Transcript Folie 1

6” Origami Module and
Front End Electronics
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
6th Open Meeting of the Belle II Collaboration
5. July 2010
6th Open Meeting of the
Belle II Collaboration
Components
Testing & Quality Check
Assembly
Module Test
Summary
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
2
6th Open Meeting of the
Belle II Collaboration
Components
Testing & Quality Check
Assembly
Module Test
Summary
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
3
6th Open Meeting of the
Belle II Collaboration
Origami Chip-on-Sensor Concept
• Principal feasibility shown in 2009 by building a prototype
module using a 4” DSSD
• Successfully tested in lab and SPS beam
• Can it be scaled up to the size required for the center
6” sensor of the outer most ladder?
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
4
6th Open Meeting of the
Belle II Collaboration
Summary 4” Origami
• Flex hybrid produced by CERN PCB workshop
– long delivery time
– almost made by hand
– not suitable for series production
• Several vias broke after soldering of electronics
components
• Only moderate bondability
• We found a Japanese company for production
of new 6” design
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
5
6th Open Meeting of the
Belle II Collaboration
Origami Design for Belle II SVD
• Present design for center module
of outermost ladder (6“ HPK sensors)
• Consists of 4 parts:
– Origami PCB (~ 450 mm long)
– PA0: pitch adapter for n-side, glued onto hybrid
– PA1, PA2: pitch adapters to connect p-side strips, wrapped around
the edge of the sensor
• Manufacturer: Taiyo Industrial Co., LTD.
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
6
6th Open Meeting of the
Belle II Collaboration
Origami Design for Belle II SVD
• Design finalized in mid of March
– needed some iterations with company
– e.g. separate PA0 from hybrid to ensure high precision,
tear drop design for vias
• First batch of PA1 and PA2 received
end of March (10 pieces each)
• 10 pieces of Origami PCB and PA0
shipped end of April
• Shipment within appointed time
• PA0 already glued onto hybrids by
company
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
7
6th Open Meeting of the
Belle II Collaboration
6“ Origami PCB and Pitchadapters
Origami PCB
PA0
• Origami PCB
– 3-layer design
– 237µm thick (nominal)
PA1
glued onto
hybrid
PA2
• PA0, PA1 and PA2
– 2-layer design
– 145µm thick (nominal)
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
8
6th Open Meeting of the
Belle II Collaboration
Components
Testing & Quality Check
Assembly
Module Test
Summary
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
9
6th Open Meeting of the
Belle II Collaboration
PA1 & PA2
•
•
•
•
Top layer (coating) and thus PAs are thicker than expected
PAs are rigid  complicates bending
Can the company use a thinner coating?
We brought one (thinner) pitch adapter made by CERN as sample
• Shorts and open measurement  no failure found
• Design issue:
–
–
–
–
–
5. July 2010
Staggering at sensor side is wrong
Does not match staggering of DSSD
Bonding is nevertheless possible
Can be used for prototype module
Layout will be corrected for next batch
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
10
6th Open Meeting of the
Belle II Collaboration
Origami PCB - Optical Inspection
•
•
•
•
Overall impression is very good
Shape of pads and lines look very precise
Gold plating of bond pads seems to be uniform
PA0 is well aligned to the footprints of the AVP
chips.
• Few minor issues:
– On one PCB the top layer of PA0 is slightly
displaced by about 88µm (~1 pitch), but still
bondable
– Drill of some vias is not centered, but within tolerance
– The second (left hand) bias pad is covered by the
solder stop layer  missing opening
– Alignment marks on two PCBs are incomplete
(partially removed), e.g. hybrid #10
 required for auto adjustment of our bonder..
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
11
6th Open Meeting of the
Belle II Collaboration
Origami PCB – Electrical Test
• We measured the connection
between the connectors and all APV
chips
• ~3300 measurements in total
• Very satisfying results:
9 of 10 board are good
• On one PCB we found a single open
via near APV #1 (p-side)
• Design issues:
– APV signals lines swapped: 0<>1, 2<>3Origami PCB
of neighboring sensor
– doesn‘t matter for prototype
covers and damages
bond wires
– Bias connection between PA0 and Origami
PCB will be covered by neighboring hybrid
– change location of pad
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
12
6th Open Meeting of the
Belle II Collaboration
Origami PCB – Soldering
• Was an annoying issue of 4” Origami
hybrids (soldered by hand @ ~350 °C)
• Broken vias after soldering, caused by
local delamintation
• Vapor phase soldering ?
– lower temperature
– Nanonics connector
• Baking of one hybrid: 48h @ 65°C
• Solder cycle w/o applied parts:
5 min. @ 230°C
• Optical inspection & electr. Test
– no visible damage
– all connections still ok
– no broken via
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
13
6th Open Meeting of the
Belle II Collaboration
Origami PCB Soldering – Nanonics Connector
• We equipped and soldered one hybrid using vapor
phase soldering  no damage of PCB
• Nanonics connector is still problematic
–
–
–
–
connector has two rows of pins
sometimes they are not at the same height
causes bad connection of some pins
resoldering of inner row is difficult
• Further investigation required
• Other connector available?
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
14
6th Open Meeting of the
Belle II Collaboration
Origami PCB Soldering – Resistor
Edge of PA0
• Resistor in front of APV
– located between two bond pads
– no space for solder stop between resistor and pads
• solder tin covers bond pad  no more bondable
• Solution for prototype:
– we glued small pads beneath the damaged one
• Future design:
– modified routing of lines between solder and bond pads
– glued PA0 acts as solder stop layer
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
15
6th Open Meeting of the
Belle II Collaboration
Origami PCB – Thermal Cycling
• Test long term thermal stress resistivity
• 84 cycles between -10 and +30 °C
• Same conditions as used for CMS
thermal tests
• Duration: ~ 2 weeks
• Thereafter we repeated optical inspection
and electrical test
• No damage detected 
Temp Sensor 1
Temp Sensor 2
T hermal Cycling
35
30
25
Temperature [°C]
20
15
10
5
0
-5
-10
-15
9.6.
5. July 2010
10.6.
11.6.
12.6.
13.6.
14.6.
15.6.
16.6. 17.6.
Date/Time
18.6.
19.6.
20.6.
21.6.
22.6.
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
23.6.
24.6.
16
6th Open Meeting of the
Belle II Collaboration
Origami PCB – Impedance of Analog Data Line
•
•
•
•
Determined diff. impedance of analog output of p-side APV #1
Measured reflexions of rectangular input signal and
adjusted Rterm until reflection disappeared
Measurement:
Zdiff = 102.6 W
Simulation:
Zdiff = 102.7 W
•
•
•
Path length: ~ 45 cm, Propagation delay (measured) = 5.46 ns
Propagation speed: c = 164 800 km/s  6.07 ns/m
Hybrid cable: Zdiff = 125 W  still no reflexions visible
Rterm << Rz
5. July 2010
Rterm = 102.6Ω = Rz
Rterm = 125Ω
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
17
6th Open Meeting of the
Belle II Collaboration
Origami PCB – Ohmic Resistance of Line
Rline = 22 W (single line)
Rline = 9.4 W  half the measured value
•
•
Measurement:
Simulation:
•
Comparison line geometry
line width
line distance
dist.to next pair
nominal [µm]
100
121
380
measured [µm]
74.3+-2.0
144.3+-1.5
410
 slightly over etched, but does not explain huge difference of R
thickness
•
•
•
•
nominal [µm]
9
measured [µm]
5
Measurements were done with the tactile sensor of our CMM
5µm copper thickness is compatible with measured resistance
Apparently, there is less copper than planned, but this is not a real problem.
Anyhow, we originally requested 5µm  less material budget
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
18
6th Open Meeting of the
Belle II Collaboration
Origami PCB – Wire Bonding
• Fully automated bonding
machine
• Bond process control
• Easy to find working
parameters
• Very uniform deformation
• Excellent bondability
Origami
Sensor
Excellent and uniform gold
coating
deformation vs. time
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
19
6th Open Meeting of the
Belle II Collaboration
Components
Testing & Quality Check
Assembly
Module Test
Summary
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
20
Assembly procedure of 6” Origami
•
•
•
We must newly develop the assembly jigs so that the size of component is entirely
different with that of 4” DSSD Origami prototype modules.
The procedures should be as simple as possible to minimize the developing and
actual assembly time (~7 weeks. limited by my stay at HEPHY), but should be
extendable whole ladder for near future production.
The procedures are developed to require basic three jigs and these supplements.
jig1
jig2
•
The Jigs are used to hand the current procedure to the next procedure with
keeping their alignment ( assured by pin&hole ). The assembly precision
is limited by the difference of pin and hole’s diameter. If we use
precise pin and linear bush( circulating boles inside wall of hole to
reduce friction ), it easily achieve several mm precision level.
•
As a results, we could assemble very smoothly on schedule using developed
5. July 2010
procedures without any problems.
jig3
21
Procedure1: Placing DSSD on the Jig1
1. Placing the DSSD on the jig1
2. Fixing DSSD with vacuum chucking
3. Removing stopper and pin
Jig1
Poral stone to vacuum-chuck
Commonly used holes to hand
Origami components to the other jigs
DSSD
5. July 2010
Attachable stopper to align the DSSD
22
Procedure2: Placing Pitch adaptors on the DSSD
1. Placing PAs on the sensor
2. Aligning PAs
5. July 2010
23
Procedure3: Placing PA-jigs and picking up the PAs
1. Placing PA-jig on the PA
2. Fixing PAs to the PA-jigs
with vacuum-chucking
3. Picking up PAs
PA-jig
Pins
PA-jig
Chucked PA
Jig1
5. July 2010
Vacuum chucking holes
24
Procedure4: Picking up PA-jigs, gluing, wire-bonding
1. Placing PAs on the PA-jigs upside down
2. Dispensing glue on the backside of PAs
3. Placing PA-jigs on the Jig1 again to glue
4. After cured, removing PA-jigs
Araldite 2011(Epoxy adhesive)
5. July 2010
5. Wire-bonding p-side of DSSD and PAs
25
Procedure5: Placing ribs on the jig2 and dispensing glue
Jig2
Support ribs
1. Placing the Support ribs in the groove
of Jig2
2. Dispensing glue top of the ribs
Grooves for ribs
5. July 2010
Poral stone
Jig2
26
Procedure6: Placing the jig2 on the jig1
Jig2
1. Placing the jig2(Ribs) on the jig1(DSSD) to glue
Jig1
2. Flipping upside down
3. Fixing DSSD to the jig2 with vacuum chucking
4. After curing the glue(~24h), removing jig2
Hole
Jig2
Pin
5. July 2010
Jig1
27
Procedure7: Placing the jig2 on the jig1
1. Placing the Rohacell on the Jig1 with as
same as procedure1
2. Fixing the Rohacell with vacuumchucking
3. Dispensing glue on the Rohacell
Jig1
5. July 2010
Rohacell
Attachable stopper for Rohacell
28
Procedure8: Placing the jig2 on the jig1
Jig1
1. Placing Jig1(Rohacell) on Jig2(DSSD) to
glue them
Jig2
2. After cured(~24h), removing Jig1
Jig2
Jig2
Jig1
Rohacell
5. July 2010
29
Procedure9:Placing the PCB on the Rohacell to align
Microscope
1. Placing the Origami PCB on the
Rohacell(Jig2) to align
2. Aligning the bonding pads of the PCB to
the bonding pad for N-side of DSSD
zzz…
Origami PCB
Jig2
Tired man
5. July 2010
Attachable support for the Origami PCB
30
Procedure10:Picking up the PCB, gluing and wire-bonding
Jig3
1. Placing the Jig3 on the PCB(jig2).
And then, picking up PCB to the
Jig3 with vacuum-chucking
PCB support
2. Dispensing glue on the backside
of the PCB
Jig2
1
Jig3
2
3. Placing the jig3 again to glue the
PCB and Rohacell on the Jig2
4. After cured(~24h), removing Jig3
5. Wire-bonding APV to n-side DSSD
Jig3
Jig2
5. July 2010
3
4
31
Procedure11:Wrapping PAs, gluing and wire-bonding
Jig2
1. Wrapping the PAs with special jigs
2. Dispensing PAs and gluing to the PCB
3. After cured, wire-bonding the PAs to the
APV.
Special jig
=> Assembly completed
We think the developed procedures are
based for the full ladder production to
extend to the lateral direction.
5. July 2010
32
6th Open Meeting of the
Belle II Collaboration
Assembly – Required Equipment
 Origami assembly toolbox 
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
33
6th Open Meeting of the
Belle II Collaboration
Assembly – Finished Module
Although the schedule was very tight, we succeeded to
finish the first module before the B2GM
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
34
6th Open Meeting of the
Belle II Collaboration
Components
Testing & Quality Check
Assembly
Module Test
Summary
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
35
6th Open Meeting of the
Belle II Collaboration
Origami Module – Sensor Characteristics
Specifications of Polysilicon Resistors
IV Curve
18 March 2010 by HPK
I [µA]
3,5
Process parameters
Serial No. Pside
1
P1
2
P1
3
P1
4
P1
5
P1
6
P1
7
P1
8
P1
9
P1
10
P1
11
P1
12
P2
13
P2
14
P2
15
P2
16
P1
17
P1
18
P1
19
P3
5. July 2010
Nside
N1
N1
N1
N2
N2
N2
N2
N3
N3
N3
N3
N2
N2
N2
N2
N2
N2
N2
N2
Poly resistor Measurment results (Mega
Ohm)
MAX
MIN
AVE
P1
17.6
15.1
16.3
P2
46.3
35.6
41.1
P3
7.1
6.4
6.8
3,0
2,5
2,0
1,5
1,0
N1
N2
N3
20.5
7.2
3.3
15.3
6.2
2.7
17.4
6.5
2.9
0,5
0,0
0
20
40
60
80
100
120
140
160
-0,5
-1,0
•
•
•
Sensor #8:
Dark current:
Rpoly:
– p-side:
– n-side:
B2HPK_10938-9239_8
3µA @ 80V
16.3 MW
2.9 MW
– confirmed by measurement
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
36
6th Open Meeting of the
Belle II Collaboration
Stripscan Sensor #8 (n-side)
•
•
•
•
•
5. July 2010
6.
I_strip = -3.31 nA
R_poly =2.76 Mohm
C_ac=186.83 pF
No pinhole
HPK: AC AL short at
strips 275/276 - could
be verified by C_ac
measurement!
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
37
6th Open Meeting of the
Belle II Collaboration
Origami – Source Test
•
•
•
•
Made a source test using 90Sr source
Module works well  all APVs ok
No I2C problems
Currently we can read out 8 of 10 APVs
(limited by existing readout system)
• Due to leak of time:
– Only one single run performed
– No cooling pipe applied
– Results are very, very preliminary!
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
38
6th Open Meeting of the
Belle II Collaboration
Source Test – Very Preliminary Results
no C1
Entrie s
Mea n
RMS
8000
7000
ndC1
Origam i6_mod1_p (Orgami 6") - Noise
512
248
1 43.8
entries []
s igm a [e]
Origa mi6 _mod 1_p (Orga mi 6") - No ise
6000
Ent ries
Mean
RMS
χ / ndf
Constant
Mean
Sigma
50
40
512
3353
621. 3
29.85 / 26
41. 3 ± 2. 6
3330 ± 20.8
439.4 ± 18.8
5000
30
Noise = 3330.4
4000
3000
20
2000
10
1000
0
100
200
300
400
no C3
Origa mi6 _mod 1_n (Orga mi 6") - No ise
s igm a [e]
0
500
s trip []
Entrie s
Mea n
RMS
2400
2200
2000
2000
3000
4000
5000
6000
7000 8000
s igm a [e]
ndC3
Ent ries
512
Mean
975. 8
RMS
144. 8
χ / ndf
27.83 / 7
Constant 189.7 ± 11.5
Mean
962.7 ± 4. 6
Sigma
96.74 ± 3. 83
180
160
140
1600
120
1400
1200
100
1000
80
800
Noise = 962.7
60
600
40
400
20
200
0
100
200
300
Average cluster noise:
p-side:
3330 e
n-side:
963 e
5. July 2010
1000
Origam i6_mod1_n (Orgami 6") - Noise
512
2 62.7
1 47.6
1800
0
0
entries []
0
400
•
•
•
•
500
s trip []
0
0
500
1000
1500
2000
s igm a [e]
Some noisy strips on both sides
Noise on p-side seems to high
Need to investigate
Requires measurements with cooling!
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
39
6th Open Meeting of the
Belle II Collaboration
Source Test – Very Preliminary Results II
s iC1
Ent ries
Mean
RMS
χ / ndf
Width
MP
Area
GSigma
1400
1200
1000
34659
3. 506e+04
1. 566e+04
620.1 / 96
4518 ± 26.0
2. 441e+04 ± 53
3. 565e+07 ± 195277
500 ± 14.5
7000
clw =
1 2
346 57
8 .8 45
2 .7 74
3+
Peak = 6.74
FWHM = 34.2%
4000
3000
400
2000
200
1000
10000
20000
30000
40000
50000
60000
70000
80000
0
90000
signal [e]
siC3
Ent ries
Mean
RMS
χ / ndf
Width
MP
Area
GSigma
5000
4000
3000
95601
3. 048e+04
1. 266e+04
488.4 / 81
2652 ± 19.1
2. 28e+04 ± 26
9. 804e+07 ± 321418
2718 ± 35.3
Peak = 23801.7
FWHM = 55.3%
1000
30000
40000
50000
60000
70000
80000
90000
signal [e]
30
40
50
60
0
SNR []
snr3
Entrie s
956 00
Mea n
2 3.63
RMS
9 .3 99
clw =
3000
1000
20000
20
4000
2000
10000
10
5000
2000
0
0
Origam i6_mod1_n (Orgami 6") - S NR
entries []
0
Origa mi6 _mod 1_n (Org ami 6" ) - Sig nal
entries []
8000
6000
Peak = 24436.5
FWHM = 74.5%
600
0
Entrie s
Mea n
RMS
5000
800
0
snr1
Origam i6_mod1_p (Orgami 6") - S NR
entries []
entries []
Origa mi6 _mod 1_p (Org ami 6" ) - Sig nal
1 2
3+
Peak = 17.08
FWHM = 53.8%
0
10
20
30
40
50
60
SNR []
• Low SNR on p-side  should be > 10 with cooling
• n-side benefits from short pitch adapter (PA0) and short strips
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
40
6th Open Meeting of the
Belle II Collaboration
Components
Testing & Quality Check
Assembly
Wire Bonding
Module Test
Summary
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
41
6th Open Meeting of the
Belle II Collaboration
Summary
•
Origami Components:
–
–
–
–
–
–
•
Assembly Procedure
–
–
–
–
•
New 6” Origami PCB and PAs manufactured by Taiyo Industrial Co., LTD.
Delivered all parts in time
Excellent quality
Only some minor issues
PA1 and PA2 should be slightly thinner to ease bending
Some design issues  will be solved with next batch
Design and production of jigs within few weeks, thanks to effort of Onuki-san
Proposed assembly procedure verified by building prototype module
In principle scalable for ladder production
Only minor modifications required, e.g. using linear bush to enhance precision
Module Performance
–
–
–
–
5. July 2010
All APVs work well
High noise level on p-side  but currently no cooling applied
More tests and analysis required
Beam test and irradiation scheduled for October 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
42
6th Open Meeting of the
Belle II Collaboration
 Thank you for your attention 
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
43
6th Open Meeting of the
Belle II Collaboration
Backupslides
5. July 2010
Christian Irmler (HEPHY Vienna), Yoshiyuki Onuki (Tohoku University)
44