Optical Leak Testing Technology October 9th, 2014 Hermetic Leak - PowerPoint PPT Presentation

Optical Leak Testing Technology October 9th, 2014

Hermetic Leak Test Methods Helium Mass Spectrometry (Fine) Bubble Leak testing (Gross) Dye Penetrant (Gross) (Destructive) Residual Gas Analysis (Destructive) Cumulative Helium Leak Detection ( Fine and Non-Batch Gross) Optical Leak Testing ( Fine and Gross Batch Testing)

Historic Limitations and Technical Issues Conventional HMS, and Bubble Leak Testing Helium absorption on the package body High bomb pressures and long cycle times One way leakers “Virtual leakers” Inability to test at the board level Repeatable leak rate test coverage in the 10 -05 cc- atm/sec range Problems with helium backfills Product contamination and masking of fine leaks during bubble testing Costs of flourinert , and environmental concerns using perfluorocarbon with bubble leak testing

Range of Leak Sensitivity (Component Level) Gross Leak -6 Fine Leak Bubble II Helium II Optical I Leak Test True Leak Rate (deceasing ) Only Optical Leak Testing covers the full required test range in one method.

Optical Leak Testing Features Simultaneous gross and fine leak inspection on up to 200 devices in a single test Shortened test cycles (eliminates He bombing) Quantitative results in cc-atm/second, He Can test PC board mounted devices One-step calibration and setup High volume production testing Results can be networked for SPC or to a pick and place machine to remove rejected devices.

Component Testing But t er erflies es Pacem em aker ers Frequency C Cont rol Di Display De y Devi vices Quart art z C Cry ryst al als Hybr brids ds Pow er er D Dev evices es Fibe ber Opt i pt ics Cust st om TO Styles

Sources of Leaks in Packages Along the weld At the glass to metal feedthrough Or through a defect in the package itself

Optical Leak Testing Optical Leak Testing Test Procedure Test Procedure

Start Test End Test Package Lid is Flat Gross or Open Lid Moves Up Fine Leaker Chamber Pressure Deforms Lid No Hermetic Change Sealed

Optical Leak Testing Schematic Diagram OPTICAL LEAK TEST SYSTEM SCHEMATIC DIAGRAM LASER WINDOW DIGITAL CCD CAMERA ---tt-- • i • ~----- • 1-200 HERMETIC SYSTEM COMPUTER PACKAGES IMAGE TEST CELL 1-5 BAR HELIUM PROCESSOR

Laser Interferometer Phase Maps Hermetic: only one fringe • 0 ··- · •·- ,: ~ - ~ . a 6 . •• • ...... Gross: no fringes. . 1, .... . :..... . :.- . :. - ,:.-... EiJ • · •··- [9] ~ No lid movement. • 1, .'•-- Fail: Many fringes 1 fringe = 0.26 um indicate lid moving up. = ½ wavelength of light

Example of Leaking Device and Relative Lid Deflection

Optical Leak Testing Optical Leak Testing Fine Leak Testing Fine Leak Testing

True Leak Rate Measurement Achieved by: • Determining lid deflection (um) due leak in the package. • Modulating the chamber pressure sinusoidally to determine lid stiffness (um/psi).

• Lid Stiffness: um/psi • Modulate the chamber pressure sinusoidally to determine lid stiffness. 60.1 psig Test 60.0 psig Pressure 59.9 psig Start End of test

• Leakage is change in package pressure. • Leakage= Lid movement (um) Lid stiffness (um/psi) • Leak Rate = - Ln Delta P final V Delta P initial t • Volume (cc) is free air volume. • t (sec) test time in seconds. • Delta P initial = test pressure. • Delta P final = test pressure – leakage.

• The measured leak rate is converted to a standard leak rate by dividing by the chamber pressure. • Leak Rate (standard) = Leak Rate (measured) Test Pressure (atm) • The output from the Optical Leak Tester is a true leak rate (cc-atm / sec) which is called “L” in MIL-STD-883.

• OLT can test to the ‘11’s L air or ‘14’s R 1 • Recent test on wafer level device. • 24 hours, 64 psig, 0.01 cc, • stiffness -0.05 um/psi. • L he = 1.04 x 10 -10 cc-atm/sec (actual) • L air = 3.87 x 10 -11 cc-atm/sec (calculated) • R 1 = 3.93 x 10 -14 cc-atm/sec (calculated) • OLT can match the test sensitivity or any leak test method available including Krypton 85.

System Operation System Operation

Testing a full matrix of pig tailed butterfly packages. Butterfly devices in process tray Operator places tray into system

Operator Chooses the Appropriate Part number I '',, • ,I Al'wavs Print Graphic Never Ptinl Graptuc Print Graphic Onty it Parts f eil Print Text Results The operator then presses Start

~ ~ ~ ~ ci! Optical L ea k Ted S~.tcm lbultc olly 9• 1J pcf) Test Results Automatic Test Analysis Test In Progress with a set Accept/Reject threshold shows 6 leaking and 4 hermetic packages. ~-- T c~ c~ t Configuration t C onfiguratio1 II . . . INq L ~12 - . . • lrll!lll:z!l'I · : ,,..,.,1,.- 1C IU'~ ~)F1'lll'nMSOI\ 1, "' I

Optical Leak Test Systems

Beta Site Production Test Data • Collected on production parts from  Collected on production parts from Vectron International, Hudson NH Vectron International, Hudson NH beginning in Jan 2003 beginning in Jan 2003 • Three different package volumes/styles  Three different package volumes/styles and four different machines and four different machines • Over 110,000 parts tested, only a very  Over 110,000 parts tested, only a very small percentage of parts failed and were small percentage of parts failed and were confirmed by conventional HMS and confirmed by conventional HMS and bubble test per TM 1014 bubble test per TM 1014 • During Beta site testing the parts were  During Beta site testing the parts were tested to equivalent air leak rate below tested to equivalent air leak rate below 1EE-08 cc/sec air. 1 EE-08 cc/sec air.

Package Details (XTAL Oscillators) Volume .03cc 5 mil thick kovar lid (7.5 X 7.5 mm) 50 psi chamber pressure, 4 minute test Test Sensitivity level set at 7.4 X10 -09 cc/sec L He Equivalent to 2.7X10 -09 cc/sec AIR

Package Details (XTAL Oscillators) Volume .1cc 5 mil thick kovar lid (7X14mm) lid stiffness 1 μ m/psi 25 PSI chamber pressure, 4 minute test Test Sensitivity level set at 9.8 X10 -08 cc/sec L He Equivalent to 3.7X10 -08 cc/sec AIR

Machine Qualification Test Data X-Tal PKG A: # Tested TM 1014 Rejects OLT Rejects (gross/fine) Machine #1 3106 12 12 Machine #3 2582 20 19 X-Tal PKG B: Machine #1 892 0 0 Machine #3 892 0 0 Totals 7742 32 31 Bottom Line: .01% of total parts tested (only one in 7742) passed using OLT but failed TM 1014 Source: Vectron Qual Data

Testing Board Mounted Devices Discrete hermetic devices that pass MIL STD 883 helium testing may fail during or after board assembly due to helium absorption. Because Optical Leak Testing is not subject to gas absorption issues, devices on assembled circuit cards can be leak tested with OLT.

Optical Leak Testing History Optical Leak Testing History MIL-STD-883 MIL-STD-883 • Included into MIL-Std-883 Method 1014 in 1995. • Included into MIL-Std-883 Method 1014 in 1995. • Updated in 2004 for Gross and Fine leak testing on • Updated in 2004 for Gross and Fine leak testing on both individual, components and board level testing. both individual, components and board level testing. • The system outputs true leak rate instead of lid • The system outputs true leak rate instead of lid deflection. deflection. • Leaking devices samples are no longer needed for • Leaking devices samples are no longer needed for calibration. calibration. • Revision J draft has been submitted to DSCC (DLA) • Revision J draft has been submitted to DSCC (DLA) in May of 2012. in May of 2012. • Failure criteria and test methodology is better defined. • Failure criteria and test methodology is better defined.

Optical leak testing using advanced laser interferometry technology  Tests for Gross and Fine leaks in one test  No Helium bombing required  No Helium absorption problems  Eliminates bubble leak testing for gross leaks  Used in production in the following industries: MEMS, communications, micro-electronics, automotive and aerospace devices and sensors, implantable medical electronics.  Mature technology with a proven track record. (Over 100 systems installed worldwide)

Presentation Contact Chris Aubertin Tel. 610-592-0167 ttrafford@NorComSystemsInc.com