burn-in & test socket workshop & test socket workshop session 5 ... 860 ph 310 ph 78 ff 23...

March 3-6 , 2002Hilton Phoenix East/Mesa Hotel

Mesa, Arizona

Sponsored By The IEEE Computer SocietyTest Technology Technical Council

Burn-in & TestSocket Workshop

IEEE IEEE

COMPUTERSOCIETY

COPYRIGHT NOTICE• The papers in this publication comprise the proceedings of the 2002BiTS Workshop. They reflect the authors’ opinions and are reproduced aspresented , without change. Their inclusion in this publication does notconstitute an endorsement by the BiTS Workshop, the sponsors, or theInstitute 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 respectivecompanies: as such, proper acknowledgement should be made to theappropriate source. Any questions regarding the use of any materialspresented should be directed to the author/s or their companies.

Burn-in & Test SocketWorkshop

Session 5Tuesday 3/05/02 1:00PM

Modeling, Analysis and Characterization

“Electrical Modeling And Contactor Performance In A RF System”Jim Adley - Johnstech International CorporationEric Leung - Johnstech International CorporationJeff Sherry - Johnstech International Corporation

“Leaded 2mm Contactors: Measuring And Modeling To 10 GHz”Tom Strouth - GigaTest Labs Orlando Bell - GigaTest LabsGary Otonari - GigaTest Labs Eric Bogatin - GigaTest Labs

Jeff Sherry - Johnstech International Corporation

“Force Measurement On Sockets And Contactors”Richard Block - Advanced Micro DevicesRafiq Hussain - Advanced Micro Devices

Technical Program

1

Electrical Modeling andContactor Performance in a RF

Test System0.5 mm Pitch BGA

Jim Adley, R&D ManagerEric Leung, R&D EngineerJeff Sherry, R&D Engineer

Johnstech International

2

Discussion Topics

Modeled Data

Measured Data

Comparative Data

Equivalent Circuit Model

Conclusion

3

Introduction: 0.5 mm PitchBGA Contactor

Cross sectional view of a contactor

4

Modeled Data

Modeling was done using Agilent HFSSsoftware

An AutoCAD drawing, including the actualstructure built by GigaTest Labs, was importedinto HFSS

Data obtained from simulation includes: Return Loss S11

Insertion Loss S21

5

Modeled Data

0.5 mm pitch BGA model from an AutoCAD file

6

Modeled Data - S11 This is data from the HFSS model of return

loss by the actual 0.5 mm BGA structuretested by GigaTest Labs

-50

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

1 2 3 4 5 6 7 8 9 10

Frequency (GHz)

Retu

rn L

oss

(dB

)

HFSS 0.5mm Pitch Adjacent BGA Contact Model

Optimum Load Board Layout of HFSS 0.5mm Pitch AdjacentBGA Contact Model

7

Modeled Data - S21 & S11 This is data from the HFSS model of return

loss by the actual .5 mm BGA structure testedby GigaTest Labs

0.5mm Pitch BGA Adjacent Contact HFSS Insertion Loss Models

-2

-1.75

-1.5

-1.25

-1

-0.75

-0.5

-0.25

0

1 2 3 4 5 6 7 8 9 10Frequency (GHz)

Inse

rtion

Loss

(dB

)


Optimum Load Board Layout of HFSS 0.5mm Pitch Adjacent BGAContact Model

-1 dB Loss Point @ 17 GHz

-1 dB Loss Point @ 13 GHz

8

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

9

Measured Data

GigaTest Labs tested a 0.5 mm pitch BallSeries contactor

GigaTest Labs used a surrogate device -Short, Open, Load, Thru (SOLT) to conducttesting

Data was measured by GigaTest Labsthrough probing from the back side of anon-optimized load board

10

Measured Data

Surrogate Circuit

11

Measured Data GigaTest Labs used a micro probe station for

measuring two adjacent contacts S-parameters

12

Measured Data

The 0.5 mm pitch BGA housing and BGAsurrogate package

13

Comparative DataHFSS Model vs. Measured Test Data for 0.5mm BGA Adjacent Contacts

-50

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

0 1 2 3 4 5 6 7 8 9 10Frequency (GHz)

Ret

urn

Loss

(dB

)


Measured Test Data from GigaTest Labs for 0.5mm Adjacent BGAContact (Non-Optimum pads)

-20 dB Measured Loss Point @ 2.95 GHz

14

Comparative Data0.5mm Pitch BGA Adjacent Contact Modeled vs. Measured Test Data

-2

-1.75

-1.5

-1.25

-1

-0.75

-0.5

-0.25

0

0 1 2 3 4 5 6 7 8 9 10Frequency (GHz)

Inse

rtion

Los

s (d

B)


Measured Test Data from GigaTest Labs for 0.5mm Adjacent BGA Contact (Non-Optimum pads)


15


Characterize the parasitic effects

Simulate the contact with an equivalentcircuit for time domain response

Integrate the contactor into a system levelsimulation to:

Reduce test cost and time

Optimize system performance

16


Measure the S-parameters for short-, open- andthru- fixtures from two adjacent contacts

Use measured data and Agilent AdvancedDesign System (ADS) for model extraction andverification

Compare measured and ADS simulatedinsertion and return loss

17


This figure shows the Equivalent Circuit Modelfor two adjacent 0.5 mm BGA contacts

18

Equivalent Circuit Model0.5mm Pitch BGA Adjacent Contact ADS Equivalent

Circuit Model vs. Measured Test Data

-60

-50

-40

-30

-20

-10

0

0 1 2 3 4 5 6 7 8 9 10Frequency (GHz)

Ret

urn

Loss

(dB

)

ADS Return Loss Model for 0.5mm Adjacent BGA Contact From EquivalentCircuit Measured Test Data from GigaTest Labs for 0.5mm Adjacent BGAContact (Non-Optimum pads)


19

Equivalent Circuit Model Phase comparison shows that the contact is

linear over the frequency modelS11 Phase Comparison for ADS 0.5mm Equivalent Circuit Model vs. Measured

Test Data

-200

-150

-100

-50

0

50

100

150

200

0 1 2 3 4 5 6 7 8 9 10

Frequency (GHz)

Ang

le (D

egre

es)

ADS Return Loss Phase Model for 0.5mm Adjacent BGAContact From Equivalent Circuit Return Loss Phase of Measured Test Data from GigaTest Labsfor 0.5mm Adjacent BGA Contact

20

Equivalent Circuit ModelDevelopment

0.5mm BGA Adjacent Contact Insertion Loss (S11) ADS Equivalent Circuit Model vs. Measured Test Data

-2

-1.5

-1

-0.5

0

0 1 2 3 4 5 6 7 8 9 10Frequency (GHz)

Inse

rtion

Los

s (d

B)

ADS Insertion Loss Model for 0.5mm Adjacent BGA Contact FromEquivalent Circuit Measured Test Data from GigaTest Labs for 0.5mm AdjacentBGA Contact (Non-Optimum pads)


21

Equivalent Circuit ModelDevelopment

S21 Phase Comparison for ADS 0.5mm BGA Equivalent Circuit Model vs. Measured Test Data

-180

-160

-140

-120

-100

-80

-60

-40

-20

0

0 1 2 3 4 5 6 7 8 9 10Frequency (GHz)

Ang

le (D

egre

es)

ADS Insertion Loss Phase Model for 0.5mm Adjacent BGA ContactFrom Equivalent Circuit Insertion Loss Phase of Measured Test Data by GigaTest Labs for0.5mm Adjacent BGA Contact

22

Conclusion For leading edge RF applications such as

0.5 mm pitch, modeling is helpful toachieve optimal system performance

Accurate equivalent circuit models canrepresent contact behavior and help indetermining complete system response

Modeling can help RF engineers save timeand money in development by correctlypredicting system results and eliminatingor reducing hardware builds and testiterations

Leaded 2 mm Contactors:Measuring and Modeling to 10 GHz

Tom Strouth, Orlando Bell, Gary Otonari, Eric BogatinGigaTest Labs, www.GigaTest.com

andJeff Sherry

Johnstech, www.Johnstech.com

BiTS 2002

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 adjacentleads that predicts the measured S parameters

Use this verified, high bandwidth model forperformance evaluation

• StrategyUse measurements of open, short, thru topologiesDe-embed the fixturingUse simplest model for accurate, 10 GHz bandwidthUse SPICE model of de-embedded contactor for

performance simulation

BiTS 2002

Slide - 5

MeasurementConfigurations

1

2

1

2

1

2

open short thru

BiTS 2002

Slide - 6

Surrogate Package:

open

shortthru

Enables configuring open, short, thruconnections for edge and corner leads

BiTS 2002

Slide - 7

Instrument Set Up

BiTS 2002

Slide - 8

Probing From the Back Side

Probing using microprobes and aProbing using microprobes and aGigaTest Probe StationGigaTest Probe Station

BiTS 2002

Slide - 9

Modeling the System withAgilent ADS

VNAVNA FixtureFixtureboardboard

Pads andPads andcontactorcontactor

SurrogateSurrogatepackagepackage

short

open

short

Lcbg

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

S1S150Z

−+=

open

short

open

short

BiTS 2002

Slide - 12

Optimized Model: OpenImpedance of one trace Coupling between traces

Solid Line is Simulated

BiTS 2002

Slide - 13

Optimized Model: Open

Reflection Transmission

Open provides mutual capacitance info

Triangles (purple): Measured

BiTS 2002

Slide - 14

Optimized Model: ShortImpedance of one trace Coupling between traces

Solid Line is Simulated

BiTS 2002

Slide - 15

Optimized Model: ShortReflection Transmission (coupling)

Short provides mutual inductance info


BiTS 2002

Slide - 16

Optimized Model: Loop-ThruReflection Transmission

Bandwidth of the model > 10 GHz


BiTS 2002

Slide - 17 Summary of ExtractedParameters

VNAVNA FixtureFixture ContactorContactor

23 pH23 pH

140 140 fFfF

89 89 fFfF 860 pH860 pH

310 pH310 pH78 78 fFfF

23 pH23 pH

140 140 fFfF 89 89 fFfF860 pH860 pH

SurrogateSurrogateNote: load board was not optimized for performance

151 151 fFfF 33 pH33 pH

25 25 fFfF133 133 fFfF

151 151 fFfF33 pH33 pH

33 pH33 pH

BiTS 2002

Slide - 18

De-Embedded Insertion andReturn Loss of Contactor

Meets specification:< -20 dB, below 10 GHz

Meets specification:> -1 dB, below 10 GHz

Return loss Insertion loss

Note: load board is not optimized for performance,Note: load board is not optimized for performance,standard contactor - standard contactor - notnot enhanced contactor enhanced contactor

BiTS 2002

Slide - 19

Transient Simulation UsingDe-embedded Contactor Model

Specs:

De-embedded contactormodel

20 psec rise time

50 Ω source, termination

Differential drive

BiTS 2002

Slide - 20

Transient Simulationwith 20 psec Rise Time

• 20 psec input rise time

• 28 psec output rise time

• 20 psec intrinsicinterconnect rise time(> 15 GHz bandwidth)

• 21 psec time delay

Note: load board is not optimized for performance,Note: load board is not optimized for performance,standard contactor - standard contactor - notnot enhanced contactor enhanced contactor

BiTS 2002

Slide - 21

Conclusions• A contactor model can be de-embedded

from S-parameter measurements

• A simple model matches measured data upto at least 10 GHz. (model could have morebandwidth)

• A model extracted from frequency domainmeasurements can be used in a transientsimulation.

Force Measurement onSockets and Contactors

2002 Burn-in and Test Socket WorkshopMarch 3 - 6, 2002

Presenter: Richard BlockRafiq Hussain

Jan 10, 2002 Force Measurement on Sockets &Contactors

2

Agenda

• Force Issues in Test and Burn-In Env.• Force Measurement Unit• Experimentation & Test Data• Conclusions• Looking Forward


3

Force Issues in Test andBurn-In Env.

• Bent Pins / Deformed Balls• Overdriving of Pogos• Crack/Chipped Die or Pkg• Pkg warping• Force Distribution


4

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


5

FMU consists of 3 pcs2 spacers1 transducer

PCB

DUT

Thermal Head


6

Board Deflection

• Motherboard deflection was necessary fortransducer 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


7

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


8

Actual Force vs. Vendor Spec

15

12

28

11

15

18

25

18

12

28

1716

25

16

0

5

10

15

20

25

30

A B (1) B (2) C D E F

Vendor

Forc

e (g

ram

s/pi

n)

Vendor Spec.Actual


9

• FMU was used for various validations– Second Experiment– Insertion study to find optimal force vs. Pogo life

Experimentation cont.

•Second run:–100,000 insertions at different force–LLCR measurements were compared to first run

•First run:–100,000 insertions were made at ideal force firstexperiment–LLCR measurements were taken periodically to see ifresistance values increased


10

Insertion Life vs. ForceVendor Insertions ave ohms Vendor Insertions ave ohms

A Low 22.02 D Low 23.3718 g/p 25 k 22.04 16 g/p 25 k 22.75

50 k 22.10 50 k 22.7075 k 22.09 75 k 22.57

B (1) Low 23.11 E Low 22.2412 g/p 25 k 23.27 25 g/p 25k 22.10

50 k 22.93 50k 22.3475 k 22.68 75k 22.30

96 k 22.14

C Low 22.82 F Low 22.2517 g/p 40 k 22.04 16 g/p 25 k 22.47

80 k 22.16 50 k 22.3075 k 22.38


11

Insertion Life vs. Force cont.Vendor Insertions ave ohms

B (1)Low 23.11

15 g/p 30 k 23.4150 k 23.2170 k 23.21

DLow 22.57

19 g/p 25 k 22.7340 k 22.40

ELow 22.24

28 g/p 25k 26.6775k 26.35

• Insertions were done ata greater force then firstseries of insertions.– ~3 to 4 g/p depending on

initial required force• Vendor D had pin

failures at this higherforce


12

Conclusions

• Insight into vendors tolerance control andmachining ability

• Confirmation of test forces on DUT withreaction forces based on complete testsetup, not individual pieces testedseparately

• More accurate qualification for compressionbased sockets


13

Looking Forward

• Die size Transducers

• Force variation across die/pkg


14

Plunger

TransducerContactor Alignment Plate

Transducer for HandlersExample


15

Package/Die Transducer Example

PCB

DUT

Thermal Head

Package Transducer

Die Transducer


16

Distribution of Force over Die/PkgExample

10

10.2

10.4

10.6

10.8

11

PSI

• Check Force across die surface– Check flatness of thermal head– Check that socket is level

burn-in & test socket workshop & test socket workshop session 5 ... 860 ph 310 ph 78 ff 23...

Documents