b

on

d

Final pinhole test on ARC

Wire bond

Thermal cycle hybrids


Majority of preparations completeMajority of preparations complete

• Many added preparations to smooth production have been completed

Test stand clam shell boxes with LED arrays

Module carrier plates Test pigtails Vienna boxes Storage racks HV crowbar for test stands

• 4 hybrid thermal cycling/test box almost complete. Expected to be ready for use in August.

Currently plan to ramp up hybrid completion task in September, taking over for CERN.


Rod LT Test Facility (RU Group)Rod LT Test Facility (RU Group)

• Peak production < 10 rods per week per site. Schedule ~ 25/month/site. Capacity ~ 45. 72 hour “burn-in” with slow thermal-cycling

• Each site will have capacity of 8 SS rods.

• Hardware all specified and in process Chest freezer, plumbing, flow meters, valves are almost

all in hand. All machined parts complete Low temp chiller and C6F14 are at FNAL – ready.

• Interfaces nicely with control PC• +20 C to –25 C in 1.5 h

Chiller DAQ system • Programming mostly complete

• Interlock software ready

• First system to be shipped to FNAL in August and fully operational by October.


Transport CrateTransport Crate

• Al Rod boxes in a larger crate. Rochester group has studied this in detail.

• Designed for cushioning >10g internally

Commercial ABS plastic cases• 4 rods/layer in ~10 layers with interleaved foam

• Rod boxes sealed in N2 filled anti-static sleeve.

Container ~ 80 kg and each rod box is 3 kg. Total weight per full shipment will exceed 220 kg

• Rochester to Helsinki with shock logger. • No problem!


Other ItemsOther Items

• Stereo module production readiness Stereo wirebond fixture parts are now being machined Stereo gantry plates now being designed Expect to be ready to make stereo modules at UCSB in mid-

September


Broken Bonds in TransportBroken Bonds in Transport

• Four shipments of modules from the USA to CERN by 3 different methods of transport:

All were damaged in similar manner. Many broken bonds – particularly between sensor and pitch adaptor.

• At UCSB we performed extreme drop tests that also resulted in sensor to sensor bonds being broken as shown above.


Reinforcement with SylgardReinforcement with Sylgard

• Sylgard is a silicon based encapsulant Used to encapsulate ALL wirebonds on innermost layer of CDF Working temperatures –50C to 200C Radiation hard Low thermal conductivity

•Applied long or short beads of this material to back sides of modules at sensor-pa joint and sensor-sensor joint


Drop Test ResultsDrop Test Results

• With minimal packaging we dropped the reinforced sensors 1 meter!

Almost all bonds broke on the control module (with no reinforcement)

No obvious deterioration of bond pull strengths for reinforced modules

Module 867 pull strengths (large glue beads)

0

2

4

6

8

10

12

0 100 200 300 400 500 600

Strip Number

Pu

ll S

tren

gth

s (g

)

initial pull test

after glue beads

after cold cycling

1 meter drop after repackaging


Effect on Cooling:Test SetupEffect on Cooling:Test Setup•Used two modules

One with full glue bead between carbon fiber and sensor

One unmodified

•Measured temperature of carbon fiber (hybrid location) and sensor (near hybrid)

Temperature measurement good to 0.03 C

Temperature Sensor Locations


Thermal Cycle (Control Module)Thermal Cycle (Control Module)

• Thermal cycle without any modifications Ran until temperature

stabilized Began cooling Once at low temperature,

began running chip until temperature stabilized

Began warm-up

• Cooling plate and sensor temperature matched well

• T=2.3 C between carbon fiber and sensor CERN cooling mock-up

using final carbon fiber and heater saw T=2.7 C

Unglued Module

-20

-15

-10

-5

0

5

10

15

20

25

30

0 100 200

Time (Minutes)

Te

mp

atu

re(C

) T(plate)

T(carbonfiber)T(sensor)


Thermal Cycle (Glue Module)Thermal Cycle (Glue Module)

•Thermal cycled module with full glue bead

Same Thermal Cycle

• T=2.0 C between carbon fiber and sensor

Glued Module

-25

-20

-15

-10

-5

0

5

10

15

20

25

30

0 50 100 150 200 250 300

Time (minutes)

Tem

per

atu

re (

C)

T(plate)

T(carbon fiber)

T(sensor)


T between Carbon Fiber and SensorT between Carbon Fiber and Sensor

Temperature difference between carbon fiber and sensor

0

0.5

1

1.5

2

2.5

3

0 50 100 150 200 250 300

Time (minutes)

T

(C) No glue

Glue

Adding full bead of glue increases sensor temperature by 0.3 C


Additional ConcernAdditional Concern• 39 modules tested

28 of which have been tested at both 250 & 550 V

• Of those 28 modules, 10 modules have entire chips with >2x regular noise All have a larger bias current

than expected from sensor probing

Four of these 10 modules have been visually scanned and we find no obvious sensor damage.

• 9-15 wire bonds have to be removed to reduce the noise of chip to normal values,

666


Module 869Module 869

Module 869

0

500

1000

1500

2000

2500

3000

3500

0 100 200 300 400 500

VoltageC

urre

nt(n

A) Sensor 1

Sensor 2

Sensor 1 +2

Module

Modules with as little as 2 A extra bias current relative to sensor probing data have exhibited extremely large noise.

Module 869 shows increased noise on first chip at ~425 V. Other chips have had increasing noise starting at 100-450 V.

Module 869


Noise Vs. VoltageNoise Vs. Voltage•~1-2 channels per module have increasing noise with bias voltage

In most cases, sensor flaws can be seen optically

•We are concerned that these channels may be less extreme examples of same problem

It is NOT clear how these modules will change with time, radiation damage, and temperature.

•Further testing, particularly radiation and long term operation, is necessary

In CDF SVX’ we pulled bonds on ALL such channels and tied those strips to ground through a capacitor.


Summary & ConcernsSummary & Concerns

• Now able to fabricate and fully test&burn-in 35 single sided modules per week. Could build and wirebond 3 times this number but we do not

have adequate burn-in capability yet.

• Rod assembly preparations well underway. First rod is to be built and tested this September. First LT Rod test stand to be ready by September

• Transportation of rods is under control. Indications are clear that breakage of bonds during transport

can be avoided. We need to agree on exact solution to start production.

• Only real concern is the problem of APV instability for noisy strip inputs. We will make ~35 more modules in US in next few weeks and

conduct extensive studies in collaboration with Karlsruhe.

Documents

US Tracker Project – CERN - July 17, 2003 - Incandela 1 US Production Readiness CMS Tracker Week 17/06/03 CERN J. Incandela University of California Santa