In every project involving PCB, there are three main steps. First comes the design process. Then comes assembling. But the work is never fully done without testing. After all, it would be a shame to release a brilliant device with a defect due to an oversight of a crucial detail.
And while some tests are carried out in the simulation software, working in a real-life environment is always going to be different. The simulation itself doesn’t cover the temperature, humidity, or slight voltage changes. These aspects apply only in the real world. The board itself has to go through laboratory tests that can model the real-life conditions.
It is also done to verify that the design is working as planned. After all, no one wants to be left with a thousand units of defective PCBs on their hands. Costs in both money and time can be disastrous. The risk involved in speeding up production without testing two or three prototypes is too great to ignore.
How do you test for PCBs?
Tests can be categorised into three main groups: visual, electrical, and functional. To suggest that testing is a single activity is far from the truth; multiple layers and phases must be completed. These are carried out at different stages of the production cycle as well. For example, a visual check is conducted on every board collectively. Defects can occur even after producing many copies; therefore, it cannot be neglected even if the prototype turned out to be perfect.
The best modern practice to follow is the “fail-fast” philosophy. It focuses on tackling every problem and defect head-on instead of leaving them to be fixed later. Failing fast means not accepting anything with a discrepancy. Small problems can cause twice the trouble if left for later, as they can lead to a complete PCB redesign when only a minor upgrade or change was intended for version 2.0. This requires a short feedback loop, so the device can be marked as defective after a single failed test. Testers should be encouraged not to stay silent following poor results, even if it means more development time is required to resolve the issue. After all, it is better to spend some time developing a great product than to lose months and materials after assembling many PCBs that do not work as intended.
Visual inspection
This is conducted as the initial test on every unit produced; both prototypes and final products must undergo this process. Although most of these tests are now performed automatically using AI vision systems, a prototype should still be inspected by a tester’s own eyes. The name of this category is self-explanatory: it is a straightforward check to ensure everything has been printed and assembled correctly. A single poorly placed pin or loose cable is more than enough to deem a PCB defective.
Electrical inspection
This phase is commonly carried out once the board has been fabricated, but prior to soldering the components. Its main purpose is to verify that all copper traces are connected exactly where intended. This also involves checking that the insulation is complete and that there are no interferences or power leaks. This stage can be likened to inspecting the internal integrity of the board, in contrast to the external check performed during visual inspection.
Functional inspection
This phase involves the imitation of the real-life environment. Tests conducted at this stage should verify the hardware limits and performance under various conditions. Above all, confirming whether the device actually functions as intended occurs here. This is the most sophisticated phase of testing for a simple reason: every project is unique. While visual testing is generally standardised and electrical inspection can be automated as a fixed sequence of tasks, functional inspection requires a degree of imagination. The conditions in which a board is expected to operate can vary significantly. For example, if the device is intended for use as a wearable, its impact resistance must be higher than that of a device fixed in a secure location.
Common PCB defects
PCBs are manufactured using automated production lines, which can lead to soldering issues. These are the most common of all flaws but are typically eliminated early in the testing process, primarily through Visual Inspection. Traces are another frequent source of issues. There are two main errors associated with trace defects: “Shorts” and “Opens”. A “Short” is an unwanted connection between traces, whereas an “Open” occurs when a trace is broken.
The board itself is only the beginning. Once it has been inspected, a second phase of testing begins following component assembly. This often reveals specific defects. “Tombstoning”, for example, is a situation where a component lifts off one pad and stands vertically resembling a tombstone. Misalignments are the second most common flaw. These may be minor, and while the circuit might still be complete, it will have a poor electrical connection. Another issue arises when a component, such as a diode, is fitted the wrong way round. Nowadays, this is also verified during optical inspection by focusing on the markings on the component bodies.
During functional tests, various issues may be identified, with delamination being one of the most common. This is the primary reason for simulating different environmental conditions, as delamination can occur under specific temperatures or humidity levels. Delamination renders a PCB unsafe and practically unusable; therefore, testing the limitations of the materials used for the board is essential.
What are the 6 types of PCB testing methods?
These are specific types of tests. Earlier, the three groups of tests were mentioned. Now it’s time to focus on the details of particular tests aimed at finding specific issues.
In-Circuit Testing
This is also known as the “Bed of Nails” or ICT for short. It is an automated test used primarily for mass production and falls under the electrical inspection category. It is carried out by pressing many tiny, spring-loaded pins into the board. These pins send and receive electrical impulses to verify several factors.
Firstly, this test covers the presence of Opens and Shorts. Secondly, it verifies whether the board’s overall logic is integrated correctly; to achieve this, it provides power to the entire PCB and inspects the outcomes. Another use for this test is component verification; it examines every component to ensure that, for example, resistors have the correct resistance.
Flying Probe Testing
While ICT connects the pins simultaneously in a static position, FPT uses a few robotic probes to manoeuvre around the board. It is a more flexible approach, as it can scan components and test points separately. This inspection is conducted as a rapid sequence in which the probes “fly” around the board to confirm the absence of defects. This sequence is based on the Gerber files or CAD data of the PCB, making it more software-driven than ICT. Its significant advantage is access; as the probes can change the angle at which they touch the board, they can reach cramped areas that would be inaccessible to the pins used in ICT.
The primary drawback of this method is time. Costs must be calculated in terms of both setup time and testing time. While ICT provides results rapidly once configured, FPT takes longer in both respects. However, once established, it costs almost nothing to adjust the design or reprogramme the system.
Automated Optical Inspection
This covers the external aspect of the visual inspection phase. High-definition cameras, working in conjunction with image processing, are the heart and brain of this process. They compare every assembled board against a “golden standard” and the design parameters from the database. The process also uses lighting from different angles and in various colours to highlight the 3D topography of the board. Pattern recognition examines component placement, labelling, and orientation to ensure there are no flaws. It also involves the examination of soldering through light reflection.
Whilst this test is a modern standard, it is not without limitations. The most obvious is its ability to check only the outer layers of the PCB. Although this is addressed by other test types, it does represent additional expense for the overall process. There is also the possibility of a “false call” due to shadows or variations in component colouring.
Burn-In Testing
This is the equivalent of a stress test for the PCB. The board is placed in a specialised chamber and powered to its maximum capacity. It is subjected to high temperatures to accelerate the chemical and physical processes that lead to failure. The board is monitored in real-time to check for malfunctions. This process can last for up to seven days. If an issue is detected, the board is marked as defective and sent for repair or scrapping.
The drawback of this method is the cost, as significant energy and time are required to conduct it properly. The main purpose of this test is to drastically reduce the chance of a product being “dead on arrival”, a phenomenon where a device breaks down just a few days after purchase. It is also essential for devices that must not fail, particularly in the military and medical sectors, where defects can be fatal.
X-Ray Inspection
This serves as a complement to the visual inspection phase. Whilst AOI inspects only the exterior, this method sees through the board. It verifies connectivity with top-down 2D scans and allows for the creation of a 3D model using scans from different angles. It checks integrity even within the internal layers of the board, detecting air bubbles or insufficient wetting. Identifying these from the outside can be impossible, and the problem might only be detected at a later stage, increasing costs significantly. This test is also useful for detecting counterfeit chips, as it can reveal if a component is empty or if the internal wiring is incorrect.
Although it can save time and costs in the long run, an X-ray system is not inexpensive. Setting it up requires lead shielding and the adherence to strict safety protocols. It also requires highly trained operators for the initial configuration. Whilst it can be automated, maintenance and the initial launch must be handled by professionals.
Functional Testing
In contrast to previous methods, FCT does not focus on whether the board was built correctly, but rather if it functions correctly. It simulates the end-user experience. It is a more flexible approach, as it initially requires a human tester to plan the requirements. While sending various protocols to check connectivity can be automated, a specialist must define those protocols. As every PCB is unique, so are its functions; therefore, there is no single, universal script for this method. As the name suggests, it covers the entire functional inspection phase. Certain processes can be automated following initial tests, but a person will always be required to create the testing plan and devise additional methods for verifying the board.
How to choose the right method?
Final tests do not necessarily have to incorporate every method; some are only required when working with highly sophisticated devices, as the cost of testing can occasionally exceed the value of the product itself. Others, however, are essential. There are no devices that have not undergone functional testing, as this is the one area that is utterly impossible to omit. The general principle for selecting the most effective testing methods is to ensure all three phases are covered, which should be sufficient in most cases.
However, if circumstances require the product to be comprehensively tested “inside and out”, it is better to devote the necessary time to ensure the setup is correct. It is far better to incur the relatively small costs associated with detecting defective units early than to lose significant capital, time, and customer trust by neglecting testing in the initial stages.
