Photodetector Timing Resolution

Burle Micro-Channel Plate Photo

Multiplier Tubes (MCP-PMTs)

H. Wells Wulsin

SLAC Group B

Winter 2005

31 March 2005 H.W. Wulsin 2

Photo Multiplier Tube Specifications

DIRC PMT 2x2 MCP-PMT

Pixel Size ~2.8 cm ~24 mm

Timing resolution (σ)

1.6 ns [single photo-electrons]

•80-100 ps [single photoelectrons]•30-50 ps [test beam: ~60 photoelectrons]

31 March 2005 H.W. Wulsin 3

End Station A Setup

PiLas laser, 635 nm (int. trigger)

Myl

ar

atte

nuat

ors

Electronics:•ADC•Single threshold discriminator •TDC (25 ps/cnt)

Quartz barsPad 0

Pad 2

Pad 1

Pad 3

Burle MCP-PMT

Path of test beam

31 March 2005 H.W. Wulsin 4

Limits to MCP Timing Resolution

Cross-talk between channels Charge-sharing between pads Electronics noise Avalanche fluctuations [focus of this analysis]

31 March 2005 H.W. Wulsin 5

Timing resolution varies with ADC

• TDC measures hit time when pulse exceeds level set by single threshold discriminator.• Size of pulse (# photoelectrons) affects the measured hit time. • More photoelectrons reduce the variability of measured hit time (avalanche fluctuations have a smaller effect).

Single photoelectron pulses (2 mV/div) ~300 photoelectrons / pulse (200 mV/div)

Threshold Threshold

1 ns

1 ns

31 March 2005 H.W. Wulsin 6

ADC Distribution of Runs 1-8, taken 11-22-04, 11-24-04

Ent

ries

ADC (counts)

31 March 2005 H.W. Wulsin 7

TDC v. ADC, Runs 1-8

ADC (counts)

ADC

TD

C (

coun

ts)

TD

C

31 March 2005 H.W. Wulsin 8

TDC Histogram, Runs 1-8en

trie

s

TDC (counts)

31 March 2005 H.W. Wulsin 9

Linear fit to Run 8

TDC = -0.0209(ADC) + 640.1

ADC (counts)

TD

C (

coun

ts)

31 March 2005 H.W. Wulsin 10

Linear correction to Run 8

ADC ADC

TDCTDC

entr

ies

TD

C

TD

C

entr

ies

σraw = 31.0 ps

σcorr = 30.0 ps

31 March 2005 H.W. Wulsin 11

Linear fit to Run 7

TDC = -0.098(ADC) + 671.5

ADC (counts)

TD

C (

coun

ts)

31 March 2005 H.W. Wulsin 12

Linear correction to Run 7

σraw = 53.5 ps

σcorr = 48.5 ps

ADC ADC

TDC TDC

TD

C

TD

Ce

ntr

ies

en

trie

s

31 March 2005 H.W. Wulsin 13

Hyperbolic fit to Runs 1-6T

DC

(co

unts

)

ADC (counts)

TDC = 632.6 + 509.8 / (ADC – 258.6)

31 March 2005 H.W. Wulsin 14

Hyperbolic correction to Runs 1-6

ADC ADC

TDC TDC

TD

Ce

ntr

ies

en

trie

sT

DC

σraw = 470 ps σcorr = 320 ps

31 March 2005 H.W. Wulsin 15

Independent fits to three data sets

590.0

610.0

630.0

650.0

670.0

690.0

710.0

25

9

28

0

30

1

32

2

34

3

36

4

38

5

40

6

42

7

44

8

46

9

49

0

51

1

53

2

55

3

57

4

59

5

ADC (counts)

TD

C (

co

un

ts)

Linear: Run 8

Linear: Run 7

Hyperbolic: Runs 1-6

31 March 2005 H.W. Wulsin 16

Optimized total fit

By Hand:TDC = 619 + 165/√(ADC – 262)

By Recursion:TDC = 619.2 + 160.7/√(ADC – 262.2)

590.0

610.0

630.0

650.0

670.0

690.0

710.0

259

278

297

316

335

354

373

392

411

430

449

468

487

506

525

544

563

582

ADC (counts)

TD

C (

cou

nts

)

Linear: Run 8

Linear: Run 7

Hyperbolic: Runs 1-6

Total Fit

ADC (counts)

TD

C (

cou

nts

)

31 March 2005 H.W. Wulsin 17

Total correction for Runs 1-8

ADC

ADC ADC

ADC

TD

CT

DC

TD

CT

DC

31 March 2005 H.W. Wulsin 18

Timing Resolution Correction: All Runs

FWHM ~ 50 counts ADC FWHM ~ 15 counts ADC

TDC (counts)TDC (counts)

en

trie

s

en

trie

s

31 March 2005 H.W. Wulsin 19

Use Run 1 (maximum attenuation): ADC counts photoelectrons

Pedestal value: 265.4 counts ADC Single photoelectron mean ADC: 2.1 counts

ADC (counts)

en

trie

s

en

trie

s

No TDC Signal TDC Signal Recordedentries

ADC - Pedestal (counts)

entries

ADC (counts)

entries No TDC Signal TDC Signal Recorded

1 photoelectron = 2.1 counts ADC

31 March 2005 H.W. Wulsin 20

Timing Resolution vs. Photoelectrons

Npe

σ (

ps)

σ (

ps)

σ = 9.08 + 214.61 / √(Npe – 0.45)

Npe

31 March 2005 H.W. Wulsin 21

Comparison to Previous Analysis

photoelectronsphotoelectrons

Time resolution vs (ADC signal)Quartz bar pad 0

10

100

1000

0 50 100 150 200 250

ADC signal [photoelectrons]S

igm

a [p

s]

Sigma 1 rawSigma 1 corrected1/sqr(ADC signal) corrected1/sqr(ADC signal) rawSigma 1 power corrected1/sqr(ADC signal) power corrected

10

100

1000

0 50 100 150 200 250

Photoelectrons

Sig

ma

(ps) σTDC raw

σTDC corr

Fit Corr

Similar resolution spectrum

Wells Josef (different data runs)

31 March 2005 H.W. Wulsin 22

Conclusions Achieved timing resolution of 30 – 50 ps for

~100 photoelectrons (Run 8) down to ~30 photoelectrons (Run 7).

Correction for timing – photon correlation most helpful in comparing disparate runs. For constant number of photoelectrons,

correction less useful. Further study:

Confirm # photoelectrons per ADC count. Examine charge-sharing between pads.

31 March 2005 H.W. Wulsin 23

Thank you to all of Group B for your help all quarter!Thanks to all Group B for lots of help this quarter!