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