构建成功的主板测试策略：第二版精华

需积分: 10 139 浏览量更新于2024-07-19 收藏 43.82MB PDF 举报

"《构建成功的板级测试策略第二版》是一本专门探讨在电子产品制造过程中如何制定和执行有效测试策略的实用指南。该书强调了测试策略与产品生命周期管理的整体关联，涵盖了从产品开发、制造到测试阶段，以及售后服务中的问题，甚至对公司的质量形象有着深远影响。作者史蒂芬·F·谢克伯格深入剖析了选择测试策略的关键因素，并提供了实用工具，帮助读者根据自身公司的理念和特性定制最适合的测试策略。书中详述了如何根据不同工程、市场和管理需求来评估和优化测试策略的有效性。作者特别关注了与VME总线（Versatile Microprocessor Environment）、VXI总线（VME Extensions for Instrumentation）等标准相关的测试方法和边界扫描架构的应用，这些技术在当时的工业界中具有重要意义。版权方面，部分章节的内容基于IEEE的多个标准，但必须注意的是，未经出版商事先书面许可，不得复制、存储或以任何形式传播这些材料，以免侵犯版权。《构建成功的板级测试策略》不仅为读者提供了理论框架，还提供了实操层面的指导，旨在帮助读者在面临各种公司规模和复杂度时，能够构建出最适合他们特定环境的高效测试策略，从而确保产品的质量和商业成功。通过阅读这本书，读者将能提升其在产品质量控制、成本优化和时间管理方面的决策能力，对于任何涉及电子设备制造和测试流程的企业来说，这是一份不可或缺的参考资源。"

What

Is a

Test

Strategy?

1.1

Why Are You

Here?

What

drives

you to the

rather daunting task

reading

textbook

board-

test strategy? Although reasons

can

vary

much

as the

manufacturing techniques

themselves,

they usually break down into some version

of the

following:

•

The

manufacturing process

getting away

from

you,

Test represents your primary bottleneck,

and

•

Test

has

become

part

of the

problem rather than part

of the

solution.

The

design-and-test process must treat

"test"

as an

ongoing activity.

Its

goal

furnish

clean product

manufacturing

designing

for

manufacturability

and

testability, while encouraging

the

highest possible product quality

and

relia-

bility.

(Product

quality

means that

functions when

leaves

the

factory. Reliabil-

ity

refers

to its

resistance

failure

in the field.)

The

purpose

manufacturing

is to

provide:

•

The

most products

•

At the

lowest possible cost

•

the

shortest time

•

the

highest possible quality

Debate

has

raged

for

years over

the

relative importance

these goals. Cer-

tainly,

test people

often

maintain that quality should

paramount,

while man-

agement prefers

look

first at

costs. Nevertheless,

company that cannot provide

enough products

satisfy

its

customers will

not

stay

business

for

very long.

Suppose,

for

example, that

you

have contracted

provide 100,000 personal-

computer (PC) motherboards over some period

time,

and in

that time

you can

deliver

50,000 perfect motherboards.

Despite

superior product quality,

if you

cannot

meet

the

contract's volume requirements,

the

customer

will

fly

into

the

arms

one or

your competitors

who

can.

Using

similar reasoning,

the

purpose

"test"

is to

maximize product

throughput, reduce warranty

failures,

and

enhance your company's reputation—

thereby

generating additional business

and

keeping

jobs

secure.

We get

there

designing

the

best, most

efficient

test strategy

for

each

specific

situation.

1.2 It

Isn't

Just

Testing

Anymore

Therein lies part

of the

problem. What

"test"?

Unless

broaden

the

concept

include more quality-assurance activities,

verifying

product

quality

through

"test"

will

soon

approach

impossible.

Inspection,

for

example,

usually considered part

manufacturing,

rather

than test. Simple human nature suggests that this perception tends

make test engi-

neers

less

likely

include

it in a

comprehensive strategy.

Yet

inspection

can

identify

faults—such

missing components without bed-of-nails access

insufficient

solder

that makes proper contact only intermittently—that conventional test

will

miss.

BUILDING

SUCCESSFUL BOARD-TEST

STRATEGY

Similarly,

design

for

testability reduces

the

incidence

some faults

and

permits

finding

others more easily. Feeding failure information back into

the

process allows adjustments

improve

future

yields.

These steps also belong

part

the

larger concept

"test."

Embracing

those steps

addition

to the

conventional

definition

"test"

allows

test people

determine more easily

the

best point

in the

process

iden-

tify

particular

fault

class. Pushing detection

certain faults further

upstream

reduces

the

cost

of finding and

repairing them.

addition,

not

looking

for

those same

faults

again downstream

simplifies

fixture and

test-program gener-

ation, shortens manufacturing cycles,

and

reduces costs.

Test

strategies have traditionally attempted

to find

every

fault

possible

each

step. Adopting that approach ensures that several steps

will

try to

identify

least

some

of the

same faults.

cost-effective

alternative would push detection

all

faults

as far up in the

process

possible, then avoid looking

for any

fault

covered

in an

earlier step later

on.

Self-test,

too, forms

part

this

strategic

approach.

Many

products

include

self-test,

usually some kind

power-on test

assure

the

user that

the

system

functioning

normally. Such tests

often

detect more than

third

possible

fault

mechanisms, sometimes much more. Which suggests

the

following

"rules"

for

test-

strategy

development:

•

Inspect everything

you

can, test only what

you

must.

•

Avoid looking

for any

problem more than once.

•

Gather

and

analyze

data

from

the

product

give

you

useful

information

that allows

you to

improve

the

process.

1.3

Strategies

and

Tactics

Test

strategies

differ

significantly

from

test tactics. In-circuit test,

for

example,

tactic. Removing manufacturing defects represents

the

corresponding strategy.

Other

tactics

for

that

strategy include manual

and

automated

inspection,

manufacturing-defects

analysis

subset

in-circuit test—see Chapter

2), and

process improvement.

strategy outlines

the

types

quality problems

you

will

likely experience,

then

describes which

those problems

you

choose

to fight

through

the

design

process during design verification, which

you

assign

test,

and

which

you

leave

for

"Let's

wait until

the

product

is in the field and the

customer

finds

it,"

The

difference

between strategies

and

tactics boils down

issues

term

and

focus.

test strategy lasts

from

product's

conception until

the

last unit

in the

field

dies. During that time,

the

manufacturer

may

resort

many tactics. Also,

tactic addresses

particular place

and

time

in the

overall

product

life

cycle.

strategy

generally

focuses

on the

whole picture.

building

test strategy,

we are

always looking

for

"digital"

answers

"analog"

problems. That

is, we

must decide whether

the

product

good

bad.

But

how

good

good?

How bad

does "bad"

have

to be

before

the

circuit

will

not

function?

What

Is a

Jest

Strategy?

Suppose,

for

example, that

the

manufacturer

of a

digital device

specifies

"0" as

less than

0.8

volts. Based

that

specification,

parametric

measurement

logic

low at

0.81 volts would

fail.

Yet

will

the

system actually

not

perceive

voltage

0.81

as a

clean "0"?

How

about 0.815 volts?

The

answer,

course,

is a

firm "It

depends."

The

situation resembles

the

century-old conundrum:

How

many

raindrops does

take

before

baseball

field is wet

enough

delay

the

game?

that

context,

the

question seems absurd.

You

can't count raindrops!

Yet at

some

point, someone must make

value judgment. Most baseball people accept

the

fact

that delaying

the

start

of a

game usually requires less rain than does stopping play

once

the

game

has

begun. Similarly,

the

question

of how

closely

circuit must

conform

published specifications

may

depend

surrounding circumstances.

The

real

question

remains:

Does

the

product

work?

product

complexity con-

tinues

skyrocket,

the

necessity

the

compromises implicit

this

approach

become glaringly apparent.

Compounding

the

challenge, issues

power consumption, heat dissipation,

and

portable-product battery

life

have required drastically reducing operating

volt-

ages

for

most digital systems.

The 5V

transistor-transistor logic (TTL) parts

of the

past have yielded

devices operating

less then

3V,

with more

come.

As a

result,

the gap

between

logic

" 1" and a

logic

"0"

narrows every day. Devices must

perform

more precisely; boards

and

systems cannot tolerate electromagnetic

inter-

ference

(EMI)

and

other noise that were commonplace

only

a few

years ago.

New

generations

test equipment must

cope

with these developments,

and

test strate-

gies

must take them into account.

1.3.1

The

First

Step

Consider

(loaded) question: What

is the

single most important considera-

tion

developing

test strategy?

The

answer

may

seem obvious.

Yet in

board-test-

strategy seminars

from

New

York City

Singapore, responses range

from

budgets

design-for-testability

to "Do we

need

test?"

to "Do we

choose in-circuit

functional

test?"

Before

facing

any of

these issues, however, designing

successful

test

strategy

requires

determining

the

nature

of the

product.

That

is,

what

are you

trying

test? What

is the

product? What does

look

like?

How

does

work?

What design technologies does

contain?

Who is

designing

it? Who is

manufac-

turing

it? Who is

testing

it? In

many organizations,

one

obstacle

arriving

at an

effective

test strategy

that

the

people involved decide

test-strategy components

and

tactics

before

answering these simple questions.

Test

engineers

do not

design products.

nobody

tells

them what

the

product

is,

how it is

designed,

and

what

it is

supposed

to do,

their decisions

may

make

sense. They might arrive

at a

correct strategy,

but

only

accident,

and it

would

rarely

represent both

the

most successful

and the

most economical approach.

Test

components

test

strategies

that

work

for one

company

product

line

may

not be

appropriate

another situation.

If you do not

know what

you are

trying

test,

you

cannot systematically determine

the

best strategy. Even

if you

find a

strategy that works, thoroughly knowing

the

product

will

likely

help

you

BUILDING

SUCCESSFUL BOARD-TEST

STRATEGY

suggest

better one. Which brings

us to the

following

definition:

successful test

strategy

represents

the

optimum blend

test methods

and

manufacturing

processes

produce

the

best-quality boards

and

systems

sufficient

number

the

lowest possible cost.

Selling-price

erosion among electronic products

and

electronic components

larger products exerts ever-increasing pressure

manufacturing

and

test

operations

keep costs down.

they have

for

more than

two

decades, personal

computers (PCs) provide

excellent case

point.

The

price

of a

particular

level

of PC

technology declines

more than two-thirds every

four

years.

Looking

at it

another way, today's

PCs are

about

times

powerful

machines

only

five

years ago,

about

the

same price. Bill Machrone,

one of the

industry's

leading analysts, describes this trend

what

calls "Machrone's Law":

The

computer

you

want

will

always cost $5000.

One

could argue

the

magnitude

the

number, which depends partly

on the

choice

printers

and

other peripherals,

but

the

idea that

stable computer-equipment budget

will

yield increasingly capable

machines

indisputable. Machrone's reputation

as an

industry prognosticator

remains intact—especially when

you

consider that

coined

the law in

1981!

Reasonably equipped

PCs

priced

under $1000 have become increasingly

common. Peripherals have reached commodity-pricing status. Even

the

micro-

processors themselves

are

experiencing price pressure.

PCs

based

on new

micro-

processors

and

other technologies rarely command

premium price

for

more than

few

months before competition

forces

prices into line. Meanwhile, product-gener-

ation

half-lives

have

fallen

less than

year, less than

months

for

some critical

subsystems

such

hard disk drives

and

CD-ROM readers

and

burners.

Even

flat-

panel liquid-crystal displays (LCDs), once exorbitantly expensive,

are

beginning

replace

conventional monitors—the last remaining tubes

common use.

Customers

are

forcing companies

to cut

manufacturing costs,

while

test costs

remain stable

best

and

rise dramatically

worst. Test costs today

often

occupy

one-third

one-half

of the

total

manufacturing cost. Every dollar saved

testing

(assuming

equivalent quality level) translates directly

to a

company's bottom

line.

For

example,

manufacturing costs represent

percent

of a

product's selling

price

reasonable number)

and

test costs represent one-third

that

percent,

then

strategy that reduces test costs

by 25

percent reduces overall manufacturing

costs

36.67 percent,

difference

3.33 percent.

If the

company

was

making

percent

profit,

its

profit increases

13.33 percent,

difference

one-third.

wonder managers want

reduce test costs

much

possible!

1.3.2 Life Cycles

successful

test strategy

is a

by-product

overall

life-cycle

management.

requires

considering:

Product development

Manufacturing

Test

What

Is a

Test

Strategy?

Service

Field returns

•

The

company's "image

quality"

Note

that only test,

field

returns,

and

service involve testing

all,

and field

returns

do so

only indirectly. Reducing

the

number

failures that

get to the

test

process

or the

number

products

that

fail

after shipment

customers

also

simplifies

test activities, thereby minimizing costs.

Test-strategy

selection goes

far

beyond merely choosing test techniques.

Design issues,

for

example, include bare-board construction.

engineer once

described

50-layer board

that

was

designed

such

a way

that

could

not be

easily

repaired.

avoid

the

very expensive scrapping

of bad

boards,

his

colleagues bor-

rowed

technique

from

designers

random access memory (RAM) components

and

large liquid-crystal-display (LCD) panels—they included redundant traces

for

most

of the

board's

internal logic paths. Paths were chosen

soft

switches driven

on-board components individually programmed

for

each

board.

Although

this solution

was

expensive,

the

board's

$100,000 price

tag

made

such

expensive choice viable, especially because

it was the

only approach that

would

work.

Without

the

redundancy,

board

yields would have been unacceptably

low,

and

repair

was

impossible. Unfortunately,

the

solution created another

problem.

The

board's

components contained specific instructions

select known-

good paths.

The

bare board

defied

testing without component-level logic. There-

fore,

the

engineers created

test

fixture

that meshed with

the

sockets

on the

board

and

mimicked

its

components.

addition

pass

fail

information,

the

test

would

identify

successful path, then generate

the

program with which

burn

the

"traffic-cop"

devices

part

of its

output.

Including

the

redundancy

as a

design

choice mandated

particular extremely complicated test strategy. Sometimes test-

strategy

choices reduce

"poor"

and

"none."

The

acceptability

particular test steps depends

whether

the

strategy

for

a new or

existing

facility,

product,

product

line,

technology.

In an

existing

facility,

there adequate

floor

space

for

expansion?

Is the

facility

already running

three

work

shifts,

or can a

change

strategy involve merely adding

shift?

Test

managers must also decide whether

design their

own

test equipment

or buy it

from

commercial vendors, whether they should

try to

"make

do"

with

existing

equipment,

and

whether

new

equipment must

be the

same type

from

the

same manufacturer

as the

installed base.

test strategy's success

also

depends

aspects

of the

overall manufactur-

ing

operation.

For

instance,

how

does

product

move

from

test station

repair

station

from

one

test station

to the

next?

Are

there conveyors

other auto-

mated handlers,

or do

people

transfer material manually? Concurrent-engineering

principles

encourage placing portions

of the

manufacturing process physically

to one

another, thereby minimizing bottlenecks

and

in-transit

product

damage. This arrangement also encourages employees

who

perform

different

parts

the job to

communicate with

one

another, which tends

increase manufactur-

ing

efficiencies

and

lower costs.

剩余349页未读，继续阅读

aphrodite_aglai

粉丝: 4
资源: 35

构建成功的主板测试策略：第二版精华

英文原版-Data Mining 2nd Edition

ipv6-second-edition-theory-protocol-and-practice-2nd-edition

System identification -a frequency domain approach-2nd edition

Building-Machine-Learning-Systems-With-Python-Second-Edition:Packt Publishing使用Python构建机器学习系统

Learning.Node.Moving.to.the.Server-Side.2nd.Edition.14919

Building-Microservices-with-.NET-Core-2.0-Second-Edition:用.NET Core 2.0构建微服务-第二版，由Packt发行

Building-Machine-Learning-Systems-with-Python-Third-edition:Packt用Python第三版构建机器学习系统的代码存储库

Learning-Java-by-Building-Android-Games-Second-Edition:Packt发行，《通过构建Android游戏学习Java，第二版》

Learn-OpenCV-4-By-Building-Projects-Second-Edition:由Packt发行，《通过构建项目学习OpenCV 4，第二版》

Addison Wesley - Building Web Applications with UML 2nd Edition

最新资源