A CLU ster COU nting D rift C hamber for ILC

1. A CLUster COUnting Drift Chamber for ILC F.Grancagnolo, INFN – Lecce

2. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- outline • Requirements for tracking at ILC • Cluster Counting • 4thConcept CLUCOU Drift Chamber • 4thConcept MUon Detector

3. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- borrowed from:

4. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- From Gluckstern: @ ILC, for B = 5 T, L = 1.5 m N = 150 , L ~ 2m (1 cm2 hex. cells) 60.000 sense wires 120.000 field wires

5. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Multiple scattering contribution (equivalent L/X0): 60.000 20 mm W sense wires → 1.8 × 10-3 (X0 = 0.35 cm) 120.000 80 mm Al field wires → 2.2 × 10-3 (X0 = 8.9 cm) 2 m gas (90% He + 10% iC4H10) → 1.5 × 10-3 (X0 = 1300 m) Equivalent L/X0 = 5.5 × 10-3 Sagitta measurement contribution (in  plane):

6. this CLUCOU Drift Chamber F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Momentum Resolution 4thConcept TPC (160 pads) 4thConcept TPC + 5 pixel planes

7. 2 cm drift tube 90%He-10%iC4H10 few x 10 gain 5 F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- CLUster COUnting Dt : time separation between consecutive ionization clusters, as a function of their ordered arrival time, for different impact parameters. (caveat: electrons!) i+1,i In this He mixture , provided that: sampling frequency of signals > 2 Gsa/s and rise (and fall) time of single electron signals < 1ns single electron counting is possible.

8. b=0 b=1 cm 10 mV 500 ns tmax=1.6 ms tfirst tlast t0 trigger signal F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- CLUster COUnting MC generated events: 2cm diam. drift tube gain = few x 10 gas: 90%He–10%iC4H10 no electronics simulated vertical arbitrary units cosmic rays triggered by scintillator telescope and readout by: 8 bit, 4 GHz, 2.5 Gsa/s digital sampling scope through a 1.8 GHz, x10 preamplifier

9. tfirst t0 Ncl! Ncl-i i-1 Ncl (Ncl-i)! (i-1)! F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- CLUster COUnting CLUster COUnting For a given set up, and a digitized pulse (tlast is constant with a spread < 20 ns) t0 = tlast–tmax gives the trigger time bf = ∫ v(t) dt first approx. of impact parameter b (c/2)2 = r2 - bf2 length of chord Ncl = c / (l(bg) ×sinq) expected number of cluster Nele = 1.6 x Ncl expected number of electrons (to be compared with counted one) {ti } and {Ai }, i=1,Nele ordered sequence of ele.drift times and their amplitudes P(i,j), i=1,Nele, j=1,Ncl probability i-th ele.  to j-th cl. Di (x) = (1-x) x probability density function of ionization along track

10. sb Fit to resolution function: (from KLOE data) l = 770 14 mm np = 12.9  0.1 cm-1 sD = 140 mm cm-1/2 st = 4.8  0.2 ns b F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- • Each cluster contributes to the measurement of the impact parameter with an • independent estimate weighted according to the Poisson nature of the process • and the electron diffusion along the drift path. • The resolution on the impact parameter, sb, improves with the addition of each • cluster beyond the first one. • It, however, saturates at a value of 30-35 mm, convolution of: • spread in mechanical tolerances (position of sense wire; gravitational sag; electrostatic displacement) • timing uncertainties (trigger timing; electronics calibration; t0) • degree of knowledge of time-to-distance relation • instability of working parameters (HV, gas temperature and pressure, gas mixture composition) Reasonable to assume sb = 50 mm per sense wire

11. experiment: • p/m = 1.3 s • theory:trunc. mean: • p/m = 2.0 sp/m = 0.5 s gas mixture = 95%He+5%iC4H10 Ncl = 10/cm beam test measurements p = 200 MeV/c CLUCOU chamber expected dNcl/dx resolution for a 2 m m.i.p. at 13 cluster/cm: s(dNcl/dx)/(dNcl/dx) = 2.0 % F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Particle Identification

12. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- 4thConcept ILC Drift ChamberLayout and assembly technique Length: 3.4 m at r = 22.5 cm 3.0 m at r = 147.0 cm Spherical end plates: C-f. 12 mm + 30 mm Cu (0.047 X0) Inner cylindrical wall: C-f. 0.2 mm + 30 mm Al (0.001 X0) Outer cylindrical wall: C-f./hex.cell. sandwich held by 6 unidir. struts 0.020 X0) Retaining ring Stiffening ring TOTAL = 0.028 X0 Gas = 0.0015 X0 Wires = 0.0040 X0

13. 4thConcept ILC Drift ChamberLayout F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Hexagonal cells f.w./s.w.=2:1 cell height: 1.00  1.20 cm cell width: 6.00  7.00 mm (max. drift time < 300 ns !) 20 superlayers, in 200 rings 10 cells each (7.5 in average) at alternating stereo angles ±72  ±180 mrad (constant stereo drop = 2 cm) 60000 sense w. 20 mm W 120000 field w. 80 mm Al “easy” t-to-d r(t) (few param.) >90% sampled volume

14. Summary for ILC F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- • A drift chamber à la KLOE with cluster counting (≥ 1GHz, ≥ 2Gsa/s, 8bit) • uniform sampling throughout >90% of the active volume • 60000 hexagonal drift cells in 20 stereo superlayers (72 to 180 mrad) • cell width 0.6 ÷0.7 cm (max drift time < 300 ns) • 60000 sense wires (20 mm W), 120000 field wires (80 mm Al) • high efficiency for kinks and vees • spatial resolution on impact parameter sb = 50 mm (sz = 300 mm) • particle identification s(dNcl/dx)/(dNcl/dx) = 2.0% • transverse momentum resolution Dp/p = 2·10-5 p 5·10-4 • gas contribution to m.s. 0.15% X0, wires contribution 0.40% X0 • high transparency (barrel 2.8% X0, end plates 5.4%/cosq X0+electronics) • easy to construct and very low cost • is realistic, provided: • cluster counting techique is at reach (front end VLSI chip) • fast and efficient counting of single electrons to form clusters is possible • 50 mm spatial resolution has been demonstrated

15. G.V. Fedotovich and the CMD-3 Collaboration F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Drift Chamber for CMD-3 Detector @ VEPP-2000, Novosibirsk

16. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- to be presented at

17. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- during CLUCOU Proposal presented @

18. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- rendering of first two superlayer stereo

19. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC ---

20. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC ---

21. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC ---

22. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- fraction of hit cells per ring highest multiplicity events: 300 cells superlayer 680 cells ≥ 300 tracks per such event average number of tracks crossing a hit cell

23. diameter x length 12.8 m x 15.4 m 7.7 m 6.4 m F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- 4thConcept layout

24. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- 4thConcept Dual Solenoid m - BARREL MUD barrel MUD E C N A D P CLUCOU chamber

25. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- m-system basic element: drift tube • radius 2.3 cm • filled with 90% He – 10% iC4H10 @ NTP • gas gain few × 105 • total drift time 2 µs • primary ionization 13 cluster/cm≈ 20 electrons/cm total • both ends instrumented with: • > 1.5 GHz bandwith • 8 bit fADC • > 2 Gsa/s sampling rate • free running memory • for a • fully efficient timing of primary ionization: cluster counting • accurate measurement of longitudinal position with charge division • particle identification withdNcl/dx

26. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Drift tube end plug detail

27. ×18 1750 tubes 70 cards 650 tubes 26 cards ×18 ×36 550 tubes 22 cards F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Modularity

28. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- 1/3 of barrel ×3 10500 tubes 420 cards

29. 1440 tubes 1632 channels 76 cards ×6 F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- 1/3 end cap

30. ×2 F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- One end cap

31. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Full muon system Barrel: 31500 tubes 21000 channels 840 cards End caps: 8640 tubes 9792 channels 456 cards Total: 40140 tubes 30792 channels 1296 cards

32. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- m+m−at 3.5 GeV/c

33. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Downscaling to SuperB

34. F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- SuperB Drift ChamberLayout and assembly technique Length × Diameter : ~ 2.8 m × 0.8 m Spherical end plates: C-f. 6 mm equivalent + 30 mm Cu (0.024 X0) Inner cylindrical wall: C-f. 0.2 mm + 30 mm Al (0.001 X0) Outer cylindrical wall: C-f./hex.cell. sandwich held by 6 unidir. struts 0.020 X0) Retaining ring Stiffening ring TOTAL = 0.023 X0 Gas = 0.0014 X0 (80%He+20%iC4H10) Wires = 0.0004 X0

35. SuperB Drift ChamberLayout F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- Hexagonal cells f.w./s.w.=2:1 cell height: ~ 1.70 cm cell width: ~ 10.mm (max. drift time < 700 ns !) 10 superlayers in 100 rings 10 cells each (7.5 in average) at alternating stereo angles ±72  ±130 mrad (constant stereo drop = 2 cm) 12000 sense w. 20 mm W 24000 field w. 80 mm Al “easy” t-to-d r(t) (few param.) >90% sampled volume

36. Summary for SuperB F. Grancagnolo. INFN – Lecce--- CLUCOU for ILC --- • A drift chamber à la KLOE with cluster counting (≥ 1GHz, ≥ 2Gsa/s, 8bit) • uniform sampling throughout >90% of the active volume • 12000 hexagonal drift cells in 8 stereo superlayers (72 to 130 mrad) • cell width ~ 1.0 cm (max drift time < 700 ns) • 12000 sense wires (20 mm W), 24000 field wires (80 mm Al) • high efficiency for kinks and vees • spatial resolution on impact parameter sb = 50 mm (sz = 400 mm) • particle identification s(dNcl/dx)/(dNcl/dx) = 2.3% • transverse momentum resolution Dp/p = 1.0·10-3 p 1.5·10-3 • gas contribution to m.s. 0.14% X0, wires contribution 0.03% X0 • high transparency (barrel 2.2% X0, end plates 2.6%/cosq X0+electronics) • easy to construct and very low cost • is realistic, provided: • cluster counting techique is at reach (front end VLSI chip) • fast and efficient counting of single electrons to form clusters is possible • 50 mm spatial resolution has been demonstrated