CFAR: Practical Reliability

Practical Reliability

Jose Cornelio Flores, Cypress Manufacturing Ltd.

Last May 26, the Council of Failure Analysis and Reliability held its Monthly Learning Session at ARCDI, Alabang together with IHS CEO/CTO Mr. Henry Tan. His talk, entitled “Practical Reliability” opened up the minds of the more than 40 participants to the essentials of Reliability Engineering and so as with Failure Analysis.

Henry tackled about practicality in life, whether it would be in the lab or at home. He said, practicality is doing something that makes sense and “cents”. He emphasized that in this modern times, practicality plays a very important role in semiconductor companies such that is should be employed in every area especially in Reliability and Failure Analysis.

The IHS CEO began his topic by giving an overview of Reliability Engineering and what it’s all a bout. The objective of a Reliability or Failure Analysis Engineer is to provide accurate and timely results so that other engineers around them could also have a better assessment on the problem, and on top of that is a good /quality report. These results came from data that were carefully collected and analyzed after being put under accelerated stress tests. It is important that the equipment used is not only capable to do the job, but also has the correct settings/parameters and properly maintained before conducting the actual test. The key item is how one chooses the right machine with the right capabilities.

Henry also stressed out that although there are a lot of machines available in the market today, some of these are, he considered, not practical for semiconductor packaging stress tests. For instance, there are moisture soak machines that have the capability of running -40^oC up to 180^oC and up to 100% RH. But based on industry standards, like JEDEC,
moisture soak conditions only require between 30^oC to 85^oC and a maximum of 85% RH.

Courtesy of IPC/JEDEC J-STD-020C – July 2004

Machines such as this, according to Henry, cause more difficulty in the long run. Though it is quite attractive at first because of its extended capabilities, some parts of the machine that gives these capabilities eventually wears down, as most of the time, these parts are also activated during operation of the machine. The worst thing to happen is if it wears down first before the other “critical” parts. This is the time when the user realizes that it was impractical because the machine becomes idle, losing a lot of productive work and causing delay, until he/she replaces/fixes the defective part which he is not getting any benefit. That’s why Henry recommends as much as possible, to acquire machines which is just enough to do the intended job.

Among the criteria that he emphasized for a practical machine are:

1.) It is able to satisfy the requirements and loading capacity set by the user (may be based on a standard or specification, i.e. JEDEC, IPC, JIS, IEEE, AEC, etc.)

2.) It’s special features will truly benefit the user and will greatly help in achieving the set requirements (i.e. HAST machine is known to be better than the PCT machine when conducting Unbiased Highly Accelerated Stress Test or Pressure Cooker Test because of its capability to automatically control parameters during ramp up and ramp down)

3.) It’s initial cost, and cost of operation and maintenance is on the “practical” level (i.e. power consumption, heat dissipation, power saving modes, etc.)

Different reliability failure modes were also discussed during the learning session. Henry enumerated some and these are summarized below.

Stress Test	Conditions	Expected Failures
Highly Accelerated Stress Test	130oC/85%RH	Corrosion, Excessive Intermetallics, Leakage due to top die ionic contamination (dissolved ionic substance from wafer Fab or die saw/attach)
Pressure Cooker Test	121oC/100%RH	Corrosion, Leakage due to ionic contamination (dissolved upon penetration of moisture through the mold compound)
Temperature Cycle Test	-65oC/150oC	Crack initiation and propagation on package and die, delamination, wire breakage (can be associated with delamination) – advised for thick packages (longer dwell time and cycles)
Preconditioning:	Bake(optional) à Moisture Soak à 3X Reflow	Delamination, Package resistance to solder heat, may induce cracks both on die and package (due to very high temperature gradient)
Thermal Shock (Liquid TCT)	-55oC/125C	Crack initiation and propagation on package and die, delamination, wire breakage (can be associated with delamination) – advised for thin packages (shorter dwell time and cycles)
Temperature Humidity Bias Test – Alternative for HAST (1000hrs equivalent)	85oC/85%RH	Corrosion, Excessive Intermetallics, Leakage due to top die ionic contamination (dissolved ionic substance from wafer Fab or die saw/attach)
EFR/LFR/HTOL/LTOL	125oC or 150oC	Fab Defects, EOS failures
Electromigration, HCI (Hot Carrier Injection)	N/A	Shorting due to metal bridging, metal via voids, Ion migration, charge densification (may heal if dry baked)
ESD-CDM/HBM and Latch-up Tests	N/A	Valid ESD failures
Solderability Testing (Dip and Look, CPT, and Wetting balance)	215oC, 235oC, or 260oC	Non-wet, Dewetting

At the end of the learning session, the participants were able to voice out their questions and inquiries relating to Practical Reliability. The session ended smoothly and was a huge success.

--------------------------

(The author is currently the Membership Committee coordinator of CFAR under ASEMEP.)