Previous Page  40 / 52 Next Page
Information
Show Menu
Previous Page 40 / 52 Next Page
Page Background

38

Chip Scale Review November • December • 2017

[ChipScaleReview.com]

Testing of automotive electronic ICs

By Davide Appello

[STMicroelectronics]

n the five years since my previous

editorial contribution [1], the most

consolidated applications segments

existing at that time evolved without major

discontinuities with respect to the offered

performances. Key trends identified, such

as functional safety and security, took place,

thereby definitively setting new standards

and methods. In contrast, new applications

driven by the booming advanced driver

assistance systems (ADAS) segment require

disruptive changes in some of the traditional

automotive paradigms. This paper provides

more details about the key factors driving

the various application segments and also

analyzes the consequences of the front-

end and back-end technologies necessary

to accomplish the relevant objectives, as

well as on the tools and methods needed to

effectively test ICs.

Key factors

With respect to automotive electronics,

and more specifically of ICs developed for

automotive applications, we have often

reported as key factors and keyword attributes

such as quality, extended temperature range

and lifecycle duration, burn-in, robust

design, and design for reliability.

Figure 1

qualitatively shows the incidence of the cost

of electronics with respect to the overall value

of a car throughout the years. To project the

number of components and overall number

of applications in the car, the reader should

also consider the significant reduction of

application service providers (ASPs) per IC

year over year as a natural consequence of

increases in volume.

The increasing quality requirements,

which translated into longer test times and

multi-temperature coverage, determined the

need for countermeasures to reduce costs as

well as the cost of test (CoT). Because of this

need, automotive testing adopted several test

cost reduction methods from other market

segments; these methods were used to push

development of more advanced technology

despite being at a lower level of maturity.

Very high parallelism probing is one example,

accompanied by high multi-site testing at the

package level. This forced the development

of customized ATE cells offering relevant

instrumentation and configuration options.

Additionally, a great deal of effort has

been spent to improve design for testability

(DFT) by pushing on few key techniques

such as memory built-in self-test (BIST),

limited pin count test ports, compression,

and more recently, logic BIST. DFT

methods have been challenged not just to

achieve very high coverage indexes, but

also to offer diagnostic capabilities to enable

high accuracy and resolution

diagnostics for reliability

learning.

Progress doesn’t end

here and indeed, as has

happened several times in

this field, we have signals

showing radical changes in

past paradigms. To be more

specific, the requirements are

not changing, instead, they

are advancing relentlessly

and without major gaps.

Table 1

gives the time

horizon for these changes

in requirements. The most

relevant and potentially

d i s r u p t i v e c h a n g e s

concern the conditions in

which requirements can

be satisfied. Many of the

methods used with a good

degree of success in the

past, appear to be no longer

sufficient to meet the goal.

Let's review how and why.

Applications

The following quote from

[2] succinctly summarizes

what has happened in the last

few years: “The auto industry

is lost in translation between

evolutionary, revolutionary

and disruptive key trends

that all need to be managed

at the same time.” Some of

the applications followed

an evolutionary path (see

Figure 2

). Powertrain,

braking, and transmission

a p p l i c a t i o n s e v o l v e d

seamlessly—progressively integrating

new features mainly driven by safety and

security requirements. These features are not

major performance-eaters and are instead

leveraging robust development methods,

verification, and traceability. Quality and

reliability performances demand, increased

despite starting from already severe

conditions.

Body and dashboard applications,

I

Figure 1:

Electro

nic content in the car.

Table 1:

IC characteristics trend: how these features will impact

test requirements.

Figure 2:

ADAS applications.