Scan string segmentation for digital test compression
View Patent ↗One may use a new technique to determine the placement of exclusive-ors in each scan string of a chip to achieve improved test vector compression, and one may combine this technique with methods to minimize the overhead of the exclusive-or logic, to eliminate clock enable logic for multiple scan strings, to minimize the changes to existing test logic insertion and scan string reordering, and to minimize the test vector compression computation time.
1. A method of compressing test data for a segmented scan string using a compression algorithm, the method comprising:
using said compression algorithm to independently check any given shift state of the scan string for a compressed solution to setting a multiplicity of care-in locations in the scan string to some predetermined states given an initial scan state;
wherein only a predetermined number, K, plus one shift locations are checked, starting at a location corresponding to the number of said multiplicity of care-in locations minus K, up to a location corresponding to the number of said multiplicity of care-in locations.
2. A method as in claim 1 , wherein K is a number between 0 and 9.
3. A method as in claim 1 , wherein K is chosen based on compression computation and a desired amount of compression.
4. A method as in claim 1 , further comprising:
setting the multiplicity of care-in locations in the scan string according to said compressed solution.
5. A method of compressing test data for a device containing multiple scan strings using a compression algorithm, the method comprising:
selecting a first scan string having a largest number of associated care-in values of the multiple scan strings;
using said compression algorithm to independently check any given shift state of the first scan string for a compressed solution to setting a multiplicity of care-in locations in the first scan string to some predetermined states given an initial scan state, wherein only a predetermined number, K, plus one shift locations are checked, starting at a location corresponding to the number of said multiplicity of care-in locations minus K, up to a location corresponding to the number of said multiplicity of care-in locations, and wherein the compressed solution results in a solution shift count;
choosing a next scan string from the multiple scan strings, for which no compressed solution has yet been obtained, and using said compression algorithm to independently check any given shift state of the next scan string corresponding to a number of shifts greater than or equal to the solution shift count for a next compressed solution to setting a multiplicity of care-in locations in the next scan string to some predetermined states given an initial scan state, to thereby obtain a next shift count corresponding to the next compressed solution;
setting the solution shift count equal to the next shift count if the next shift count is not equal to the solution shift count; and
repeating said choosing and said setting until all scan strings have been chosen.
6. A method as in claim 5 , further comprising:
setting the multiplicity of care-in locations in each of the multiple scan strings according to a compressed solution for all scan strings that corresponds to the solution shift count.