① May I ask about the LU experiment, Q part class 2? I have noticed that many reports have conducted it at 3ea high temperature and 3ea room temperature. Is there any industry opinion on this?
The worst-case scenario for LU is at high temperatures, so it’s sufficient only to conduct the test at high temperatures. This involves using three voltage and three current excitations, which will be applied to the same unit for both voltage and present excitations.
② May I ask, if there is a suspicion of delamination between PI1 and PI2 in WLCSP 2P2M, can it be detected by SAT?
In theory, it should be possible to detect delamination using Tmode. If the delamination is not too small in area or too subtle (for example, if the delamination in the Z direction is less than 0.1um), it should be detectable. However, if there is suspicion of delamination, it’s recommended to perform a cross-sectional analysis regardless of whether or not it is detected by SAT. This may pose some difficulty.
③ May I ask, during reliability testing of a single IC device, PCT is often conducted? However, according to the official websites of Amkor and SPIL, they do not perform PCT and instead use UHAST. What are the reasons for this difference, and is PCT more destructive to WLCSP products than UHAST?
PCT involves exposing the device to 100% humidity, which is not commonly encountered in everyday life and imposes stronger stress than uHAST. Therefore, many companies now prefer to use uHAST instead of PCT for reliability testing.
④ A customer found a hole in the plastic package during high-temperature aging testing of the board. After decapsulation, there were no burn marks on the chip. Has anyone encountered a similar case?
If there are no burn marks, it’s likely that the failure was caused by moisture. If it’s not a batch issue, it’s possible that the device was opened and left in the air for too long before being tested at high temperatures. In such cases, it may be possible to remedy the situation by low-temperature baking before testing.
⑤ What is the effective shelf life of plastic packaging materials after being reflowed?
The effective shelf life of plastic packaging materials after reflowing varies depending on the type of material used. Some materials have a shelf life of 24 hours, while others can last up to three days. It’s important to refer to the manufacturer’s specifications for the specific material being used. In general, it’s not recommended to use the material beyond three days.
⑥ If both defective and good inventory products are analyzed for hotspots without opening the package and hotspots are identified, but it’s still unclear whether they are abnormal, what other solutions can be used to analyze the situation? What can hotspots on both good and defective products indicate?
Firstly, confirm whether the products are indeed good or not, because the thermal hotspot analysis reflects on a certain surface of the product, and the product is composed of many material structures. Differences in materials at the same point may lead to anomalies, so it is necessary to compare the good and anomalous products after opening the package. The differences in structures, including performance differences, need to be analyzed. This will require the use of probe testing. In addition, there is also the possibility that the so-called good products may not actually be good. In this case, worse condition testing can be performed to verify the products.
⑦ What factors can affect the authenticity of hotspots made from InGaAs? Why is removing the metal layer more accurate or easier to find?
If there is too much metal layer, it can block the brightness of the hotspots, so sometimes it is recommended to make backside hotspots. InGaAs can be summarized in two points: first, the metal layer is not transparent to infrared, and second, the MOS transistor emits light when it operates in the cutoff region.
⑧ The device passed ESD testing before, but after assembly, some pins stopped working, and now there is a crack when the device is opened. Could this crack be caused by ESD, or is there another reason for it?
ESD energy is relatively small and can cause localized momentary high temperatures, but generally does not cause the entire chip to break. If only one chip is affected, it may be due to the existence of microcracks. More chips can be tested to verify this. Additionally, the degree of ESD damage-fuse can be observed to see if it is sufficient to cause a crack. If ATE testing is performed after ESD testing, it could potentially worsen the damage and even lead to cracks. Cracks or peeling that appear to be caused by stress release or mechanical stress, such as stress caused by dicing, should also be considered.
⑨ Which products require LTOL testing?
LTOL (Low-Temperature Operating Life) testing is similar to HTOL (High-Temperature Operating Life) testing but with lower-temperature furnace settings. LTOL is often used to identify failures caused by hot carriers and to test memory devices or sub-micron-sized devices. Related standards for LTOL testing include JESD22-A108B: Temperature, Bias, and Operating Life and JESD47K: Stress-Test-Driven Qualification of Integrated Circuits.
Most reliability studies indicate that devices degrade more severely at higher temperatures. However, the opposite is true for hot carrier injection effects, where the effect becomes more pronounced at lower temperatures. Therefore, aging acceleration is more severe at lower temperatures in the case of LTOL.
The testing conditions for LTOL, as specified in JESD47, typically reference Tj<50℃, but a temperature setting of Ta=-40℃ is commonly used to ensure that the operating temperature remains below 50℃. The results obtained from LTOL testing are applicable to a relatively wide range of applications. In automotive validation, the operating ambient temperature range typically starts at Ta=-40℃.
