Modern life has become more and more dependent on ICs. From consumer mobile phones to the electric vehicle industry, ICs are inseparable. In the future, IC products will develop towards processes with smaller sizes and finer circuits. However, as quality requirements become more and more stringent, The standards and requirements for reliability are becoming more and more detailed. Because the process and steps of manufacturing IC are very complex and there are many variables that affect the performance of the product, reliability testing is very important. As far as product reliability is concerned, the general definition It is how long it takes for it to fail under normal use.
IC Three stages of product failure
usually IC Depending on the failure rate, products can be divided into three characteristic stages, which is also known as the bathtub curve of reliability. ( picture 1) :
figure 1 bathtub curve

- The first stage Infant Mortality : Represents the process in which the failure rate of a product drops rapidly from a high point in the early stages of use. The reason may be due to poor product design or process design. Since it accounts for a small part of the total population, you will see Steeply declining failure curve , as mentioned above, because the manufacturing process is becoming more and more complex, how to reduce the early failure rate to 0 This is no longer possible, and customers don’t want to receive this product, so how to filter out these premature products usually involves Using electric field and temperature to do accelerated life experiments, the so-called burn-in process, if the wafer foundry can use electrical testing during the wafer manufacturing stage, it can also screen out these early defective products.
- second stage Normal life : That is the stage of actual use of the product. The failure rate at this stage is the lowest. It may be caused by short-term failures caused by electrical interference, random external defects, or even some incomplete filtration in the early stages. product caused
- The third stage is Wear-Out , under normal use of the product, the actual product will slowly reach the limit of use, and the failure rate will begin to increase. Finally, based on the collected data, we will define the reliability life and confirm whether it can meet customer needs.
IC Common product reliability issues, the main differences are: IC Components and IC Mainly packaging, IC Common component problems such as oxide layer collapse (oxide breakdown) , which may cause leakage current, even short circuit or partial line failure. In addition, hot carrier injection (Hot carrier injection, HCI) ,by MOSFET As a switching element, if the electron carriers escape from the normal flow path, the characteristics of the switch will change and an unstable situation will occur. Therefore, it can be analyzed such as critical voltage and conductivity. (transconductance) The decay process of other parameters over time; as for packaging reliability, most of it is caused by packaging materials and manufacturing processes, so it is common to use temperature cycle testing, using high and low temperature cycle testing to see whether there is degumming or changes in electrical properties. In addition, high-speed and high-voltage testing is also one of the common tests, and it is also commonly seen in vehicle reliability, such as IC Package testing is similar to reliability testing.

Reliability analysis using statistical methods
The following mainly introduces JMP Reliability acceleration analysis platform, this platform can obtain the life of product failure under more severe conditions, and can calculate the life of failure under normal conditions. Here is an example Statistical Methods for Reliability Data by Dr. Meeker and Dr. Escobar, John Wiley & Sons, 1998 , we want to understand the collapse of dielectrics and do accelerated experiments under given different voltages. (100.3, 122.4, 157.1, 219.0, kV) , want to know about the 50 kV What is the lifespan under? First, the lifespan and voltage are presented in the form of a linear axis ( picture 2) , it can be observed that as the voltage becomes stronger, the life span becomes shorter, and the dispersion of life span under each voltage is obviously different.
figure 2 Linear coordinates

Then convert the axes to logarithms - Logarithmic presentation ( picture 3) , the dispersion range under each voltage will become close, so when accelerated experiments are used for analysis, the data needs to be converted before linear model analysis can be applied. Then when estimating the life under normal conditions using voltage, it will be more accurate.
image 3 logarithm - Logarithmic coordinates

Using the reliability accelerated analysis platform, it can be observed that at voltage = 219kV When , a few points deviate from the fitted straight line, but most of them still fall on the line. Therefore, using logarithmic normal distribution can obtain good fitting results under a fixed slope. ( picture 4) ,
Figure 4 Distribution fitting

And according to JMP The selected model is a three-parameter regression model. Voltage will significantly affect the performance of lifespan, although the models with different position and scale parameters are not much different from the regression model. ( picture 5) , you can collect a few more sets of data for verification later, and finally store the prediction formula. ( picture 6) ,
Figure 5 Model selection results


picture 6 Forecast formula

If the model is verified correctly, you can add voltage = 50kV The distribution curve under ( picture 7) ,learn 10% , 50% , 90% The position where the quantile fails, or enter the corresponding value directly on the form ( picture , or with Profile( picture 9) It can be concluded what the probability of failure is.
Figure 7 Scatter plot

Figure 8 Failure probability data table

Figure 9 Profile

When fitting the model, data conversion can make it easier to use linear models to estimate, reducing the deviation or excessive variation in curve model estimation. In addition, after data scale conversion, it is easier to observe and confirm the data fitting situation. JMP And you can compare the model differences of different effects and select a suitable model. Finally, the estimation formula and Profile For forecasting purposes.

Reference information:
- semiconductor IC Product Reliability Statistical Physics and Engineering Second Edition
- JMP online video : https://www.jmp.com/en_in/events/ondemand/best-practices-in-reliability-data-analysis/accelerated-life-testing.html
- source : http://reliawiki.com/index.php/Inverse_Power_Law_(IPL)-Lognormal_Model
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.