The semiconductor industry chain is very long, but from the perspective of chip testing, the chip manufacturing process can be divided into wafer manufacturing, wafer testing, chip packaging, finished product testing and other stages. The wafer test is completed by ATE and Prober; The finished product test is completed by ATE with high and low temperature impact airflow meter.

From R&D to mass production, testing is the key necessary link

Because chip testing accounts for a relatively small proportion of the total chip manufacturing cost, in the range of several percent, it is usually ignored by many chip design companies. But in fact, chip testing is the key link from R&D to mass production. First of all, we guarantee the quality of our products through wafer testing and finished product testing. Secondly, the quality of the test plan directly affects the yield and the test cost.

Chip testing is a relatively complex work, which involves test equipment, test procedures, test parameters, test conditions, test methodology, new product mass production and introduction process and system, chip function and performance verification, test data analysis and many other aspects. To develop a high-quality test plan, developers need to be familiar with the functions of the chip under test, the performance of the test equipment used, have a sound theoretical knowledge of semiconductor testing, and have rich practical experience in mass production test development.

The focus of the test project is the development and implementation of the test program, including three main ATE test programs: the QUAL program for reliability testing, the CHAR program for chip verification, and the CP and FT program for mass production.

2、 Different expectations for ATE test plan

First, the expectation of reliability test on ATE test plan (QUAL)

First, very fast lock test program (ASAP). Before and after the chip reliability test, we need to use ATE program to collect the drift of chip test results on key parameters. Some reliability test projects take a long time to test. Only by locking the ATE QUAL test program as early as possible can we start the reliability test, so as to know whether the reliability design of the chip meets the expectations as soon as possible.

Second, sufficient coverage ("Good Enough" test coverage). Without sufficient coverage, the program will lose the monitoring of the test results of some key parameters of the chip, thus ignoring the potential quality risk.

Third, accuracy. We need a certain accuracy to track the drift of key parameters, and the accuracy of the test should be able to quantitatively reflect the change of the trend.

Is the result of multiple measurements of the same parameter of the same chip consistent with that of Gauge R&R – Repeatability&Reproducibility? Are the results consistent on different test machines? Are the test results obtained by using different test fixtures on the same machine consistent? Are the test results at different stations on the same machine and fixture consistent? These are the data we need to collect and analyze.

ThermoTST TS760, developed by Chengdu Medium and Low Temperature Production Co., Ltd., is a precise high and low temperature impact airflow instrument with a wider temperature range of - 70 ℃ to+225 ℃, providing a very advanced temperature conversion test capability. Temperature conversion from - 55 ℃ to+125 ℃ for 10 seconds; It meets the requirements of various production environments and engineering environments through long-term multi-condition verification. TS760 is pure mechanical refrigeration, without liquid nitrogen or any other consumable refrigerant. It is used with ATE to conduct reliability tests such as characteristic analysis, high and low temperature temperature change test, temperature impact test, failure analysis, etc.

Secondly, the expectation of the verification test on the ATE test scheme (CHAR)

First, complete coverage. CHAR test program is the largest and most complete ATE test program in a project. Its function is to make the most complete detection of each ip and parameter of the chip under different test conditions. The test results obtained by CHAR program under different process cores, different voltages and different temperatures help us to conduct a "comprehensive physical examination" of the chip, namely characterization across "PVT".

Second, it is very accurate. Compared with the QUAL test program, we have higher expectations for the accuracy of the CHAR test results. Because CHAR program is not used for mass production and is not sensitive to test time, in order to analyze the chip performance more accurately, we can achieve the most accurate results that ATE instrument can achieve by setting more accurate measurement gear, taking more sample points for averaging, and reserving more wait time/setting time.

Third, stability. Like the QUAL program, we also need stable ATE test results.

Finally, the expectation of production test on ATE test plan (CP/FT)

First, the lowest cost. In any large-scale production, the production cost must be our top priority. Therefore, we need to use "the most appropriate, not the most expensive, high-end ATE test equipment" to develop the mass production test plan.

Second, very stable (High Gauge R&R). Among the three ATE test procedures, the "mass production test procedure" has the highest requirement for stability. This is because the stability of the test results greatly affects the yield of mass production, which is also indirectly converted into the production and manufacturing costs of the chip, and also affects the production capacity and delivery of the product.

Third, sufficient test coverage. We use sufficient coverage, not complete coverage, because of the test cost. It is common for ATE CHAR program to complete the test data collection of a chip in minutes, tens of minutes, or even hours. Obviously, this program cannot be used for mass production because it is too expensive! The hourly charge for a set of ATE testing equipment varies from $10 to $200. If a chip test takes dozens of minutes, it must sell at a sky-high price in order to break even. So we have to make trade-offs between test coverage and test time.