Best Paper Award

A 20-GHz Differential Push-Push VCO for 60-GHz Frequency Synthesizer toward 256 QAM Wireless Transmission in 65-nm CMOS


[Trans. Electron., Vol.E100-C, No.6 June 2017]

  Recently, demand for high-speed wireless communication is maintaining explosive growth due to the increasing use of wireless devices. For example, streaming 4K videos, instantly synchronizing personal clouds, and connecting AR/VR sets. The 60-GHz frequency band provides promising solutions for such applications. Wireless transceivers operating in the 60-GHz band usually employ high-order quadrature-amplitude-modulation (QAM) for achieving as high a communication data-rate as possible.

  In order to satisfy the requirement for supporting high-order QAM, wireless transceivers should obtain both a high signal-to-noise-ratio (SNR) and low local oscillator (LO) phase noise. The SNR mainly depends on system linearity and noise figure. The phase noise of a LO mainly depends on VCO. Thus, the VCO phase noise sometimes has a dominant influence on transceiver performance, and it is desirable to design a VCO with low phase noise.

  The authors propose a 20-GHz differential push-push VCO for 60-GHz 256-QAM wireless transceiver. According to the harmonic injection lock theory, the 20-GHz VCO is designed at a sub-harmonic frequency of 60-GHz for injection lock. Thus the 60-GHz LO phase noise is dominated by the 20-GHz VCO, for which the tank quality factor is higher than at 60 GHz. Some previous work has been done, and 64-QAM can be supported. In order to further explore high data-rates and low phase noise for employing the 256-QAM, the authors propose the 20-GHz VCO based on 10-GHz QVCO with a tail-filter and push-push architecture. The paper addresses 10-GHz QVCO LC tank optimization, and the analysis results show that 10-GHz QVCO with a tail filter has 4 dB phase noise improvement over a standalone 10-GHz VCO.

  This paper also includes a differential push-push architecture and design considerations that enables differential output for push-push operation. Since the conventional push-push VCO always has a single-ended output, it is not suitable for injection circuits. The proposed 20-GHz VCO has a measured phase noise of -113.8 dBc/Hz at a 1MHz offset. The normalized phase noise to 60-GHz LO is capable of 256-QAM modulation, and the phase noise performance shows 8 dB improvement compared with other state-of-the-art millimeter-wave VCOs.

  In summary, this paper addresses the role of a VCO phase noise performance in 60-GHz wireless transceiver design and proposes a 20-GHz VCO employing 10-GHz QVCO and differential push-push architecture. The proposed VCO shows significant phase noise improvement and provides a solution for 256-QAM wireless transceivers at 60 GHz. Therefore, this paper is deserving of the IEICE Best Paper Award.