Burn-in & Test Socket Workshop - BiTS Workshop

Burn-in & Test
Socket Workshop
IEEE
March 3-6 , 2002
Hilton Phoenix East/Mesa Hotel
Mesa, Arizona
IEEE
COMPUTER
SOCIETY
Sponsored By The IEEE Computer Society
Test Technology Technical Council

COPYRIGHT NOTICE
• The papers in this publication comprise the proceedings of the 2002
BiTS Workshop. They reflect the authors’ opinions and are reproduced as
presented , without change. Their inclusion in this publication does not
constitute an endorsement by the BiTS Workshop, the sponsors, or the
Institute of Electrical and Electronic Engineers, Inc.
· There is NO copyright protection claimed by this publication. However,
each presentation is the work of the authors and their respective
companies: as such, proper acknowledgement should be made to the
appropriate source. Any questions regarding the use of any materials
presented should be directed to the author/s or their companies.

Burn-in & Test Socket
Workshop
Technical Program
Session 5
Tuesday 3/05/02 1:00PM
Modeling, Analysis and Characterization
“Electrical Modeling And Contactor Performance In A RF System”
Jim Adley - Johnstech International Corporation
Eric Leung - Johnstech International Corporation
Jeff Sherry - Johnstech International Corporation
“Leaded 2mm Contactors: Measuring And Modeling To 10 GHz”
Tom Strouth - GigaTest Labs Orlando Bell - GigaTest Labs
Gary Otonari - GigaTest Labs Eric Bogatin - GigaTest Labs
Jeff Sherry - Johnstech International Corporation
“Force Measurement On Sockets And Contactors”
Richard Block - Advanced Micro Devices
Rafiq Hussain - Advanced Micro Devices

Electrical Modeling and
Contactor Performance in a RF
Test System
0.5 mm Pitch BGA
Jim Adley, R&D Manager
Eric Leung, R&D Engineer
Jeff Sherry, R&D Engineer
Johnstech International
1

Discussion Topics
▼ Modeled Data
▼ Measured Data
▼ Comparative Data
▼ Equivalent Circuit Model
▼ Conclusion
2

Introduction: 0.5 mm Pitch
BGA Contactor
▼ Cross sectional view of a contactor
3

Modeled Data
▼ Modeling was done using Agilent HFSS
software
▼ An AutoCAD drawing, including the actual
structure built by GigaTest Labs, was imported
into HFSS
▼ Data obtained from simulation includes:
▼ Return Loss S 11
▼ Insertion Loss S 21
4

Modeled Data
▼ 0.5 mm pitch BGA model from an AutoCAD file
5

Modeled Data - S 11
▼ This is data from the HFSS model of return
loss by the actual 0.5 mm BGA structure
tested by GigaTest Labs
0
-5
-10
-15
Return Loss (dB)
-20
-25
-30
-35
-40
-45
HFSS 0.5mm Pitch Adjacent BGA Contact Model
Optimum Load Board Layout of HFSS 0.5mm Pitch Adjacent
BGA Contact Model
-50
1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
6

Modeled Data - S 21 & S 11
0
-0.25
▼ This is data from the HFSS model of return
loss by the actual .5 mm BGA structure tested
by GigaTest Labs
0.5mm Pitch BGA Adjacent Contact HFSS Insertion Loss Models
-1 dB Loss Point @ 17 GHz
InsertionLoss (dB)
-0.5
-0.75
-1
-1.25
-1 dB Loss Point @ 13 GHz
-1.5
-1.75
-2
HFSS 0.5mm Pitch Adjacent BGA Contact Model
Optimum Load Board Layout of HFSS 0.5mm Pitch Adjacent BGA
Contact Model
1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
7

Modeled Data
▼ HFSS Capabilities
▼
▼
▼
▼
▼
▼
▼
Contact design parameters
Load board design effects
Device pad interactions
Expected performance
Effects of tolerances
Interaction between components in system
(device, contactor, handler, etc.)
Trends
8

Measured Data
▼ GigaTest Labs tested a 0.5 mm pitch Ball
Series contactor
▼ GigaTest Labs used a surrogate device -
Short, Open, Load, Thru (SOLT) to conduct
testing
▼ Data was measured by GigaTest Labs
through probing from the back side of a
non-optimized load board
9

Measured Data
Surrogate
Circuit
10

Measured Data
▼ GigaTest Labs used a micro probe station for
measuring two adjacent contacts S-parameters
11

Measured Data
▼ The 0.5 mm pitch BGA housing and BGA
surrogate package
12

Comparative Data
0
HFSS Model vs. Measured Test Data for 0.5mm BGA Adjacent Contacts
-5
-10
Return Loss (dB)
-15
-20
-25
-30
-35
-40
-45
-20 dB Measured Loss Point @ 2.95 GHz
HFSS 0.5mm Pitch Adjacent BGA Contact Model
Measured Test Data from GigaTest Labs for 0.5mm Adjacent BGA
Contact (Non-Optimum pads)
-50
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
13

Comparative Data
0
0.5mm Pitch BGA Adjacent Contact Modeled vs. Measured Test Data
-0.25
-0.5
Insertion Loss (dB)
-0.75
-1
-1.25
-1.5
-1.75
HFSS 0.5mm Pitch Adjacent BGA Contact Model
-1 dB Measured Loss Point @ 10.3 GHz
Measured Test Data from GigaTest Labs for 0.5mm Adjacent BGA Contact (Non-Optimum pads)
-2
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
14

Equivalent Circuit Model
▼ Characterize the parasitic effects
▼ Simulate the contact with an equivalent
circuit for time domain response
▼ Integrate the contactor into a system level
simulation to:
▼
▼
Reduce test cost and time
Optimize system performance
15

Equivalent Circuit Model
▼ Measure the S-parameters for short-, open- and
thru- fixtures from two adjacent contacts
▼ Use measured data and Agilent Advanced
Design System (ADS) for model extraction and
verification
▼ Compare measured and ADS simulated
insertion and return loss
16

Equivalent Circuit Model
▼ This figure shows the Equivalent Circuit Model
for two adjacent 0.5 mm BGA contacts
17

Equivalent Circuit Model
0.5mm Pitch BGA Adjacent Contact ADS Equivalent
Circuit Model vs. Measured Test Data
0
-10
Return Loss (dB)
-20
-30
-40
-50
-20 dB Measured Loss Point @ 2.95 GHz
ADS Return Loss Model for 0.5mm Adjacent BGA Contact From Equivalent
Circuit
Measured Test Data from GigaTest Labs for 0.5mm Adjacent BGA
Contact (Non-Optimum pads)
-60
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
18

Equivalent Circuit Model
▼ Phase comparison shows that the contact is
linear over the frequency model
200
S 11 Phase Comparison for ADS 0.5mm Equivalent Circuit Model vs. Measured
Test Data
150
Angle (Degrees)
100
50
0
-50
ADS Return Loss Phase Model for 0.5mm Adjacent BGA
Contact From Equivalent Circuit
Return Loss Phase of Measured Test Data from GigaTest Labs
for 0.5mm Adjacent BGA Contact
-100
-150
-200
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
19

Equivalent Circuit Model
Development
0
0.5mm BGA Adjacent Contact Insertion Loss (S 11 ) ADS Equivalent Circuit
Model vs. Measured Test Data
-0.5
Insertion Loss (dB)
-1
-1.5
-1 dB Measured Loss Point @ 10.3 GHz
ADS Insertion Loss Model for 0.5mm Adjacent BGA Contact From
Equivalent Circuit
Measured Test Data from GigaTest Labs for 0.5mm Adjacent
BGA Contact (Non-Optimum pads)
-2
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
20

Equivalent Circuit Model
Development
0
S 21 Phase Comparison for ADS 0.5mm BGA Equivalent Circuit Model vs.
Measured Test Data
-20
-40
Angle (Degrees)
-60
-80
-100
-120
-140
ADS Insertion Loss Phase Model for 0.5mm Adjacent BGA Contact
From Equivalent Circuit
Insertion Loss Phase of Measured Test Data by GigaTest Labs for
0.5mm Adjacent BGA Contact
-160
-180
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)
21

Conclusion
▼ For leading edge RF applications such as
0.5 mm pitch, modeling is helpful to
achieve optimal system performance
▼ Accurate equivalent circuit models can
represent contact behavior and help in
determining complete system response
▼ Modeling can help RF engineers save time
and money in development by correctly
predicting system results and eliminating
or reducing hardware builds and test
iterations
22

Leaded 2 mm Contactors:
Measuring and Modeling to 10 GHz
Tom Strouth, Orlando Bell, Gary Otonari, Eric Bogatin
GigaTest Labs, www.GigaTest.com
and
Jeff Sherry
Johnstech, www.Johnstech.com

Slide - 2
Outline
• Contactors
• Fixturing
• Measurement set up
• Modeling process
• Results
• Using the model for simulation
BiTS 2002

Slide - 3
32 lead MLP2 Contactor
BiTS 2002

Slide - 4
Analysis
• Goal
Create an equivalent circuit model for two adjacent
leads that predicts the measured S parameters
Use this verified, high bandwidth model for
performance evaluation
• Strategy
Use measurements of open, short, thru topologies
De-embed the fixturing
Use simplest model for accurate, 10 GHz bandwidth
Use SPICE model of de-embedded contactor for
performance simulation
BiTS 2002

Slide - 5
Measurement
Configurations
open short thru
1
2
1
2
1
2
BiTS 2002

Slide - 6
Surrogate Package:
Enables configuring open, short, thru
connections for edge and corner leads
thru
short
open
BiTS 2002

Slide - 7
Instrument Set Up
BiTS 2002

Slide - 8
Probing From the Back Side
Probing using microprobes and a
GigaTest Probe Station
BiTS 2002

Slide - 9
Modeling the System with
Agilent ADS
short
open
L cbg
short
VNA
Fixture
board
Pads and
contactor
Surrogate
package
BiTS 2002

Slide - 10
Methodology
• Measure S parameters of calibration vias in bare fixture
board
• Extract model for just the fixture
• Select two adjacent corner leads (longest leads)
• Measure open, short, thru for the pair of leads
• Extract model of contactor pins and surrogate package
• Use de-embedded contactor circuit model to simulate
performance
BiTS 2002

Slide - 11
Corner Leads (worse case):
Open/short Measurements
short
Z
=
1+
50
1−
S
S
open
open
short
BiTS 2002

Slide - 12
Optimized Model: Open
Impedance of one trace
Coupling between traces
Solid Line is Simulated
BiTS 2002

Slide - 13
Optimized Model: Open
Reflection
Transmission
Triangles (purple): Measured
Open provides mutual capacitance info
BiTS 2002

Slide - 14
Optimized Model: Short
Impedance of one trace
Coupling between traces
Solid Line is Simulated
BiTS 2002

Slide - 15
Optimized Model: Short
Reflection
Transmission (coupling)
Triangles (purple): Measured
Short provides mutual inductance info
BiTS 2002

Slide - 16
Optimized Model: Loop-Thru
Reflection
Transmission
Triangles (purple): Measured
Bandwidth of the model > 10 GHz
BiTS 2002

Slide - 17
Summary of Extracted
Parameters
23 pH
89 fF
860 pH
151 fF
33 pH
140 fF
23 pH
78 fF
140 fF 89 fF
310 pH
860 pH
133 fF
151 fF
25 fF
33 pH
33 pH
VNA
Fixture
BiTS 2002
Contactor
Note: load board was not optimized for performance
Surrogate

Slide - 18
De-Embedded Insertion and
Return Loss of Contactor
Return loss
Insertion loss
Meets specification:
< -20 dB, below 10 GHz
Meets specification:
> -1 dB, below 10 GHz
Note: load board is not optimized for performance,
standard contactor - not enhanced contactor
BiTS 2002

Slide - 19
Transient Simulation Using
De-embedded Contactor Model
Specs:
De-embedded contactor
model
20 psec rise time
50 Ω source, termination
Differential drive
BiTS 2002

Slide - 20
Transient Simulation
with 20 psec Rise Time
• 20 psec input rise time
• 28 psec output rise time
• 20 psec intrinsic
interconnect rise time
(> 15 GHz bandwidth)
• 21 psec time delay
Note: load board is not optimized for performance,
standard contactor - not enhanced contactor
BiTS 2002

Slide - 21
Conclusions
• A contactor model can be de-embedded
from S-parameter measurements
• A simple model matches measured data up
to at least 10 GHz. (model could have more
bandwidth)
• A model extracted from frequency domain
measurements can be used in a transient
simulation.
BiTS 2002

Force Measurement on
Sockets and Contactors
2002 Burn-in and Test Socket Workshop
March 3 - 6, 2002
Presenter: Richard Block
Rafiq Hussain

Agenda
• Force Issues in Test and Burn-In Env.
• Force Measurement Unit
• Experimentation & Test Data
• Conclusions
• Looking Forward
Jan 10, 2002
Force Measurement on Sockets &
Contactors
2

Force Issues in Test and
Burn-In Env.
• Bent Pins / Deformed Balls
• Overdriving of Pogos
• Crack/Chipped Die or Pkg
•Pkg warping
• Force Distribution
Jan 10, 2002
Force Measurement on Sockets &
Contactors
3

Force Measurement Unit
• Simple Design
– 1 transducer
– 2 spacers
– 1 digital display (giving real-time force
readouts to 0.1 lb accuracy)
• Location
– Replaced support under PCB and Socket
Jan 10, 2002
Force Measurement on Sockets &
Contactors
4

DUT
Thermal Head
FMU consists of 3 pcs
2 spacers
1 transducer
PCB
Jan 10, 2002
Force Measurement on Sockets &
Contactors
5

Board Deflection
• Motherboard deflection was necessary for
transducer to take measurements
– Motherboard was 62 mils thick
– Transducer max deflection was 3 mils for 250lbs
– Deflection did not create any noticeable error in
force readout
• 2 lb, 5 lb, and 10 lb weights were used for confirmation
Jan 10, 2002
Force Measurement on Sockets &
Contactors
6

Experimentation
• FMU was used for various validations
– First Experiment
– Actual force (g/pin) vs. Vendors Spec (g/pin)
• LLCR measurements were taken at various forces
– Passing tests consisted of “no opens pins”
• Both Shorted and Thermal Vehicle pkgs were used
Jan 10, 2002
Force Measurement on Sockets &
Contactors
7

Actual Force vs. Vendor Spec
30
28
28
25
Vendor Spec.
Actual
25
25
Force (grams/pin)
20
15
10
15
18
12
12
11
17
15
16
18
16
5
0
A B (1) B (2) C D E F
Vendor
Jan 10, 2002
Force Measurement on Sockets &
Contactors
8

Experimentation cont.
• FMU was used for various validations
– Second Experiment
– Insertion study to find optimal force vs. Pogo life
•First run:
–100,000 insertions were made at ideal force first
experiment
–LLCR measurements were taken periodically to see if
resistance values increased
•Second run:
–100,000 insertions at different force
–LLCR measurements were compared to first run
Jan 10, 2002
Force Measurement on Sockets &
Contactors
9

Insertion Life vs. Force
Vendor Insertions ave ohms Vendor Insertions ave ohms
A Low 22.02 D Low 23.37
18 g/p 25 k 22.04 16 g/p 25 k 22.75
50 k 22.10 50 k 22.70
75 k 22.09 75 k 22.57
B (1) Low 23.11 E Low 22.24
12 g/p 25 k 23.27 25 g/p 25k 22.10
50 k 22.93 50k 22.34
75 k 22.68 75k 22.30
96 k 22.14
C Low 22.82 F Low 22.25
17 g/p 40 k 22.04 16 g/p 25 k 22.47
80 k 22.16 50 k 22.30
75 k 22.38
Jan 10, 2002
Force Measurement on Sockets &
Contactors
10

Insertion Life vs. Force cont.
• Insertions were done at
a greater force then first
series of insertions.
– ~3 to 4 g/p depending on
initial required force
• Vendor D had pin
failures at this higher
force
Vendor
Insertions ave ohms
B (1)
Low 23.11
15 g/p 30 k 23.41
50 k 23.21
70 k 23.21
D
Low 22.57
19 g/p 25 k 22.73
40 k 22.40
E
Low 22.24
28 g/p 25k 26.67
75k 26.35
Jan 10, 2002
Force Measurement on Sockets &
Contactors
11

Conclusions
• Insight into vendors tolerance control and
machining ability
• Confirmation of test forces on DUT with
reaction forces based on complete test
setup, not individual pieces tested
separately
• More accurate qualification for compression
based sockets
Jan 10, 2002
Force Measurement on Sockets &
Contactors
12

Looking Forward
• Die size Transducers
• Force variation across die/pkg
Jan 10, 2002
Force Measurement on Sockets &
Contactors
13

Transducer for Handlers
Example
Plunger
Contactor Alignment Plate
Transducer
Jan 10, 2002
Force Measurement on Sockets &
Contactors
14

Package/Die Transducer Example
Die Transducer
Thermal Head
DUT
Package Transducer
PCB
Jan 10, 2002
Force Measurement on Sockets &
Contactors
15

Distribution of Force over Die/Pkg
Example
11
10.8
10.6
10.4
PSI
10.2
10
• Check Force across die surface
– Check flatness of thermal head
– Check that socket is level
Jan 10, 2002
Force Measurement on Sockets &
Contactors
16