TestWay™ - Test Point Savers

Introduction

For PC Board using through-hole technology, it is used to get a full physical accessibility with one test access per net. Testability was maximal: commandability, observability, test efficiency and troubleshooting tasks were optimal.

With the drastically increase in term of density report most of the time this interesting configuration is no longer true and leads to the following conclusion: It is more and more difficult to reach all of the nets for any test access.

For most of the test strategies, this lack of access directly leads to fault coverage lost and of course to test efficiency lost.
To face this problem, two complementary approaches have been developed:

  • To improve test interface design in order to get fine pitch bed of nails (50 mils)
  • Reduce the number of required test points to get a good test coverage (measure of test efficiency)

Test Points

In most of the design cycle strategies, test accesses are chosen after the layout phase. It is performed according to classic constraints such as connector pads, via diameter, isolation… For a given net, when no test access can be defined according to those constraints, this net is classify as inaccessible.

For today’s PC Boards, this way of choosing test accesses without any relationship to the fault coverage, only using geometrical criteria is at least strange and not adequate with the testing challenge the electronic industry have to face.

To provide an alternative and better solution to test accesses choice, TestWay provides a full and structured approach from schematic capture to layout:

  • Test efficiency analysis according to test accesses identified at the layout stage or declared at the schematic capture.
  • Proposal to reduce the number of test points necessary to get a good fault coverage.
  • Proposal to increase test efficiency by adding a limited number of test points.
  • Clusters identification: Defining devices groups to be tested together using a reduced number of accesses.

In a traditional design cycle, TestWay allows to analyze what should be the consequences of a lost of physical accessibility for the global test efficiency. By considering existing test accesses, TestWay determines the test coverage (nets and devices that is not tested or with a low efficiency) and proposes additional test points in order to get a significant increase of test efficiency.

In a concurrent engineering design cycle, TestWay identifies soon during the schematic capture just the necessary test points to get after the layout phase the desire accessibility for a good test efficiency.

Reduce the number of accesses

According to the fact that one test access per net is no longer possible to obtain only a virtual access may allow to keep an acceptable fault coverage.

coverage=f(access)

Fig. 1: Coverage = f(Physical or virtual access)

A virtual access is available each time it is possible to affect a logic state by propagation in an up stream chain (commandability) and to get a measurement result by propagation in a down stream chain (observability).

Virtual access

Fig. 2: Virtual access

Up and down stream chains must be designed so that signal propagation is easy. According to this, 1149.1 Boundary-Scan standard test strategy provides a virtual access to each of the device pins.

Interconnect Test Analysis

A net is testable when it is possible to affect a logic state through a first channel and to get a measurement result by a second one.

Several configurations may take place:

  • Output state generation through a device, measurement through a tester channel.
  • State generation through the JTAG chain, measurement through a tester channel.
  • State generation through a tester channel, measurement through the JTAG chain.
  • State generation through the JTAG chain measurement from an other cell in the JTAG chain.

TestWay-ITA (Interconnect Test Analyzer) module performs nets classification in order to affect to each of them a priority level for test access allocation.

ITA groups

Fig. 3: Interconnect Test Analysis

Surrounding nets analyze allows to classify each net in one of the following nine groups:

  1. Ground nets
  2. Power nets
  3. Unused pins
  4. Pins only connected to a pull resistor
  5. Boundary-Scan testability bus
  6. PURE Boundary-Scan nets
  7. PARTIAL Boundary-Scan nets
  8. Tester access
  9. Non-scan nets with no tester access

To adjust the analyze, each group is split in sub groups each with a predefined priority level for test access allocation. TestWay identifies unnecessary test points and provides also suggestion about what should be modify in the design to reduce the number of necessary test accesses.

Device Test Analysis

TestWay-DTA module is intended to verify according to the overall test accesses available that it is possible to apply a logic state on each device inputs and to measure the resulting states at least on one output.

If it is possible to drive all of the inputs the theoretical test efficiency may be given by the following formula: DTA formula

DTA module then proposes to try to increase the test fault coverage by adding a limited number of new test accesses. Adding a test access allows to increase up stream device observability and down stream devices commandability. Greatest priority is given to commandability increase which will have as a secondary effect observability increases too.

A better commandability allows to drive much more devices and a better observability leads to a higher fault coverage with a higher diagnostic accuracy.

Report presents steps for maximizing the number of testable devices using the minimum number of additional test points. The way of working is the following: “If you add N physical accesses, you will test M devices more”. This process is implemented until an optimum test coverage is reached.

Clusters analysis

DTA module identify clusters in partitioning the board in independent functional blocks.
A cluster is often a group of non-JTAG devices whose inputs and outputs are coming from accessible points.

  • Cluster is a device block which will be tested together at the same time with the same vectors using a limited number of test accesses.
  • Internal net accesses to this kind of block is not so much important to obtain a good fault coverage. Fault dictionary use allows to keep a high level of default localization.
  • Device type analysis is an important parameter to define easy testable clusters.
  • Boundary-Scan clusters have this particular characteristic to have some of their access points connected to partial boundary scan nets.

Cluster borders identification leads to interconnection file automatic generation which contains cluster description and its primary accesses. If a cluster is composed of programmable devices, RAM, PROM memories or low complexity devices, test vectors and fault dictionary generation should be done automatically using adequate existing tools.

PC Board fault coverage

Using fault coverage computed by ITA and DTA, TestWay shows the test efficiency of the overall PC Board using colors on electrical diagram. This feature allows the users to quickly identify how adding supplementary test access or removing existing test points may improve the test coverage of its PC Board.

A selection of our customers:

A selection of our customers
Apr 24-26

NEPCON China Shanghai 2019 – stand 1N13

Apr 24 - Apr 26
Shanghai
Chine
Mar 26

ElectroTest EXPO 2019

Mar 26
Newcastle-under-Lyme, Staffordshire
United Kingdom
Jan 26-31

IPC APEX Expo 2019

Jan 26 - Jan 31
San Diego
United States
Nov 28

SMTA Silicon Valley Expo

28 Nov 2018
San Jose, CA, USA
Nov 13-16

Electronica Munich 2018

13 Nov 2018 - 16 Nov 2018
Munich
Nov 7

PCB Carolina 2018

7 Nov 2018
NC 27606, USA
Oct 23-24

Enova Paris 2018

23 Oct 2018 - 24 Oct 2018
Paris
France