Research On Flip Chip Ultrasonic Scanning Detection And Image Reconstruction Method

Flip chip technology with solder joints interconnection has been widely applied to electronic packaging,in which the bare die is flipped over and placed down to the substrate.Flip chip technology provides decreased package size,greater I/O density,lower signal delay,and higher speed of signal propagation.However,package interfaces are easily deformed because of thermal expansion coefficient mismatch of chip and substrate,which will result in internal defects of flip chip such as solder bump missing,voids,and cracks,etc.These defects lead to a pernicious influence on the effectiveness of flip chips.Therefore,it is necessary to develop efficient techniques for the inspection of solder bumps in flip chips.In this paper,an intelligent diagnosis system for solder bumps based on the scanning acoustic microscopy(SAM)has been developed.But the original SAM images of the flip chip samples are low resolution,which makes it difficult to detect the defects of flip chips.Therefore,we will apply the super-resolution technique to reconstruct the original SAM images.Following is the main content:The original SAM image is reconstructed based on sparse representation method.It includes two stages: dictionary training and reconstruction.Solder bumps are then segmented according to the gradient matrix of the SAM image,and the statistical features are extracted and adopted for bump classification.LM-BP algorithm is applied to recognize the defective bumps.We present convolutional neural network method for SAM image superresolution.Our method directly learns an end-to-end mapping between the low/high-resolution images.The mapping is represented as a deep convolutional neural network that takes the low-resolution image as the input and outputs the high-resolution one.It includes patch extraction and representation,non-linear mapping and reconstruction.After that,the same image post-processing and recognition algorithm are used for inspection of solder bump.The experimental results prove the feasibility of the proposed method for defects inspection of flip chips.