Radiation-Hardness of VCSELs & PINs
Richard Kass
The Ohio State University
A. Adair, W. Fernando, K.K. Gan, H.P. Kagan, R.D. Kass,
H. Merritt, J. Moore, A. Nagarkar, S. Smith, M. Strang
The Ohio State University
M.R.M. Lebbai, P.L. Skubic
University of Oklahoma
B. Abi, F. Rizatdinova
Oklahoma State University
OUTLINE
Introduction/ATLAS pixel detector
Radiation Hardness of VCSELs
Radiation Hardness of PINs
Summary
R. Kass
ICTPP09
The Current ATLAS Pixel Detector
ATLAS: an LHC detector designed to study 14 TeV pp collisions
Pixel detector is inner-most system -> Radiation damage is the issue
Present Pixel Detector:
ATLAS’s Inner most charged particle tracker
Measures (x,y,z) to ~30 mm
Pixel detector is based on silicon
Pixel size 50mm by 400 mm ~80 million pixels
Radiation hardness is an issue
must last ~ 10 years
A pixel module contains:
1 sensor (2x6cm) ~40000 pixels
16 front end (FE) chips 2x8 array
Flex-hybrid
1 module control chip (MCC)
There are ~1744 modules
~1.85m
Detector upgrades planned:
New Inner layer (“IBL”) & later a new pixel detector for Super-LHC
R. Kass
ICTPP09
Present Pixel Opto-link Architecture
Current optical link of pixel detector transmits signals at 80 Mb/s
Opto-link separated from FE modules by ~1m
transmit control & data signals (LVDS) to/from modules on micro twisted pairs
Use PIN/VCSEL arrays
Use 8 m of rad-hard/low-bandwidth SIMM fiber fusion
spliced to 70 m rad-tolerant/medium-bandwidth GRIN fiber
a Simplify opto-board and FE module production
a Sensitive optical components see lower radiation level than modules
a PIN/VCSEL arrays allow use of robust ribbon fiber
~80m
~1m
optoboard
VCSEL:
VDC:
PIN:
DORIC:
R. Kass
Vertical Cavity Surface Emitting Laser diode
VCSEL Driver Circuit
PiN diode
Digital Optical Receiver Integrated Circuit
ICTPP09
optoboard holds VCSELs, VDCs, PINS
Radiation Dosage at SLHC
VCSEL/PIN of current pixel detector are mounted
on patch panel (PP0) instead of directly on the pixel module
a much reduced radiation level compared with module
VCSEL/PIN for pixel detector at SLHC will be mounted
further away from pixel module
a expected dosage at r = 37 cm for 3,000 fb-1
with 50% safety factor:
u silicon: 7.2 x 1014 1-MeV neq/cm2
u GaAs: 2.8 x 1015 1-MeV neq/cm2
Assuming radiation damage scales with
Non-Ionizing Energy Loss (NIEL)
R. Kass
ICTPP09
Real Time Monitoring in T7 Beam Test
24 GeV proton beam
2009 Beam Tests used a simple system
Real time monitoring of PIN current & optical power.
VCSEL arrays
laser spot = beam spot
Control Room
R. Kass
PIN diode arrays
ICTPP09
VCSEL arrays
850 nm VCSEL Irradiation
2006-7:
12-channel VCSEL arrays were irradiated to SLHC dosage
AOC 2.5 Gb/s (obsolete), 5 Gb/s, 10 Gb/s
ULM 5 Gb/s, 10 Gb/s
Optowell 2.5 Gb/s
insufficient time for annealing during irradiation
2008:
AOC 5 Gb/s, 10 Gb/s
Optowell 2.5 Gb/s
MPO connector
MPO adaptor
Opto-pack
2009:
AOC 10 Gb/s
goal: 20 arrays
actual: 6 arrays due to manufacturer problem
R. Kass
ICTPP09
AOC 10 Gb/s VCSEL (2008)
irradiation
annealing
7.6 x 1015 1-MeV neq/cm2
Optical power recovery by annealing is slow
Almost recover the initial power after extended annealing
VCSEL produces more power at lower temperature
R. Kass
ICTPP09
AOC 10 Gb/s VCSEL (2009)
irradiation
annealing
4.4 x 1015 1-MeV neq/cm2
w/o long
twisted/coil
ed fiber
145 mW
Good optical power for 6 arrays irradiated
Await return of arrays to OSU for annealing/characterization
need to irradiate a sample of 20 arrays in 2010
R. Kass
ICTPP09
2008 PIN Irradiation
Gb/s
GaAs (4.4 x 1015 1-MeV neq/cm2)
Responsivity (A/W)
Pre
Post
ULM
AOC
4.25
5.0
0.50
0.60
0.09
0.13
Optowell
3.125
0.60
0.17
2.5
0.50
0.28
Taiwan
1.0
0.55
0.21
Hamamatsu S5973
1.0
0.47
0.31
Hamamatsu S9055
1.5/2.0
0.25
0.20
Hamamatsu G8921
Si (7.5 x 1014 1-MeV neq/cm2)
Irradiated 2 arrays or several single channel devices
of each type
Hamamatsu devices: low bandwidth but more radiation hard
Irradiated 20 Optowell arrays in 2009
R. Kass
ICTPP09
PIN Responsivity vs Bias Voltage
(2008)
4.4 x 1015 1-MeV neq/cm2
Pre-irrad
Optowell
Optowell
AOC
ULM
Responsivity doesn’t depend on bias voltage before irradiation
Can increase responsivity with higher bias after radiation
R. Kass
ICTPP09
PIN Responsivity vs Bias Voltage
Optowell
Can fully recover pre-irradiation responsivity with large bias
voltage
Need to look at pulse shape at high bias voltage
R. Kass
ICTPP09
Eye Diagram at High Bias Voltage
Optowell
Test limited to 1 Gb/s @ 40 V due to carrier board limitation
Eye diagram looks reasonable
a need more detailed characterization
R. Kass
ICTPP09
Results on Optowell PIN Arrays
20 Optowell PIN arrays irradiated August 2009
Good responsivity after irradiation
average responsivity after irradiation ~0.3 A/W
60
Channels
50
40
Analysis of additional ten arrays
complicated by beam misalignment
Need more detailed study, including
eye diagram after cool down.
10 arrays
Pre-Irrad
Post-Irrad
8.1 x 1015 1-MeV neq/cm2
30
20
10
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Responsivity (A/W)
R. Kass
ICTPP09
Summary
AOC 10 Gb/s arrays have good optical power after irradiation
VCSEL produces more power as room temperature decreases
Need to repeat irradiation with large sample in 2010
Hamamatsu PINs are slow but more radiation hard
Optowell PIN arrays have good responsivity after irradiation
Can increase responsivity with higher bias voltage after radiation
Will irradiate a large sample of AOC PIN arrays in 2010
AOC plans to release high-speed PIN arrays in 2010
R. Kass
ICTPP09
extra slides
R. Kass
ICTPP09
Real Time Monitoring in T7 Beam Test
2006-8 test of opto-board system used loop-back setup
Compare transmitted and decoded data
Measure minimum PIN current for no bit errors
Measure optical power
25m optical
Opto-board
fiber cable
bi-phase marked
optical signal
clock
PIN
DORIC
data
decoded data
VCSEL
VDC
decoded clock
VCSEL
VDC
In control room
In beam
Bit error test setup
at CERN’sT-7 beamline
24 GeV protons
Two VCSEL arrays
from same vendor
per opto-board
R. Kass
Signal routed back to
opto-baord via test board
attached to 80-pin
connector & test board
ICTPP09
Opto-Chip setup