Ageing

1
Ageing
DØSMT

• Eric Kajfasz (CPPMarseille)
• D0 Workshop, Beaune, June 17, 2003

2
D0SMT ageing ...

• HDI failures
• Limited access gt HDI means Low Mass Cable
  HDI chips
• Excess Noise
• Usual noise between 1.5 and 3 ADC counts
• Excess noise seem to affect only Micron
  F-Wedges
• Irradiation
• Detector was built to sustain 2 fb-1

F11-1-2
Example of a worst case
3
SMT Design
North 1/2-cylinder
South 1/2-cylinder
12 F-Disks
6 Barrels
4 H-Disks

• F-Disks sensor vendors
• 3,5,8 and 10 Eurysis
• 8 others Micron

4
Low Mass Cables
3 forward F-Disks
3 (F-DiskBarrel) modules
Pigtail
HDI
Ladder
South Half Cylinder
SVXII R/O chips
5
Conclusions ...
80/3MCLK cables
Adapter card
Low-massCLK cables
½-cylinder
6
HV
LV
Fuse Panel
Bulk LV Power Supplies
I B
Interface Board Crates (8x18)
Cathedral
Around Iteraction Region
CLKs
19-30 High Mass Cable (3M/80 conductor)
3/6/8/9 Chip HDI
Adaptor card
8 Low Mass Cables
3/6/8/9 Chip HDI
CLKs
CLKs
25 High Mass Cable (3M/50 conductor)
Cathedral
Horse Shoe
Sensor
I,V,T Monitoring
17 twisted pair cables
Optical Link 1Gb/s
SEQ
Sensor
SEQ Controller
Sequencer Crates (6x20)
HV breakout box
Serial Command Link
VTM
VRBC
VRB
S B C
Platform
25 twisted pair cables
VRB Crates (12x10)
HV fanout 1gt4
VME Crates (4x3)
HV Crates (8x62x4)
MCH2
Monitoring SDAQ
PDAQ (L3)
MCH3
MCH2
PowerPCs and Single Board Computers are accessed
thru Ethernet
7
HDI failures
Would not bet my shirt on an extrapolation of
these
8
HDI failures
9
Effects on physics
Take with a grain of salt these are run IIb
studies
10
Noise in North Barrels

• Usual noise between 1.5 and 3 ADC counts
• Look at the fraction of channels with a noise
  above 4 or 10 ADC counts

05/15/2003
10/01/2001
OK
11
Noise in South Barrels
05/15/2003
10/01/2001
OK
12
Noise in F-disks
NOT OK For Micron wedges
10/01/2001
05/15/2003
13
Noise in F-disks
NOT OK For Micron wedges
10/01/2001
05/15/2003
14
Noise H-disks
OK But keep an eye on them
05/15/2003
10/01/2001
15
Irradiation

• Main effects of irradiation(see R. Lipton,
  D0note-4077 for details)
• Increase of leakage current
• Increase in shot noise
• Not the show stopper.
• Increase in trapped and surface charges in
  insulating layers
• lowered threshold for junction breakdown
  (micro-discharges)
• Expect to limit the lifetime of Micron detectors
• Change in effective impurity concentration
• Remove donors and form acceptors in bulk silicon
• n-type inverts to p-type after about 300 KRads
• Depletion voltage decreases until inversion,
  then increases with dose.This in turn induces
• coupling capacitor breakdown
• micro-discharges

16
Irradiation
Coupling Capacitors Breakdown
broken capacitors
Vdep (V)
Voltage (V)
Dose (Mrads)
_at_ 2.1 MRads
Noise (ADC)
Noise (ADC)
Dose (Mrads)
2fb-1 _at_ layer 1
Bias Voltage (V)
17
Conclusion Diamonds (not Silicon) Are Forever!

• HDI failures
• Difficult to conclude on a real trend
• Anywhere between 13 and 100 disabled channels by
  2009
• Following months will help us understand a bit
  more how it goes
• Increasing Excess Noise
• With present trend
• Seem to affect Micron F-wedges only
• Some (large?) fraction of these wedges will
  become useless in the next few years
• Irradiation (R. Lipton, D0note-4077)
• Significant loss of channels in layers 12 above
  3.6 fb-1 (Vd150V)
• All channels in layers 12 dead above 4.9 fb-1
  (Vd200V)
• Wise thoughts of the day
• Need Run IIa MC studies to better understand
  effect on physics
• Because of all of the above, it is very risky to
  count on SMT to behave sufficiently well by 2009,
  not to impede our physics program
• Build the upgrade, if we have the funding to do so