Reasons of EOS damage during production lines or applications

As semiconductor process technology becomes more and more advanced, operating voltages become lower and lower, and system functions become more complex, EOS (Electrical Over Stress) is more likely to invade the internal system, causing the damage area to increase. For years, EOS has been one of the reasons of the highest product failure rates in the technology industry. However, in terms of solution time and cost, the cost of the occurrence of EOS is too expensive. Therefore, in most applications, failed components in the product will be provided to suppliers and the suppliers will use professional test and measurement and analysis to determine the failure phenomenon and find out the root cause of the failure and implement effective solutions. Otherwise, the long-term reliability of the entire product is likely to be affected.     

As shown in Figure 1, the causes of EOS damage can be divided into three major categories. The first category is EOS events in natural environments that mainly occur outdoors or with products connected to outdoor cables. For example, routers connected to outdoor network cables, DVR monitoring hosts connected to outdoor cameras, or TVs connected to outdoor antennas may be directly attacked by lightning strikes during bad weather, or indirect lightning strikes may be induced into the wires and enter into the product. They are all the reasons for EOS damage. In addition, many systems will complete EOS protection designs for outdoor cable interfaces. However, if the system is poorly grounded and the lightning strike energy cannot be released enough, it may be released to the connected system through the interface (such as HDMI) connected to other systems. In this case, it is very likely to cause EOS damage of the interface connecting the two systems.     

The second category is EOS events during system operation: This category of events may not only occur when consumers use the system, but may also occur during factory production line testing or assembly stages. Common phenomena include unstable test current or voltage; damaged test fixtures, skewed connector pins or contacts leading to wrong holes; reverse connection of connectors, live switch, hot swapping, etc. Hot swapping means suddenly changing the current while power is on, which may produce dangerous induced voltages that affect system operation. Moreover, when plugging or unplugging, it is easy to cause instability in the power supply or signal in the cable, such as overshoot or undershoot. A more serious situation may occur when two devices with different ground potentials are connected. EOS damage may occur due to excessive voltage difference or large current, and sparks may even be observed. In order to avoid EOS damage caused by hot swapping, many industries use First Mate Last Break (FMLB) grounding points on connectors, which can be implemented by simply extending the grounding pin. When using such a connector, the ground pin will first touch the system ground and finally disconnect during the unplug process, thus ensuring that there is a 100% reliable ground potential to refer to.     

In addition, special attention needs to be paid to whether there is unstable voltage at the power supply. The power supply infrastructure in developing countries may not be sound enough, so it is easy for the power supply voltage to be unstable, leading to large energy surges and causing EOS damage. The EOS events that may be encountered during the operation of the above systems can be reduced through the system's heat dissipation design, voltage stabilization design, and effective OVP or TVS protection solutions to reduce the damage caused by EOS to the product.     

The third category is EOS events caused by transient noise. Such phenomena may exist in system coupling RF, stray EMI or electromagnetic pulses, etc. to cause EOS phenomena. For example, when using an antenna or radar, if there is noise received by the antenna, appropriate protection circuits and filter circuits need to be designed in the system, otherwise the noise amplified by the circuit may cause permanent damage to the detector circuit. When the power supply of the distribution box or the relay switch in the circuit is switched, high-frequency surges may also directly damage the system. Therefore, in circuits with large loads or when the equipment is sensitive, adding protective components or using voltage stabilizing equipment should be considered. In addition, if there are components in the system that are prone to noise, such as motors or inverters, and without proper shielding or isolation, the original high-frequency surges effected by the inductance and capacitance of the cables may turn into a low-frequency EOS surge[1]at the back end, which causes back-end products to be damaged by EOS.

According to statistics, the most common causes of EOS include hot swapping, overvoltage, power surges and component welding errors. Most damage does not occur during the component manufacturing process, but during the PCB/module assembly process (about 30 %) or during applications (about 40%) [2]. In addition, when analyzing EOS problems, the most important thing is to fully describe the occurrence and quantitative data, including how the anomaly was discovered? Can it be reproduced? Application scenario? The possible path and energy size of the surge, and even the situation of IC damage (package melting/bonding burnout/chip burnt), etc. These information can make EOS analysis more accurate. For example, in the actual case below, after collecting damaged samples, it was found that most of the IC burns were caused by the meltdown of a single component. After using surge equipment to reproduce the same burn phenomenon, it was then learned that the damage was caused by an 8/20us surge energy of about 7A. Therefore, it is recommended to use protective components with an IPP greater than 7A to provide system protection, which can improve more than 60% of the current EOS problems, as shown in Figure 2. This is the concept of thinking of TVS as a fuse. If the TVS is damaged, you can use the recurrence method to know how much surge energy is encountered, and then select a more correct TVS. As long as appropriate TVS protection components are added to the port or power supply during design, protection can be done in advance to avoid system product damage or crashes, resulting in losses and increased costs. It can also reduce RMA issues and enhance the company's brand image.

Figure 2: EOS damage reproduced when using Surge machine

Amazing Microelectronic Corp has advanced electrostatic discharge protection design technology. Especially for products that require EOS protection, we have developed surge protection components in different packages and specifications, which can be selected by customers according to product needs, as shown in Figure 3. Amazing Microelectronic Corp uses its own patented technology to launch a series of protection components, providing 0603 solutions for different applications. The single ESD tolerance exceeds 30kV and has EOS protection capabilities. Among them, the 0603 package size is only 1.6 mm x 1.0mm and the thickness is only 0.5 mm, which can meet the needs of new products for small packages. With the trend of new technologies in the future, Amazing Microelectronic Corp will continue to develop protective components that are more in line with market customer needs, and can also provide perfect solutions for customized products.