A 0.35-V Untrimmed BJT-Based Temperature Sensor Achieving −1.0/+0.86 ° C (p-p) Inaccuracy From –40 ° C to 120 ° C for Cold-Start Temperature Awareness in Energy-Harvesting AIoT Nodes
IEEE Transactions on Circuits and Systems I: Regular Papers · 2026-11 · 73 (11) · 7583-7596
DOI: 10.1109/tcsi.2026.3713052 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems I: Regular Papers, 2026-11. Verfasst von Haochen Yu und 4 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben. Gefördert von: New Generation Artificial Intelligence National Science and Technology Major Project, China, Science and Technology Development Fund, Macau, SAR, University of Macau.
Abstract (Originalsprache)
This work presents a 0.35-V, 169.5-nW temperature sensor for intermittently powered energy-harvesting Artificial Intelligence of Things (AIoT) nodes, enabling early temperature awareness during cold start before the main rail becomes available. A dual-mode switched-capacitor BJT front-end realizes dynamic self-biasing and reduces the average static bias power of the sensing branch. An interleaved charge pump with a proportional-to-absolute-temperature (PTAT)-biased low-voltage oscillator and a voltage-tripling bootstrap network ensures reliable ultra-low-voltage operation and improves ripple tolerance. A voltage-to-frequency readout with a two-stage comparator provides direct digital output without a conventional ADC. To reduce first-order complementary-to-absolute-temperature (CTAT) nonlinearity under tight power constraints, we integrate a leakage-assisted CTAT-linearity compensation scheme into the switched-capacitor network. Fabricated in a 65-nm CMOS process, the sensor achieves a peak error of −1.0/ $+ 0.86~^{\circ }$ C without trimming from $- 40~^{\circ }$ C to $120~^{\circ }$ C, a 26.6-mK resolution, and a supply-induced temperature deviation below $0.4~^{\circ }$ C.
Förderung
- New Generation Artificial Intelligence National Science and Technology Major Project, China
- Science and Technology Development Fund, Macau, SAR
- University of Macau
- National Natural Science Foundation of China
- Hetao SZ-HK S&T Innovation Cooperation Zone Project
Zitieren
APA
Yu, H., U, C., Zhang, Y., Martins, R. P., & Lam, C. (2026). A 0.35-V Untrimmed BJT-Based Temperature Sensor Achieving −1.0/+0.86 ° C (p-p) Inaccuracy From –40 ° C to 120 ° C for Cold-Start Temperature Awareness in Energy-Harvesting AIoT Nodes. IEEE Transactions on Circuits and Systems I: Regular Papers, 73(11), 7583-7596. https://doi.org/10.1109/tcsi.2026.3713052
Vancouver
Yu H, U C, Zhang Y, Martins RP, Lam C. A 0.35-V Untrimmed BJT-Based Temperature Sensor Achieving −1.0/+0.86 ° C (p-p) Inaccuracy From –40 ° C to 120 ° C for Cold-Start Temperature Awareness in Energy-Harvesting AIoT Nodes. IEEE Transactions on Circuits and Systems I: Regular Papers. 2026;73(11):7583-7596. doi:10.1109/tcsi.2026.3713052
BibTeX
@article{Yu2026,
author = {Yu, Haochen and U, Chi-Wa and Zhang, Yin and Martins, Rui P. and Lam, Chi-Seng},
title = {A 0.35-V Untrimmed BJT-Based Temperature Sensor Achieving −1.0/+0.86 ° C (p-p) Inaccuracy From –40 ° C to 120 ° C for Cold-Start Temperature Awareness in Energy-Harvesting AIoT Nodes},
journal = {IEEE Transactions on Circuits and Systems I: Regular Papers},
year = {2026},
volume = {73},
number = {11},
pages = {7583--7596},
doi = {10.1109/tcsi.2026.3713052}
}
EUVIMED