ZDD installation and performance BESIII Collaboration Meeting December
ZDD installation and performance BESIII Collaboration Meeting, December 2011 A. Calcaterra, for the ZDD group (LNF+TO)
Talk outline • ZDD performance in Frascati, Spring 2011 – Cosmic rays – Single-electron beam • ZDD installation, August 2011 • ZDD tests with cosmic rays at IHEP, Fall 2011 • Outlook – Short term – Next weeks 2011/11/30 A. Calcaterra 2
The ZDD in the East area Fiber bundles (2 m) Gearbox for minical movement W screen γ beam axis ZDD: Pb/Sci. Fi Array, scintillating material 60% of total (in volume), two modules and down the beam) dimensions: 14 x 4 x 6 cm 3 2011/11/30 A. Calcaterra (up 3
ZDD module segmentation 4 cm 14 cm 5 4 3 10 9 8 1 2 6 7 beam Each sector is sent to a PM, sectors 1&2 (6&7) are sent to the same PM (for now) 2011/11/30 A. Calcaterra 4
BTF test beam at LNF (may 16 -22 2011) One (out of two) ZDD module tested at BTF with 450, ~300, ~200 Me. V e- bunches (Ne-=1, 2, 3) Final Pb-scifi ZDD module, bundles guides, PM’s, TDC, at the moment not FADC but ADC caen V 792 N Small scintillator (60 x 11 x 4) mm 3 used to trigger and selectrons impact point 2011/11/30 A. Calcaterra 5
Setup in Cosmic rays (LNF) 2011/11/30 A. Calcaterra 6
Setup in Cosmic rays Top “finger” scintillator, (11 x 5 x 50) mm 3 Bottom “paddle” scintillator Trigger coincidence rate = (2 -3 minutes)-1 2011/11/30 A. Calcaterra 7
Purpose of cosmic rays DAQ • 2 types of data taking: integrated charge (QDC CAEN V 792 N) and lineshape (Flash ADC CAEN V 1721). Timing information is also present (TDC CAEN V 1190) but not systematic. • QDC data used for: – inter-channel calibration – resolution studies – absolute scale, comparing cosmics to single-electron • FADC data used for Me. V/m. V calibration 2011/11/30 A. Calcaterra 8
Run 224: QDC cts for each PM 2011/11/30 A. Calcaterra 9
HV correction table for “up” calorimeter Ch Ped(p. C) Peak(p. C) Delta(p. C) G/Gref old. V new. V 0 15. 962 38. 968 46. 011 0. 978 1. 434 1. 427 1 15. 611 63. 125 47. 514 0. 947 1. 442 1. 426 2 13. 941 63. 929 49. 988 0. 900 1. 432 1. 401 3 15. 491 60. 51 45. 019 1. 000 1. 349 4 15. 059 37. 434 44. 75 1. 006 1. 399 1. 401 5 15. 793 55. 505 39. 712 1. 133 1. 405 1. 443 6 14. 558 58. 434 43. 876 1. 026 1. 451 1. 459 7 13. 895 64. 744 50. 85 0. 885 1. 427 1. 391 2011/11/30 A. Calcaterra 10
Find passing tracks (left side) 2011/11/30 A. Calcaterra 11
4 Q passing tracks (“up” calorimeter) 2011/11/30 A. Calcaterra 12
Absolute scale and E/E • According to Montecarlo a passing cosmic track leaves 16 Me. V of energy in the scintillator • QDC scale (when “Happy Box” is used) is approximately 170 p. C / 16 Me. V = 11 p. C / Me. V. • For E/E we find 20/180 = 11%, consistent with simulation (no shower fluctuations). 2011/11/30 A. Calcaterra 13
Number of photoelectrons • Assuming a PM gain ≈ 1. 2· 106, the number of photoelectrons per cosmic track is: • The factor 2 in the denominator is due to the “Happy Box” • Approximately 1/7 th of these 443 p. e. develop in the first PM, 2/7 th in each of the other 3 PMs 2011/11/30 A. Calcaterra 14
Photoelectrons/Me. V (cosmics) • According to Montecarlo 16 Me. V are deposited in the scintillator by cosmic track, 50 Me. V in all. This means: • 28 photoelectrons/Me. V in the fibers (14 cm) • 9 photoelectrons/Me. V in the whole calorimeter 2011/11/30 A. Calcaterra 15
ZDD as a new luminometer(luminosity monitor) Frascati cosmic ray test shows: The time resolution is 0. 97 ns, which meets the requirement of a luminometer(<4 ns) The signal width is only 5. 2 ns, so dead time is very little. • • • The old luminometer on east side of BESIII was uninstalled and replaced by ZDD signal is fanned out as luminometer. The electronics and DAQ for luminometer are kept the same. Tested by the noise, the new luminometer system is working properly. Its performance as a luminometer shall be checked under colliding mode of BEPCII. Slide by Xue Zhen 2011/11/30 16
Tests at the BTF Single-electron beam from Frascati Beam Test Facility trigger (6. 0 x 1. 1 x 0. 4)cm 3 finger scintillator • The BTF: few-electrons, 50 Hz pulses from DAFNE Linac • Minicalorimeters rotated, fibers vertical • Trigger on AND of RF signal and «finger» • Data taken mostly with QDC (FADC electronics available only at the end of our beam time) 2011/11/30 A. Calcaterra 17
Purpose of single-electron DAQ • Study response to single electrons of different energies • Study • resolution with single electrons • absolute scale factor, Me. V/p. C • photoelectron statistics • How does these data compare to cosmicrays ones? 2011/11/30 A. Calcaterra 18
QDC 8 Q, 450 Me. V, PM equalized 38. 43/294 = 13% 36. 84/296 = 12. 4% 2011/11/30 A. Calcaterra 19
Absolute scale • According to Montecarlo a 450 Me. V electron leaves 12%· 450 = 54 Me. V of energy in the scintillator • If the absolute scale from cosmics is right, we should see 54 Me. V· 5. 5 p. C/Me. V=297 p. C (“Happy Box” X 2 preamp was not in use yet) • …. and we do! Perfect! 2011/11/30 A. Calcaterra 20
Number of photoelectrons • Assuming a PM gain ≈ 1. 2· 106, the number of photoelectrons at 450 Me. V is: • According to MC (12%· 450 = 54 Me. V deposited in the scintillator) this means 30 photoelectrons per Me. V (28 in cosmics) • How are these divided among the strata? 2011/11/30 A. Calcaterra 21
Run 4 at 450 Me. V 32/300=10% 150/300=50% 86/300=30% 2011/11/30 31/300=10% A. Calcaterra 22
The ZDD installation war! Arrival at IHEP 2011/11/30 A. Calcaterra 23
The ZDD installation war! 2011/11/30 A. Calcaterra 24
The ZDD installation war! 2011/11/30 A. Calcaterra 25
The ZDD installation war! 2011/11/30 A. Calcaterra 26
The ZDD installation war! Thanks Mario and Zhen for a fantastic effort! 2011/11/30 A. Calcaterra 27
Cosmic rays at IHEP • There is no external trigger • Auto-generated trigger: ≥ 2 out of 8 Flash. ADC channels must have a minimum below some threshold (baseline-3 cts = baseline-12 m. V) 2011/11/30 A. Calcaterra 28
Analysis still very preliminary • Fitting the waveforms we obtain for each channel a minimum and a time of minimum – Noise shows up mainly in 2 -hits events (peak times differ randomly) – 3 -hits events are much cleaner • …still, very difficult to define a passing track, due to the very small solid angle • No control over the track length, many different track lengths in the samples 2011/11/30 A. Calcaterra 29
Tue 2011/11/22, 6. 5 hours of data Horizontal scale ns, 1 bin=2 ns. We plot Dt if majority=2, or the biggest time difference out of 3 (majority=3) or out of 6 (majority=4) 2011/11/30 A. Calcaterra 30
Next steps (very soon) • Use BESIII signals into our own DAQ instead of auto-generated one but read data onto our separate PC. • Implement the TDC (100 ps resolution) • Estimate data size in real running conditions – Expected size = 1 byte per channel per BESIII trigger per window size (1 sample/2 ns) – 200 ns window 16 bytes*100 samples 2011/11/30 A. Calcaterra 31
Next steps (not much later) • After data size is known (and accepted )… • …and we demonstrate that the data may be useful… • …then, we will finally ask to incorporate VME readout into BESIII general dataflow • At this point, help will be needed from expert BESIII DAQ- and offline-reconstruction persons 2011/11/30 A. Calcaterra 32
Conclusions • The ZDD has been designed, built, and tested in record-time • The design is sound and the performances are acceptable given the relatively little time and effort left for data taking and analysis • The installation was hard but the final result is satisfactory • Next stop……. physics results 2011/11/30 A. Calcaterra 33
A final thought of wisdom “One knows very well that, in reducing ideals to practice, great latitude of tolerance is needful; very great” T. Carlyle 2011/11/30 A. Calcaterra 34
Spares 2011/11/30 A. Calcaterra 35
The HV correction • Let’s choose a “target” amplification (45 p. C) • In column 5 we find the factor to correct for • According to • We compute 2011/11/30 A. Calcaterra 36
PM HV calibration • Initial 8 -channel equalization done on the basis of Hamamatsu individual datasheets • For all events, find pedestals and peaks • Choose one channel as normalization • Assuming (Hamamatsu datasheets) • Invert formula and find new Vn’s 2011/11/30 A. Calcaterra 37
FADC scale calibration 2011/11/30 A. Calcaterra 38
Run 228, cosmics, QDC data Ch Ped (p. C) Peak (p. C) Signal (p. C) /E(%) 0 16. 1 (27. 2± 0. 3) 11. 1 (4. 2± 0. 3) 38% 1 15. 7 (36. 2± 0. 9) 20. 5 (7. 2± 1. 0) 35% 2 14. 1 (35. 9± 1. 0) 21. 9 (6. 8± 0. 7) 31% 3 15. 6 (37. 6± 0. 7) 22. 0 (5. 5± 0. 8) 25% 2011/11/30 A. Calcaterra 39
Run 228, cosmics, FADC data 2011/11/30 Ch Peak (m. V) 0 (142± 10) 1 (323± 23) 2 (383± 25) 3 (316± 57) A. Calcaterra (m. V) /E (%) Peak in channels (63± 20) 44 2 1, 2, 3 is a factor higher than (109± 34) 38 in channel 0, just as (102± 40) expected. 27 This is a (162± 91) 51 good thing! 40
Conversion factors at 1. 4 k. V Ch Peak (p. C) Peak (m. V) Factor (m. V/p. C) Factor (m. V/Me. V) 0 11. 1 142 12. 8 70. 4 1 20. 5 323 15. 8 86. 9 2 21. 9 383 17. 5 96. 25 3 22. 0 316 14. 4 79. 2 • Run 228: no “Happy Box”: 5. 5 p. C/Me. V of scintillator-deposited energy • QDC+FADC cosmics data: unfortunately, a small sample (1 day only) 2011/11/30 A. Calcaterra 41
- Slides: 41