7.10 An 8-to-28GHz 8-Phase Clock Generator Using Dual-Feedback Ring Oscillator in 28nm CMOS

Multi-phase clock generation is among the most critical building blocks in high-speed wireline transceivers. The integrated jitter and phase error of the clocks often dictate the maximum available SNDR at higher frequencies. While the classic method of dividing the differential clocks from a phase locked loop provides excellent phase accuracy with low jitter [1]–[8], it requires the phase locked loop to generate high frequency clocks at multiples of the output frequency <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\mathrm{f}_{\text{out}}\times \mathrm{N})$</tex>. In CMOS processes with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{f}_{\mathrm{T}}$</tex> of around 200-300GHz, the combination of phase locked loop and divider becomes increasingly costly, if not infeasible, in 224Gb/s PAM-4 applications and beyond [9]–[11]. Quadrature VCOs [12]–[14] and poly-phase filters [15]–[17] partly mitigate the issue, but they are sensitive to device mismatches and more importantly, only provide 4 clock phases. The delay locked loop (DLL) consisting of chains of inverters conceptually expands the input clock to any desired number of phases, at the cost of jitter accumulation and higher phase errors due to mismatch. Cascading a ring oscillator injection locked by the DLL suppresses phase error as the ring oscillator is an autonomous circuit with intrinsically strong feedback [18]–[20]. Whereas the cascading configuration offers low jitter and low phase error at the same time, the maximum operation frequency is still limited by the minimum loop delay in the ring oscillator, the frequency of which is determined by <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1/(2\mathrm{m}\times \mathrm{t}_{\mathrm{d}})$</tex> with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$t_{d}$</tex> being the delay of an inverter and m being the number of inverters.

Authors

Institutions

Publication Details

Published
2025-02-16
DOI
https://doi.org/10.1109/isscc49661.2025.10904739
Citations
3
Primary Topic
Advancements in PLL and VCO Technologies
Type
article
Field-Weighted Citation Impact
2.19
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

7.10 An 8-to-28GHz 8-Phase Clock Generator Using Dual-Feedback Ring Oscillator in 28nm CMOS

Shuai Liu, Hao Xu, Weitao He, Na Yan et al.
3 citations
Advancements in PLL and VCO Technologies
2.19
article

7.10 An 8-to-28GHz 8-Phase Clock Generator Using Dual-Feedback Ring Oscillator in 28nm CMOS

Shuai Liu, Hao Xu, Weitao He, Na Yan, Junjie Gu, Yechen Tian
article en
3 citations

Abstract

Multi-phase clock generation is among the most critical building blocks in high-speed wireline transceivers. The integrated jitter and phase error of the clocks often dictate the maximum available SNDR at higher frequencies. While the classic method of dividing the differential clocks from a phase locked loop provides excellent phase accuracy with low jitter [1]–[8], it requires the phase locked loop to generate high frequency clocks at multiples of the output frequency <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\mathrm{f}_{\text{out}}\times \mathrm{N})$</tex>. In CMOS processes with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{f}_{\mathrm{T}}$</tex> of around 200-300GHz, the combination of phase locked loop and divider becomes increasingly costly, if not infeasible, in 224Gb/s PAM-4 applications and beyond [9]–[11]. Quadrature VCOs [12]–[14] and poly-phase filters [15]–[17] partly mitigate the issue, but they are sensitive to device mismatches and more importantly, only provide 4 clock phases. The delay locked loop (DLL) consisting of chains of inverters conceptually expands the input clock to any desired number of phases, at the cost of jitter accumulation and higher phase errors due to mismatch. Cascading a ring oscillator injection locked by the DLL suppresses phase error as the ring oscillator is an autonomous circuit with intrinsically strong feedback [18]–[20]. Whereas the cascading configuration offers low jitter and low phase error at the same time, the maximum operation frequency is still limited by the minimum loop delay in the ring oscillator, the frequency of which is determined by <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$1/(2\mathrm{m}\times \mathrm{t}_{\mathrm{d}})$</tex> with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$t_{d}$</tex> being the delay of an inverter and m being the number of inverters.

Fudan University (CN)
Affordable and clean energy
Openalex Percentile: Top 13%
Advancements in PLL and VCO Technologies
2.19
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.