Gyu-Seob Jeong

Silicon photonic receiver and transmitter operating up to 36 Gb/s for λ~1550 nm

We present the hybrid-integrated silicon photonic receiver and transmitter based on silicon photonic devices and 65 nm bulk CMOS interface circuits operating over 30 Gb/s with a 10(-12) bit error rate (BER) for λ ~1550nm. The silicon photonic receiver, operating up to 36 Gb/s, is based on a vertical-illumination type Ge-on-Si photodetector (Ge PD) hybrid-integrated with a CMOS receiver front-end circuit (CMOS Rx IC), and exhibits high sensitivities of -11 dBm, -8 dBm, and -2 dBm for data rates of 25 Gb/s, 30 Gb/s and 36 Gb/s, respectively, at a BER of 10(-12). The measured energy efficiency of the Si-photonic receiver is 2.6 pJ/bit at 25 Gb/s with an optical input power of -11 dBm, and 2.1 pJ/bit at 36 Gb/s with an optical power of -2 dBm. The hybrid-integrated silicon photonic transmitter, comprised of a depletion-type Mach-Zehnder modulator (MZM) and a CMOS driver circuit (CMOS Tx IC), shows better than 5.7 dB extinction ratio (ER) for 25 Gb/s, and 3 dB ER for 36 Gb/s. The silicon photonic transmitter achieves the data transmission with less than 10(-15) BER at 25 Gb/s, 10(-14) BER at 28 Gb/s, and 6 x 10(-13) BER with the energy efficiency of ~6 pJ/bit at 30 Gb/s.

An Optimum Loop Gain Tracking All-Digital PLL Using Autocorrelation of Bang–Bang Phase-Frequency Detection

An all-digital phase-locked loop with a bang-bang phase-frequency detector (BBPFD) that tracks the optimum loop gain for minimum jitter is proposed. The autocorrelation of the output of BBPFD indicates whether the bang-bang PLL operates in the nonlinear regime or the random noise regime. An adaptive loop gain controller continuously evaluates the autocorrelation of the BBPFD output and adjusts the loop gain to make the autocorrelation zero. The digital loop filter operates at higher than the reference clock frequency to reduce the loop latency and to mitigate the resolution of the digitally controlled oscillator. The prototype chip has been fabricated in a 65-nm CMOS process. The core consumes 5 mW at 2.5 GHz and exhibits root-mean-square jitter of 1.72 ps.

A 22 to 26.5 Gb/s Optical Receiver With All-Digital Clock and Data Recovery in a 65 nm CMOS Process

This paper presents a 22 to 26.5 Gb/s optical receiver with an all-digital clock and data recovery (AD-CDR) fabricated in a 65 nm CMOS process. The receiver consists of an optical front-end and a half-rate bang-bang clock and data recovery circuit. The optical front-end achieves low power consumption by using inverter-based amplifiers and realizes sufficient bandwidth by applying several bandwidth extension techniques. In addition, in order to minimize additional jitter at the front-end, not only magnitude and bandwidth but also group-delay responses are considered. The AD-CDR employs an LC quadrature digitally controlled oscillator (LC-QDCO) to achieve a high phase noise figure-of-merit at tens of gigahertz. The recovered clock jitter is 1.28 ps rms and the measured jitter tolerance exceeds the tolerance mask specified in IEEE 802.3ba. The receiver sensitivity is 106 and 184 for a bit error rate of 10-12 at data rates of 25 and 26.5 Gb/s, respectively. The entire receiver chip occupies an active die area of 0.75 mm2 and consumes 254 mW at a data rate of 26.5 Gb/s. The energy efficiencies of the front-end and entire receiver at 26.5 Gb/s are 1.35 and 9.58 pJ/bit, respectively.

A 20-Gb/s 1.27pJ/b low-power optical receiver front-end in 65nm CMOS

This paper describes a CMOS interface circuit for silicon photonics. 20-Gb/s operation of an optical receiver front-end circuit is demonstrated using an optical signal applied to the optical front-end. The transimpedance amplifier (TIA) is based on an inverter with resistive and inductive feedback for low power consumption and frequency compensation. A negative capacitance generation is employed in the limiting amplifier (LA) for bandwidth extension. The combined TIA and LA block exhibits a transimpedance gain of 78 dBΩ and a bandwidth of 11 GHz. The TIA and the LA block consume 1.3 mA and 24 mA at 1 V supply voltage, respectively.

A 0.36 pJ/bit, 0.025 mm 2 , 12.5 Gb/s Forwarded-Clock Receiver With a Stuck-Free Delay-Locked Loop and a Half-Bit Delay Line in 65-nm CMOS Technology.

This paper describes a power and area-efficient forwarded-clock (FC) receiver and includes an analysis of the jitter tolerance of the FC receiver. In the proposed design, jitter tolerance is maximized according to the analysis by employing a delay-locked loop (DLL) based de-skewing. A sample-swapping bang-bang phase-detector (SS-BBPD) eliminates the stuck locking caused by the finite delay range of the voltage-controlled delay line (VCDL), and also reduces the required delay range of the VCDL by half. The proposed FC receiver is fabricated in 65-nm CMOS technology and occupies an active area of 0.025 mm2. At a data rate of 12.5 Gb/s, the proposed FC receiver exhibits an energy efficiency of 0.36 pJ/bit, and tolerates 1.4-UIpp sinusoidal jitter of 300 MHz.

A 0.36 pJ/bit, 12.5 Gb/s forwarded-clock receiver with a sample swapping scheme and a half-bit delay line

A 12.5Gb/s forwarded clock receiver based on a DLL with a bang-bang PD is presented. The stuck locking is detected and averted by swapping edge and data samples at the output of the PD. Moreover, required delay range of the VCDL is reduced by half with the proposed sample swapping scheme. The prototype chip exhibits the power efficiency of 0.36pJ/bit and occupies 0.025mm2. Due to the wide jitter tracking bandwidth of DLL and the inherent jitter correlation between data and forwarded clock, the proposed receiver exhibits outstanding jitter tolerance whose corner frequency is higher than 300 MHz.

A 20 Gb/s 0.4 pJ/b Energy-Efficient Transmitter Driver Utilizing Constant-

This paper describes a transmitter driver based on a CMOS inverter with a resistive feedback. By employing the proposed driver topology, the pre-driver can be greatly simplified, resulting in a remarkable reduction of the overall driver power consumption. It also offers another advantage that the implementation of equalization is straightforward, compared with a conventional voltage-mode driver. Furthermore, the output impedance remains relatively constant while the data is being transmitted, resulting in good signal integrity. For evaluation of the driver performance, a fully functional 20 Gb/s transmitter is implemented, including a PRBS generator, a serializer, and a half-rate clock generator. In order to enhance the overall speed of the digital circuits for 20 Gb/s data transmission, the resistive feedback is applied to the time-critical inverters, which enables shorter rise/fall times. The prototype chip is fabricated in a 65 nm CMOS technology. The implemented driver circuit operates up to the data rate of 20 Gb/s, exhibiting an energy efficiency of 0.4 pJ/b for the output swing of 250 mVpp,diff.

{\rm G}_{\rm m}

A 10 Gb/s PLL-based forwarded clock receiver is implemented in 65-nm CMOS technology. The proposed architecture uses a hybrid PLL-based deskew which is combined with a PLL and a DLL. The PLL provides jitter filtering and the DLL performs deskewing by shifting the divided clocks. Since the operating frequency of the DLL is low, the power consumption of the DLL can be reduced. The measurement results show that the rms jitter of the recovered clock is only 576.7 fs which is quite low for a ring-oscillator-based PLL. The receiver chip occupies an active area of 0.0136 mm2 and consumes 22.1 mW at the data rate of 10 Gb/s.

Bias

This brief illustrates the design of an inductorless high-speed clock generator. Compared to inductance-capacitance (

A 10 Gb/s hybrid PLL-based forwarded clock receiver in 65-nm CMOS

LC