검색 상세

A Bandgap Reference with Single-Slope Curvature Correction

  • 강민생 (Minsaeng Kang, 서강대학교 일반대학원)

초록(요약문)

Battery-limited IoT sensor nodes increasingly demand reference circuits that can provide both ultralow-power operation and high accuracy over a wide temperature range. In particular, wake-up receivers (WuRx) rely on continuously active analog front-ends, making nW-range power consumption essential for long-term battery life. At the same time, the sensing, control, and communication blocks of IoT nodes depend critically on a stable and accurate reference voltage, as the accuracy of all downstream sensor outputs ultimately traces back to the BGR acting as their ground-truth reference. These conflicting requirements expose the fundamental design challenge in modern IoT platforms: achieving high temperature stability without sacrificing the extreme power efficiency required for always-on operation. However, previous bandgap reference (BGR) designs fail to meet both requirements simultaneously. Prior curvature-compensated BGRs report excellent temperature coefficients (TC), but they typically consume tens to hundreds of microwatts, making them unsuitable for WuRx-based or battery- constrained IoT systems. Conversely, low-power BGRs that target sub-µW consumption often omit curvature compensation entirely, resulting in large TC degradation and insufficient accuracy for modern sensor workloads. This clear and persistent trade-off between power and performance highlights the urgent need for a fundamentally different curvature-compensation approach that is compatible with nW-range operation. To overcome these limitations, this work proposes a single-slope curvature-compensation architecture that uses only two comparators to continuously track temperature and generate a temperature- dependent compensation current for the bandgap core. By directly comparing PTAT and CTAT voltages, the proposed system accomplishes curvature correction with minimal active circuitry, enabling substantial reductions in both power and area. The lightweight temperature-tracking mechanism eliminates the need for complex analog processing, ensuring compatibility with nW-range operation while maintaining stable compensation across the full temperature range. Furthermore, dynamic element matching (DEM) is incorporated into both the current mirror and the PTAT generator to suppress mismatch-induced non-PTAT errors, which traditionally dominate in low- power BGR designs. This mismatch-averaging approach significantly enhances stability, enabling one-point trimming to fully correct the remaining TC and reference-voltage inaccuracy. As a result, the proposed scheme effectively resolves the long-standing power–accuracy trade-off and establishes a robust path toward highly integrated, ultralow-power BGRs for next-generation IoT sensor nodes.

more

목차

Chapter 1 : Introduction 9
1.1 Target Application 9
1.2 Main Specifications 10

Chapter 2 : Fundamentals of BGR and Curvature Correction 11
2.1 Temperature Compensation Techniques 11
2.1.1 Complementary-To-Absolute Temperature (CTAT)
2.1.2 Proportional-To-Absolute Temperature (PTAT)
2.1.3 Temperature Compensation
2.2 Sub-threshold Biased BGR 14
2.2.1 Single-Stage Gate Coupled PTAT Generator
2.2.2 Multi-Stage Gate Coupled PTAT Generator
2.3 Curvature Correction 17

Chapter 3 : Proposed Architecture 22
3.1 Motivation of the Design 22
3.2 Proposed Curvature Correction Circuit 22
3.2.1 Low Power Reference Temperature Generator
3.2.2 Single-Slope Curvature Correction Method
3.3 Operating Principles 24

Chapter 4 : Circuit Implementation & Performance Results 27
4.1 Circuit Implementation 27
4.1.1 Curvature Compensation Circuit
4.1.2 Bandgap Reference Core Circuit
4.2 Performance Results 31
4.2.1 One-Point Trimming Simulation Results
4.2.2 Monte-Carlo Simulation & Power Sensitivity
4.2.3 Performance Comparison

Chapter 5 : Conclusion 35
References 36
Acknowledgements 38

more
반출 Meta View 목록