Electronics components, printed circuit boards and complete system assemblies are often produced through contract manufacture. Buyers may opt for what appears to be the most cost effective EMS, however a well-priced offer can quickly turn out to be a cost trap or have fatal consequences in terms of product quality. On closer examination, a higher priced EMS solution can be more economical, particularly if it includes a quality management system and test strategy to minimise consequential costs or liability risks.
When considering an EMS partner’s offer, take time to examine additional costs such as waste production, repairs, upgrades and warranty claims, which can be more difficult to estimate in advance. The ‘rule of ten’ suggests that the cost of repair for a faulty module will increase by a factor of ten for each additional process step it undergoes. It is therefore crucial to detect errors early and use the data collected to improve the process.
A combination of different test methods is almost always worthwhile. Testing uncovers existing errors and potential design flaws early on. Through specific tests, it should be possible to correct errors in a timely manner, to optimise vulnerable components early and to adapt production processes in advance, therefore reducing the reject rate. At the same time, this will reduce the cost of troubleshooting and repair, as well as recalls and warranty claims.
Future testability should be an important focus area during product design. ‘Design for manufacturability’ and ‘design for test’ methodology can improve diagnostic capabilities and reduce the number of potential errors. At this early stage it is possible to influence the design and adapt test requirements, such as setting additional test points.
It also helps optimise the manufacturing process in advance and by skilfully selecting components, you can reduce or simplify any future repairs and minimise later susceptibility to errors.
After completing the design and development phase, products should ideally be tested within their individual life cycles. The test strategy of a prototype would, for example, be quite different from that used in volume production. The nature of the products will also determine the traceability requirements and the level of documentation required for test results. To manufacture a product efficiently, the individual tests must be perfectly matched. This places high demands on the competence of the test engineer as well as on the available test methods.
In the development and implementation of integrated test strategies, electronic engineering and manufacturing services (E2MS) providers often have good experience, because these companies do not just test during production. Rather, they get involved in the planning and development phase, which would include the design of a product-specific test strategy.
An important parameter is first pass yield (FPY), which is often calculated via a functional test. The FPY indicates how many modules are error-free after running through all steps in the process.
For example, with a production quality of 50 defects per million opportunities on a module with 10,000 solder joints and without previous testing, an FPY of 60 per cent would be expected. Through the limited diagnostic possibilities of a functional test alone, repair will be time intensive and require a deep understanding of the product’s functionality. A FPY of 98 per cent should be sought to avoid affecting the production process unnecessarily.
In practice, significantly lower FPY values are common, possibly because manufacturing service providers are called in after the completion of the design phase. In these cases certain tests cannot be realised at all due to an unfavourable design. Against this backdrop, holistic thinking E2MS partners are increasingly providing ‘design for excellence’ (DfX), which allows for smooth production and encompasses ‘design for test’ (DfT).
With a good test strategy, several test methods will be systematically combined and tailored to the individual assembly’s requirements. Individual testing decisions will be based on the technical characteristics of the product, as well as lot size, economic factors or other characteristics within the production process. Even an individual company’s circumstances and product liability issues must be considered.
A good test strategy will combine the strengths of different test methods in order to achieve high fault coverage with minimal overlap. The use of automatic optical inspection (AOI), automated x-ray inspection (AXI), in-circuit test (ICT), functional test systems and flying probe tests is widespread. Which tests are used in which production step, however depends on many factors. An experienced test engineer will develop individual testing strategies and with the help of test data, can optimise the manufacturing process in future.
Choosing a match
In order to achieve this goal, the test engineer must understand the test methodology thoroughly and understand the product to be manufactured, as a whole, and in detail. They also need to have visibility of the whole production process, to assess risks and calculate costs.
Companies awarding contracts to EMS or E2MS companies should always assess their test expertise, which will be explained differently by every company. Generally a service provider should have performed all standard tests and they should be able to recommend the most sensible test strategy for their customers. In the mid/low volume, high-mix area, there is no one size fits all solution and decisions of specific test methods should always be made on a case by case basis.