We propose a design for test and diagnosis architecture that simplifies the process of identifying failures within a failing sub-module, without explicitly identifying the faults responsible for that failure. The extra hardware eliminates the post-processing step needed to process the failure logs from the tester and automatically determines the sub-modules inside a core that must be replaced. The proposed methodology reduces the turnaround time for locating a failing module or sub-module and also facilitates testing and diagnosis at the same time. Experimental results are presented using industrial test-cases to validate the effectiveness of the proposed method.