Post on 16-Jan-2016
description
23/01/2007 ID Software Workshop 1
SCT performance from the SR1 endcap data
Satoru MimaOkayama University
ID Software Meeting23-Jan-2007
- Noise Study -
23/01/2007 ID Software Workshop 2
Contents -Noise Study -
1. Trigger rate scan2. Noise threshold scan3. Edge detection mode On/Off4. Cabling5. Summary
Online - Sergey Burdin Online Monitoring - Martin White Offline - Satoru Mima
23/01/2007 ID Software Workshop 3
Noise study - very important issue
Top Left Quadrant of the SCT EC-C 246/247 Modules
Hybrid Temperature ~ 17℃
SCT EC modules have 3 main types: Outer: 56.5–71.8 mm x 123.1 mm
Pitch: 70.8 – 90.3 μm Middle: 56.1–75.2 mm x 118.7 mm
Pitch: 70.3 – 94.8 μm Inner: 43.8–55.8 mm x 73.9 mm
Pitch: 54.4 – 69.5 μm
Sergey Burdin
EndCap-C
Module production phaseSR1 Cosmic ray testUnderground commissioning
Compare Barrel vs EC-C
23/01/2007 ID Software Workshop 4
1. Trigger Rate Scan EC-C
Run Number Trigger Rate7702 1 kHz7703 10 kHz7704 50 kHz7705 100 kHz
1 kHz 10 kHz 50 kHz 100 kHzLayer 0 3.78 (1.87) 3.58 (1.76) 3.57 (1.75) 3.41 (1.69)Layer 1 3.87 (2.21) 3.86 (2.15) 3.87 (2.19) 3.78 (2.09)Layer 2 5.04 (2.53) 4.76 (2.35) 4.74 (2.37) 4.73 (2.37)Layer 3 4.49 (3.12) 4.17 (3.04) 4.48 (2.87) 4.23 (2.92)Layer 4 4.50 (2.84) 4.46 (2.49) 4.11 (2.59) 4.09 (2.68)Layer 5 4.16 (2.77) 4.16 (2.63) 3.92 (2.68) 4.12 (2.61)Layer 6 5.53 (3.46) 5.29 (3.24) 5.33 (3.37) 5.25 (3.30)Layer 7 5.47 (2.47) 5.47 (2.72) 5.40 (2.71) 5.31 (2.73)Layer 8 3.66 (2.04) 3.69 (2.01) 3.60 (1.96) 3.58 (1.88)
Runs 7702 to 7705 were all performed with 1 fC thresholds, but with different trigger rates:
The mean and the RMS of the occupancy distribution does not apparently change with trigger rate:
The table shows the noise occupancy mean (RMS), scaled by 10E5
Martin White
Sergey Burdin
Online
Online Monitoring
23/01/2007 ID Software Workshop 5
Noise Occupancy of EC-Crun7705(1kHz) vs run7702(100kHz)
No trigger rate dependence
23/01/2007 ID Software Workshop 6
2. Noise threshold scan EC-C
Run Number Threshold (fC)7707 0.907711 0.957702 1.007710 1.057708 1.107709 1.20
All 10kEvent
A plot of noise Occupancy versus threshold. We fit Noise Occupancy with Error Function.
Good fitting result
23/01/2007 ID Software Workshop 7
Noise threshold scan EC-C
separated into Eta(Outer,Middle,Inner).
ENC(e-)
The distribution of Equivalent Noise Charge (ENC) value.
ENC (Outer) : 1570 e- (Middle) : 1550 e- (Inner) : 1110 e-
Zoom
per chip
per module
23/01/2007 ID Software Workshop 8
Comparison of ENC of Barrel and EndCap-C
ENC(Barrel) : 1580e-
EndCap-C
ENC(Outer) : 1570 e-
(Middle) : 1550 e-
(Inner) : 1110 e-
ENC(e-)Same ENC for Barrel and EndCap-C
23/01/2007 ID Software Workshop 9
Comparison of ENC for EC-C
My analysis resultRefer to “The ATLAS SemiConductor Tracker End-Cap Module” (2006) NIM Paper
SR1 Comic-Ray testModule Production
ENC(Outer) : 1570 e-
(Middle) : 1550 e-
(Inner) : 1110 e-
Only 50-100electron increase in ENC
23/01/2007 ID Software Workshop 10
3. Edge Detection Mode In physics mode the edge detect mode was ON
<config>0 3 0 1 0 0 0 0 0 0 1</config> Higher occupancy Sensitive to redundant clock?
Occupancy (Edge detect ON)
1.00E-08
1.00E-07
1.00E-06
1.00E-05
1.00E-04
1.00E-03
1.00E-02
1.00E-01
1.00E+00
1 21 41 61 81 101 121 141 161 181 201 221 241
Module
Calibration
Physics
S.BurdinObserved factor 3-4 increase in NO
Online
23/01/2007 ID Software Workshop 11
Increased Occupancy in the Endcap when switching between Edge Mode and Edge Detect
Re-cap of what the Front-End readout looks like:
DAC
Binary PipelineComparato
r
PerAmp+Shaper
Threshold Voltage
Edge-Detect circuit
Edge-Detect Enable
Readout Buffer
Sensor
Test-Input
Threshold Voltage
t
t
v
“Shaped” input pulse to
Comparator
“Logic” output of comparator
Δt Δt
Output pulse duration of comparator:
Δt ≈ Time-over-threshold of input pulse to comparator
Data Compression modes:
00 – Hit
01 – Level
10 – Edge
11 - Test
A. Greenall & J. VossebeldS.Burdin
23/01/2007 12
Noise occupancy Edge Detection Mode ON/OFF
Run 7757Edge DetectON
Run7702Edge DetectOFF
Inner
Middle
Outer
5E-65E-6
Noise OccupancyNoise Occupancy
Large Mean valueAnd dispersion
0 0
23/01/2007 ID Software Workshop 13
Time bin (Edge Off ) 20KEvent
Edge Detection Mode OFF
Time bin (Edge On)20KEvent
000 001 010 011 100 101 110 111
Time Bin
000 001 010 011 100 101 110 111
Time Bin
ON
Random, short, much smaller than 40MHz Noise
23/01/2007 ID Software Workshop 14
4. Cabling for EC-C at SR1
Module in Disk 3, Inner side had disappeared.
Inner
Middle
Outer
Fixed by Shaun Roe, Sergey Burdin
Disk
Missing Disk 3
Fixed!
(I use same cosmic run)
23/01/2007 ID Software Workshop 15
5. Summary
No trigger rate dependence (1kHz-100kHz) ENC=1600e- (spec.<1500e-) for Barrel and EndCap-C (Outer/Middle)
Inner ENC=1100e- (Quiet) Edge Detection Mode has been analyzed
Observed NO increase in Time bin = 001,010,100. Cabling map problem fixed
-> Looking forward to underground commissioning!