A realistic fault model and test algorithms for static random access memories

Abstract

Testing static random access memories (SRAM's) for all possible failures is not feasible. We have to restrict the class of faults to he considered. This restricted class is called a fault model. A fault model for SRAM's is presented based on physical spot defects, which are modeled as local disturbances in the layout of an SRAM. Two linear test algorithms are proposed, that cover 100% of the faults under the fault model. A general solution is given for testing word oriented SRAM's. The practical validity of the fault model and the two test algorithms are verified by a large number of actual wafer tests and device failure analyses.

A systematic method for modifying march tests for bit-oriented memories into tests for word-oriented memories

“…The traditional method of converting a BOM test into a WOM test [1], [2], [3], by repeating the BOM test for each complementary pair of DB values, can only be applied using the uCFst fault model. In addition, it was inefficient because the BOM test, designed for single-cell faults and interword faults, was repeated d=2 times such that the total test time became the product of the "(BOM test time) * (d=2)."…”

“…In the case of idempotent CFs (CFids), disturb CFs (CFdsts) [8], and state CFs (CFsts) [1], the fault will not be detectable with algorithms for BOMs, as shown below. Any BOM test can be converted into a WOM test to detect single-cell and interword faults, as follows:…”

“…Read operations read the B bits simultaneously; write operations write data into the B bits, whereby the data to be written into each bit can be specified independently of the data for the other bits. The value of a B-bit word is referred to as the data background (DB) [1].…”

“…The tests for WOMs to detect coupling faults (CFs) [2], [3], [1] used different types of DBs, such as the Walking 1/O and the Bridging Fault patterns. The disadvantages of the repeated application of the BOM test, together with these types of DBs, are test time inefficiency and limited fault coverage for intraword CFs.…”

New data-background sequences and their industrial evaluation for word-oriented random-access memories

“…Traditionally, WOMs have been tested by repeated application of BOM tests, where different data backgrounds (DBs) are used during each iteration [5], [15]- [17], [19]. The disadvantages of this methodology are test time inefficiency and limited fault coverage of coupling faults (CFs) between cells within the same word, which are called intraword CFs.…”