Upload
cecilia-hampton
View
216
Download
0
Embed Size (px)
Citation preview
Mass production testing of the front-end ASICs for the ALICE
SDD system
L. Toscanoa, R. Arteche Diazb,d, S. Di Libertob, M.I. Martíneza,c, S.Martoiua, G. Mazzaa, M.A. Mazzonib, F. Meddib, A. Rivettia,
F. Toselloa, G.M. Urciuolib and R. Wheadona
a INFN Sezione di Torino, Torino, Italyb INFN Sezione di Roma, Roma, Italy
c UNAM, Mexico City, Mexicod CEADEN, Havana, Cuba.
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
Outline
•SDD overview
•Front-end ASICs
•Test criteria and test set-up
•Test results
•Conclusions
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
The Inner Tracking System
The ITS: six layers of silicon detectors
Two layers of hybrid pixels Two layers of silicon drift
detectors Two layers of double sided
microstrips
The SDD layers Will have 260 SDDs Total area: 1.3 m2
About 130.000 electronic channels
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
SDD mechanical support
The SDD mechanical support is composed by carbon-fibre cones and cylinders:
Carbon-fibre structures called ladders form two layers: • 3rd with 14 units• 4th with 22 units
Ladders supported by rings SDD modules on ladder:
• detector• front-end electronics
End ladder cards• HV boards• LV boards• data compression • interface with the DAQ
Microcables• aluminium on Upilex• TAB-style bonding
rings
Ladders
cones
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
The front-end module
The silicon drift detector • 256 anodes each side
Two front-end hybrids: • based on two ASICs• 8 PASCAL and AMBRA
Two LV boards: • power supply • LVDS to CMOS • parameter monitoring
One HV board:• filter capacitors• voltage divider
The basic building block of SDD system is the module, composed by:
detector
front-end hybrids
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL
•64 preamplifiers •64 channel SCA with 256 cells/channel
•32 10-bit ADCs•internal test-pulse generator•JTAG protocol•custom low power digital interface
PASCAL is a mixed-mode ASIC with 64 input channels:
Preamp Analog memorySAR ADC
PreampPreamp Analog memoryAnalog memorySAR ADCSAR ADC
Preamp Analog memorySAR ADC
PreampPreamp Analog memoryAnalog memorySAR ADCSAR ADC
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA
• baseline subtraction (6-bit registers)• channel masking• compression (10 to 8 bits)• parity check (RAM & registers)• multiplexing• fully differential digital I/Os• addressable JTAG interfaceAMBRA compression curve
0
64
128
192
256
0 256 512 768 1024
Din(10 bits)
Do
ut(
8 b
its
)
AMBRA is a control, pre-processing and buffering ASIC:
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
The 8-inch wafer
• For cost reasons both ASICs fabricated using the same set of masks
• IBM Commercial technology
• Both ICs implemented in 0.25 m CMOS
• Each 8-inch wafer hosts about 146 testable couples PASCAL and AMBRA AMBRA
PASCAL
Preselection and characterization of AMBRA and PASCAL:
•On-wafer tests
•2 test procedures
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL test procedure (1)
1. Power consumption test: measures Iadc, Ianlg and Idig without clock measures repeated with clock at 40MHz values in a reference range
2. ADC frequency test: at 40MHz or at 20MHz programmable via
JTAG ADC current value reduced at 20MHz
3. AM frequency test: at 40MHz or at 20MHz programmable via
JTAG internal test pulse generator measures test pulse signal position
The PASCAL chips test procedure consists of 6 tests:
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL test procedure (2)
4. Noise test: measures the mean and rms for all channels performed at the low gain value parameters must be in range for all 64 channels
5. Baseline test: sets a sweep of baseline values via JTAG tests the baseline DAC linearity gain values must be in range for all 64 channels
6. Signal test: sets a sweep of test pulse values via JTAG checks the calibration DAC linearity gain values must be in range for all 64 channels
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA test procedure
The AMBRA test procedure consists of one power consumptions test + 4 functional tests:
1. Power consumptions test: checks the I dd at 40MHz
checks Idde
2. Logic test A: handshake signals from/to
PASCAL data bus and flow control
signals multi event buffer mode
3. Logic test B: 10-to-8-bit compression baseline subtraction registers
4. Memory test: 4 memory buffers parity bit
5. JTAG test: W\R all JTAG
registers JTAG lines from/to
PASCAL
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
Test system set-up
Agilent E3631A Power
supplies
Agilent 16702B Logic
Analysis System
Test board
PC
Probe station
Chuck stage
Probe card
Vanl&Vdig CMOS signals
LVDS signals + Val
GPI B bus TCP/ I P
Wafer
y x
z
Minimal changes in the hardware:
2 full custom probe cards2 different test boards
Agilent 16702B Logic Analysis System for data acquisition (TCP/IP).
Test system controlled by a PC running dedicated software in LabView for data analysis
2 Agilent E3631A power supplies (GPIB)
The test system set-up is based on a semiautomatic Cascade Microtech probe station with motorized chuck stage for 8-inch wafers (GPIB):
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA and PASCAL probe cards
Custom low power digital interface between AMBRA and PASCAL
On board digital buffers pseudo LVDS to LVDS standard conversion
Are required:
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL selection criteria
The PASCAL chips are considered:
• Good, all six tests are passed with no errors
• Grey, the first three tests are passed, at least one of the last three is not passed
• Bad, all other ones
LabView front panel for PASCAL tests
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA selection criteria
The AMBRA chips are preselected by this classification:
• Good, power consumption test and all four functional tests are passed with no errors
• Grey, power consumption test and functional tests are passed with no errors and no more than 128 memory bits are stuck
• Bad, all other ones
LabView front panel for AMBRA tests
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL power test results (1)
ADC and analogue current values of 143 good PASCAL chips in wafer ALEQARX
Mean = 11.95 mA
= 0.24 mA (2.0%)
Mean = 68.22 mA
= 3.09 mA (4.5%)
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL power test results (2)
Digital current values for 143 good PASCAL chips in wafer ALEQARX with clock at 40MHz:
• Mean value = 52.7mA
• = 0.41 mA (0.8%)
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
0
500
1000
1500
2000
2500
0,89 0,9 0,91 0,92 0,93 0,94 0,95 0,96 0,97
gain
Gain
gain 0.90000
Minimum 0.95000
Maximum 5069 Channels 0.92543
Mean 0.0073
Std Dev 0.00010
Std Error
chan
nels
5069 good PASCAL channels
On one wafer
Mean value = 0.925
= 0.007 (0.76%)
Typical voltage gain values for baseline DAC
PASCAL baseline test results
27743 good PASCAL channels
On 4 wafers
Mean value = 0.92
= 0.01 (1.13%)
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL Signal test results
0
100
200
300
400
500
39,6 42 44,4 46,8 49,2 51,6
gain-good
Gain (mV/fC)
gain-good
40.450Minimum
51.290Maximum
5012channels
46.317Mean
1.6Std Deviation
0.023Std Error
C
H
A
N
N
E
L
S
5012 good PASCAL channels
On one wafer
Mean value = 46.3 mV/fC
= 1.6 mV/fC (3.5%)
Typical gain values for test pulse DAC
29233 good PASCAL channels
On 4 wafers
Mean value = 46.87 mV/fC
= 1.94 mV/fC (4.1%)
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA power test results (1)
Values of I dd of 63 good AMBRA chips on wafer SK8MQHX:
• Mean value = 47.46 mA
• = 0.41 mA (0.86%)
0
5
10
15
20
46,2 46,6 47 47,4 47,8 48,2 48,6
Idd (mA)
# A
MB
RA
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA power test results (2)
0
5
10
15
20
25
30
33 33,5 34 34,5 35 35,5 36 36,5 37
Idde (mA)
#AM
BRA
Values of Idde of 63 good AMBRA chips on wafer SK8MQHX:
• Mean value = 34.9 mA
• = 0.5 mA (1.43%)
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
PASCAL test map
105 151 152 153 154 155 156 157 158 159 160
135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97
56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
7 8 9 10 11 12 13 14 15 16 17 18 19
1 2 3 4 5 6
Wafer AXEQFNX
Good = 119 -> 81%
Grey = 15 -> 10.2%
Bad = 13 -> 8.8%
Tot = 147
First 12 wafers:
Good = 89%
Grey = 6%
Bad = 5% Good
Grey
Bad
Damaged
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
AMBRA test map
155 156 157 158 159 160 161
142 143 144 145 146 147 148 149 150 151 152 153 154
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
27 28 28 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
1 2 3 4 5 6 7 8 9 10 11
Wafer AXEQFNX
Good = 114 -> 75.5%
Grey = 13 -> 8.6%
Bad = 24 -> 15.9%
Tot = 151
Good
Grey
Bad
Damaged
First 12 wafers:
Good = 73%
Grey = 10%
Bad = 17%
11th Workshop, Heidelberg, Sept 12-16,2005L. Toscano - INFN Torino
A mass production testing system was required to test 2400 AMBRA + 2400 PASCAL for 260 SDD modules
The system is working efficiently at INFN in Rome Typical testing times are 510 seconds for PASCAL and
220 seconds for AMBRA 89% good PASCAL chips and 73% good AMBRA chips on
12 wafers
Conclusions
1.8 wafers/month
2800 PASCAL and 2400 AMBRA in February 2006
22 wafers to be tested to meet the requirements