Abstract
This work presents a high-gain and wideband 150-GHz LNA featuring a single-to-differential (S2D) topology. The S2D topology eliminates the need for a lossy balun in the input matching network, resulting in an inherent improvement of the achievable noise figure (NF). The proposed four-stage LNA is implemented based on high-gain and wideband Gmax cores. For a compact realization, the first-stage single-ended (SE) Gmax core uses lumped elements (L-C-L) within its embedding network. In addition, a pseudo-simultaneous noise and input matching (p-SNIM) technique is used to mitigate the tradeoff between gain and noise characteristics. Differential wideband Gmax cores in the subsequent stages use a coupled-line-based embedding network (CLEN) without any explicit capacitors. It helps simplify the dc biasing and generally enhances the tolerance to process variations with a compact form factor. Implemented in a 40-nm CMOS technology, the proposed LNA achieves a peak power gain of 20.0 dB, a 3-dB bandwidth of 18.8 GHz, a minimum NF of 4.9 dB, and a maximum OP1dB/IP1dB of −7.7/−22.4 dBm, respectively, while dissipating dc power of 19.6 mW. To the authors’ best knowledge, this work is the first implementation of an S2D LNA operating above 100 GHz using CMOS technologies.
Original language | English |
---|---|
Pages (from-to) | 1973-1984 |
Number of pages | 12 |
Journal | IEEE Journal of Solid-State Circuits |
Volume | 60 |
Issue number | 6 |
DOIs | |
State | Published - 2025 |
Keywords
- Amplifier
- CMOS
- D-band
- G
- LNA
- single-to-differential (S2D)
- sub-THz