DCH Summary G Finocchiaro For the DCH group

DCH Summary G. Finocchiaro For the DCH group XIV Super. B Meeting LNF, 1 October 2010

Topics Discussed in This Meeting • • Lab activities Background studies Cell design optimization DCH Readout architecture 01/10/2010 G. Finocchiaro 2

LAB activities @LNF – Prototype 1 Continuing cosmic ray data taking with “Proto 1” MT track PROTO 1 track – 6 x 4 BABAR–like hex cells – external tracker Al-Mylar with ≤ 80 mm extrapolation accuracy windows – Campaign with He + various i. C 4 H 10/CH 4/C 2 H 6 mixtures and HV / threshold settings 01/10/2010 G. Finocchiaro 10 cm lead pcut~150 Me. V/c 3

Space-time correlations Example: 60%He-40%C 2 H 6 @HV=2010 V • Measured time vs. extrapolated impact parameter – t vs. R • Fit R vs. t relationship using Chebyshev polynomials of order (up to) 5 – Time range rescaled to [-1, +1] interval Drift velocity 01/10/2010 G. Finocchiaro 4

Space-time relations fits (cont. ) • Internal cells are symmetric • Border cells show some L-R asymmetry L+R RIGTH LEFT 60%He-40%C 2 H 6 01/10/2010 G. Finocchiaro HV=2010 V 5

Track fit residuals & Spatial resolution – 60%He-40%C 2 H 6 HV=2010 V 01/10/2010 G. Finocchiaro 6

NEW PRESENTED @ ELBA Gas Mixture Comparison 80%He-20%i. C 4 H 10 90%He-10%i. C 4 H 10 95%He- 5%i. C 4 H 10 35%He-65%CH 4 51%He-49%CH 4 67%He-33%CH 4 79%He-21%CH 4 60%He-40%C 2 H 6 73%He-27%C 2 H 6 79%He-21%C 2 H 6 01/10/2010 r(g/cm 3) x 1 04 6. 3 4. 0 4. 9 4. 1 3. 3 2. 7 6. 4 4. 9 4. 2 X 0(m) NP(cm-1) NT(cm-1) tmax(ns) s(mm) 810 1410 1010 1260 1690 2270 780 1080 1310 21 13 9 26 20 15 11 19 15 12 47 26 15 37 29 21 15 49 36 30 420 550 700 210 300 350 470 550 600 650 • tmax and space resolution depend on other parameters. Figures shown are preliminary and indicative • Results for i. C 4 H 10 and CH 4 –based gas mixtures to be updated with analysis software used for C 2 H 6 gases G. Finocchiaro 130 170 200 220 200 130 130 7

Drift Tubes for Cluster Counting • Two square-section tubes (24 mm side) built to study cluster counting issues in simplified environment. • One tube (400 mm long), equipped with 250 MHz BW preamplifier placed on top of our telescope CINJ = 1. 8 p. F Gain ≈ 5 m. V/f. C Noise ≈ 1900 erms @ CIN = 3 p. F • Waveform digitized with the DRS 4 switched capacitor array (http: //drs. web. psi. ch/) – 4 channels, 1024 channels each @ 5 GSa/sec max. – remaining channels read 3 cells of Proto 1 • Dump waveforms when at least one of them exceeds 25 m. V In spite of x 3 higher preamp bandwidth, a noise level similar to proto 1 (~1. 4 m. V RMS) is obtained with a careful shielding (…) 01/10/2010 G. Finocchiaro 8

Counting Clusters… • Adequate preamp bandwidth is (clearly) crucial to count clusters (1 st attempt at) Peak finding 85%He-15%C 2 H 6 INPUT SIGNAL 1024 channels, 2 Gsample/sec n. P=10. 2/cm TARGET SIGNAL (cluster) Position of peaks from finding algorithm 01/10/2010 G. Finocchiaro 9

Lab Activity @TRIUMF: Wire Aging Tests • Goal is to verify that the chamber will survive the Super. B lifetime. • Test proposed materials usingle cell, as per Boyarski. e. g. bare Al wire. • Primarily a test of Malter effect (field wire aging). • Uses 55 Fe both to age the wires and to characterize performance. SCHEMATIC size of single electron peak increases due to Malter effect as chamber ages A. M. Boyarski, Nucl. Instr. Meth. A 535, 632 (2004) picoammeter monitors total deposited charge 01/10/2010 G. Finocchiaro 10

Charge collected by aging chamber in 55 Fe events low pulse-height events due to interactions near edge of cell Higher gain (3. 3 x) — 55 Fe peak off scale to precisely measure the 1 e- peak auger events in gold 55 Fe 01/10/2010 G. Finocchiaro normalized noise/cosmic spectrum ADC channel 11

Lab Activity @TRIUMF: Cluster Counting • Single-cell 2. 7 m long drift tube to test the feasibility of detecting individual clusters as they drift to the sense wire. aluminized-mylar windows – dispersion and attenuation as a function of distance from preamp. – Start with 55 Fe, ~170 e- in Ba. Bar gas (He: Iso 80: 20, no water). • aim to use UV laser eventually, to generate small, triggered pulses – First preliminary results – example: rise time as a function of distance from preamp 01/10/2010 G. Finocchiaro 12

Topics Discussed in This Meeting • • Lab activities Background studies Cell design optimization DCH Readout architecture 01/10/2010 G. Finocchiaro 13

“Super. B” layer configuration in Full. Sim 01/10/2010 G. Finocchiaro 14

Occupancies vs. energies Dana Lindemann (Mc. Gill) 01/10/2010 G. Finocchiaro 15

Shield Geometry 01/10/2010 G. Finocchiaro 16

01/10/2010 G. Finocchiaro 17

Occupancy from Large-angle Bhabha’s with Fast. Sim D. Swersky (Mc. Gill) 01/10/2010 UNSHIELDED G. Finocchiaro 18

Issues to be solved • Background rate depends on GEANT 4 step length – incorrect treatment of Coulomb diffusion in low-density material? • Belle-II estimate much smaller background rate from radiative Bhabha’s and much larger from Touschek than we do (H. Nakayama’ talk on Tuesday 28 th) 01/10/2010 G. Finocchiaro 19

Topics Discussed in This Meeting • • Lab activities Background studies Cell layout optimization with Garfield DCH Readout architecture 01/10/2010 G. Finocchiaro 20

Garfield studies of cell shape and superlayer transitions Chris Hearty, Philip Lu 21

Transitions between stereo superlayers 22

Stereo-stereo transition: Impact on Reconstruction 23

Axial-stereo transition 01/10/2010 G. Finocchiaro 24

Axial-stereo transition: impact on reconstruction 01/10/2010 G. Finocchiaro 25

Topics Discussed in This Meeting • • Lab activities Background studies Cell design optimization DCH Readout architecture 01/10/2010 G. Finocchiaro 26

The trigger burst case G. FELICI (LNF) Data transfer optimization Ch n 32 samples Trigger 1 Ch n+i 32 samples Trigger 2 Ch n+i 32 samples Trigger 3 FEE data transfer optimization : download the 3 events as a single “big” event § Pro : Front-End data transfer optimization § Cons : § events are overlapped in the same data frame (further elaboration required to split single events) § data frames do not have the same lengths (L 1 trigger occurrences) Event management optimization Ch n 32 samples Trigger 1 Ch n+i 32 samples Trigger 2 Ch n+i 32 samples Trigger 3 Single event data transfer : readout each event separately § Pro : § Front-End events have the same size § FEX can be applied while reading the event § Readout procedure provides event de-randomization § Cons : § Partial (previous) L 1 event re-reading G. Felici LNF-Super. B Workshop – September 2010

A pushing-mode FE readout architecture Trigger Latency Time + n samples (Dual-port memory) Sampled data n 2 1 0 32 word block data readout RO buffer DATA RO SM (FEX) Pushing mode E E E v v v 3 2 1 Read ADDR Concentrator board (counter value @ L 1) - Latency ADDR Sampling clock COUNTER (0 – n) ADDR L 1 (synchronized by sampling clock) L 1 FIFO Fi. FO Empty Fi. FO Read Data (counter value @ L 1) NB : minimum trigger spacing > sampling period (≈ 36 ns) A simulation of the readout architecture will be carried out in the next weeks G. Felici LNF-Super. B Workshop – September 2010

Summary • Good progress in many areas including: – study of gas mixtures with prototypes – evaluation of background for different shields and DCH geometries – Definition of optimal cell layout – Aging tests • Effort starting on: – Cluster counting R&D – Design of readout chain • DCH trigger still to be tackled 01/10/2010 G. Finocchiaro 29