Effect of Solder Joint Degradation on RF Impedance

University of MarylandCopyright © 2008 CALCE

1Center for Advanced Life Cycle Engineering12th Workshop on Signal Propagation on Interconnects

Effect of Solder Joint Degradation on RF Impedance

SPI’200812th

Workshop on Signal Propagation on InterconnectsMay 15, 2008

Daeil Kwon, Dr. Michael H. Azarian, and Prof. Michael G. Pecht

Center for Advanced Life Cycle Engineering (CALCE)University of Maryland

College Park, MD 20742

[email protected]



MotivationAccompanying the growing use of microprocessors and signal processors in electronics has come the need to reliably transmit signals having frequencies of several hundred megahertz or more across interconnects.

Projected off-chip frequency for high-performance interconnects [1].

0

10

20

30

40

50

60

70

2006 2008 2010 2012 2014 2016 2018 2020 2022 2024

Year

Off

-Chi

p Fr

eque

ncy

[GH

z]

ITRS 2007 Roadmap



Signal Propagation at High Frequencies•

The skin depth refers to the distance below the surface of a conductor at which the current density falls to 1/e (about 37%).

•

At frequencies over about 100 MHz, skin depth becomes less than a tenth of interconnect size.

0

1

10

100

1000

100k 1M 10M 100M 1G 10G

Frequency (Hz)

Skin

Dep

th (μ

m) Thickness of 1-oz copper

Diameter of solder ball

Eutectic tin-lead

Copper



Interconnect Degradation Often Starts from the Surface of the Interconnect and Propagates Inward

This observation generally applies to a variety of failure mechanisms, including:

•

Fatigue (thermal or vibration)•

Mechanical over-stress

•

Creep•

Corrosion



Detection of Solder Joint Degradation•

Due to the skin effect, a small crack at the surface of a solder

joint can directly

influence signal integrity, which may reduce the performance of high speed electronic products.

•

Traditional methods for monitoring interconnect reliability do not adequately detect initial stages of degradation.

•

RF impedance provides early warning of failure and is a better means of assessing the lifetime of products used for high speed electronics.

DC resistance RF impedance

Impe

danc

e

ZRF

Timetf

-∆tRF

tf

-∆tDC

tf

Time

Res

ista

nce

RDC

tf

-∆tDC

tf

Failure criterion

∆

tDC ∆tRF



Time Domain Reflectometry (TDR)

LPFImpulse SMT low pass

filter (LPF)on 50 Ohm controlledimpedance board

•

TDR reflection coefficient (Γ) is the ratio of the incident and reflected impulse caused by impedance discontinuities in the circuit [2].

•

In the time domain, any discontinuities due to impedance mismatches within the circuit are seen as discrete peaks.

•

TDR reflection coefficient is used as a measure of RF impedance.

0

0

reflected L

incident L

P Z ZP Z Z

−Γ = =

+

•

-1 ≤ Γ ≤ 1; Γ is dimensionless, and can be conveniently reported in milliunits, mU

•

Γ=1 when ZL

=∞; Γ=-1 when ZL

=0- ZL

: the impedance of device under test- Z0

: characteristic impedance of the circuit (50 Ohm)

Vector network analyzer (VNA)

Time

TDR

resp

onse

(Γ)

0



Experimental Setup

LPF

RF RF+DC

DC

Bias-Tee

DC

RF RF+DC

Bias-Tee

VNA

Cyclic loadTytron

250

•

Controlled mechanical force is applied and measured using an MTS Tytron

250 to perform fatigue tests on an impedance-

controlled test vehicle.•

RF and DC measurements are automated for continuous monitoring during the fatigue tests.

Port 1 Port 2



Test Conditions•

Equipment –

MTS Tytron

250 (load unit)

–

Agilent E8364A vector network analyzer (RF measurement)–

Keithley

2010 digital multimeter

(DC measurement)

•

It was verified that operation of the MTS Tytron

250 did not introduce measurable electrical noise into the RF impedance measurement.

Applied shear force (blue, command)

Measured shear force (red, response)

•

Test variables–

Monitoring RF frequency: 500 MHz ~ 6 GHz

–

Shear force: 40±10 N–

Load frequency: 0.25 Hz

–

Data acquisition interval: every 30 sec



0

100

200

300

400

500

600

700

8 8.5 9 9.5 10 10.5 11

Signal transit time (ns)

TD

R r

efle

ctio

n co

effic

ient

(mU

)

Sample with intact solder joint

Sample with failed solder joint

TDR reflection coefficient at the failure site

TDR Reflection Coefficient MeasurementsTDR reflection coefficients at the failure site were extracted from the collected TDR data to examine their changes.



0

10

20

30

40

50

0 200 400 600 800 1000 1200 1400

Test time (min)

TD

R r

efle

ctio

n co

effic

ient

(mU

)

3

4

5

6

7

8

DC

res

itanc

e (o

hm)

DC resistance

TDR reflection coefficient at failure site

Comparison between RF and DC ResponsesIn the fatigue test, RF response increased by 25% of the initial

value

before the failure, while DC resistance remained almost constant.



20

22

24

26

28

30

1240 1250 1260 1270 1280 1290 1300 1310

Test time (min)

TD

R r

efle

ctio

n co

effic

ient

(mU

)

3

4

5

6

7

8

DC

res

ista

nce

(Ohm

s)TDR reflection coefficient at the failure site

5% increase of the initial value



DC resistance

Quantification of Early Response of RF Impedance5%, 10%, and 15% increase of the initial TDR value were observed

at 36.5, 3.5, and 1 minute(s) prior to the failure, respectively.



150

152

154

156

158

160

0 50 100 150 200 250 300 350

Test time (Min)

TD

R r

efle

ctio

n co

effic

ient

(mU

)

3

4

5

6

7

8

DC

res

ista

nce

(Ohm

s)

TDR reflection coefficient at the failure site

DC resistance

The test was stopped

Intermediate Stage of Solder Joint DegradationA program was written to stop the fatigue tests, in order to allow failure analysis, at intermediate stages of solder joint degradation.



Cross-Sectioning Direction and Plane of Observation

The partially degraded solder joint was cross-sectioned along the direction of the signal trace.

Board

SMT low pass filter Solder joint

Copper pad

Cross-sectioningdirection

Plane of cross-section



Failure Analysis of Degraded Solder Joints (1)

Crack

Slide 15

Shear forcedirection

Slide 16

Solder joint

Low pass filter

Copper pad



Solder joint Copper pad

Low pass filter

Crack

Shear forcedirection





Solder joint Copper pad

Low pass filter



Conclusions•

RF impedance changes are detectable prior to changes in DC resistance during solder joint degradation.

•

A crack extending only partway across a solder joint is sufficient to cause an increase of RF impedance but no change in DC resistance.

•

RF impedance provides a more accurate assessment of the reliability of high speed electronic products in response to solder joint degradation.



Potential Benefits

•

A more sensitive and non-destructive measurement technique for reliability monitoring of electronic assemblies.

•

A basis for accelerated test failure criteria which are better correlated to failures of high speed electronic products.

•

The ability to localize failure sites.•

Substantial savings in operational and repair costs through –

condition-based maintenance,

–

reduction of unplanned down-time, and–

reduced incidence of “no trouble found”

failures due to intermittent

contact behavior.



References1.

“International Technology Roadmap For Semiconductors”, Assembly and Packaging Chapter, Tables AP2a and AP2b, 2007.

2.

“TDR Fundamentals: For Use with HP-54120T Digitizing Oscilloscope and TDR,”

Hewlett-Packard Application Note 62, 1988.

Effect of Solder Joint Degradation on RF Impedance

Documents