Definition and classification of welding quality issues:
In SMT applications, the quality of the weld of the product can be described by the following definitions.
“A certain degree of mechanical and electrical performance can be maintained during the intended use environment, method and lifetime.”
In this definition, "use environment" refers to the occasions used, such as indoor or outdoor, stationary workbench or moving vehicles, and the temperature and humidity of the environment; "mode" mainly refers to the power-on mode of operation. For example, it will switch many times a day (such as mobile phones, computers, MP3s, automotive electronics, etc.), or basically does not shut down after power-on (such as communication station, home phone, power supply protection, etc.); "lifetime" refers to It is the expected life of the product. These will be different because of the industry situation and the positioning of the company, and the design department must be defined, so the above definition says "design intent." “Maintaining a certain degree” refers to an acceptable level of failure or failure, such as a 1% product failure, or a 20% reduction in performance quantification.
From the above definitions, it is pointed out that we often do not take good care of in daily production, that is, the "life" of the product. Due to the detection technology, as well as the limitations of cost and knowledge, there are not many users who can take care of this in the current SMT user group. So it is not difficult for us to see 'not durable' products on the market.
For companies that want to do a good job of SMT, the definition of quality must include two categories. It is 'zero hour quality' and 'reliability' (or 'life'). 'Zero hour' means that the usage time is zero. That is the quality performance at the time of delivery. If you do not consider the impact of packaging, transportation, inventory, etc., it is the quality of the manufacturer when it is shipped, and the quality of the work is checked by FT (function test), calibration, etc. Since the customer will complain about the return after receiving the defective product, the general manufacturer has a better understanding of this aspect. However, the performance of 'reliability' may not be quantified with sufficient detailed records and data.
In addition to distinguishing between 'zero hour' and 'lifetime' qualities, the quality of the weld can also be divided into 'weld spot' and 'non-weld spot' or 'material' quality. The term 'solder' quality refers to whether the reflow solder joint can maintain its mechanical and electrical bonding properties steadily over its lifetime and under ambient conditions. In reflow soldering, the entire product, including all devices and substrates on the PCBA, will go through high temperatures, and poor or unsuitable high temperature control can damage these materials, which requires engineers to research and process 'non-welding. Point 'quality. Typical non-solder quality issues such as bursting or delamination of the device package, material melting, and the like.
The quality of the solder joints needs to meet several external and internal factors. The external conditions have the following three points:
1. Sufficient and good wetting; 2. Appropriate solder joint size; 3. Good shape.
Sufficient and good wetting is an important indicator of the 'weldability' condition. An unwetted solder joint is difficult to have enough IMC formation, which indirectly tells us that the solder quality is poor. One caveat here is that although there are signs of wetting indicating solderability, it does not fully indicate that the IMC is qualified. The degree or condition of IMC formation is the key to determining the reliability of solder joints. This is an important limitation of the visual inspection capability.
The size of the solder joint directly determines the mechanical strength of the solder joint and the ability to withstand fatigue fracture and creep. In reflow soldering technology, the material of solder joints is generally derived from the amount of solder paste printed. In the case of unsatisfactory matching with the material of the soldered end of the device, large solder joints can sometimes also serve as a buffer quality problem. From the above point of view, we hope that the solder joints are too large. However, too large solder joints can also cause problems. For example, it affects the inspection of wetting, and it is easy to cause process problems such as tin absorption and bridging, and may even shorten the life of electromigration failure.
The contour of the solder joint is also important. Since the stresses on the various parts of the solder joint structure are not the same during use, the 'sink size' factor mentioned above must also be considered together with this 'profile' factor. For example, a 'small tin' appeared on the airfoil pin 'toe', which did not appear to be serious in the 'heel' position.
The three main aspects of the internal structural factors of solder joint quality should be guaranteed.
1. a suitable intermetallic alloy layer;
2. Enriched internal structure of the solder joint;
3. The microcrystalline structure inside the solder joint.

The formation of the intermetallic alloy IMC is the key to determining the mechanical strength of the solder joint. Different metals form IMCs with different combinations of components, and their strengths vary. Therefore, it is an important task to ensure the quality of the matching device, PCB pad plating metal and solder paste metal. After selecting the appropriate material, the next problem is to make the IMC a good thickness by the control of the welding process. When the IMC is not formed, we call the solder joint 'small solder' and its structure is not strong. However, since the IMC itself is a fragile metal, once formed too thick, the solder joints are easily broken in the IMC structure. Therefore, controlling the thickness of the IMC has become a focus in the welding process.
The inside of the solder joint must be 'real'. Since solder paste and PCB materials may emit gas during the reflow soldering process, if the appearance of the solder joints is properly qualified, the inside of the solder joint may be inflated due to the emission of these gases, and some large and small ones appear. Stomata. The performance of the solder joint is actually similar to the case of 'small solder joints', and reliability is threatened.
The microcrystalline structure of the solder joint is affected by the heating temperature, time, and thermal cooling rate. Different thickness structures also have different fatigue resistance. This problem is not very significant in traditional tin-lead. However, after entering lead-free technology, there have been reports of sensitivity to certain alloy materials. When selecting a lead-free material, the user should give the necessary evaluation considerations according to his or her own situation.
'Non-weld point' quality aspects. The heat resistance of the materials we care about (devices and PCBs). As a user, we generally request this information from the supplier at the time of selection in the DFM (Manufacturability Design) process. At present, the more popular approach of suppliers is to provide users with a standard similar to 'reflow curve', which indicates the temperature and time limits for users to follow. In fact, this practice needs to be improved. Because the device is not a single material, it is a 'product' that has been processed from different materials, structural designs and processes. At present, this heat resistance index description method does not control and guarantee quality very accurately. I will provide more detailed explanations in future articles. What readers should know now is that we must have a heat resistance indicator to follow and control our welding process.
Judging the quality of solder joints:
At present, most of the methods used in the industry to check the welding results are MVI (visual) or AOI (automatic optical inspection), combined with ICT (online electrical test) and FT (functional test). The former belongs to the 'appearance' test. Although some process problems can be detected, it cannot cover all appearance failure modes. The ability to use a microscope for manual visual inspection is strong. However, the relationship between speed and cost has not been adopted. Although the AOI speed efficiency is good, the detection rate is not ideal. The latter two tests are electrical tests rather than process tests. That is to say, the process problem must be serious enough to cause electrical problems or differences in the detection, and the process problem can be detected by the two methods. For example, process problems with too small solder joints have not caused electrical problems or differences for most of the time. Process problems like this cannot be identified or detected.
Whether it is the appearance inspection of the former or the electrical detection of the latter, they have not been able to have a high detection rate for the life of the solder joint. Earlier we talked about the external and internal structural factors of quality. These common practices lack the ability to test internal structural factors. So strictly speaking, our inspection technology is currently unable to provide sufficient quality assurance. I will talk more deeply about quality assurance issues in future articles.
Reflow soldering temperature curve:
Before we discuss some of the reflow soldering process failures, let's review the reflow soldering curve of the reflow soldering process. So that we can correspond to the failure mode later. Readers who want to know more about the reflow process can refer to the 2004 "Rework Tips and Technology Integration Considerations for Reflow Soldering Technology" article.
The time/temperature curve of a typical reflow soldering process is similar to that shown in Figure 1 below.
As can be seen from the figure, the entire reflow soldering process can be divided into five processes. That is:
1. Warming up
2. Constant temperature (also known as preheating or volatilization)
3. Welding
4. welding
5. cool down
The purpose of the temperature rise in the first step is to bring the temperature of each point on the PCBA into operation as soon as possible without damaging the product. The so-called working state begins to volatilize the solder paste components that do not help soldering.
The second process, as indicated by its three names (constant temperature, volatilization, preheating), has three functions. One is constant temperature, which is to provide enough time for the temperature of the cold spot to 'chasing' the hot spot. The closer the temperature of the solder joint is to the hot air temperature, the slower the rate of temperature rise. We use this phenomenon to bring the temperature of the cold spot closer to the hot spot temperature. The purpose of making the hot and cold point temperature close is to reduce the amplitude of the peak temperature difference when entering the soldering and soldering area, and to control the quality of the solder joint and ensure consistency. The second function of the constant temperature zone is to volatilize the chemical components that are no longer used in the solder paste. The third effect is to avoid being subjected to too much thermal shock when entering the next reflow process and facing high temperatures.
The fluxing process is when the active material (flux) in the solder paste functions. The temperature and time at the moment provide the activation conditions required for the flux to clean the oxide.
When the temperature enters the weld zone, the heat supplied is sufficient to melt the metal particles of the solder paste. Generally, the materials used in the solder joints and PCB pads of the device have a higher melting point than the solder paste, so the starting temperature of this region is determined by the solder paste characteristics. For example, in the case of 63Sn37 solder paste, this temperature is 183oC. After the temperature rises above this temperature, the temperature must continue to rise and remain for a sufficient period of time to provide sufficient wettability of the molten solder paste, as well as to form an appropriate IMC between the solder terminals of the devices and the PCB pads.
The final cooling zone function, in addition to allowing PCBA to return to room temperature for post-process operation, the cooling rate can also control the microcrystalline structure inside the solder joint. This affects the life of the solder joint.
Relationship between reflow soldering process faults and curves:
Each of the five reflow soldering processes mentioned above has its effect, and the associated failure modes are different. The key to dealing with these process problems is to understand them and how to judge the relationship between failure modes and processes.
For example, the first heating process, if caused by improper settings, may be caused by 'gas explosion', 'solder ball caused by splashing tin', 'material is damaged by thermal shock' and so on. The second section of the 'constant temperature' process may cause different problems. The failure mode of this process may be 'hot collapse', 'connected tin bridge', 'high residue', 'solder ball', 'wetness', 'stomach', 'tombstone' and so on. During the welding process, the specific temperature zone of the furnace is responsible for processing a certain time period or process in the curve, but we generally cannot see the actual process of welding (the current furnace does not provide these measures or convenience, even if some furnaces install glass windows) Also because the solder joint is small, the window is far away and cannot be clearly observed). Only the results after welding can be seen. If we want to solve the problem, we must have the ability to infer the relevant process from the failure mode. To do this, in addition to the need for good observation and capture capabilities, we must first have a very specific understanding of the principles of various fault phenomena (Note 1).
Failure mode of reflow soldering:
In a typical reflow soldering PCBA assembly process, the reflow soldering process is often used by the user for inspection and quality control points. The problems observed here, although not all due to the reflow soldering process, are also related to the failure to set or control the reflow process. To solve problems effectively and thoroughly, we must give research and control to these failure modes, including reflow and non-reflow processes, both online and offline. This can play the role of technology integration.
If we focus on the 'reflow soldering process', the common failure modes are as follows.
1. Poor or insufficient wetting;
2. Solder welding / weak soldering (including due to insufficient heat, see Note 3);
3. Insufficient reflow (solder not fully melted);
4. Shift / fly material (including 'tombstone');
5. Tin/tin reduction;
6. Tin loss (causing less tin or open welding);
7. Bridging / short circuit / continuous tin;
8. Tin ball / tin ball;
9. 'popcorn' effect;
10. Thermal damage to the device;
11. Air holes or vacuum holes appear in the solder joints;
12. Rough solder joints;
13. Cracks or breaks on the surface of the solder joint;
14. Secondary melting (appears in the mixing process or double reflow process).
In addition to the 'virtual welding/weak welding' of item 2, the 'thermal damage' of part 10, and the 'venting hole' failure mode of item 11, it is related to 'zero time fault' and 'appearance' of. In the failure mode associated with 'reliability' or 'lifetime' failure, we have additional descriptions. This is to define the three failure modes of items 2, 10 and 11 above by the destruction (or test) mode in use. The commonly used modes are as follows:
1. Fatigue fracture
2. Creep rupture;
3. Tensile (Note 4);
4. Anti-cutting (Note 4);
5. Earthquake resistance
6. Anti-impact.
The 14 failure modes of 'zero time failure' have a certain relationship with the six failure modes of 'reliability'. Under the appropriate DFR (reliability design) and DFM (manufacturability design), if we can guarantee the 14 failure modes of 'zero time failure', we can guarantee the 'reliability' of the product to a large extent. . It is this relationship that allows us to guarantee reliability through more feasible production quality management and inspection.
Failure mode analysis and solution case:
In SMT technology, all failure modes are not caused by a single factor. It is the main work of the user-based process engineer to find out the factors of each failure mode and conduct research. Let's take a look at an example of a fault principle. I hope that through this case, readers can better understand the concept of technology integration application.
Let us take the first problem of poor or insufficient wetting. The cause of this failure involves various factors such as material type or characteristics, packaging, inventory, logistics handling, and process. In the perspective of supply chain or industrialization, it involves the design department, suppliers, warehouse logistics department, and production process departments. In the requirements of technical integration management, these departments must define their respective responsibilities and ensure that they do their own work. This will prevent problems from happening. The so-called cooperation between the two refers to the definition of work indicators through technical principles and cost-benefit considerations. So at