A Novel Tunable Low-Noise Active Inductor
A novel tunable Low-Noise Active Inductor (LN-AI) with a small area is proposed. It mainly consists of four modules: the gyrator-based feedback loop, the single-ended negative resistance, the noise suppression branch, and the current mirror structure. Through the coordination of four core circuit modules and the combination of external voltage regulation terminals, on the one hand, it achieves a wide range of independent tuning of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$Q$</tex> peak value relative to inductance value, as well as obtaining different inductance values at different frequencies, while maintaining excellent performance with corresponding <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$Q$</tex> peak values that remain basically unchanged; On the other hand, it effectively suppresses the input noise of active inductors. The verification results indicate that the Active Inductor can achieve a wide range of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{Q}$</tex> peak value adjustment from 67 to 810 at a high frequency of 8.3 GHz, while the corresponding inductance value only varies slightly from 58.79 nH to 58.95 nH; At high frequencies of 5.5 GHz, 7.7 GHz, and 8.3 GHz, the inductance values are 21.3 <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{nH}, 27.4 \text{nH}$</tex>, and 40.5 nH, respectively, with a rate of change of 90.1 %, while the corresponding <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$Q$</tex> peak values are 311.9, 311.8, and 313.1, with a rate of change of only 0.4 %; The input reference noise voltage is <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$4.9 \text{nV} / \sqrt{} \mathbf{H z}$</tex> at <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1 \mathbf{G H z}$</tex>, which is a decrease of 1.9 <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{nV} / \sqrt{} \text{Hz}$</tex>, approximately 27.9 %, compared to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$6.8 \text{nV} / \sqrt{} \text{Hz}$</tex> before the addition of noise suppression branch. In addition, LN-AI achieves a small area in the layout.
Authors
- S. Reynold Chu
- Hui Zhang (ORCID: https://orcid.org/0000-0001-6415-7005)
- Xue Wang (ORCID: https://orcid.org/0000-0003-1149-9700)
- Jiao-Jiao Song (ORCID: https://orcid.org/0009-0002-8667-7794)
- Jian Bo Cao
Institutions
- Beijing University of Technology (CN)
Publication Details
- Published
- 2025-07-11
- DOI
- https://doi.org/10.1109/eicct65471.2025.11099846
- Citations
- 1
- Primary Topic
- Advancements in PLL and VCO Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.73
Funders
- National Natural Science Foundation of China
- China Postdoctoral Science Foundation
- Beijing Municipal Commission of Education
- Natural Science Foundation of Beijing Municipality