2018Unpublished venueRequires access

A DDS-Driven ADPLL Chirp Synthesizer with Ramp-Interpolating Linearization for FMCW Radar Application in 65nm CMOS

Liheng Lou, Kai Tang, Zhongyuan Fang, Bo Chen, Ting Guo, Zhe Liu, Yuanjin Zheng

Open publisher page 10 citations

Abstract

The paper presents a wideband, low-power chirp synthesizer for Ku-band FMCW radars. The DDS-driven ADPLL chirp synthesizer generates chirps up to 2GHz bandwidth. A low-power DDS is employed to generate a stepped reference frequency for ADPLL to ensure the long term chirp linearity, while the ADPLL fulfill the ramp-interpolating linearization by generating a narrow band chirp within the step frequency. The entire chirp benefits from the phase domain modulation and mitigates ADPLL loop bandwidth variation to achieve high linearity and low power. A novel adaptive ramping control scheme is adopted to ensure the chirp rate is constant after combining the narrow band chirp segments. By this two-stage chirp generation, a 2-GHz chirp is achieved with low relative frequency RMS error. Fabricated in a 65nm CMOS, the synthesizer generates a wideband chirp from 13.9GHz to 15.9GHz with configurable rate from 0.25 to 4GHz/ms, and consumes 53.8mW.

About this research paper

What this paper is about

The paper presents a wideband, low-power chirp synthesizer for Ku-band FMCW radars. The DDS-driven ADPLL chirp synthesizer generates chirps up to 2GHz bandwidth. A low-power DDS is employed to generate a stepped reference frequency for ADPLL to ensure the long term chirp linearity, while the ADPLL fulfill the ramp-interpolating linearization by generating a narrow band chirp within the step frequency. The entire chirp benefits from the phase domain modulation and mitigates ADPLL loop bandwidth variation to achieve high linearity and low power. A novel adaptive ramping control scheme is adopted to ensure the chirp rate is constant after combining the narrow band chirp segments. By this two-stage chirp generation, a 2-GHz chirp is achieved with low relative frequency RMS error. Fabricated in a 65nm CMOS, the synthesizer generates a wideband chirp from 13.9GHz to 15.9GHz with configurable rate from 0.25 to 4GHz/ms, and consumes 53.8mW.

Why it matters

OpenAlex reports 10 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The paper presents a wideband, low-power chirp synthesizer for Ku-band FMCW radars. The DDS-driven ADPLL chirp synthesizer generates chirps up to 2GHz bandwidth. A low-power DDS is employed to generate a stepped reference frequency for ADPLL to ensure the long term chirp linearity, while the ADPLL fulfill the ramp-interpolating linearization by generating a narrow band chirp within the step frequency. The entire chirp benefits from the phase domain modulation and mitigates ADPLL loop bandwidth variation to achieve high linearity and low power. A novel adaptive ramping control scheme is adopted to ensure the chirp rate is constant after combining the narrow band chirp segments. By this two-stage chirp generation, a 2-GHz chirp is achieved with low relative frequency RMS error. Fabricated in a 65nm CMOS, the synthesizer generates a wideband chirp from 13.9GHz to 15.9GHz with configurable rate from 0.25 to 4GHz/ms, and consumes 53.8mW.

Key concepts: Chirp, Chirp spread spectrum, Electronic engineering, CMOS, Direct digital synthesizer, Bandwidth (computing), Wideband, Frequency synthesizer

Related papers

Back to paper searchBrowse research topicsOriginal source
A DDS-Driven ADPLL Chirp Synthesizer with Ramp-Interpolating Linearization for FMCW Radar Application in 65nm CMOS — Research Paper | ScholarLens