Research On Test Excitation Compression Method Based On Compatibility Transform

The development level of the chip has reached the ultra-deep submicron level,and the number of integrated transistors on the circuit has also doubled,which will increase the difficulty of chip testing.The increase of test stimulus makes the automatic tester need a larger memory to store the test data,but the memory of the automatic tester is limited,so it is very important to effectively compress the test stimulus.Aiming at the problem of excessive test stimulus and difficulty in compressing test data during the testing process,this paper proposes a test stimulus compression method based on compatibility transformation.In this paper,the transformation of the original test vector set is divided into the main component set and the residual set,and the three parts of the original test vector set,the main component set and the residual set are processed respectively.For the compatibility fusion of the original test set vector,this paper defines the compatibility relation directed graph,and constructs the compatibility fusion method based on the compatibility relation directed graph to reduce the scale of test vectors.In this paper,the cellular automata(Cellular Automata,CA)is used as the hardware vector structure,and the principal component set is constructed after screening.Aiming at the problem of filling the X bits in the residual set,based on the compatibility relationship proposed in this paper,the X bits are filled in the direction that is conducive to the compatibility of the residual set and the principal component set.The filled residual set is encoded and compressed and stored on the automatic tester.When the test is implemented,the main component set and the decompressed residual set are XORed back to the original test set and applied to the circuit under test.Compared with directly encoding and compressing the original test set,this article transforms it into encoding and compressing the residual set,which can significantly improve the compression rate of the encoding.Experiments show that the design can effectively compress the test excitation.The scale of the test set after compatibility processing is reduced by an average of 16.63% compared to the original test set.Choosing cellular automata as the hardware structure reduces the chip area by 21.74%compared with the linear feedback shift register,and the residual set compression rate The average is up to 75%,which is a 20.3% increase in compression rate compared to compressing the original test vector set.