Impact of 0201 Components on Automated X-Ray Inspection 2005 - - PowerPoint PPT Presentation

Slide 1 EBTW 2005, Tallinn, Estonia

EBTW05

Impact of 0201 Components

• n Automated X-Ray

Inspection

2005 European Board Test Workshop

CHWEE LIONG TEE

Intel

• Corporation

Kulim, Malaysia

Slide 2 EBTW 2005, Tallinn, Estonia

EBTW05

Agenda

Objective of Study Design of Experiment (DOE) Summary Backup Overview

Slide 3 EBTW 2005, Tallinn, Estonia

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).

Slide 4 EBTW 2005, Tallinn, Estonia

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

Slide 5 EBTW 2005, Tallinn, Estonia

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

Slide 6 EBTW 2005, Tallinn, Estonia

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

Slide 7 EBTW 2005, Tallinn, Estonia

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.

Slide 8 EBTW 2005, Tallinn, Estonia

EBTW05

Test Time

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.

Capacitor Average test time (s) Resistor Average test time (s) Total Views FOV 23.825 23.725 65 400 20.425 20.725 27 650 19.325 18.975 18 800

Slide 9 EBTW 2005, Tallinn, Estonia

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.

4660 315 800 4365 138 650 3874 52 400

Resistor panel (number of calls) Capacitor panel (number of calls) FOV

Slide 10 EBTW 2005, Tallinn, Estonia

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

• f focus for the

secondary side.

Slide 11 EBTW 2005, Tallinn, Estonia

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

• perator to determine if the call was real.

Slide 12 EBTW 2005, Tallinn, Estonia

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.

Slide 13 EBTW 2005, Tallinn, Estonia

EBTW05

Comparison of Threshold Settings Based on FOV

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.

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.

Slide 14 EBTW 2005, Tallinn, Estonia

EBTW05

Comparison of Threshold Settings Based on FOV

Resistor-loaded Panel

0.499 0.066 0.565 800 1.068 0.12 1.188 650 1.384 0.094 1.478 400 Difference Min Open Signal Threshold on Bare Reflow Panel Min Open Signal Threshold on Loaded Panel FOV

Slide 15 EBTW 2005, Tallinn, Estonia

EBTW05

Threshold Setting Comparison Base

• n FOV
• 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.

1.162 800 1.134 650 1.348 400 Pad Thickness Threshold FOV

Capacitor-loaded Panel

Slide 16 EBTW 2005, Tallinn, Estonia

EBTW05

Effect of Pad & Stencil Opening Sizes

• n 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

• pening sizes were studied.

1:1 0.5

• 1. 10x13x8
• 2. 12x12x10
• 3. 12x12x12
• 4. 12x13x10
• 5. 12x13x12
• 6. 12x13x8
• 7. 12x14x10
• 8. 12x14x8
• 9. 14x13x8

Stencil Openings Pad Sizes

(Width x Length x Inter-pad Distance)

Slide 17 EBTW 2005, Tallinn, Estonia

EBTW05

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

1.313 1.282 14 x 13 x 8 1.352 1.179 12 x 14 x 10 1.320 1.170 12 x 14 x 8 1.134 0.834 12 x 13 x 12 1.407 1.107 12 x 13 x 10 1.287 1.116 12 x 13 x 8 1.182 0.913 12 x 12 x 12 1.187 0.926 12 x 12 x 10 1.219 0.974 10 x 13 x 8

Average Min Open Signal Threshold

1:1 1:0.5 PAD SIZES

Effect of Pad & Stencil Opening Sizes on Threshold Settings

Resistor-loaded Panel

Slide 18 EBTW 2005, Tallinn, Estonia

EBTW05 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.

Stencil Opening: 0.5 Stencil Opening: 1.0

Effect of Pad & Stencil Opening Sizes on Threshold Settings

Resistor-loaded Panel

Slide 19 EBTW 2005, Tallinn, Estonia

EBTW05

Effect of Pad & Stencil Opening Sizes on Threshold Settings

Capacitor-loaded Panel

1.76 1.70

14 x 13 x 8

1.93 1.75

12 x 14 x 10

1.92 1.75

12 x 14 x 8

1.70 1.67

12 x 13 x 12

1.79 1.66

12 x 13 x 10

1.72 1.60

12 x 13 x 8

1.80 1.76

12 x 12 x 12

2.02 1.75

12 x 12 x 10

1.27 1.22

10 x 13 x 8

Average Pad Thickness Threshold

PAD SIZES 1:1 1:0.5 STENCIL OPENINGS

Slide 20 EBTW 2005, Tallinn, Estonia

EBTW05 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

• penings.

Effect of Pad & Stencil Opening Sizes on Threshold Settings

Capacitor-loaded Panel

Slide 21 EBTW 2005, Tallinn, Estonia

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.

Slide 22 EBTW 2005, Tallinn, Estonia

EBTW05

Backup

Slide 23 EBTW 2005, Tallinn, Estonia

EBTW05

Capacitor and Resistor X-ray Images

Slide 24 EBTW 2005, Tallinn, Estonia

EBTW05

Explanation of Minimum 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.

Slide 25 EBTW 2005, Tallinn, Estonia

EBTW05

Explanation of Min Open Signal Threshold

For resistors, the min open signal is the difference between the green and yellow regions.

Slide 26 EBTW 2005, Tallinn, Estonia

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

Slide 27 EBTW 2005, Tallinn, Estonia

EBTW05

Explanation of Pad Thickness Threshold

Description: The pad thickness threshold sets the expected thickness in mils of solder for the pad for opaque components.