Data Analysis with SEDana
Salland Engineering BV the Netherlands
Using SEDana for statistical analysis of ATE tester data.
Introduction
In a semiconductor production environment, it is mandatory to know if the produced products are according specifications. With Automated Test Equipment, the semiconductor industry products are tested, and decided if it failed or pass.
During the test process a lot of data is collected and saved for statistical analysis. By analysing the data production errors can be detected at an early stage and corrected accordingly.
With SEDana it is possible to do statistical analysis on data, which is generated by Automated Test Equipment. SEDana is a stand alone aplication which can be used by Engineers on the work floor.
This document informs you about the usabilitiy of SEDana and how this application can be used for statistical analysis of ATE tester data.
Statistical Process Control
SEDana uses methods which are part of Statistical Process Control (SPC).
SPC is a widely used set of methods for analysing production information and using them for production control.
The combination of reading and analysing ATE tester data, combined with SPC analysis, makes SEDana the ideal combination for the ATE industry.
Click here to read more about SPC
Using SEDana for statistical analysis
When Automatic Test Equipment is used, a lot of measured values and other information is stored for later analysis. In the early days, only binning information was stored, it was time consuming to store more information because of the cost and speed of info storage.
Nowadays it is much more easier and cheaper to store data. More and more complete datasets of measurement data is stored for later analysis.
In combination with the improved calculation power of modern computers and SEDana, application Engineers have the possibility to do this analysis 'realtime' on the test floor.
SEDana is the connection between ATE data and a complete printed or catchy digital report.
Just load raw data into SEDana, create a report with the tables, graphics and diagrams you need and within minutes you can start analysing.
CSV and STDF formatted input
SEDana uses STDF or CSV formatted input.
The STDF format is a standard format which is specified by Teradyne. The STDF format is a defacto standard widely used in the ATE industry.
CSV is known as comma separated file format. It could be handled by most spreadsheet programs.
Showing Raw Data
When a dataset is selected and loaded into SEDana this data is visible in the raw data table.
With optional filter settings it is possible to select only important data.
In the rows tests are displayed and in the columns the measured device data.
Measurements that are below the low limit are displayed in blue, and above the high limit are displayed in red. Layout could be customized.
Measurements with the Alarm bit set can optionally be displayed with a different background color.


Capability:
Statistical analysis starts with validating available data.
Validating data means that the engineer can say something about the stability of the process and the usability of the data for further analysis.
In the Capability table all necessary information is displayed to give these answers.
Statistic values like Mean, Median and Average are calculated, also typical Process Control values like Cp and Cpk are available.
Variables are colour coded for visual information about the level.
To give additional information, each test could be displayed in a separate histogram or line graph for detailed analysis.
With the Capability table the engineer could take conclusions on issues like:
 Is the data within specifications,
 is the production process stable
 is the spread of the data distributed normally

Repeatability:
The variability of the measurements obtained by one appraiser while measuring the same part repeatedly is called the repeatability of a gage. This is also considered the 'inherent variability' of the gage. The repeatability is commonly expressed as an index representing the width of an interval containing 99% of the measurements.
Assuming a normal distribution, 99% of the area under the probability density function is contained between 2.575 and +2.575 standard deviations.
In the Repeatability table the engineer could see the variation of each test.
The parameters for calculating the repeatability are typical values for the ATE industry. If needed these parameters can be changed. 


Reproducibility
Reproducibility is the variation in the average of measurements made by different appraisers, using the same instrument, measuring the identical characteristics on the same part.
SEDana has the possibility for analysing multiple files. With this feature it is possible to combine different appraisers and show the results into a reproducibility table or reproducibility plot.



Range table

The Range Method is a gage study that will provide a quick approximation of measurement variability. This method will only provide the overall picture of the measurement variability. It does not decompose the variability into repeatability and reproducibility. Due to the fact that it's quick, the Range Method is to be used during the first iterations in optimizing GR&R during test program development.
The range methods in SEDana are practical solutions to analyse the data of the used appraisers. When needed, it is possible to change limits, to see what the effects are on calculated results. The multiple Plots window gives one overview of the most important diagrams:
 the Box plot
 the Scatter plot
 the Delta plot
 the Range plot.
If needed, diagrams can be enlarged or shown in separate windows.

Binning
At the end of each test, the binning gives the information about which device has passed and which one has failed.
The Bin Pareto gives a clear view about how the devices binned.
The 3D Wafermap Presentation of the binning is possible when Wafer coördinate information is available.


Create customized statistical reports with SEDana
SEDana is a quick and easy toolbox for statistical analysis of ATE data.After the analysis is done, it is important to get a report for further use and information of the measured data and the conclusions which are connected to it.
The report generator adds all selected tables and graphs to one report. With this generator it is possible to create both digital and printed reports. The layout of the report is customizable.
It is possible to save a report layout and use the same settings for other datasets.
The Project option of SEDana creates a user friendly environment with predefined tables and graphs of actual raw data. This makes SEDana also a valid tool for people who don't have the knowledge for statistical analysis.


Summary
SEDana bridges the gap between raw tester data and statistical reports, which could be used for SPC analysis.
The power of SEDana is speed, simplicity and improved effciency. It brings 'real time' statistical analyis direct on the test floor.
With SEDana test engineers and production engineers have better tools for process analysis, process control and quality improvement.
SEDana is a product of Salland Engineering the Netherlands. Click here for SEDana product information.
