透過您的圖書館登入
IP:44.192.53.34
  • 學位論文

利用雙線性反饋移位暫存器達到低功耗內嵌式自我測試

Dual Linear Feedback Shift Register for Low Power BIST

指導教授 : 蔡文錦 曾王道

摘要


儘管線性反饋移位暫存器 (LFSR reseeding) 架構是一種用於測試資料量壓縮的有效方法,但它可能會導致功率消耗過大。在本論文中,一開始我們描述了一種基於LFSR reseeding的壓縮方法,該方法也使用兩個LFSR來生成壓縮的測試資料,同時減少在掃描移位操作期間發生的不必要位元切換。以下是說明本論文提出的三種方法。 首先,我們所提出的雙線性反饋移位暫存器架構中,第一個LFSR稱為selection LFSR,另一LFSR則稱為compensation LFSR。儘管我們提出的dual LFSR reseeding方法需要用到兩個LFSR,但是與傳統的單一LFSR方法相比,可以透過犧牲一些壓縮效果來實現顯著的功耗降低。此外,所提出的方法所指定的位元數遠小於過去dual LFSR相關的方法,所需的LFSR長度和seed長度都比過去的方法更小。 其次,我們提出的另一種方法,也是使用兩個不同的LFSR並行運行以生成用於AND/OR組合電路的一對來源測試樣式,這一對獨立隨機的測試樣式經由AND/OR組合電路運算後,都具有比原始測試樣式低25%的位元切換機率。 在基於掃描的測試過程中,過多的功率消耗始終是VLSI設計面臨的嚴峻挑戰。有許多研究的發表是透過關閉部份不使用的子掃描鏈的方法,藉此來達到減少功率消耗的效果。因此至關重要的是要識別出觀察細胞將它們都集中 (cluster) 在某些子掃描鏈中,或是將那些較不常使用到的掃描細胞都集中到某些子掃描鏈中。在本論文中,我們將掃描鏈集中與停止的技術相結合,以達到在抓取週期中將停止掃描鏈的數量最大化的效果。

並列摘要


Although the LFSR reseeding scheme is an efficient method for test data compression, it may cause excessive power consumption. In this thesis, firstly, we describe an LFSR-reseeding-based compression method, which also uses two LFSRs, to generate the compressed test data and at the same time to reduce the unnecessary transitions occurred during the scan-shift operation. Although the proposed dual-LFSR reseeding method requires two LFSRs, compared to the traditional single LFSR method, significant power reduction can be achieved by sacrificing some compression effect. Besides, since the number of specified bits targeted by the proposed method is smaller than those by dual-LFSR related methods, the required LFSR lengths and seed volume are therefore smaller than them. Secondly, we propose another approach to use two distinct LFSRs which are operated in parallel to generate the pair of source sequences for AND/OR composition because both AND and OR compositions of two independent random signals have 25% lower transition probability than the original signals. Excessive power consumption is always a serious challenge for VLSI design during scan-based testing. Disabling partial scan chains during testing for minimizing unnecessary power consumption is a popular technique which has been implemented in many research works. Therefore, it is critical to recognize the observing scan cells for the desired fault effects and then properly cluster them into scan chains as dense as possible. In this thesis, we combine techniques of scan chain clustering with scan chain disabling to maximize the number of disabled scan chains in the capture cycle.

並列關鍵字

LFSR BIST test data compression low power

參考文獻


[ 1 ] S. Mitra and K.S. Kim, “X-compact: an efficient response compaction technique,” IEEE Trans. Comput.-Aided Design Integr. Circuits Syst, vol. 23, no. 3, pp. 421-432, Mar. 2004.
[ 2 ] J. Rajski, J. Tyszer, M. Kassab, and N. Mukherjee, “Embedded deterministic test,” IEEE Trans. Comput.-Aided Design Integr. Circuits Syst, vol. 23, no. 5, pp. 776-792, May 2004.
[ 3 ] S. Mitra and K.S. Kim, “XMAX: X-tolerant architecture for MAXimal test compression,” in IEEE Int Conf. Computer Design, 2003, pp. 326-330.
[ 4 ] B. Koenemann, C. Banhart, B. Keller, T. Snethen, O. Farnsworth, and D. Wheater, “A SmartBIST variant with guaranteed encoding,” in 10th Asia Test Symp., 2001, pp. 325-330.
[ 5 ] N.A. Touba, “Survey of test vector compression techniques,” IEEE Design & Test of Computers, vol. 23, no. 4, pp. 294-303, Apr. 2006.

延伸閱讀