May_EDFA_Digital

55 / 58

edfas.org
55
ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 19 NO. 2 GUEST COLUMNIST
BASIC KNOWLEDGE REQUIRED OF AN FA ENGINEER
David Burgess, Accelerated Analysis
INTRODUCTION
Failure analysis (FA) may be the best choice an engi-
neer canmake. A career in FA provides continuously chal-
lenging projects. Failure analysis also requires detailed
knowledge of topics not sufficiently covered by engineer-
ing or physics programs. For me, and probably for other
analysts, these observations simply explain why failure
analysts would not trade jobs with anyone. We can’t
imagine trading the growth inherent in solving different
problems with repetitive jobs or testing or designing
similar items.
The ISTFA symposium is a good source of up-to-date
FA information that is hard to find anywhere else. The
2016 ISTFA event included a Panel Discussion address-
ing the required attributes of the “Next Generation of FA
Engineer.” [1]
The Panel Discussion noted that an FA engineer must
have the capability to handle several techniques to local-
ize defects in increasingly complex devices. He/she must
be knowledgeable about deprocessing techniques. He/
shemust understand design, layout, fabrication, test, and
application. In addition, communication skills are neces-
sary for interactingwith the personnel associatedwith all
the aforementioned areas. Frank Altmann’s summary of
the Panel Discussion indicated that it provided thoughtful
discussion about “the high level of technical knowledge as
well as high flexibility and communication skills” required
of failure analysts.
The consensus of the Panel Discussion attendees was
that a potential analyst requires a daunting list of prereq-
uisites. I must agree, but few, if any, of the best analysts
I know enjoyed all of the recommended background.
Several lack most of the academic requirements. Clearly,
future analysts will have a head start if their academic
background includes FA tools and awider range of related
topics.
Perhaps it is too obvious to mention, but the Panel
Discussiondid not identify problemsolving as a necessary
part of FA training. Problem solving is a key asset and a
strong point for the best analysts I know. There are a few
other characteristics that all the best analysts have in
common:
1. All have a solid academic foundation in at least a
few of the areas listed as being necessary for a future
analyst.
2. All have problem-solving basics at the core of their
analysis approach. Failureanalysis is problemsolving.
Therefore, basic problem questions—“what, when,
where, and how much”—guide every FA.[2]
3. All are motivated to fully understand the theory,
mechanics, and limitations of the FA tools they use.
They want to understand how the output of the tool
applies to the specific sample at hand.
4. This is an extension of No. 3. When limitations of
existing FA tools block the progress on an FA problem,
all look for ways to extend the capability of the
tool. (Consider, for example, backside probing and
nanoprobing.)
5. This also is anextensionof No. 3.Whenpresentedwith
a possible “cause” or “defect,” all delay celebration
until the possible cause is verified in some way. At
the least, the possible causemust adequately answer
the applicable “what, when, where, and how much”
questions.
MENTORING
The learning curve of a young failure analyst can
be enriched by working with an experienced analyst.
Typically, a young engineer brings the latest technology
to the team. The experiencedmentor provides some hard-
to-get perspective.
For example, every new analyst receives the advice
“Don’t assume.” However, the undisputed advice comes
without explanation. Actually, analysts assume every day.
Whenever available evidence has two or more equally
probable explanations, one is assumed. Additional data