MAX3950EGK Rev. A RELIABILITY REPORT FOR MAX3950EGK PLASTIC ENCAPSULATED DEVICES October 14, 2003 MAXIM INTEGRATED PRODUCTS 120 SAN GABRIEL DR. SUNNYVALE, CA 94086 Written by Reviewed by Jim Pedicord Quality Assurance Reliability Lab Manager Bryan J. Preeshl Quality Assurance Executive Director Conclusion The MAX3950 successfully meets the quality and reliability standards required of all Maxim products. In addition, Maxim’s continuous reliability monitoring program ensures that all outgoing product will continue to meet Maxim’s quality and reliability standards. Table of Contents I. ........Device Description II. ........Manufacturing Information III. .......Packaging Information IV. .......Die Information V. ........Quality Assurance Information VI. .......Reliability Evaluation ......Attachments I. Device Description A. General The MAX3950 deserializer is ideal for converting 10Gbps serial data to 16-bit wide, 622Mbps parallel data in SDH/SONET and DWDM applications. Operating from a single +3.3V supply, this device accepts CML serial clock and data inputs, and delivers low-voltage differential-signal (LVDS) clock and data outputs for interfacing with high-speed digital circuitry. The MAX3950 is available in the extended temperature range (-40°C to +85°C) in a 68-pin QFN package. The typical power dissipation is 900mW. B. Absolute Maximum Ratings Item Positive Supply Voltage (VCC) CML Input Voltage Level LVDS Output Voltage Level Operating Temperature Range Storage Temperature Range Lead Temperature (soldering, 10s) Continuous Power Dissipation (TA = +85°C) 68-Pin QFN Derates above +85°C 68-Pin QFN Rating -0.5V to +5.0V (VCC - 0.8V) to (VCC + 0.5V) -0.5V to (VCC + 0.5V) -40°C to +85°C -55°C to +150°C +300°C 2800mW 43.5mW/°C II. Manufacturing Information A. Description/Function: +3.3V, 10.7Gbps 1:16 Deserializer with LVDS Outputs B. Process: GST4-F60 C. Number of Device Transistors: 4800 D. Fabrication Location: Oregon, USA E. Assembly Location: Korea F. Date of Initial Production: October, 2000 III. Packaging Information A. Package Type: 68-Pin QFN (10 x 10) B. Lead Frame: Copper C. Lead Finish: Solder Plate D. Die Attach: Silver-filled epoxy E. Bondwire: Gold (1.2 mil dia.) F. Mold Material: Epoxy with silica filler G. Assembly Diagram: Buildsheet # 05-7001-0460 H. Flammability Rating: Class: UL94-V0 I. Classification of Moisture Sensitivity per JEDEC standard JESD22-A112: Level 3 IV. Die Information A. Dimensions: 115 x 99 mils B. Passivation: Si3N4 (Silicon nitride) C. Interconnect: Au D. Backside Metallization: None E. Minimum Metal Width: Metal1: 1.2; Metal2: 1.2; Metal3: 1.2; Metal4: 5.6 microns (as drawn) F. Minimum Metal Spacing: Metal1: 1.6; Metal2: 1.6; Metal3: 1.6; Metal4: 4.2 microns (as drawn) G. Bondpad Dimensions: 5 mil. Sq. H. Isolation Dielectric: SiO2 I. Die Separation Method: Wafer Saw V. Quality Assurance Information A. Quality Assurance Contacts: Jim Pedicord (Manager, Reliability Operations) Bryan Preeshl (Executive Director of QA) Kenneth Huening (Vice President) B. Outgoing Inspection Level: 0.1% for all electrical parameters guaranteed by the Datasheet. 0.1% For all Visual Defects. C. Observed Outgoing Defect Rate: < 50 ppm D. Sampling Plan: Mil-Std-105D VI. Reliability Evaluation A. Accelerated Life Test The results of the 150°C biased (static) life test are shown in Table 1. Using these results, the Failure Rate (λ) is calculated as follows: λ= 1 = MTTF 1.83 (Chi square value for MTTF upper limit) 192 x 9823 x 135 x 2 Temperature Acceleration factor assuming an activation energy of 0.8eV λ = 3.59 x 10-8 λ = 3.59 F.I.T. (60% confidence level @ 25°C) This low failure rate represents data collected from Maxim’s reliability qualification and monitor programs. Maxim also performs weekly Burn-In on samples from production to assure reliability of its processes. The reliability required for lots which receive a burn-in qualification is 59 F.I.T. at a 60% confidence level, which equates to 3 failures in an 80 piece sample. Maxim performs failure analysis on rejects from lots exceeding this level. The Burn-In Schematic (Spec.# 06-6948) shows the static circuit used for this test. Maxim also performs 1000 hour life test monitors quarterly for each process. This data is published in the Product Reliability Reports (RR-1M & RR-B3A). B. Moisture Resistance Tests Maxim evaluates pressure pot stress from every assembly process during qualification of each new design. Pressure Pot testing must pass a 20% LTPD for acceptance. Additionally, industry standard 85°C/85%RH or HAST tests are performed quarterly per device/package family. C. E.S.D. and Latch-Up Testing The HF76Z die type has been found to have all pins able to withstand a transient pulse of +/-1500V, per Mil-Std883 Method 3015 (reference attached ESD Test Circuit). Latch-Up testing has shown that this device withstands a current of ±250mA. Table 1 Reliability Evaluation Test Results MAX3950EGK TEST ITEM TEST CONDITION Static Life Test (Note 1) Ta = 150°C Biased Time = 192 hrs. FAILURE IDENTIFICATION SAMPLE SIZE NUMBER OF FAILURES DC Parameters & functionality 135 0 Moisture Testing (Note 2) Pressure Pot Ta = 121°C P = 15 psi. RH= 100% Time = 168hrs. DC Parameters & functionality 77 0 85/85 Ta = 85°C RH = 85% Biased Time = 1000hrs. DC Parameters & functionality 77 0 DC Parameters & functionality 77 Mechanical Stress (Note 2) Temperature Cycle -65°C/150°C 1000 Cycles Method 1010 Note 1: Life Test Data may represent plastic DIP qualification lots. Note 2: Generic process/package data. 0 Attachment #1 TABLE II. Pin combination to be tested. 1/ 2/ Terminal A (Each pin individually connected to terminal A with the other floating) Terminal B (The common combination of all like-named pins connected to terminal B) 1. All pins except VPS1 3/ All VPS1 pins 2. All input and output pins All other input-output pins 1/ Table II is restated in narrative form in 3.4 below. 2/ No connects are not to be tested. 3/ Repeat pin combination I for each named Power supply and for ground (e.g., where VPS1 is VDD, VCC, VSS, VBB, GND, +VS, -VS, VREF, etc). 3.4 Pin combinations to be tested. a. Each pin individually connected to terminal A with respect to the device ground pin(s) connected to terminal B. All pins except the one being tested and the ground pin(s) shall be open. b. Each pin individually connected to terminal A with respect to each different set of a combination of all named power supply pins (e.g., VSS1, or VSS2 or VSS3 or VCC1 , or VCC2 ) connected to terminal B. All pins except the one being tested and the power supply pin or set of pins shall be open. c. Each input and each output individually connected to terminal A with respect to a combination of all the other input and output pins connected to terminal B. All pins except the input or output pin being tested and the combination of all the other input and output pins shall be open. TERMINAL C R1 R2 S1 TERMINAL A REGULATED HIGH VOLTAGE SUPPLY S2 C1 DUT SOCKET SHORT TERMINAL B Mil Std 883D Method 3015.7 Notice 8 TERMINAL D R = 1.5kΩ C = 100pf CURRENT PROBE (NOTE 6)