EE290H F05 Spanos

Control Charts for Attributes

P-chart (fraction non-conforming)

C-chart (number of defects)
U-chart (non-conformities per unit)
The rest of the “magnificent seven”

Lecture 11: Attribute Charts 1

EE290H F05 Spanos

Yield Control

100 100

80 80

60 60
Yield

40 40

20 20

0 0
0 10 20 30 0 10 20 30

Months of Production

Lecture 11: Attribute Charts 2

EE290H F05 Spanos

The fraction non-conforming

The most inexpensive statistic is the yield of the production

line.
Yield is related to the ratio of defective vs. non-defective,
conforming vs. non-conforming or functional vs. non-
functional.
We often measure:
• Fraction non-conforming (P)
• Number of defects on product (C)
• Average number of non-conformities per unit area (U)

Lecture 11: Attribute Charts 3

EE290H F05 Spanos

The P-Chart
The P chart is based on the binomial distribution:

n x n-x
P{D = X} = p (1-p) x = 0,1,...,n
x
mean np
variance np(1-p)
the sample fraction p= D
n
mean p
variance p(1-p)
n

Lecture 11: Attribute Charts 4

EE290H F05 Spanos

The P-chart (cont.)

p must be estimated. Limits are set at +/- 3 sigma.

m
mean p
Σ pi
p=
i=1
variance p(1-p)
m nm

(in
(in this
this and
and the
the following
following discussion,
discussion, "n" "n" isis the
the number
number ofof
samples
samples in in each
each group
group and
and "m"
"m" isis the
the number
number of of groups
groups
that
thatweweuseusein
inorder
orderto todetermine
determinethe
thecontrol
controllimits)
limits)

Lecture 11: Attribute Charts 5

EE290H F05 Spanos

Designing the P-Chart

In general, the control limits of a chart are:
UCL= µ + k σ
LCL= µ - k σ
where k is typically set to 3.
These formulae give us the limits for the P-Chart (using the
binomial distribution of the variable):

p can be estimated:
Di
pi = i = 1,...,m (m = 20 ~25)
p(1-p) n
UCL = p + 3 n m

LCL = p - 3
p(1-p)
n
Σ pi
i=1
p=
m
Lecture 11: Attribute Charts 6
EE290H F05 Spanos

Example: Defectives (1.0 minus yield) Chart

0.5

% non-conforming 0.4 UCL 0.411

0.3
0.232
0.2

0.1
LCL 0.053
0.0
0 10 20 30
Count
"Out
"Out of
of control
control points"
points" must
must be
be explained
explained and and eliminated
eliminated before
before we
we
recalculate
recalculatethe
thecontrol
controllimits.
limits.
This
Thismeans
meansthat
thatsetting
settingthe
thecontrol
controllimits
limitsisisan
aniterative
iterativeprocess!
process!
Special
Specialpatterns
patternsmust
mustalso
alsobe
beexplained.
explained.
Lecture 11: Attribute Charts 7
EE290H F05 Spanos

Example (cont.)
After the original problems have been corrected, the
limits must be evaluated again.

Lecture 11: Attribute Charts 8

EE290H F05 Spanos

Operating Characteristic of P-Chart

In order to calculate type I and II errors of the P-chart we
need a convenient statistic.
Normal approximation to the binomial (DeMoivre-Laplace):

if n large and np(1-p) >> 1, then

(x-np) 2
n 1 -
P{D = x} = p x (1-p) (n-x) ~ e 2np(1-p)
x 2πnp(1-p)

In other words, the fraction nonconforming can be treated as

having a nice normal distribution! (with μ and σ as given).
This can be used to set frequency, sample size and control
limits. Also to calculate the OC.

Lecture 11: Attribute Charts 9

EE290H F05 Spanos

Binomial distribution and the Normal

bin 10, 0.1 bin 100, 0.5 bin 5000 0.007

4.0
65 50

60 45
3.0
55 40
35
2.0 50
30
45
1.0 25
40
20
35
0.0 15

As
Assample
samplesize
sizeincreases,
increases,the
theNormal
Normalapproximation
approximationbecomes
becomesreasonable...
reasonable...

Lecture 11: Attribute Charts 10

EE290H F05 Spanos

Designing the P-Chart

Assuming that the discrete distribution of x can be
approximated by a continuous normal distribution as
shown, then we may:
• choose n so that we get at least one defective with
0.95 probability.
• choose n so that a given shift is detected with 0.50
probability.
or
• choose n so that we get a positive LCL.
Then, the operating characteristic can be drawn from:

β = P { D < n UCL /p } - P { D < n LCL /p }

Lecture 11: Attribute Charts 11

EE290H F05 Spanos

The Operating Characteristic Curve (cont.)

The OCC can be calculated two distributions are
equivalent and np=λ).

p = 0.20,
LCL=0.0303,
UCL=0.3697

Lecture 11: Attribute Charts 12

EE290H F05 Spanos

In reality, p changes over time

(data from the Berkeley Competitive Semiconductor Manufacturing Study)

Lecture 11: Attribute Charts 13
EE290H F05 Spanos

The C-Chart
Sometimes we want to actually count the number of defects.
This gives us more information about the process.
The basic assumption is that defects "arrive" according to a
Poisson model:
e-c cx
p(x) = x = 0,1,2,..
x!
μ = c, σ 2 = c
This assumes that defects are independent and that they
arrive uniformly over time and space. Under these
assumptions: UCL = c + 3 c
center at c
LCL = c - 3 c

and c can be estimated from measurements.

Lecture 11: Attribute Charts 14

EE290H F05 Spanos

Poisson and the Normal

poisson 2 poisson 20 poisson 100

10 130
30 120
8
110
6
20 100
4 90

2 80
10
70
0

As
Asthe
themean
meanincreases,
increases,the
theNormal
Normalapproximation
approximationbecomes
becomesreasonable...
reasonable...

Lecture 11: Attribute Charts 15

EE290H F05 Spanos

Example: "Filter" wafers used in yield model

Fraction Nonconforming (P-chart)
0.5

UCL 0.454

0.4
Fraction Nonconforming

0.3 ¯ 0.306

0.2

LCL 0.157

0.1
0 2 4 6 8 10 12 14

Defect Count (C-chart)

200
Number of Defects

100 UCL 99.90

Ý 74.08

LCL 48.26

0
0 2 4 6 8 10 12 14
Wafer No

Lecture 11: Attribute Charts 16

EE290H F05 Spanos

Counting particles

Scanning a “blanket” monitor wafer.

Detects position and approximate size of particle.

y
x

Lecture 11: Attribute Charts 17

EE290H F05 Spanos

Scanning a product wafer

Lecture 11: Attribute Charts 18

EE290H F05 Spanos

Typical Spatial Distributions

Lecture 11: Attribute Charts 19

EE290H F05 Spanos

The Problem with Wafer Maps

Wafer maps often contain information that is very
difficult to enumerate

A simple particle count cannot convey what is happening.

Lecture 11: Attribute Charts 20

EE290H F05 Spanos

Special Wafer Scan Statistics for SPC applications

• Particle Count
• Particle Count by Size (histogram)
• Particle Density
• Particle Density variation by sub area (clustering)
• Cluster Count
• Cluster Classification
• Background Count

Whatever
Whatever we
we use
use (and
(and we
we might
might have
have to
to use
use more
more
than
thanone),
one),must
mustfollow
followaaknown,
known,usable
usabledistribution.
distribution.

Lecture 11: Attribute Charts 21

EE290H F05 Spanos

In Situ Particle Monitoring Technology

Laser light scattering system for detecting particles in

exhaust flow. Sensor placed down stream from
valves to prevent corrosion.
Laser
chamber

to pump

Detector

Assumed to measure the particle concentration in

vacuum
Lecture 11: Attribute Charts 22
EE290H F05 Spanos

Progression of scatter plots over time

The endpoint detector failed during the ninth lot, and was
detected during the tenth lot.

Lecture 11: Attribute Charts 23

EE290H F05 Spanos

Time series of ISPM counts vs. Wafer Scans

Lecture 11: Attribute Charts 24

EE290H F05 Spanos

The U-Chart
We could condense the information and avoid outliers by
using the “average” defect density u = Σc/n. It can be
shown that u obeys a Poisson "type" distribution with:
μu = u, σ2u = u
n
so
UCL =u + 3 u n
LCL =u - 3 u n
where u is the estimated value of the unknown u.
The sample size n may vary. This can easily be
accommodated.

Lecture 11: Attribute Charts 25

EE290H F05 Spanos

The Averaging Effect of the u-chart

poisson 2 average 5
5.0
10
4.0
8
3.0
6

4 2.0

2 1.0

0 0.0
Quantiles Quantiles

Moments Moments

By
Byexploiting
exploitingthe
thecentral
centrallimit
limittheorem,
theorem,ififsmall-sample
small-samplepoisson
poissonvariables
variables
can
canbe
bemade
madeto toapproach
approachnormal
normalbybygrouping
groupingandandaveraging
averaging

Lecture 11: Attribute Charts 26

EE290H F05 Spanos

Filter wafer data for yield models (CMOS-1):

0.5
Fraction Nonconforming (P-chart)
UCL 0.454
0.4

Fraction Nonconforming 0.3 ¯ 0.306

0.2

LCL 0.157

0.1
0 2 4 6 8 10 12 14

200
Defect Count (C-chart)
Number of Defects

100 UCL 99.90

Ý 74.08
LCL 48.26

0
0 2 4 6 8 10 12 14

6
Defect Density (U-chart)
5
Defects per Unit

4
UCL 3.76

3
× 2.79

2
LCL 1.82

1
0 2 4 6 8 10 12 14
Wafer No
Lecture 11: Attribute Charts 27
EE290H F05 Spanos

The Use of the Control Chart

The control chart is in general a part of the feedback loop
for process improvement and control.

Input Output
Process

Measurement System
Detect
Verify and assignable
follow up cause

Implement Identify root

corrective cause of problem
action
Lecture 11: Attribute Charts 28
EE290H F05 Spanos

Choosing a control chart...

...depends very much on the analysis that we are

pursuing. In general, the control chart is only a small
part of a procedure that involves a number of statistical
and engineering tools, such as:
• experimental design
• trial and error
• pareto diagrams
• influence diagrams
• charting of critical parameters

Lecture 11: Attribute Charts 29

EE290H F05 Spanos

The Pareto Diagram in Defect Analysis

Typically, a small number of defect types is responsible
for the largest part of yield loss.
The most cost effective way to improve the yield is to
identify these defect types.

figure 3.1 pp 21 Kume

Lecture 11: Attribute Charts 30

EE290H F05 Spanos

Pareto Diagrams (cont)

Diagrams by Phenomena

• defect types (pinholes, scratches, shorts,...)

• defect location (boat, lot and wafer maps...)
• test pattern (continuity etc.)

Diagrams by Causes

• operator (shift, group,...)

• machine (equipment, tools,...)
• raw material (wafer vendor, chemicals,...)
• processing method (conditions, recipes,...)

Lecture 11: Attribute Charts 31

EE290H F05 Spanos

Example: Pareto Analysis of DCMOS Process

DCMOS Defect Classification
100

80

60
Percentage

occurence
cummulative
40

20

others
ontamination

sed contacts
ss problems

se particles
ern bridging
evious layer

s scratches

Though the defect classification by type is fairly easy, the

classification by cause is not...
Lecture 11: Attribute Charts 32
EE290H F05 Spanos

Cause and Effect Diagrams

(Also known as Ishikawa,fish bone or influence diagrams.)

figure 4.1 pp 27 Kume

Creating such a diagram requires good understanding of

the process.

Lecture 11: Attribute Charts 33

EE290H F05 Spanos

An Actual Example

Lecture 11: Attribute Charts 34

EE290H F05 Spanos

Example: DCMOS Cause and Effect Diagram

Past Steps Parametric Control Particulate Control

rec. handling cassettes
inspection SPC
SPC automation
calibration equipment
monitoring
cleaning
Defect
skill
chemicals
vendor
experience loading
utilities
shift transport vendor
filters
boxes
Operator Wafers Handling Contamination Control

Lecture 11: Attribute Charts 35

EE290H F05 Spanos

Example: Pareto Analysis of DCMOS (cont)

DCMOS Defect Causes
100

80

60
percentage

occurence
cummulative
40

20

0
loading
equipmnet

others
inspection
utilities

smiff boxes

Once classification by cause has been completed,

we can choose the first target for improvement.
Lecture 11: Attribute Charts 36
EE290H F05 Spanos

Defect Control
In general, statistical tools like control charts must be
combined with the rest of the "magnificent seven":
• Histograms
• Check Sheet
• Pareto Chart
• Cause and effect diagrams
• Defect Concentration Diagram
• Scatter Diagram
• Control Chart

Lecture 11: Attribute Charts 37

EE290H F05 Spanos

Logic Defect Density is also on the decline

Y = [ (1-e-AD)/AD ]2

Lecture 11: Attribute Charts 38

EE290H F05 Spanos

What Drives Yield Learning Speed?

Lecture 11: Attribute Charts 39