A novel semiconductor test equipment concept: automatic test equipment with computational intelligence technique (ATE-CIT)

Abstract

Absfrod ~ Semicundrrr./w ouromurir resf quipmenr (ATE) anul.vser /he response .from the semiconductor chip hosed on U sef of pre-dejinrd res! parrerns and re.s/ conditions. and mark.s the chip os good or bad. This se/ qf Iesrs (parrerns and condition.) is either niunrwlly developed hv engineers or generared vi" circuir-simulation rook. The pmces.~ qfpeneraiing a sei of w,orsr case re.m (I,arrerns and condirionr) i.s vew rime consunling. usual1,v triol and error for drferenl res/ combinationsfimn a long iterative loop during /he design (silicon) anal-vsis phase. The niujor disadvantage is tho1 ATE can no/ learn. mnnipulare and opriniize h. v i l d f b a x d on previous tesfs rrperiences. In /his paper. we propo.wd a computarional inrelligence rechnique ( O r ) with ATE concepr. such rhar recl responses can he descrihed bv .firzq logic. learned by neural network. and resb can he optimized ouroniaricallj, b. v generic algorirhnr. Our experimenral resulrs denionstrare on e.xcellenr efjcienc?; using A TE-CIT during /he design analysis phase.

Embedded Ch based automatic testing environment for electronic products

“…These range from check-out and fault diagnosis of such diverse UUTs as aircraft auto land systems, weapon launchers, radar trackers, missile guidance and control systems, power amplifiers and printed circuit boards [3]. A lot of ATE systems have been developed and acted a very important role in many fields, such as aircraft automatic testing [5] [6], open test software architecture [7] [8], the next generation of ATE [9], Aviation Maintenance [10], Avionics Test Systems [11], medical equipment [12], semiconductor automatic test equipment [13], test script [14][15] [16][17] [18], Automatic Test Markup Language (ATML) [19], and Utilizing Portable Computing Devices in the ATE [20].…”

Computational intelligence semiconductor ate - ultra-short worst case test

“…However, this method will not work for a complex ATE test pattern and test condition coding, because it may not produce valid functional test sequences. Therefore, we propose to use a state-machine rule checker concept similar to [14][15][16] for the GA crossover process in a Multiple Individual Evolution (GA-MIE) method to combine two or more different chromosomes. The idea is similar to evolving many different species at the same time, where the species are not related to each other, and the evolution process only impacts each species individually.…”

“…Test program software does not contain adaptive interoperable reasoning tools. Computational intelligence technologies [8][9][10][11][12][13][14][15] provide significant improvements in diagnostic capability. Genetic algorithms and other artificial intelligence technologies have provided new methods for analyzing and interpreting test results.…”