Fast Timing: future prospects

ccsem.infn.it
  • No tags were found...

Fast Timing: future prospects

Fast Timing:

future prospects

K. Doroud,

C. Williams and A. Zichichi

Wednesday, June 26, 13

1


Fast timing: what is it

Devices/experiments that benefit from a timing

resolution better than 50 ps (σ).

And at this moment fast timing is a hot topic

why

Wednesday, June 26, 13

2


fast timing: a hot topic

the identification of rare events implies High luminosity

accelerators.

Pile-up of up to 200 collision events per bunch crossing at

the upgraded LHC.

Wednesday, June 26, 13

3


CMS event:

the green towers indicate the energy of the photons

The tracking detectors can be used

to assign charged tracks to individual

vertices: but what about the

Wednesday, June 26, 13

photons

4


So we need to build a pre-shower detector with precise

timing just in front of calorimeter

N.B. need time resolution(σ) of 5-20 ps !

Photo detector

Cascade of electrons

and photons

small

scintillating

crystals

Silicon photomultiplier

Incoming photon

Wednesday, June 26, 13

5


Time of Flight techniques can alleviate the pile-up

problem, but:

Current state of the art; ALICE TOF ~ 75ps (sigma)

The goal of a calorimeter with 10 ps (sigma) TOF

resolution is beyond the current state of the art

Wednesday, June 26, 13

6


An example of excellent synergy between

particle and medical physics:

they both needs exactly the same detector

working with cutting edge performance

Wednesday, June 26, 13

7


Positron-emission tomography

(PET)

Injection of a positron-emitting radionuclide (chemical substances such as

glucose, carbon, or oxygen used naturally by the particular organ or tissue

during its metabolic process) .

emission and annihilation of the positron with an electron: production of two

back-to-back 511 keV gammas.

Detection of these gammas and find the position: construct line between these

two space points.

Construction of a 3D image from these lines.

Wednesday, June 26, 13

8


So we have to create a 3-D image ... and that requires a

detector that surrounds the patient - i.e. need a large

detector with ultra-precise timing

Conventional

TOF

TOF for rejecting background events (Compton scattering and random coincidences).

TOF for improving image S/N

Ultra-precise TOF gives direct 3D information

Wednesday, June 26, 13

9


Current commercial TOF-PET has ~ 500ps CTR (10 cm)

State of the art (in our lab) ~ 250 ps FWHM (4 cm)

270!

260!

250!

4x4 SiPM array coupled to the crystal array of 3.5x3.5x15 mm3

in coincidence with 2x2x10 mm3 LYSO crystal

CTR ( ps)!

240!

230!

220!

210!

200!

72.8! 73! 73.2! 73.4! 73.6! 73.8! 74! 74.2!

SiPM bias Voltage (Volt)!

But to make a significant improvement one needs a FWHM of “100 ps”

i.e. 16 mm position resolution along line of flight

Wednesday, June 26, 13

10


Can we approach this 100 ps limit

Concept of differential readout and the NINO asic (this asic

was developed a decade ago as the ‘obvious solution’ for

TOF arrays with MRPC : nowadays people are beginning to

realise the importance of this development)

Silicon photomultiplier

Wednesday, June 26, 13

11


Differential or single-ended readout

Basic schematic of all detectors

Negative electrode

+

-

input impedance

Positive electrode

all signals in all detectors are due to the movement of charge

between two readout electrodes. Equal and opposite signals are

produced on these two electrodes and current flows through

the external circuit.

Wednesday, June 26, 13

12


one element fires

Channel 1

Channel 1

Channel 2

Channel 3

input impedance

input impedance

input impedance

Single-ended

readout pumps

spikes of current

into common

ground. The

ground is the

reference for all

the discriminators.

Channel 2

Channel 3

input impedance

input impedance

input impedance

detector ground

Wednesday, June 26, 13

Channel 4

single

ground of

front-end electronics

ended

input impedance

Result: crosstalk

and additional

timing jitter.

differential

13

Channel 4

input impedance

Signal current flows through

front-end and back to where it

came from. No current injected

into common ground.

Result: No crosstalk and no

extra electronic jitter


So our message is:

differential readout is essential to minimise

cross-talk and common mode noise (in the ground)

and is absolutely necessary for precise timing

So for fast timing: one needs

(a) fast differential amplifier/discriminator

(i.e. the NINO)

(b) detectors that generate a differential signal

Wednesday, June 26, 13

14


The best/only differential discriminator ASIC so far !

The NINO

Wednesday, June 26, 13

15


The Silicon photomultiplier

SPAD

~ 4000 SPADs (single photon avalanche diode)

• operated in Geiger mode; thus single photon generates detectable signal

Wednesday, June 26, 13

16


22

Na point

source

2 #mes 4x4 crystal array of 3.1x3.1x15 mm 3

each coupled to a monolithic SiPM array.

CTR ~ 350 ps FWHM

Bias Voltage = 73.6 V on both MPPCs

NINO Threshold =80 mV

Where we were:

January 2013

Counts

70

60

50

40

30

20

FWHM=350.37 ps

1

FWHM=341.09 ps

70

140

FWHM=354.49 ps

160 FWHM=335.92 ps

D4/D1 D3/D2 D2/D3 D1/D4

60

120

140

50

100

120

100

40

80

80

30

60

60

20

40

40

Counts

Counts

Counts

10

10

20

20

0

-8800 -8600 -8400 -8200 -8000 -7800

delay (ps)

0

-8800 -8600 -8400 -8200 -8000

delay (ps)

0

-9000 -8800 -8600 -8400 -8200 -8000

delay (ps)

0

-8800 -8600 -8400 -8200 -8000 -7800

delay (ps)

Counts

Counts

100 FWHM=333.86 ps

90 FWHM=506.6 ps

C4/C1 C3/C2 C2/C3

100

160

80 C1/C4

80

140

70

80

120

60

60

100

50

60

80

40

40

40

60

30

Counts

Counts

20

120

FWHM=358.60 ps

delay (ps)

20

40

20

180 FWHM=333.86 ps delay (ps)

20

10

0

-8800 -8600 -8400 -8200 -8000

delay (ps)

0

-9200 -9000 -8800 -8600 -8400 -8200

delay (ps)

0

-10800-10600-10400-10200-10000

0

-7000-6800-6600-6400-6200-6000

Wednesday, June 26, 13


Where we are now

4x4 discrete SiPM array with differential readout

Counts

140

s

120

FWHM=247.8 ps

Counts

140 FWHM=263.5 ps

120

100

100

80

80

60

60

40

40

20

20

0

0 500 1000 1500 2000 2500 3000

delay (ps)

0

0 500 1000 1500 2000 2500 3000

delay (ps)

Counts

120 FWHM=244.43 ps

100

80

60

Counts

120

100

80

60

FWHM=220.1 ps

40

40

20

20

Wednesday, June 26, 13

0

0 500 1000 1500 2000 2500 3000

delay (ps)

0

0 500 1000 1500 2000 2500 3000

delay (ps)


Why not improve TOF resolution to 100 ps FWHM

The Crystals:

high light yield, fast, dense etc

but in general high refractive index (1.82 for LYSO) and difficult to get light out

The SiPM

As shown previously - need differential detector.

Need to minimise dark count noise ( we believe that we have the solution to that)

The Readout electronics

Differential discriminator/amplifier such as NINO but needs improvement (we believe

we will have a significant improvement in timing with a redesign)

also: we need funding for this project

Wednesday, June 26, 13

19


Summary

In the last decade we have seen significant progress in the field of

fast timing:

This is due to the MRPC, the NINO ASIC and Silicon PhotoMultipliers

Outlook:

Large TOF arrays with 20 ps time resolution with MRPCs

Pre-shower detectors with ultra-precise timing so that photons

can be associated to the correct interaction point.

TOF-PET with better than 100 ps FWHM time resolution so that

whole body scanners can be built with suitable sensitivity

( screening of prostate and pancreas cancer).

Wednesday, June 26, 13

20

More magazines by this user
Similar magazines