Transcript nomerotski

CCD-based Pixel Detectors by LCFI
Andrei Nomerotski (U.Oxford) on behalf of LCFI collaboration
May 7 2006, UK HEP Forum
Outline
 LCFI Collaboration
 Pixel Sensors and their Readout
Column-Parallel CCDs
 Storage Pixels : ISIS

 Mechanical Studies
 Summary
1
Andrei Nomerotski
LCFI :
Linear Collider Flavour
Identification
Valencia
Goals : Development of technologies and
algorithms for the ILC vertex detector
2
Andrei Nomerotski
Pixel Sensors
 Traditionally LFCI develops CCD-based sensors
 Builds on successes of the SLD vertex detector
3
The VXD3 upgrade
vertex detector: 96
large CCDs, 307 Million
pixels (1996)
Andrei Nomerotski
Vertex Detector for ILC
Main requirements:
 Excellent point resolution (3-4 μm), ~1 Gigapixel 20x20 μm
 Low material budget ( 0.1% X0 per layer)
 Low power dissipation
 Moderate radiation hardness ( 20 krad/year)
 Tolerance to Electro-Magnetic Interference (EMI)
 Operation in 5T magnetic field
 Fast Readout – The Challenge
4
Andrei Nomerotski
The Challenge
LC Beam Time Structure: 0.2 s
337 ns
2820x
0.95 ms = one train
What readout speed is needed?


If read once per train : occupancy ~200 hits/mm2 : too slow
Need to read once accumulated occupancy ~10 hits/mm2



=> 20 times per train = 50µs/MPixel
Fastest commercial CCDs ~ 1 ms/MPixel
Two approaches
1.
2.
Parallel Readout of traditional CCD: CPCCD – information leaves the
sensor as fast as it can
Storage Sensors : each pixel has a ‘memory’ filled up during
collisions and read out between trains at slow rate: ISIS
technology – information is stored in the sensor
5
Andrei Nomerotski
Column Parallel CCD

Simple idea : read out a vector instead of a matrix  Readout time
shortened by orders of magnitude
BUT

Despite ‘parallel processing’ readout rate is still challenging : 50 MHz clock
moves charge 2500 times in 50 µs. 2500x20 µm = 50 mm

Every column needs own amplifier and ADC  requires readout chip
M
M
N
N
“Classic CCD”
Readout time 
NM/fout
6
Column Parallel
CCD
Readout time = N/fout
Andrei Nomerotski
Column Parallel CCD
 Readout Chip is a difficult but clearly feasible problem


Geometrically concept of columns is similar to the silicon strips strip detectors have complex readout chips integrated in ladder
However density of channels and 0.1% Xo constraint requires bumpbonding – non-trivial anyway
 Difference wrt Strips: to move the charge need to clock all
columns of the CCD simultaneously

Simple exercise :





typical capacitance of CCD sensor : 100nF
50MHz  10 ns rise time
Clock current = 100 nF x 2 V / 10 nsec = 20A !
Voltage drops 20 A x 0.1 Ohm = 2 V
Inductance of 1mm long bond wire = 1 nH :
corresponds to 0.3 Ohm at 50 MHz
 Driving a full area CPCCD is a major challenge!


Need a special high current clock driver
Need to be extra careful with the design of clock distribution
7
Andrei Nomerotski
CPCCD : LCFI R&D Milestones

Established proof of principle for small area
sensors : CPC1
Established proof of principle for readout chip :
CPR1, developed and produced more sophisticated
CPR2

Moved on to large area sensors : CPC2


1.
2.
3.
Need to handle the problem of clock driver
Design dedicated clock driver : CPD1
Find ways to reduce the CCD capacitance
Find ways to reduce the required clock voltage
Next slides: Results from prototypes
8
Andrei Nomerotski
CPC1 Bump-bonded to CPR1
 CPC1 : Two phase CCD, 400 (V)  750 (H)
pixels, 20 μm square;
 CMOS readout chip (CPR1) designed by the
Microelectronics Group at RAL:
0.25 μm process
 Charge and voltage amplifiers matching
the outputs of CPC1
 Correlated double sampling
 5-bit flash ADCs and 132-deep FIFO per
column
 Everything on 20 μm pitch
 Size : 6 mm  6.5 mm
 Manufactured by IBM
Bump-bonded CPC1/CPR1 in a test PCB
 Bump-bonded by VTT (Finland) using solder
bumps
9
Andrei Nomerotski
CPC1/CPR1 Performance
5.9 keV X-ray hits, 1 MHz column-parallel readout
Voltage outputs, noninverting (negative
signals)
Charge outputs,
inverting (positive
signals)
Noise 60 e-
Noise 100 e-
K.Stefanov RAL
First time e2V CCDs have been bump-bonded
 High quality bumps, but assembly yield only 30% :
mechanical damage during compression suspected
 Differential non-linearity in ADCs (100 mV full scale) :
addressed in CPR2
Bump bonds on CPC1
under microscope
10
Andrei Nomerotski
Next Generation CPCCD Readout Chip – CPR2
Bump bond pads
Voltage and charge amplifiers
125 channels each
Analogue test I/O
Digital test I/O
5-bit flash ADCs on 20 μm pitch
CPR1
Cluster finding logic (22
kernel)
CPR2
Sparse readout circuitry
FIFO
CPR2 designed for CPC2
 Results from CPR1 taken into account
 Numerous test features
 Size : 6 mm  9.5 mm
Wire/Bump bond
pads
Steve Thomas, RAL
 0.25 μm CMOS process (IBM)
 Manufactured and delivered February 2005
11
Andrei Nomerotski
CPR2 Test Results
Test clusters in
Sparsified output
 Parallel cluster finder with 22
kernel
 Global threshold
 Upon exceeding the threshold,
49 pixels around the cluster are
flagged for readout
●Tests on the cluster finder: works!
● Several minor problems, but chip is usable
● Design occupancy is 1%
● Cluster separation studies:
Errors as the distance between the
clusters decreases
 Reveal dead time
Tim Woolliscroft, Liverpool U
Tim Woolliscroft, Liverpool U
12
 Many of the findings have already
been input into the CPR2A design
Andrei Nomerotski
Next Generation CPCCD : CPC2
No
connections
this side
Clock bus
Charge
injection
Three different chip sizes
with common design:
CPC2-70 : 92 mm  15 mm
image area
Extra pads for clock
monitoring and
drive every 6.5 mm
Image area
Standard
Field-enhanced
Standard
Temperature
diode on
CCD
Four 2-stage SF in
adjacent columns
Four 1-stage and 2stage SF in adjacent
columns
Main clock
wire bonds
Main clock
wire bonds
CPR1
CPR2
 CPC2-40 : 53 mm long
 CPC2-10 : 13 mm long
 Compatible with CPR1 and
CPR2
 Two charge transport
sections
 Choice of epitaxial layers
for different depletion depth:
100 .cm (25 μm thick) and 1.5
k.cm (50 μm thick)
13
Andrei Nomerotski
CPC2 + ISIS1 Wafer
ISIS1
5” wafers
 One CPC2-70 : 105 mm  17 mm
total chip size
 Two CPC2-40 per wafer
CPC2-70
 6 CPC2-10 per wafer
 14 In-situ Storage Image
Sensors (ISIS1)
CPC2-40
 3 wafers delivered
CPC2-10
14
Andrei Nomerotski
CPC2-40 in MB4.0
Transformer
CPR1/CPR2 pads
Clock monitor pads
Transformer drive for CPC2
Johan Fopma, Oxford U
 “Busline-free” CCD: the whole image area serves as a distributed busline
 50 MHz achievable with suitable driver in CPC2-10 and CPC2-40
 First clocking tests have been done
15
Andrei Nomerotski
CPC2: First Results
55Fe
spectrum from CPC2-10
at 1 MHz
K.Stefanov RAL
●
First 55Fe spectrum at 1 MHz, -40 C, reset every pixel
16
Andrei Nomerotski
Clock Drivers for CPC2
Brian Hawes, Oxford U
Transformer Driver
 Requirements: 2 Vpk-pk at 50 MHz over 40 nF
(half CPC2-40);
 Planar air core transformers on 10-layer PCB,
1 cm square
 Parasitic inductance of bond wires is a major
effect – fully simulated;
IC driver: CPD1
 Transformer is bulky: IC driver could be a
better solution;
 Design of the first CPCCD driver chip
(CPD1) has started, manufacture in June
 CPD1: 2-phase CMOS driver chip for 20
Amp current load at 25 MHz (L2-L5 CCDs)
 0.35 μm process, size  3 x 8 mm2
 32 W peak power but 0.5% duty cycle
Thermal and electromigration issues seems
17
to be under control
Andrei Nomerotski
Next Steps for CPCCD
 Evaluate performance of CPC2 bump-bonded to
CPC2/CPR2
 Designing with e2V test devices to study how
to reduce CCD capacitance and how to reduce
clock voltage

Theoretically can achieve factor of 4 reduction in C
 Design of CPC3 will depend on results of these
tests
18
Andrei Nomerotski
Radiation Damage Effects in CCDs: Simulations
Signal density of trapped electrons in 2D
●Full 2D simulation based on ISETCAD developed
● Trapped signal electrons can be
counted
● CPU-intensive and time consuming
Simulation at 50 MHz
Operating
window
● Simpler analytical model also used,
compares well with the full simulation
● Window of low Charge Transfer
Inefficiency (CTI) between -40 C
and 0 C
● Will be verified by measurements
on CPC2
L. Dehimi, K. Bekhouche (Biskra U); G. Davies, C. Bowdery,
A.Sopczak (Lancaster U)
19
Andrei Nomerotski
Storage Pixels
 Industry analogy is “Burst mode” : capturing a limited
number of images at short intervals


20
Burst mode imagers are available commercially (ex. DALSA) with
rates up to 100MHz : the rate is limited by the charge transfer
between neighboring cells
Memory is implemented as a CCD register associated with an
imaging pixel : whatever one can fit in an area of one pixel
2003: Dart bursting a ballon : 100 consecutive frames at 1M frame/sec
Andrei Nomerotski
Storage Pixels as Particle Detectors
 ILC requirements : capture charge every 50 us
20kHz – no problem
Challenges :
 Used as particle (not visible light) detector –
need efficient charge collection from the
whole area
 Need to fit 20 cell CCD register into 20x20
square micron pixel (together with photogate
and some logic)
 Need a more complicated than pure CCD
process
21
Andrei Nomerotski
In-situ Storage Image Sensor : ISIS
Reset transistor
transfer
photogate
gate
storage
pixel #1
storage
pixel #20
reset gate
output
gate
Source follower
VDD
Row select
transistor
row select
sense node (n+)
To column load
n+
buried channel (n)
p+ well
p+ shielding implant
reflected charge
Charge collection
reflected charge
High resistivity epitaxial layer (p)
 Charge is collected into a photogate
 Each pixel has its own 20-cell CCD register : store raw
charge during collisions
 Increased resistance to RF
 Column-parallel readout during quiet time at ~1 MHz: much
reduced clocking requirements
22
Andrei Nomerotski
5 μm
In-situ Storage Image Sensor (ISIS)
Global Photogate and Transfer gate
Additional ISIS advantages:
ROW 3: CCD clocks
On-chip logic
ROW 2: CCD clocks
On-chip switches
ROW 1: CCD clocks
~100 times more radiation hard
than CCDs – less charge transfers
 Easier to drive because of the
low clock frequency: 20 kHz during
capture, 1 MHz during readout
 ISIS combines CCDs, active pixel
transistors and edge electronics in one
device: specialised process
ROW 1: RSEL
Global RG, RD, OD
 Development and design of ISIS is
more ambitious goal than CPCCD
RG RD
OD RSEL
Column
transistor
 “Proof of principle” device (ISIS1)
designed and manufactured by e2V
Technologies
23
Andrei Nomerotski
The ISIS1 Cell
1616 array of ISIS cells with 5-pixel
buried channel CCD storage register each;
 Cell pitch 40 μm  160 μm, no edge logic
(pure CCD process)
 Chip size  6.5 mm  6.5 mm
Output and reset transistors
OG RG
OD
RSEL
Column
transistor
OUT
Photogate aperture (8 μm square)
CCD (56.75 μm pixels)
24
Andrei Nomerotski
Tests of ISIS1
Tests with Fe-55 source
K.Stefanov RAL
 The top row and 2 side columns are not protected and collect
diffusing charge
 The bottom row is protected by the output circuitry
 ISIS1 without p-well tested first and works OK
25
 ISIS1 with p-well has very large transistor thresholds,
permanently off
Andrei Nomerotski
Mechanical Options
Target of 0.1% X0 per layer
(100μm silicon equivalent)
 Unsupported Silicon



Longitudinal tensioning provides stiffness
No lateral stability
Not believed to be promising
 Thin Substrates




Detector can be thinned to epitaxial layer (~20 μm)
Silicon glued to low mass substrate for lateral stability
Longitudinal stiffness still from moderate tension
Beryllium has best specific stiffness
 Rigid Structures


Foams look very promising
Will start to investigate shell structure supports
26
Andrei Nomerotski
Laser Survey System
 Laser displacement
meter

Z precision ~ 1 µm
 2D motorised stage

X-Y precision < 1 µm
 Ladder in cryostat:

∆T ~ 100C
 Fast:


1D scan < 1 minute
Scan during cooling
E.Johnson RAL
27
Andrei Nomerotski

Ladder testing with Be and Carbon
Fibre
Beryllium substrate


Minimum thickness 0.15% X0
Good qualitative agreement
from FEA models and
measurement


Better CTE match than Be
~0.09% X0, no rippling to
<200K
 lateral stability insufficient
Silicon
Tension
Beryllium
28
 Carbon Fibre substrate
Glue
J.Goldstein RAL
Andrei Nomerotski
Rigid Structures: Foams
 Properties:



Open-cell foam
Macroscopically uniform
No tensioning needed
 3% RVC prototype




Sandwich with foam core
0.09% X0
Mechanically unsatisfactory
Working on glue application
 8% Silicon Carbide prototype



Single-sided: substrate + foam
0.14% X0
3-4% believed possible
20 µm silicon
1.5 mm silicon carbide
29
Andrei Nomerotski
Silicon Carbide Foam
Glue “pillars” (right plot) are better
than thin glue layer (left plot)
30
Andrei Nomerotski
 Gas cooling test
stand established
Cooling Studies
Cold nitrogen flow
 Model of 1/4
detector
Measurements from
test stand:

Pg /
6 G.Leithall RAL
4
 CFD simulation of
the same setup

In agreement with
mesurements
 Next step: more
detailed comparison
20 litres / min
17.5 litres / min
15 litres / min
12.5 litres / min
10 litres / min
7.5 litres / min
5 litres / min
2
0
-5
31
0
-2
5
10 15 20 25
Tsurface-Tgas / K
S.Yang, Oxford
Andrei Nomerotski
Summary
 LCFI is a viable and growing collaboration to develop
technologies and algorithms for VD
 First generation sensors extensively studied
 Column parallel CCD principle proven
 First results from the second generation of sensors and
readout chips




Detector-scale CCDs
Sparsified readout
Developing advanced clock drivers
First prototypes of storage devices
 Mechanics : 0.1% X0 ladders seems achievable
32
Andrei Nomerotski