Please use this identifier to cite or link to this item: https://dspace.iiti.ac.in/handle/123456789/5698
Title: NBTI stress delay sensitivity analysis of reliability enhanced Schmitt trigger based circuits
Authors: Shah, Ambika Prasad
Vishvakarma, Santosh Kumar
Keywords: CMOS integrated circuits;Delay circuits;Field effect transistors;Integrated circuit design;Nanostructured materials;Negative bias temperature instability;Reliability;Reliability analysis;Sensitivity analysis;Thermodynamic stability;Threshold voltage;Timing circuits;Appropriate designs;Benchmark suites;Circuit performance;Critical path delays;Design for reliability;pMOS transistors;Schmitt trigger;Threshold voltage degradation;Trigger circuits
Issue Date: 2019
Publisher: Elsevier Ltd
Citation: Shah, A. P., Vishvakarma, S. K., & Cotofana, S. (2019). NBTI stress delay sensitivity analysis of reliability enhanced schmitt trigger based circuits. Microelectronics Reliability, 102 doi:10.1016/j.microrel.2019.06.083
Abstract: Negative Bias Temperature Instability (NBTI) in PMOS transistors results in increased transistor threshold voltage, is considered the major contributor to circuit performance degradation and to alleviate its effect appropriate design and lifetime measures are required. In this paper, we concentrate on a design-time solution, i.e., the replacement of CMOS inverters by more reliable counterparts, i.e., Schmitt Trigger (ST) and NMOS only Schmitt Trigger with Voltage Booster (NST-VB). We first compare the three candidates implemented in 32 nm CMOS technology concerning delay variation. Our results indicate that, after three years of NBTI stress, NST-VB exhibits an almost negligible delay shift of 0.47%, while ST and CMOS inverter experience a delay shift of 7.2% and 5.32%, respectively. Subsequently, we extend the scope and assume the ISCAS'89 s27 circuit as a discussion vehicle. Our evaluations indicate that after 3-year stress time, the critical path delay of the s27 CMOS, ST, and NST-VB based implementations increases by 105.1 ps, 185.2 ps, and 94.2 ps, respectively. To put things into a better perspective, we introduce the Inverse Power Area Reliability Product (IPARP) as compound reliability metric. Our analysis indicates that the normalized IPARP values for ST and NST-VB implementations are 0.062 and 1.903, respectively, compared to CMOS implementation. © 2019 Elsevier Ltd
URI: https://doi.org/10.1016/j.microrel.2019.06.083
https://dspace.iiti.ac.in/handle/123456789/5698
ISSN: 0026-2714
Type of Material: Journal Article
Appears in Collections:Department of Electrical Engineering

Files in This Item:
There are no files associated with this item.


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Altmetric Badge: