Fusion of Ultrasound and X-ray Data for Automatic Inspection of Flip Chip and BGA Solder Joints

1
Fusion of Ultrasound and X-ray Data for Automatic
Inspection of Flip Chip and BGA Solder Joints
Ryan Yang 27/02/2009
2
Presentation Outline

• Introduction
• Acoustic Micro Imaging
• X-ray Imaging
• Image Registration
• Image Fusion
• Conclusion

3
Introduction

• Solder joint reliability is a primary concern in
  the assembly of all Electronic components and
  products.
• The importance of solder joint reliability became
  more emphasized in recent years as a result of
  three factors

1) The shift from leaded to lead-free solders in
semiconductor industry.
2) Shrinking in die size as well as solder balls
dimension.
3) The emergence of fine-pitched area array
packages that employ hundreds of solder joints
for electrical connection.
4
Introduction
5
Introduction
Hidden Solder Joints !?
6
Introduction

• Acoustic Micro Imaging and X-ray imaging are
  principle Non-Destructive Testing techniques.
• Both techniques can penetrate through the
  component to image the hidden solder joints.
• X-ray inspection and AMI are complementary, each
  technology has distinct discriminating features
  and is good at inspecting certain defects.

7
Introduction
X-ray Inspection AMI Inspection
Voids Bridging / short Broken wire Missing Ball / Open Insufficient reflow Voids Delaminations Popcorn Crack Disbonds

• AMI is an effective approach for detecting
  gap-type defects due to strong reflections of
  ultrasound at a solid-air interface.
• X-ray inspection is able to identify volumetric
  defects which are hard to detect by AMI.

8
Introduction

• Penetration of AMI through several layers of
  dissimilar materials is a big challenge for AMI,
  whereas X-ray penetration is good but without the
  discrimination accuracy.
• Inspection of flip chip and BGA solder joints
  still remains a significant challenge to current
  testing techniques
• FUSION of ultrasound and X-ray data for flip chip
  and BGA solder joints provides a novel way to
  interpret and analyse the image of the solder
  joints and potentially increases the resolution
  of very small dimensions

9
Introduction

• When two complementary techniques are combined,
  they could be helpful in reinforcing certain
  evaluations, improving feature measurement
  resolution, technique can also applied to other
  fields.
• Combining multiple image modalities to provide a
  single, enhanced picture is offering an added
  value and more informative data to the processor
  in order to developing an automated inspection
  system. (Smith, 2005)
• In future, other techniques such as MRI. Infrared
  and AFM could be added.

10
Acoustic Micro Imaging

• Acoustic Micro Imaging (AMI) is makes use of the
  properties of ultrasonic waves which range from
  5MHz to 400MHz.
• Ultrasonic waves are generated by a piezoelectric
  transducer and propagate through an object.
• When the wave travels through the object, it may
  be scattered, reflected and absorbed with respect
  to the differences between acoustic properties of
  materials.

11
Acoustic Micro Imaging
(Images adapted from Sonoscan Inc)
Different imaging modes are used for locating
certain defects.
12
Acoustic Micro Imaging
13
X-Ray Imaging

• X-ray microscope imaging uses electromagnetic
  radiation in the soft X-ray band to produce
  images of very small objects.
• When X-rays pass through a materials, it
  experience a variety of scattering interaction.
  These interactions lead to energy attenuation and
  the energy is detected by a Charge Coupled
  Devices (CCD).
• X-rays imaging is a contrast imaging technique
  where high density materials lead to higher
  attenuation and hence produce darker image than
  those with less density or thickness.

14
X-Ray Imaging
15
Image Registration

• The essential step in the fusion process is to
  bring the X-ray and C-scan images into spatial
  alignment, known as registration.
• Image Registration is the process of overlaying
  two or more images of the same scene taken at
  different times, different viewpoints or
  different sensors.
• The registration geometrically align two images
  or transform different set of data into one
  coordinate system.

16
Image Registration

• Registration is have been widely used in

17
Image Registration

• Open Source Registration tools
• ITK-Insight software consortium
• AIR- Roger P. Woods, M.D., UCLA School of
  Medicine
• FLIRT FMRIB centre, University of Oxford
• DROP - Technische Universität München (TUM) ,
  Germany
• BunwarpJ - Arganda-Carreras , Universidad
  Autonoma de Madrid
• No pre-processing and mainly developed for
  biomedical images

18
Image Registration
Figure 7 Feature Detection
Figure 8 Feature Matching (Images adapted from
Zitova,2003 )
19
Image Registration
Figure 9 Transformation Model Estimation
Figure 10 Image Resampling and
Transformation (Images adapted from Zitova,2003 )
20
Image Registration
Figure 11 Original Images
Figure 12 Processed Images
21
Image Registration

• Point based methods and Least Square
  Approximation
• The transformation that aligns the corresponding
  fiducial points will interpolate the mapping from
  these points to other points in the view.

22
Image Registration

• Compute the weighted centroid of the fiducial
  configuration in each space

• Compute the weighted fiducial covariance matrix

23
Image Registration

• Perform singular value decomposition (SVD) of H

• Finally,

24
Image Registration
Figure 14 LabVIEW Program for Computing Point
Based Method
25
Image Fusion

• The term Image Fusion generally implies the
  intelligent combination of multi-modality sensor
  imagery for the purpose of providing an enhanced
  single view of a scene with extended information
  content. (Smith, 2005)
• Fundamental Standard of fusion
• The fused image should preserve all salient
  information of source images.
• The fusion process should not introduce any
  artefacts or inconsistencies into the fused
  image.
• Undesirable features (noise) should be suppressed
  in the fused image.

26
Image Fusion
Fusion method Description
Maximum amplitude Compare pixel value and extract maximum amplitude
Integration Use AND operator to fuse images
Weighted averaging Evaluate weight (probability) of defect detection based on knowledge of the images
Kalman filter Combine signals to enhance information by recursive filter
Bayesian analysis Decision making between hypothesis
Dempster-Shafer Decision making with belief intervals
Fuzzy Logic Fuse image follow fuzzy rules. Used when no mathematical relationship of images is available
Multi Resolution Method Extract the salient features at several levels of image decomposition from coarse to fine
Table 1 Table of Fusion Method
27
Image Fusion
Maximum Amplitude and Weighted Pixel Averaging

• Common Fusion Algorithm approaches

• Disadvantages
• Also suppresses salient features
• Low contrast
• washed-out appearance

• Advantages
• Easy implemented
• Fast to execute
• Suppressing noise

28
Image Fusion

• Multi-Resolution Methods
• Extract the salient features at several levels of
  image decomposition from coarse to fine
• Pyramidal Schemes
• Gaussian Pyramid
• Laplacian Pyramid
• Wavelet Schemes
• Colour Fusion

• Advantages
• Produce sharp, high-contrast images

• Disadvantages
• Reserve unwanted features
• Further Assessment is required

29
Image Fusion
Pyramidal Schemes
30
Image Fusion

• Reduce operation
• Expand operation

31
Image Fusion

• Wavelet Schemes
• Discrete Wavelet Transform

32
Image Fusion

• Rescaling is usually done in power of two

33
Conclusion

• Increasing the Solder Joints reliability can
  increase the product life time and increases
  customer quality.
• Reduces potential warranty costs.
• Image fusion provide a new method to keep
  inspection of hidden solder joints in line with
  the rapid reduction in component size
• Improves the ability to inspect smaller
  dimensions seen in newer packaging. May also
  improve the inspection of area array parts such
  as BGA which contain interposer
• Image fusion remains a challenging technology and
  its application in electronic inspection is less
  mature and required additional research and
  assessment

34
References
ZHANG, G.M., HARVEY, D.M. and BRADEN, D.R. (2006)
X-ray Inspection and Acoustic Micro Imaging
Applied to Quality Testing of BGA Solder Joints
A Comparative Study, 2nd GERI Annual Research
Symposium GARS 2006, Liverpool, UK, 15th June
2006 KAPUR, A. and et al (2002) Fusion of
Digital Mammography with Ultrasound A phantom
Study, Proc of SPIE The international Society
of Optical Engineering, 4682, p.526-537 SEMMENS,
J.E. (2000) Flip Chis and Acoustic Micro
Imaging An overview of Past Application, Present
Status, and Roadmap for the Future. Proceedings
of ESREF conference, Dresden, Germany, October
2000 ZITOVA, B. and FLUSSER, J. (2003) Image
Registration Methods A Survey, Image and Vision
Computing vol. 21, p977-1000 SMITH, M.I. and
HEATHER, J.P. (2005) A review of image fusion
technology in 2005 Thermosense XXVII.
Proceedings of the SPIE, Vol. 5782, pp. 29-45
35
Thank You for your Attention!