As automated test equipment becomes part of the overall electronics assembly process, DFT must include not only traditional hardware usage issues, but also knowledge of the diagnostic capabilities of the test equipment.
Design for test (DFT, design for test) is not the job of one person, but a group of representatives from Design Engineering, Test Engineering, Manufacturing and Purchasing. Design engineering must specify functional products and their tolerance requirements. Test engineering must provide a strategy to achieve the highest first pass rate (FPY) with the lowest cost and minimal rework. Manufacturing and quality departments must provide input on production costs, what has been done in similar products in the past and what has not been done, and the help of DFV (design in batches) to increase production. The purchasing department must provide information on the availability of components, especially reliability information. Test and procurement departments must jointly purchase components on the in-vehicle test hardware to ensure they are available and easy to implement. Test systems are often used as sensors to collect historical data for process improvement, which should be the goal of the quality team. Therefore, these functions should be done before placing/removing any node selections.
Preparation and understanding are key before developing a strategy for a test environment. Parameters that affect the testing strategy include:
Accessibility. Full contact and large test pads are always goals in the manufacture of printed circuit boards. There are generally four reasons why full access cannot be provided:
1. The size of the circuit board. The design is smaller; the problem is the "extra" footprint of the test board. Unfortunately, most design engineers consider accessibility to test soldering on a printed circuit board (PCB) less important. The situation is quite different when the product has to be debugged by a design engineer due to the inability to use the simple diagnostics of an in-circuit tester (ICT). If full access is not available, testing options will be limited.
2. Function. Performance loss in high-speed designs affects board performance, but gradually reduces the impact on product testability.
3. Board size/number of nodes. This is when the physical board size cannot be tested on any existing equipment. Fortunately, this problem can be solved by increasing the budget for new test equipment or using external test equipment. When the number of nodes is larger than the existing ICT, the problem is more difficult to solve. The DFT team must understand the testing methods that enable the manufacturing department to produce a good product with the least amount of time and money. Embedded Self Test, Boundary Scan (BS) and Function Block Test can do this. Diagnostics must support the unit under test (UUT); this can only be achieved through a solid understanding of the test method used, the test equipment and capabilities available, and the failure spectrum of the manufacturing environment.
4. Not using, complying with or understanding DFT rules. Historically, DFT rules have been implemented by engineers or groups of engineers who understand the manufacturing environment, process and functional testing requirements, and component technology. In the real world, this process is lengthy and requires communication between design, computer-aided design (CAD), and testing. This ubiquitous repetitive work is prone to human error and is often rushed under market pressure. Today, the industry has begun to use automatic "productivity analyzers" to evaluate CAD files using DFT rules. When using a contract manufacturer (CM, contract manufacturer), multiple sets of rules can be categorized. The advantages of this method are regular continuity and error-free product evaluation.
DFT teams should be aware of existing testing strategies. As OEMs come to rely on more and more CMs, the equipment used varies from factory to factory. Too many or too few tests may be used if the manufacturer's process is unclear. Existing test methods include: manual or automated visual testing, which uses vision and comparison to confirm component placement on the PCB. There are several ways to implement this technique:
1. Artificial vision is the most widely used in-line testing method, but due to increased production capacity and shrinking circuit boards and components, this method has become infeasible. Its main advantages are low upfront cost, no test fixtures, and the main disadvantages are high long-term cost, discontinuous defect detection, difficult data acquisition, and no electrical testing and visual limitations.
2. Automatic Optical Inspection (AOI) is a newer method of determining manufacturing defects and is usually used before and after reflow soldering. It's a non-electric, no-fixture in-line technology that uses "learn and compare" programming to minimize acceleration times. Autovision is better for polarity, component presence, and absence, as long as the latter's components are similar to the original "learned" components. Its main advantages are easy-to-follow diagnostics, fast and easy program development, and no fixtures. The main disadvantage is that the short-circuit identification ability is poor, the failure rate is high, and no electrical test is carried out.
3. Automatic X-ray Inspection (AXI) is currently the only method for inspecting Ball Grid Array (BGA) and occluded solder ball quality. It is a non-electrical, non-contact technology that can detect defects in the early process and reduce work in process (WIP). Advances in this area include pass/fail data and component-level diagnostics. There are now two main AXI methods: two-dimensional (2d), looking at the full board, and three-dimensional (3d), taking multiple images at different angles. Its main advantages are unique BGA welding quality and embedded part inspection tools, no fixture cost. The main disadvantages are slow speed, high failure rate, difficulty in rework solder joint detection, high cost of single board, and long program development time.
4. The manufacturing defect analyzer (MDA) is a good tool for high volume/low mix environments where testing is only used to diagnose manufacturing defects. Repeatability between testers is an issue when residual reduction techniques are not used. Also, MDA does not have digital drivers, so functional testing of components or firmware on the programming board is not possible. The test time is shorter than the visual time, and the MDA can catch up with the takt speed of the production line. This method uses a bed of needles, so diagnostic output is possible.
Its main advantages are lower upfront costs, lower work-in-process costs, lower programming and program maintenance costs, higher outputs, easy follow-up diagnostics, and fast full-short and open-circuit testing. The main drawbacks are the inability to confirm that the bill of materials (BOM) conforms to the unit under test (UUT), no digital confirmation, no functional test capability, no firmware calls, and usually no test coverage indication. , board and board line-to-line repeatability, fixture cost and usage issues.
ICT will identify manufacturing defects and test analog, digital and mixed-signal components to ensure they meet specifications. Many devices have the ability to program onboard memory, including serial numbers, pass/fail, and lineage data. Some devices make the procedure easier. Multi-version testing and firmware conversion are easily accomplished by embedding the tool in an easy-to-use Graphical User Interface (GUI) and storing the code in a special file. of. Some devices have sophisticated instruments that confirm the functionality of the UUT and interface with commercially available instruments. Today's test equipment has an embedded computer-aided design (CAD) interface and a non-multiplexing environment to reduce development time. Finally, some testers provide in-depth UUT coverage analysis to clarify which components are or are not being tested.
The main advantages of ICT are low cost of single board testing, strong digital and functional testing capabilities, high output, strong diagnostic capabilities, fast and thorough short and open testing, firmware programming, high defect coverage, and easy programming. The main disadvantages are fixtures, programming and debugging time, fixture costs, anticipated expenses, and usage issues.
5. In recent years, due to the improvement of mechanical precision, speed and reliability, flying probe testers have been widely used. In addition, market requirements for fixtureless test systems required for rapid changeover, prototyping, and low-volume manufacturing make flying probe testing an ideal test option. The best probe solution provides learning capabilities and BOM testing, which automatically increases monitoring during testing. The probe software should provide an easy way to load CAD data, as X-Y and BOM data must be used during programming. Since node accessibility may be incomplete on the veneer side, the test generation software should automatically generate non-repetitive split programs.
The probe uses vectorless techniques to test the connections of digital, analog and mixed-signal components; this should be done through capacitive plates, which the user can use on both sides of the UUT.
The main advantages of the flying probe tester are that it is the fastest time-to-market tool, automatic test generation, no fixture cost, good diagnostics and easy programming. The main drawbacks are low yields, limited digital coverage, fixed asset costs and usage issues.
6. Functional testing, arguably the earliest automated testing principle, has revived its importance. It is a specific board or specific unit and can be done with a variety of devices. To give a few examples:
Final product testing is the most common method of functional testing. Testing the final unit after assembly is costly and reduces operational errors. However, diagnosis is non-existent or difficult, which increases the cost. Only test the final product, if automated testing does not provide software or hardware protection, there is a chance of damaging the product. Final product testing is also slow and often takes up a lot of space. This method is generally not used when criteria must be met, as it usually does not support parametric measurements.
The main advantages of final product testing are the lowest initial cost, one-time assembly, product and quality assurance. Its main disadvantages are low diagnostic resolution, slow speed, high long-term cost, damage to the FPY, circuit board or machine without detecting a short circuit, high maintenance cost, and no parametric detection capability.
The latest thermal models are usually placed in various stages of assembly, not just in final testing. It is superior to final product testing in terms of diagnostics, but is more expensive due to the need to set up specialized test cells. If program debugging only tests a specific board, the mockup may be faster than final product testing. Unfortunately, due to the lack of protection, the test bench can be damaged if a short circuit is not diagnosed during the previous process.
Its main advantage is the low initial cost. The main disadvantages are low space efficiency, high maintenance cost of test equipment, short-circuit damage to the test unit, and no parametric test capability.
Software-controlled commercial instruments are often referred to as "rack and stack" testing because the instruments are purchased separately and then connected. The software for synchronizing devices is usually fully customizable. Commercial instruments are inexpensive compared to integrated solutions and, if done correctly, allow stand-alone UUTs to be effective. But such "homebrew" systems are often slow, and engineering changes and production site support are difficult because these applications are poorly documented.
Its main advantages are protection against UUT damage, faster output, smaller footprint requirements and independent/industry acceptable calibration. Its main disadvantages are time-consuming, difficult to support, and update and use in remote facilities.
Commercial, custom integrated systems couple software and hardware on test platforms such as IEEE, VXI, Compact-PCI, or PXI. Documentation, software support and standard manufacturing concepts make these systems easy to use and support. Upfront costs are higher than in-house build plans, but are adjustable due to higher performance, output, and repeatability. It is also easy to support in the field and during new product development.
Its main advantages are fast output, small footprint required, easiest support and reset, optimal repeatability, and the ability to provide independent industrial acceptance. The main disadvantage is the high initial cost.
Non-contact testing methods such as lasers are the latest developments in PCB testing technology. This technology has been proven in the bare board space and is being considered for testing on filled boards. This technology uses line-of-sight only, unshielded access to detect defects. At least 10ms per test, fast enough for a large production line.
Its main advantages are fast production, no need for fixed equipment, line-of-sight/non-coverage channels; the main disadvantages are low test efficiency, high initial cost, and many maintenance and usage issues.
English
Chinese