Geiger Mode - grapes-3

Download Report

Transcript Geiger Mode - grapes-3

Introduction to Silicon Photo-Multiplier
PANKAJ RAKSHE
* Ref. – Raghunandan Shukla , “Introduction to Silicon Photo-Multiplier”, WAPP2012
Why ?
Sensitive Photo-detectors like PMT’s , HPD’s are of great interest to
scientific community due to their use in examining processes
emitting very low photon flux.
For example,
 In GRAPES-3 they are used to detect scintillation light from fibres .
 PET scanners
 Florescence imaging of the cells
2
Let’s have a look at Classical PMT
 Gain ~ 106
 Size : Dia 1”, Length 6”
 Operating Voltage ~ 2000 V
 Quantum Efficiency ~ 30 %
 Response time ~ 2 ns
 Effected by Magnetic fields
 Incoming photon knocks out electron from photo cathode
 Emitted electron is focused on electrode
 Electrons are multiplied by series of dynodes due to presence of electric field
3
Avalanche Photo Diode :
 A solid state Photo-detector
 Very small in size
 Generally used in reverse bias mode for
Typical Application Circuit
photo-detection
 Gain depends upon biasing voltage
Typical reverse I-V characteristics
4
APD : Mode of operations
Avalanche Photo Diode (APD) can be used in two modes
Proportional Mode
 APD is biased under it’s breakdown
voltage
 Output pulse height is proportional
to the number of incident photons
 Gain < 1000
 Basically functions as Amplifier
Proportional
Mode
Geiger
Mode
Geiger Mode
 APD is slightly above its breakdown
voltage (~10%)
 Output pulse height is independent
of number of incident photons
 Gain ~ 106
 So, basically in Geiger mode APD
functions as a Discriminator
5
Hybrid Photo-Diode
* Photonis HPD Catalog
Comparison of Detectors
APD
PMT
HPD
 Large Gain (106)
 Low Gain (~100)
 Small Gain (~1000)
 High Cost
 Low Cost
 Cost as High as PMT
 Sensitive to Magnetic
 Insensitive to Magnetic
 Insensitive to Magnetic
Field
 Bulky
Field
 Small Size (Solid State
Device)
Field
 Small Size (Solid State
Device)
* S. R. Dugad in WAPP2012
Geiger Mode Operation
 A very high electric field is created in depletion region of the diode by applying
high reverse bias voltage.
 Incident photons knocks out a carrier ( electron or hole).
 This carrier gains enormous amount of energy accelerating through high
electric field
 Accelerated carrier imparts energy to more carriers coming in its way (Impact
Ionization)
 This process goes on and thus an Avalanche of carriers is formed.
 This avalanche has to stopped in order to limit the current and avoid damage of
the device.
8
Which mode to choose ??
Geiger Mode
Proportional Mode
 Output is proportional to the number  Output is a digital signal, indicating
presence or absence of photon.
of input photons.
 Large dynamic range
 Information about input photons is lost.
 Can not detect single photon
 Can detect single photon
 Low gain
 High Gain
We want..
A device with good features from both the modes i.e device with large
dynamic range and high gain
Solution :
Silicon Photo Multiplier ( SiPM)
9
Silicon Photo-Multiplier (SiPM)
 Compact Device
1-3 mm
 Operating voltage (30-120V)
 Resolution - Single photon detection
 Response time – ~100 ps
1-3 mm
 High gain - 106
 High Quantum Efficiency – 90%
 High Photon Detection Efficiency –
60%
 Immunity to Magnetic Field
10
SiPM
 SiPM is a 2-D array of Avalanche Photo
Diode’s (APD’s) , all resistively coupled
together.
 SiPM is generally biased above its
breakdown voltage , called as Geiger mode.
 Each pixel (APD) acts as a binary device,
indicating presence or absence of photon.
Device as whole gives analog signal
indicating number of pixels fired.
 Typical size of each APD (pixel) is 50 µm ×
50 µm and a typical gain of ~ 106
11
Passive Quenching – Simple !
 Once the avalanche develops, it is
important to stop the current build-up,
so that device is not damaged and it is
ready for next detection.
Current
Pulses
Diode
Voltage
 It is achieved with the help of series
resistance connected
 As the current through APD (pixel)
increase, voltage drop across series
resistance increases
 This effectively decreases voltage across
diode and quenches the Avalanche
Proportional
Mode
Geiger
Mode
Pulse Output
Recharge
12
SiPM Operation
Incident Photon knocks out an electron.
Electron is accelerated in high electric field at
junction generated due high reverse bias
High energy (accelerated) electron imparts energy
to electrons in its path and thus due to such
multiplication avalanche occurs
This avalanche is quenched by high series
resistance and the device is brought back to its
operating conditions
13
Typical SiPM pulse..
The signal presents 2 components:
1. Avalanche current reproduced at the output by parasitic capacitor
2. slow component due to the recharge of the diode capacitance
(Recovery time ~70ns)
14
SiPM Specifications
1) Quantum Efficiency :
𝑄𝐸 =
𝑁𝑜.𝑜𝑓 𝑃ℎ𝑜𝑡𝑜𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑛𝑠
𝑁𝑜.𝑜𝑓 𝑖𝑛𝑐𝑖𝑑𝑒𝑛𝑡 𝑝ℎ𝑜𝑡𝑜𝑛𝑠
QE for SiPM ( Silicon) is more than 90%
2) Photon Detection Efficiency :
𝑃𝐷𝐸 = 𝑄𝐸 × 𝐺𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐𝑎𝑙 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 × 𝐵𝑟𝑒𝑎𝑘𝑑𝑜𝑤𝑛 𝑝𝑟𝑜𝑏𝑎𝑏𝑖𝑙𝑖𝑡𝑦
Due to Resistors used for quenching and other features like Guard rings, some dead area is
introduced.
𝐴𝑐𝑡𝑖𝑣𝑒 𝑎𝑟𝑒𝑎
𝐺𝑒𝑜𝑚𝑒𝑡𝑟𝑖𝑐𝑎𝑙 𝐸𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 =
𝑇𝑜𝑡𝑎𝑙 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑡ℎ𝑒 𝑑𝑒𝑣𝑖𝑐𝑒
Breakdown probability depends upon applied bias voltage and it can be increased to almost
100 %
15
SiPM Specifications
3) Linearity
SiPM behaves as a perfectly linear device under low light flux.
Dynamic Range = 1 to Number of of pixels
2 p.e
3 p.e
4 p.e
1 p.e
5 p.e
0 p.e
Individual photon peaks well resolved !
* Hamamtsu MPPC user manual
16
Timing Resolution
 Timing information shown for SiPM includes laser pulse width and delay due
to electronics
 Timing resolution ~ 100 ps
 For comparison timing resolution of one of the best PMT is shown
* P. Buzhan et al. / NIM in Physics Research A 504 (2003) 48–52
17
So far we have seen that SiPM offers exceptionally
good features like small size, high dynamic range, high
QE, fast response and immunity to magnetic field etc.
But..
SiPM is not perfect !
18
SiPM Noise– Dark Counts
( Primary Noise)
 Due to thermal energy, valance band
carrier enters into conduction band
and gives rise to an avalanche.
 Such pulses look exactly like genuine
photon event and thus can not be
distinguished.
 Typical dark count rate ~ few MHz at
room temperature
 Dark counts increase as ambient
temperature increases.
* Claudio Piemonte, FNAL ,October 25th 2006
19
SiPM Noise– After Pulsing (Noise -2)
Carrier Trapping and Delayed Release  Afterpulsing
 During avalanche, carriers are
trapped due to defects in the
crystal and released after some
time.
 This leads to generation
secondary pulse following primary
pulse.
 At lower temperature trap
lifetime increases
* S.Cova, A.Lacaita, G.Ripamonti, IEEE EDL (1991)
20
SiPM Noise- 3 – Cross Talk
Hot-Carrier Luminescence
A. Lacaita et al, IEEE TED (1993)
105 avalanche carriers  1 emitted photon
Counteract:
• Optical isolation between pixels
• Avalanche charge minimization
F.Zappa et al, ESSDERC (1997)
21
 Geiger mode: above breakdown
Gain
SiPM Biasing
 VS = VBR + Over-Voltage
 Gain dependant on amount of Over-voltage
applied
Biasing Voltage (V)
* Bajarang Sutar
 But, Breakdown Voltage is dependent on
temperature
 Effectively, gain will also change! (3-5%/˚C)
 Over-Voltage should be constant for
constant gain
 𝑽𝑺 𝑻 = 𝑽𝑩𝑹 𝑻 + 𝑶𝒗𝒆𝒓 𝑽𝒐𝒍𝒕𝒂𝒈𝒆
* S. R. Dugad and K. C. 22Ravindran
Solutions to Temperature Dependency
Constant Temperature
 Indoor Applications - applicable
 Temperature control of detectors in
outdoor environment is not possible
e.g. GRAPES-3 experiment containing
400 detectors in large area outdoor
field of about 25000 m2 with
temperature variations 5 - 30 ˚C
Temperature Dependant Biasing
Conditions
 The Bias Voltage of SiPM can be
controlled for changing the operating
point
 Temperature dependent Power
Supply that will keep over-voltage
constant (i.e. Gain constant)
23
Approaches To Temperature Compensation
Method
Used
Dark
Current
Control
Author
Miyomoto
et al.
Li et al.
Year
Remarks
2009 Dark Current and Temperature relation approximated to
exponential function and use of Thermistor (similar relation) for
compensation
2012 Dark current and Bias Voltage relation used for designing the
voltage controlled current sink that changes bias condition
Bencardino 2009 Use of Temp. to voltage converter and a Op-amp based circuit for
et al.
changing the bias return potential w.r.t. temperature variations
Bias
Voltage
Control
Licciulli et
al.
2013 Blind SiPM used as a temperature sensor and amplitude of Dark
pulses of Blind SiPM is maintained constant using Op-amp based
feedback circuit for constant gain of other SiPM in parallel
Gil et al.
2011 External Input of the power supply is controlled by a microcontroller based system to change the output voltage w.r.t. temp.
Dorosz et
al.
2013 LabVIEW based feedback system for controlling the power supply
output
Limitations
1. Uses Expensive
Commercial
Power Supply
2. Limited to
one/two
channels
3. External control
required for
temperature
compensation
24
A Temperature Compensated Power
Supply for Silicon-Photomultiplier
25
Specifications
Sr.
No.
Parameter
Value
1.
Output Voltage
0 to 100V
2.
Temperature Reading Resolution
3.
Temperature Compensation Factor
4.
Number of Channels
5.
Output Voltage Resolution
10 mV
6.
Maximum Current Limit per Channel
100 µA
7.
Full Scale Leakage Current per Channel
40 µA
0.1 ˚C
10 to 100 mV/˚C
8
26
Block Diagram
High Voltage
Generation
(Voltage
Multiplier Chain)
PC
USB
Temperature
Sensors
Control Unit
(Microcontroller)
Current Sense
DAC
Voltage
Regulation
Scheme
SiPM
27
Prototype Power Supply (SiPM-PPS-v1)
28
Features of SiPM Power Supply
 Programmable Output Voltage (0 - 100 V) with resolution of ~12 mV
 Programmable Temperature Compensation Factor (~12 – 100 mV/˚C)
 In-built Data Acquisition System for recording Temperature and Leakage
Current via USB
29
0.04113% for 10% change in line voltage
0.03% for 20% change in line voltage
Ripple = 5 mVP-P
No load to Full load (100uA) regulation
0.6025%
Stable within 10 mV for 0 – 100 V
30
Compensation Algorithm Verification
Comp. factor = 240 mV/˚C
Comp. Factor = (0.029x2-1.4x) /˚C
31
SiPM Experimental Setup
32
SiPM Test: Without Compensation
Gain Variation of about 4 %/˚C
33
SiPM Test: With Compensation
Gain Variation of 0.8 %
34
In Conclusion
 Without compensation gain variation
nearly 4 %/˚C
 With compensation gain variation
0.08 %/˚C
 Output Ripple of 5 mV with the
Resolution of 12.5 mV in 100 V with
temperature correction at each 0.2 ˚C
change
35
Compared to Keithley 6487 Unit
 Voltage Source and Pico-ammeter
SiPM Programmable Power Supply
 Temperature Compensation Feature
 Small Size and Light Weight (portable)
 Multi-channel (8/16 channels)
 Cost – Inexpensive !!!
(₹ 750/channel) which is 400 times less
36
SiPM-PPS-v2
• Number of channels = 16
• Better output resolution = 6.25 mV
• DAC resolution = 14 bits
• Better current sensing resol. = 1 nA
• ADC resolution = 16 bits
• Provision for expandability
• I2C for multiple boards
37
Specifications Comparison
Sr.
No.
Parameter
Required Value
Prototype Board
SiPM_PPS v2
0 to 100V
0 to 100 V
0 to 100 V
8
8
16
1.
Output Voltage
2.
Number of Channels
3.
Output Voltage Resolution
10 mV
12.5 mV
6.25 mV
4.
Maximum Current Limit per Channel
100 µA
100 µA
100 µA
5.
Full Scale Leakage Current per Channel
40 µA
40 µA
35 µA
6.
Temperature Reading Resolution
0.1 ˚C
0.1 ˚C
0.1 ˚C
7.
Temperature Compensation Factor
12.5 to 100 mV/˚C
12.5 to 100 mV/˚C
6.25 to 100 mV/˚C
38
Summary
 SiPM has some outstanding features that in some respect could replace PMT in
near future.
 Newer versions of SiPM are being developed to overcome its limitations.
 The temperature dependence is one of the major limitation preventing the use
of SiPM in outdoor applications. The programmable temperature compensated
power supply is useful for operating the SiPM in different environmental
conditions.
39
Acknowledgment
 Prof. S. R. Dugad (TIFR)
 Prof. P. D. Khandekar (VIIT)
 Mr. Sergey Los (FNAL)
 Prof. C. S. Garde (VIIT)
 Mr. Raghunandan Shukla (TIFR)
 Prof. S. K. Gupta (TIFR)
 Ms. Sarrah Lokhandwala (TIFR)
40