Machine Learning

Thursday, September 4, 2014

Role of Built-In Self-Test [BIST] in Fault detection

We can easily feel the rapid change in the technology day by day. As we all trying to digitalize everything to make our life much and more simpler. Various developing companies are trying their best to develop user friendly gazettes that must be so small as well as cost efficient that can be easily come under the common man reach. Developers are trying to put lots n lots of components on small integrated chip. Hence, Developers had introduced a concept of ‘Built-in Self-Test’, to test the proper functioning of these chips components. So that, finally when that IC comes in the hand of common man in form of gazette, he can make its uninterrupted use.

Basically, concept of Built-In Self-Test came into picture, when an idea of not using an external circuitry just for testing the proper functioning of all IC components strikes in developers mind. According to this concept, any external circuit is not needed to test the IC. We use to develop an algorithm in such a manner, that will itself check the outputs of a test pattern generator, with this idea, that on applying that TPG output as an input of any circuit under test, we’ll get the desired result or not. BIST is not limited up to this only but also capable of finding our design’s efficiency, in terms of fault coverage, that how much number of faults our design has detected. Fault refers to those inputs, on which our CUT doesn’t give desired output.

It’s very simple approach of BIST, Design for test is developed in which, a comparator used to detect the fault, by comparing the patterns of faulty CUT with non faulty or reference CUT. In this sequence, a slight change is done in the functionality of CUT. Then, finally comparator used to match each bits of both reference circuit and faulty CUT. Previously knowing the total number of faults, this DUT can easily find out the fault finding efficiency of whole design. 

Basic Architecture of BIST
                                                                               
                                                                                

                                                                                                              Author - Akash Kumar
(Research Associate at Silicon Mentor)

No comments:

Post a Comment