The sampling method standards that my country has formulated so far for SMT patch processing are:
Random sampling method using random number dice
1, Sampling inspection guidelines
Sampling inspection procedures and charts for the rate of nonconforming products (corresponding to ISO3951)
Attribute sequential sampling inspection plan (applicable to continuous batch 2 product sampling inspection on production where inspection costs are expensive)
Single-level and multi-level continuous sampling procedures and sampling tables (applicable to the inspection of products moving on the conveyor belt)
The measurement standard type one-time sampling inspection procedure and table for the rate of non-conforming products
Means of measurement standard type one-time sampling inspection procedures and tables
Counting standard one-time sampling extreme inspection procedure and sampling table for the rate of nonconforming product
Sampling inspection and procedures for skip batches
Small batch count sampling inspection procedure and sampling table for the rate of nonconforming products
Selective sampling inspection procedure and sampling table by attributes
Sampling procedure and sampling table for PCBA product quality supervision
Product quality enumeration sampling inspection procedure for one-time supervision
Standards such as the sampling inspection procedures and sampling tables for the measurement of the average value of product quality.
These sampling method standards respectively make clear provisions for the sampling inspection of enterprises and the sampling inspection methods of national industry and local quality supervision.
Sampling inspection is to randomly select an appropriate amount of product samples from a batch of products (total) for inspection for PCBA quality inspection, and then compare the inspection results with the judgment standard to determine whether the product is qualified or need to be subject to random inspection and then a judgement A quality inspection method.
In the past, percentage sampling inspection methods were generally used. my country has also been using the percentage sampling inspection method adopted by the former Soviet Union in the 1940s. This test method assumes that the sample is always proportional to the population. Therefore, keep a fixed ratio between the number of sample samples and the population of inspection batches, such as 5%, 0.5%, and so on. However, in fact, there is the irrationality of a large number of strict and small batch widths. That is to say, even products of the same quality are subject to different treatments due to the number of inspection batches, and as the total number of inspection batches increases, Even if sampling at a certain percentage, the number of samples is quite large, which cannot reflect the economic advantages of sampling inspection. Therefore, this sampling inspection method has been phased out.
After studying the method of percentage sampling inspection, people learned that the root cause of the unreasonable percentage sampling inspection method is that the sampling plan was not designed according to the scientific method of mathematical statistics. Therefore, a series of statistical sampling inspection or statistical sampling inspection programs based on probability theory and mathematical statistics have been gradually researched and designed, and a standard sampling inspection program has been formulated. In 1949, American scientists Dodge and Romig first published the "Checklist for Primary Sampling and Secondary Sampling"; in 1950, the US military standard MIL-STD-105D is the world's representative sampling inspection method standard for attributes; Japan has successively formulated A series of sampling inspection method standards such as JIS Z9002 and JIS Z9015 have been established; Britain, Canada and other countries have successively formulated sampling inspection method standards; ISO and IEC have respectively formulated international standards for sampling inspection methods, such as ISO2859, IEC410 and so on. Practice has proved that when the above-mentioned sampling inspection method standards are applied to product quality inspection, although there is also the possibility of misjudgment, that is, there are usually manufacturer risks and user risks, but the appropriate sampling inspection plan can be selected to change The risk of this misjudgment is controlled within the scope of people's requirements and is in line with the objective and practical needs of social production and use. Therefore, it has been widely promoted in various countries around the world, replacing the original unreasonable percentage sampling inspection method.
Full inspection is to inspect each product in a batch of products one by one. After picking out the unqualified products, it is considered that the rest are all qualified products. Although this quality inspection method is suitable for large-scale PCBA electromechanical equipment products with small production batches, most PCB products with larger production batches, such as electronic PCBA components products, are not applicable. When the product output is large, the inspection items are many or the inspection is more complicated, it will inevitably cost a lot of manpower and material resources to carry out the full inspection. At the same time, it is still inevitable that false inspections and missed inspections will occur. And when the quality inspection is destructive, such as the life test of TV sets, the strength test of PCB material products, etc., all inspections are even more impossible.