The ALFA test beam 2008 *** - atlas - Desy

The ALFA test beam 2008 *** A short summary and outlook K.Hiller, K Hiller DESY Zeuthen on behalf of the test beam crew Bruno Allongue, Allongue Francis Anghinolfi, Anghinolfi Anatoli Astvatsatourov, Astvatsatourov Pierre Andreas Barrillon, Barrillon Georges Blanchot, Andre Braem , Antonio De Aguiar E Pestana De Morais, Beniamino Di Girolamo, Sebastien Franz, Per Grafstrom, Maurice Haguenauer, Wieslaw Iwanski, Krzysztof Korcyl, Alessandro Mapelli, Alberto Marcalo Da Palma, Dieter Notz, Patrick Michel Puzo, Thomas Schneider, Hasko Stenzel Stenzel, Tomas Sykora , Dennis Petschull, Petschull Uli Koetz, Koetz Vit Vorobel, Vorobel Matthieu Heller

Æ Some images from the last run Æ The beginning in 251 ÆThe measurements in H8 - data review - man power - lessons for the next iteration - coming analysis scenario ÆSummary & outlook

The last run: August 28, 8:00 (1) ALFA, mapping of X Y position y [mm]

50 60

40

50

30

40

20

30

10

20

0

10

−10 −25 −20 −15 −10 −5

Run 60145 42 kevts central position X/Y = -3.5 / -0.1 HV = 900 V Gain = 16 Thermo Th =8

0

5

10

15

20 25 0 x [mm]

Nearly full detector coverage

Fibers vs Layer: event _0



fibers

fibers

fibers

The last run: August 28, 8:00 (2) 60

60

60

50

50

50

40

40

40

30

30

30

20

20

20

10

10

10

8

10

12

14

16

18

20 layers

2

4

6

8

10

12

14

16

18

2

20 layers

4

6



Fibers vs Layer: event _6 fibers

6

fibers

4

fibers

2

60

60 50

50

40

40

40

30

30

30

20

20

20

10

10

10

4

6

8

10

12

14

16

18

20 layers

2

4

6

8

10

12

14

16

18

20 layers

all events indicate track projections noisy & dead MAPMTs/PMFs rather th high hi h level l l off noise i hits hit & cross ttalk lk sometimes indications of mapping problems

10

12

14

16

18

20 layers

?

60

50

2

8

2

4

6

8

10

12

14

16

18

20 layers

Preparations in 251 long period of motherboard debugging until August 6 followed by 5 day of DAQ runs from August 7 - 11 LED instead of fiber detector in pot Lund motherboard with latency ~ 200nsec, power supply l ffailure il PMFs 1-8 with LVDS, 9-23 CMOS all Kapton rows in R/O HV: mostly 950 V, some data with 850, 900 V verify MAPMT and fiber mapping S-curve measurements with DAC = 250, …3000 runs on [email protected], l id @ tb tl04 h passwd: LumiDaq

Results from 251: mapping check HistPmPins_MD_9

MD MAPMT Fired Pins for PMF #9 8

59027

7

59029

6

57889

5 4

59029

8

46

7

29

6

34

59030

59031

59029

59029

59030

59030

59032

59029

59029

59032

59027

59028

59029

59030

59032

59032

59030

59030

59027

59031

59030

59028

59028

59031

59029

5

90

49

38

47

42

59007

59029

59029

59030

59029

59027

59030

59028

4

342

198

306

346

371

3

59028

59032

59031

2

59029

59030

1

59031

1

9

noisy

52

138

266

364

476

59029

59027

59030

59029

59027

59028

2

786

667

732

697

672

535

594

531

59030

59031

59029

59027

59031

59031

59030

1202

1092

1205

1141

1054

1067

994

1104

2

3

4

5

6

1

2

3

4

5

6

HistPmPins_MD_11

12

12

11

18

129

135

113

128

4

644

480

483

621

623

527

417

3

724

726

820

766

772

590

2

840

821

804

818

679

667

1

1672

1357

1497

1355

1379

1337

1194

1

2

3

4

5

6

69

6

1876

497

384

426

370

384

325

1306

245

5

6403

4748

4324

4327

5066

3747

3906

4019

639

4

22635

18876

17893

18918

19555

14463

14149

12892

698

849

3

23484

19817

21160

21810

21148

17479

15385

14515

740

768

2

23819

21964

20284

19666

16495

18654

16639

14962

1193

1

32514

29431

29786

28988

26715

26441

24357

20180

1

2

3

4

5

6

30

21

4

2

5

7 8 MAPMT Fired Pins

HistPmPins_MD_21

MD MAPMT Fired Pins for PMF #21

21 2

278

Entries Mean x Mean 266y RMS x RMS y

12

65

552

493

449

394

100

92

112

196


8

747

437

Entries Mean x Mean65 y RMS x RMS y

467

440

302

147

86

242

204

242

109

23

8

258

266

218

105

14

1

7

21

1

4

21

194

208

196

7

450

51

217

206

163

216

6

442

39

236

237

188

223

5

516

22 255

233

230

107

1

55

246

215

188

235

4

462

312

276

237

87

3

8

7

43

205

215

169

198

3

480

280

263

198

93

14

5

11

2

533

271

318

281

96

30

6

21

957

747

750

659

364

232

143

109

1

2

3

4

5

6

3

3

41

170

209

203

231

1

43

51

69

125

206

299

305

290

1

2

3

4

5

6

1


HistPmPins_MD_23


57

52

61

87

Entries Mean x Mean61y RMS x RMS y

13886 4.393 2.179 2.43 1.771

307

267

207

218

T

9

4

8

3

7

2

6

1

5

13

12

X

18

23

17

22

16

21

11

15

20

10

14

19

HistPmPins_MD_20

35

21

7

66

3

2

4

4

5

7

58

7

227

192

259

163

53

2

1

6

113

8

5

3

5

9

7

85

6

270

163

168

129

54

6

1

5

181

37

28

26

26

28

27

98

5

254

142

138

4

341

152

121

184

175

131

99

167

4

292

172

3

370

168

194

204

165

142

131

252

3

284

2

389

202

210

184

180

150

148

269

2

1

945

859

979

1000

985

1000

970

1098

1

1

2

3

4

5

6


14914 2.779 4.151 1.681 2.641



11 MAPMTs covered ½ by black tape with unique pattern

HistPmPins_MD_22


60

5

23

466

86

3

2

42

7843 6.118 4.426 1.484 2.46

405

3

2

1

304

Entries 713784 Mean x 4.182 Mean 944y 2.539 RMS x 2.267 RMS y 1.38

1105

34

1

270


28

5

103

29188 4.405 2.342 2.326 1.382

HistPmPins_MD_12

7

20

4

62


21999 4.367 2.195 2.308 1.325


8

18

40

1


9

71

44

268

451

115

1

78

501

13

1

2

639

130

3

3

1

608

179

4

2

4

621

5

3

6

597

11

5

3

4

660

54

2

22

3

6

2

7

23

59029

36

6

2

10


23

59027

33

7

21

59030

7

36

24

59031

8

8

19

59028


8

HistPmPins_MD_10


Entries 3776719 Mean x 4.501 59033 Mean y 4.5 RMS x 2.291 RMS y 2.291 59027

59028


3

122

60

240

166

55

6

1

6

172

147

158

53

4

2

313

151

186

178

76

11

3

5

489

310

416

264

142

64

61

32

1

2

3

4

5

6

dummy

6


7947 2.628 4.164 1.504 2.495

this pattern was found in the MAPMT pixel plots after reconstruction (not yet ?)

Results from 251: S-curves

Pedestals only

Threshold Th h ld th thermo = 3 sufficient ffi i t to cut most electronic noise Efficiency cy

S-curve, 0

at thermo = 7..8 7 8 low light signals disappear

1

0.8

only rudimental measurements possible due to instability of DAQ

0.6

0.4

With LED

0.2

Î ss-curve curve measurements can only be used for qualitative conclusions

0 1

2

3

4

5

6

7

8

9 Voltage

In the H8 test beam area Before installation …

… with ALFA

Tunnel view with Si-telescope

Data Overview: 3 periods 1) Lund-1 data - malfunctioning motherboard, latency ~ 200nsec due to power supply failure - PMFs 1-8 with LVDS, 9-23 CMOS - eye y tracking g efficiency y ~ 10% - ALFA - fiber-silicon DAQ synchronization stable, few events offset 2) Orsay data - all PMFs with LVDS ((?)) - parameter scan: * HV: 950, 900, 850 V * DAC: 1000, 1500, 2000, 2500, 3000 * Gain: 8, 16, 24, 32, uniformity y corrections * Kaptons rows in R/O: 1+2, 4+5, 3+5, 2+4, 2+5, 1+5 * X/Y: 0.0/0.0, -0.1/10.0, -0.1,15.0, -20.0/0.0, +20.0/0.0, +22.0/0.0, - fiber-silicon DAQ synchronization stable, few events offset

150 000 triggers 150.000

650.000 triggers

3) Lund-2 data - Lund motherboard with latency ~ 1000nsec, power supply failure cured - PMFs 1-4 with LVDS, 6-7 with CMOS, 8-10 LVDS, 11-23 CMOS - reduced p parameter scan: * HV: 900, 950 V * Gain: 16, uniformity corrections * DAC: 1000, 1500, 2000, 2500 * X/Y: -2.9/+10.0, -6.0/+10.0, -17.9/+5.0, +6.0,/+5.0, +11.0/+10.0, -3.5/+5.0, -3.5/-0.1 - eye tracking efficiency ~ 80% - fiber-silicon synchronization bad, each spill add. 0(5) Si-events 370.000

triggers

Man Power Situation up to last ~ 2 days mostly working g in expert p mode majority of shifters left before stable running Î no real benefit f from f available man power

Many thanks to all:

Lund-2 Lund-1

Orsay

Experts E t always l on call, ll shifters dealing with frequent changes in soft + hardware

Lessons learned: a few examples Mechanics: detector installation tool (bottom-up version !!!) shielding hi ldi off overlap l ttrigger i til tiles effective dark box Æ see full list from Andre & Thomas Electronics: HV-cabling with access to PMFs, w/o short circuits motherboard: many problems understood understood, final version ? trigger electronics: completely missing ! Æ see full list from Francis Software: DAQ software rather outdated ALFA reconstruction software at veryy beginning g g how to get best benefit from fiber metrology in case of plate-to-plate coordinate jitters ? g point to improve are test beam data Æ best starting

Hard to imagine as final cabling

An analysis scenario 1) Si – tack reconstruction use the existing RD42 package for alignment and Si-track reconstruction Karlheinz, Karlheinz Antonio Antonio, Alberto 2) ALFA reconstruction: check / adjust mapping tables select and fit track candidates, different models possible calculate offsets and rotations of individual detector layers calculate resolution,, efficiency, y, noise for all parameter p samples p Anatoli, Alessandro, Dennis, Hasko, Matthieu, Sebastien, Karlheinz 3) ALFA versus Si-tracks resolution, efficiency, noise with clear reference track all people as before 4) GEANT3/4 simulation Vit, Tomas, Antonin Î man power is i an proposall to t be b discussed di d and d agreed d Î response time depends much on man power and experience !

Where we are ? We expect from this test beam measurements a realistic feedback about the data quality of a nearly final Roman Pot as it will be installed in the tunnel in terms of: resolution & efficiency of fiber tracking trigger efficiency handling experience, detector installation etc. test under pre-vacuum need and feedback of cooling options test of alignment for max. use of staggering

At the end we collected about 1.2 million triggers, but: most of them not with final electronics (faulty Lund board, board 2 Orsay boards) due to lack of time the most realistic data taking was compressed in 2 – 3 days, originally planned 6 + 20 days ! not all parameter options well measured with sufficient statistics most technical aspects with temporary workarounds, workarounds no tunnel solutions

My personal summary: 1) we got very important feedback in many technical aspects 2) we will learn a lot from the coming data analysis about data quality and where improvements are needed 3) the th step t ffrom proto-2 t 2 tto th the ttunnel-version l i iis ttoo llarge, tto kkeep th the risk i k on a reasonable level a next iteration of prototype is needed before installation