EBTW05 Impact of 0201 Components on Automated X-Ray Inspection 2005 European Board Test Workshop CHWEE LIONG TEE Intel � � � � Corporation Kulim, Malaysia EBTW 2005, Tallinn, Estonia Slide 1
EBTW05 Agenda � Objective of Study � Design of Experiment (DOE) � Summary � Backup � Overview EBTW 2005, Tallinn, Estonia Slide 2
EBTW05 Overview � Trend of using 0201 components is increasing � Driving force for using 0201 components is miniaturization of consumer products Passive device trends (Source: Prismark). EBTW 2005, Tallinn, Estonia Slide 3
EBTW05 Objective Study how 0201 components will impact Automated X-ray Inspection (AXI): � Test time � False calls � Escapes � Comparison of threshold based on Field of View (FOV) � Effect of different pad sizes on threshold settings EBTW 2005, Tallinn, Estonia Slide 4
EBTW05 Design of Experiment � Two types of boards: � 0201 Resistor-loaded board � 0201 Capacitor-loaded board � Three component arrangements: � Horizontal � Vertical � 45 degrees angle � Two types of stencil openings: � 1:1 � 1:0.5 � Non functional board EBTW 2005, Tallinn, Estonia Slide 5
EBTW05 Design of Experiment Used nine different pad sizes and inter-pad distances. Dimensions are in mils (width x length x inter-pad distance). � 10x13x8 � 12x12x10 � 12x12x12 � 12x13x8 � 12x13x10 � 12x13x12 � 12x14x8 � 12x14x10 � 14x13x8 EBTW 2005, Tallinn, Estonia Slide 6
EBTW05 Test Time � Critical when you are running High Volume Manufacturing � Automated X-ray Inspection (AXI) machine used is capable of inspecting seven field of views (FOV). � Three FOVs were studied—800, 650, 400. � Units for FOV are in mils � Resistor algorithm was used to inspect resistor components. � Chip algorithm was used to inspect capacitor components. EBTW 2005, Tallinn, Estonia Slide 7
EBTW05 Test Time Resistor Average test Capacitor Average FOV Total Views time (s) test time (s) 800 18 18.975 19.325 650 27 20.725 20.425 400 65 23.725 23.825 � Using the same FOV resulted in same test times. � Using different FOVs generated different number of views, which impacted the test times significantly. � Using different algorithms did not affect the test time. EBTW 2005, Tallinn, Estonia Slide 8
EBTW05 False Calls � Defined as no defect associated with the indictment. � 400 FOV was chosen for the study because it had the lowest total number of calls. Capacitor panel Resistor panel FOV (number of calls) (number of calls) 400 52 3874 650 138 4365 800 315 4660 EBTW 2005, Tallinn, Estonia Slide 9
EBTW05 False Calls � False calls are higher on resistor panels than on capacitor panels. � Factors contributing to higher false calls: � Only one panel was used to develop and fine-tune the program. � Real defects for resistors were also higher than for capacitors. � Some images were out of focus for the secondary side. EBTW 2005, Tallinn, Estonia Slide 10
EBTW05 Escapes � Defined as the failure to capture real defects. � 3 of each panel type were fault- injected with 100 defects— missing, tombstone, skewed, short, and open. � All panels were run at 400 FOV. � Fewer escapes were recorded for resistor panels because more calls were made on them compared to capacitor panels, enabling operator to determine if the call was real. EBTW 2005, Tallinn, Estonia Slide 11
EBTW05 Escapes � The AXI machine was still able to capture the defects by adjusting the threshold settings which is a challenge to a programmer. � It is important to have a bare reflow panel as a reference to ensure the test program is able to fail all the pins in the bare reflow panel. EBTW 2005, Tallinn, Estonia Slide 12
EBTW05 Comparison of Threshold Settings Based on FOV Objective of this study is to find the optimum FOV taking into consideration beat rate and false calls. Setting the right threshold has significant impact on false calls and escapes. Resistor-loaded Panels � Parameter studied is the minimum open signal threshold. � The average minimum open signal threshold of resistor panel is compared to the bare reflow panel. � The minimum open signal threshold should be set so that all the solder joints of the bare reflow panel will fail. � From the table below, the largest difference in the minimum open signal threshold between loaded and bare reflow panels is obtained using 400 FOV followed by 650 FOV and 800 FOV. � Based on this, 400 is the recommended maximum FOV for resistors. EBTW 2005, Tallinn, Estonia Slide 13
EBTW05 Comparison of Threshold Settings Based on FOV Resistor-loaded Panel FOV Min Open Signal Min Open Signal Difference Threshold on Loaded Threshold on Bare Panel Reflow Panel 400 1.478 0.094 1.384 650 1.188 0.12 1.068 800 0.565 0.066 0.499 EBTW 2005, Tallinn, Estonia Slide 14
Threshold Setting Comparison Base EBTW05 on FOV Capacitor-loaded Panel � Parameter studied is the pad thickness threshold. � The highest pad thickness reading is recorded using 400 FOV. � Based on this, 400 is the recommended maximum FOV for capacitors. FOV Pad Thickness Threshold 400 1.348 650 1.134 800 1.162 EBTW 2005, Tallinn, Estonia Slide 15
Effect of Pad & Stencil Opening Sizes EBTW05 on Threshold Settings Objective of this study is to find out the pad and stencil opening sizes that will provide the biggest separation in threshold readings relative to the bare reflow panel. The impact of nine different 0201 pad sizes and two stencil opening sizes were studied. Pad Sizes Stencil Openings (Width x Length x Inter-pad Distance) 1. 10x13x8 1:1 2. 12x12x10 0.5 3. 12x12x12 4. 12x13x10 5. 12x13x12 6. 12x13x8 7. 12x14x10 8. 12x14x8 9. 14x13x8 EBTW 2005, Tallinn, Estonia Slide 16
Effect of Pad & Stencil Opening Sizes on EBTW05 Threshold Settings Resistor-loaded Panel The average min open signal is the average of the differences between the min open signal of the loaded board and bare reflow board . STENCIL OPENINGS 1:0.5 1:1 PAD SIZES Average Min Open Signal Threshold 10 x 13 x 8 0.974 1.219 12 x 12 x 10 0.926 1.187 12 x 12 x 12 0.913 1.182 12 x 13 x 8 1.116 1.287 12 x 13 x 10 1.107 1.407 12 x 13 x 12 0.834 1.134 12 x 14 x 8 1.170 1.320 12 x 14 x 10 1.179 1.352 14 x 13 x 8 1.282 1.313 EBTW 2005, Tallinn, Estonia Slide 17
Effect of Pad & Stencil Opening EBTW05 Sizes on Threshold Settings Resistor-loaded Panel Stencil Opening: 0.5 Stencil Opening: 1.0 � Overall, the min open signal threshold differences are within the lower and upper control limits for both stencil openings and pad sizes. � Stencil 1.0 is recommended, because of a greater threshold difference for the min open signal . � All pad sizes studied are within the control limit. EBTW 2005, Tallinn, Estonia Slide 18
Effect of Pad & Stencil Opening EBTW05 Sizes on Threshold Settings Capacitor-loaded Panel STENCIL OPENINGS 1:0.5 1:1 PAD SIZES Average Pad Thickness Threshold 10 x 13 x 8 1.22 1.27 12 x 12 x 10 1.75 2.02 12 x 12 x 12 1.76 1.80 12 x 13 x 8 1.60 1.72 12 x 13 x 10 1.66 1.79 12 x 13 x 12 1.67 1.70 12 x 14 x 8 1.75 1.92 12 x 14 x 10 1.75 1.93 14 x 13 x 8 1.70 1.76 EBTW 2005, Tallinn, Estonia Slide 19
Effect of Pad & Stencil Opening EBTW05 Sizes on Threshold Settings Capacitor-loaded Panel � Stencil opening 1.0 is recommended because of a higher pad thickness threshold. � Pad size impacts the solder profile as measured by the pad thickness. � Pad size 10x13x8 gives the lowest readings for both stencil openings. EBTW 2005, Tallinn, Estonia Slide 20
EBTW05 Summary � AXI can detect 0201 faults. � FOV setting will impact test time, false calls, and escapes. � For resistors and capacitors, 400 FOV is recommended. � It is a good practice to have a bare reflow panel for setting proper thresholds. � Pad size and stencil openings impact the solder profile. � Stencil opening of 1 is recommended because it gives a clearer threshold separation between loaded and bare reflow for resistors. � Care needs to be taken to ensure clear images which may involve slowing test speed, using smaller FOV, placing more laser surface mapping; however, this will impact test time. � Balancing test time, false calls, and escapes is a great challenge for test developers. EBTW 2005, Tallinn, Estonia Slide 21
EBTW05 Backup EBTW 2005, Tallinn, Estonia Slide 22
EBTW05 Capacitor and Resistor X-ray Images EBTW 2005, Tallinn, Estonia Slide 23
Explanation of Minimum EBTW05 Open Signal Threshold Resistor minimum open signal : This threshold sets the minimum acceptable Open Signal. It is the thickness difference in mils of solder between the peak fillet (blue box) and the pad under the component (yellow box). Units are in mils. EBTW 2005, Tallinn, Estonia Slide 24
EBTW05 Explanation of Min Open Signal Threshold For resistors, the min open signal is the difference between the green and yellow regions. EBTW 2005, Tallinn, Estonia Slide 25
EBTW05 Affect of Pad Thickness on Threshold Pad thickness is different for resistors and capacitors because the algorithm looks at different regions of interest (ROI). Capacitor ROI Resistor ROI EBTW 2005, Tallinn, Estonia Slide 26
Recommend
More recommend
