Prober and Probe Card Analyzer Performance Under Load

Raymond Kraft, Ph.D.Applied Precision, Inc.

South West Test Workshop5 June 2001San Diego, CA USA

2

Presentation Agenda

• Motivation• Loads and Deflection Mechanisms• Probe Tip Mechanics• Loaded Prober and Probe Card Analyzer (PCA)

Performance• Summary and Conclusions

3

Motivation

• Increasing Loads + Tighter Accuracy – No matter how stiff we design stages, if we’re driven

to scrutinize closely enough, we’ll always see the effects of deflection under load• Deflection can impact yield• ⇒ Deflection under load increasingly important

• Presentation Objective:– Understand manifestations of deflection under load to

aid problem identification– Suggest means of reducing effects of deflection under

load

4

Normal Loads and Deflections

F

Chuck

Probe Card

Center of Stiffnessx

z Undeflected

Deflected

• Direction: Perpendicular to wafer and probe array surface

• Origin: Probe tip stiffness• Generally linear with overtravel• Vertical probe technology can be non-linear• z-direction compliance ⇒ z-direction deflection• Deflection: ∆z = F cz

5

Normal Loads and Deflections• Normal load effects on scrub

– z-deflection reduces overtravel ⇒ shorter scrubs– Normal force reduced– Overtravel can be adjusted to account for expected compliance – Adjustment errors produce longer or shorter scrubs

Overcompensated Overtravel

UndercompensatedOvertravel

6

Transverse Loads and Deflections• Transverse Loads

– Direction: In-plane– Origin: Probe tip friction– Symmetry: Transverse loads largely cancel on

average– Potential for deflection-induced errors with

asymmetric probe orientations

N1 N2

F2F1N1 ≅ N2 F1 ≅ F2

Probe

7

Torsional Loads and Deflections• Origin: Center of load

does not pass through center of stiffness

• ⇒ Rotation about center of stiffness– Torque: T = F l– Rotation: θ = F l crot

• Effects:– Dependent on center of

stiffness – Lateral deflection - often

dominant– Vertical deflection

F

x

z

l

θ

Probe Card

Chuck

Center of Stiffness

8

Center of Stiffness OffsetNormalized Offset Relative to Center of Force

Goal: Minimize Offset

0.20 .4

0 .60 .8

1

0 .2

0 .4

0 .6

0 .8

1-1

0

1

X S tag e Lo c atio nY S tag e Lo c atio n

0.20 .4

0 .60 .8

1

0.2

0 .4

0 .6

0 .8

1-1

0

1

X S tag e Lo c atio nY S tag e Lo c atio n

X Center of Stiffness Offset

Largest moment arms ⇒ Deflection Potential

Y Center of Stiffness Offset

9

Chuck Deflection from Torsion

050

100150

200

050

100150

2000

1

2

X Array S izeY Array S ize

Small ArraysNarrow Arrays

Large Arrays

• Factors– Array size– Number of pins on chuck– Center of load relative to stiffness

3σ Normalized Scrub Alignment Variation Under Load vs Array Size

10

Probe Geometries• Cantilever Probes

– Longitudinal deflection• Stiff axis• 1:1 deflection into scrub

– Lateral deflection• More flexible axis• Probes tend to drag

somewhat - reduces sensitivity to deflection effects

• <1:1 deflection into scrub– 0.3:1 common

Top View

Realized Total Scrub

LateralLongitudinal

OT Scrub

-Stage Long DefRealized Lat Scrub

-Stage Lat Def

11

Cantilever Probe Lateral Deflection

0 0.1 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7-1

-0 .9

-0 .8

-0 .7

-0 .6

-0 .5

-0 .4

-0 .3

-0 .2

-0 .1

Co e ffic ie nt o f Kine tic Fric tio n

Rat

io o

f lat

eral

scr

ub le

ngth

to la

tera

l def

lect

ion

k=0.05k=0.1k=0.2k=0.3k=0.4k=0.8k=1.5k=3.0k=6.0

Typical

Decreasing Lateral Stiffness

More DraggingLess Sensitive

Less DraggingMore Sensitive

Stiffness (gf/mil)

Effects of Friction and Lateral Probe Stiffness on Lateral Deflection

12

Probe Geometries

FlexibleFl

exib

le

• Vertical Probes– Relatively compliant with respect to in-plane forces

(e.g. friction)– Probes are “dragged” with chuck as it deflects– Deflection induced error governed by high probes– Highly planar arrays forgiving of deflection even

though loads can be high

13

Vertical Probes: Deflection Effects

First Touch

High probe

Chuck

Last Touch

Last touch at fraction

of total load

DeflectionAlignment Error

14

Vertical Probe Deflection

0 0.1 0.2 0 .3 0 .4 0 .5 0 .6 0 .7-1

-0 .9

-0 .8

-0 .7

-0 .6

-0 .5

-0 .4

-0 .3

-0 .2

-0 .1

0

Co e ffic ie nt o f Kine tic Fric tio n

Rat

io o

f scr

ub le

ngth

to d

efle

ctio

nk=0 .05k=0.1k=0.2k=0.3k=0.4k=0.8k=1.5k=3.0k=6.0

Decreasing Stiffness

More DraggingLess Sensitive

Less DraggingMore Sensitive

Effects of Friction and Probe Stiffness on Deflection

Stiffness (gf/mil)

15

PCA: Cantilever Deflection Estimation

• Co-located, opposing probes

– Co-located ⇒ Equiv loads, Equiv deflections for PCA– Equal and opposite variations in scrub length– Equal and opposite variations in scrub angle– Data can be used to estimate deflection

• Deflection ⇒ bimodal distribution of scrub length and/or scrub angle– Dependent on compliance in each axis

16

Cantilever: Statistical Distribution

0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.40

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

P la na riz e d S c rub Le ngth

Pro

babi

lity

Den

sity

Norma l Dis tributionDa ta

High Stage Compliance in Scrub DirectionNormalized Scrub Length Distribution

Long ScrubsShort Scrubs

17

Cantilever: Statistical Distribution

0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.40

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

P la na riz e d S c rub Le ngth

Pro

babi

lity

Den

sity

Norma l Dis tributionDa ta

Low Stage Compliance in Scrub DirectionNormalized Scrub Length Distribution

Unimodal Distribution

18

Prober: Cantilever Scrubs

R

T

L

L R

T

• Rotation under torsional load ⇒ lateral deflection• Wafer scrub example

– Center of stiffness at wafer center– 1 x 2 DUT

19

Deflection Compensation

• Modeling & Estimation– Model chuck compliance as a function of location

• Experimentally determine load/displacement influence coefficient map

• Estimate influence coefficients based on known component values and FEA

– Model probe loads • Nominal probe gram force parameter and overtravel• Calculate total load and load centroid

– ⇒ Estimate Chuck Deflection

20

Deflection Compensation

• Passive Compensation– Post-process results, and remove estimated

deflection from measurements – Possible only for PCA’s

• Active Compensation– Actively translate chuck using deflection estimates – Estimate real-time loads and deflections– Any prober compensation needs to be active

21

Deflection Compensation

• Deflection Compensation Example

0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.40

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

P la na riz e d S c rub Le ngth

Pro

babi

lity

Den

sity

Norm a l Dis tributionDa ta

0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.40

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

P la na riz e d S c rub Le ngth

Pro

babi

lity

Den

sity

Norm a l Dis tributionDa ta

Compensated

Un-Compensated

Normalized Scrub Length

22

Conclusions

• Ways to reduce probe scrub errors:– Cantilever and Microspring Technology

• Reduce probe gram force and overtravel• Minimize distance from chuck center of stiffness to

center of load• Optimize card orientation: align probes with least

compliant axis if possible• Reduce chuck compliance

23

Conclusions

• Ways to reduce probe alignment errors– Vertical Technology

• Tighten probe card planarity if possible• Minimize distance from chuck center of stiffness to

center of load• Reduce probe gram force• Minimize chuck compliance• Overtravel not a significant factor

24

Conclusions

• Deflection under load increasingly important issue

• Deflection can be estimated from standard test data, and quantified

• Deflection may be minimized proactively in a number of ways

• Deflection effects may be minimized via compensation