A differential photoconductive AND gate with Be-dopedlow-temperature-grown InGaAs-InAlAs MQW MSM-PD's

A novel photoconductive AND gate that overcomes the problems of: 1) the long tail of photoconductive switches and 2) signal leakage via switch capacitance (signal feedthrough) is proposed. We use Be-doped low-temperature (LT)-grown InGaAs-InAlAs MQW metal-semiconductor-metal photodetectors (MSM-PD's) to get a shorter turn-off time and propose a differential AND gate to cancel the signal feedthrough. A comparison between LT-grown MSM-PD's and those fabricated by ion implantation shows that the LT-grown ones are ultrafast with a full width at half maximum of 5.3 ps and are suitable for low-bias operation. It is experimentally confirmed that the differential AND gate completely cancels the signal feedthrough in the picosecond region. The differential AND gate: with the LT MSM-PD's achieves return-to-zero (RZ) 20 Gb/s AND operation     

Design, fabrication, and performance of high-speed monolithicallyintegrated InAlAs/InGaAs/InP MSM/HEMT photoreceivers

A detailed study of the performance of monolithically integrated photoreceivers based on metal-semiconductor-metal (MSM) photodetectors (PD's) and HEMT's is undertaken. Two different stacked-layer approaches to integrating MSM-PD's with HEMT's are investigated, and the performance of detectors and HEMT's for each approach is compared. The structure with the MSM layers grown on top of the HEMT layers exhibited the best overall performance. A physics-based MSM model is developed and incorporated into microwave circuit design software; excellent agreement between circuit simulations and measured frequency responses is demonstrated. To evaluate the effects of MSM electrode geometry and detector area on photoreceiver performance, photoreceivers with MSM interelectrode spacings of 1, 1.5, and 2 μm were fabricated and characterized. The electrical amplifier used in the photoreceivers is a two-stage, variable-transimpedance amplifier with a common-gate HEMT as the feedback path. By adjusting the DC voltage applied to the gate of this feedback HEMT, transimpedances ranging from 55.8 to 38.1 dBΩ, with corresponding -3 dB cutoff frequencies from 6.3 to 18.5 GHz, were measured experimentally. Excellent noise performance has been measured, with average input noise current spectral densities of 7.5, 8, and 12 pA/Hz^1/2 obtained for bandwidths of 6.3, 8, and 13.7 GHz, respectively. A packaged receiver has been tested at 5 Gb/s and an open eye pattern obtained     

Performance of In0.53Ga0.47As and InPjunction field-effect transistors for optoelectronic integratedcircuits. II. Optical receiver analysis

For pt.I see ibid., p.957-65. The receiver under study consists of an In_0.53Ga_0.47As p-i-n photodiode and an In_0.53Ga_0.47As or InP JFET transimpedance preamplifier. For this study, the two-region JFET model developed in pt.I is extended to include the dependence of receiver noise on transistor design. The authors find that the channel doping should be small enough to avoid shot noise due to the onset of tunneling current between the gate and drain, where as it must be large enough to provide adequate gain. Also, they show that the receiver sensitivity is not a strong function of input FET gate width. Hence, for circuits with high device density, the gate width and the FET power dissipation can be an order of magnitude less than for those structures currently investigated, thereby incurring a sensitivity penalty of only 1 dB as against wide-gate transistors. Optimized receivers using either InP or In_0.53Ga_0.47As JFETs are found to have comparable sensitivities     

2．5Gb／s PIN－HEMT光接收机噪声精确模拟

本文给出了完整的ＰＩＮ和ＨＥＭＴ器件噪声等效电路模型，对２．５Ｇｂ／ｓＰＩＮ－ＨＥＭＴ光接收机噪声和灵敏度进行了精确的分析计算，并讨论了低频闪烁噪声对灵敏度退化的影响，光接收机噪声实测结果与模拟分析结果数据符合很好。     

Multichannel CPFSK coherent optical communications systems

A detailed theoretical analysis of multichannel coherent CPFSK communications systems is presented. The analysis accounts for the crosstalk between adjacent channels, the intersymbol interference and correlation between noise samples stemming from the limited IF bandwidth the non-Gaussian statistics of the noise at the decision gate, and the impact of the laser phase noise. It is found that the IF bandwidth needed to avoid intersymbol interference is 2.2 bit rates for a modulation index m=1; it is larger for other modulation index values. For m=1, receiver sensitivity is within 1 dB of the shot noise limit, and the electrical domain channel spacing can be as small as 2.05 bit rates with 1-dB sensitivity penalty. The foregoing conclusions are valid for a negligibly narrow linewidth; the degradation due to phase noise is shown to be modest as long as the linewidth does not exceed 1% of the bit rate if m=1. Larger linewidth can be tolerated if the modulation index is larger than unity     

液氮制冷的AlGaN/GaN HEMT太赫兹探测器阵列特性研究

为充分发挥AlGaN/GaN高电子迁移率晶体管 (High-Electron-Mobility Transistor, HEMT)太赫兹探测器阵列的高电子迁移率优势,文中研究了HEMT太赫兹探测器阵列在77 K下的探测特性。使用液氮杜瓦为降温主体搭建了适用于焦平面 (Focal-Plane Array, FPA)芯片的低温系统,实现了对焦平面芯片常温与低温下的对比测试。温度从300 K降到77 K时,探测器阵列像元的平均响应度提高近3倍,平均噪声有小幅增大,340 GHz时平均噪声等效功率 (Noise Equivalent Power, NEP)从45.1 pW/Hz1/2降低到了19.4 pW/Hz1/2,灵敏度提高两倍以上。与硅透镜耦合的单元探测器相比,阵列像元的灵敏度提升仍有较大空间。主要是由于各像素点最佳工作电压的不一致,导致在给定统一工作电压下像元间的响应度和噪声都表现出较大的离散性,文中讨论了降低最佳工作电压离散度的可能解决方案。     

Photoconductor receiver sensitivity

The various contributions to photoconductor (PC's) noise are calculated and are used to determine the sensitivity of digital photoconductor receivers for use in lightwave communication systems. We find that Johnson noise is the most significant source of noise current up to bit rates as high as 4 Gbit/s, above which FET channel noise becomes dominant. In comparing the results obtained for ideal photoconductive receivers with receivers employing p-i-n photodetectors, we find that the sensitivities of both circuits are comparable, provided that low-capacitance p-i-n receivers are employed. In contrast, we find that avalanche photodiode receivers have higher sensitivities than either photoconductor or p-i-n receivers over the entire bit-rate range considered. It is concluded that equalization necessary for photoconductor receiver operation at high bit rates due to a limited gain-bandwidth product significantly degrades the sensitivity of the receiver.     

Probability of error for optical heterodyne DPSK system with quantum phase noise

An expression is derived for the probability of error in an optical heterodyne DPSK system, subject to shot noise and quantum phase noise of the optical sources. Results are given for the maximum linewidth of the transmit and local optical sources to satisfy a given degradation in optical receiver sensitivity.     

A comparative study of the noise performance ofaluminium-gallium-arsenide/gallium-arsenide high electron mobilitytransistors with and without superconducting gate electrodes

The noise performance of an AlGaAs high electron mobility transistor (HEMT) with a 1 μm vanadium/titanium superconducting gate electrode is compared to an otherwise identical nonsuperconducting titanium gate HEMT. At a frequency of 1 GHz and at a temperature below its critical temperature, the superconducting gate HEMT achieved a noise temperature of 21 K. Under these conditions the HEMT with the Ti gate electrode demonstrated a noise temperature of approximately 70 K. This factor of three reduction in noise temperature is due to the reduced gate resistance of the V/Ti superconducting gate. This is the first demonstration of noise reduction in an HEMT using a low-temperature superconducting gate electrode     

Monolithic four-channel photodiode/amplifier receiver array integrated on a GaAs substrate

A four-channel optoelectronic integrated receiver array operating at the wavelength near 850 nm has been fabricated on a single GaAs substrate by using metal-semiconductor-metal (MSM) photodiodes (PD's) and metal-semiconductor field-effect transistors (MESFET's). The largest integration scale for a monolithic receiver and uniform characteristics among circuit channels have been achieved due to the structural simplicity and the process-compatibility of this array. Also, an extremely small capacitance of MSM-PD, 0.10 pF, has lead us to obtain a high-speed operation up to a bit rate of 1.5 Gbit/s, NRZ, and a low noise characteristic exhibiting an equivalent input noise current as small as 5 pA/Hz^1/2. These results have indicated the suitability of MSM-PD/MESFET's circuits for large-scale multichannel optoelectronic integration of receivers.     

A semidistributed HEMT model for accurate fitting and extrapolationof S-parameters and noise parameters

A model is described for a low noise millimeter-wave HEMT device. It takes account of the distributed nature of the gate and drain electrodes by dividing the active region of the device into a number of slices. Each slice is modeled as an intrinsic HEMT with thermal noise sources and the slices are connected together through lossy reactances. The parameters of the first slice are made different from those of the remaining slices, in order to account for the inevitable differences in the field distribution in the gate feed region. The model parameters have been optimized numerically to fit the manufacturer's measured S-parameters and all four noise parameters, for a commercially available HEMT chip. A good fit has been achieved simultaneously to all of these parameters, and the model therefore provides a reasonable basis for extrapolation to higher frequencies. The significance of the distributed gate effect and the unequal slice effect is assessed by comparing the best fit achievable when these effects are not included     

High-frequency MESFET noise modeling including distributed effects

A microwave field-effect transistor (FET) noise analysis is presented including distributed effects caused by the wave propagation along the width of the gate. Using this model the noise characteristics of submicron gate MESFETs at frequencies beyond 20 GHz are evaluated. It is ascertained that in the case of well-designed quarter-micron low-noise MESFETs, distributed effects may be neglected. Common lumped approximations, on the other hand, are shown to produce noticeable deviations. An improved lumped model is proposed. The analysis presented can also be used with high-electron-mobility transistor (HEMT) devices after introducing adequate geometry and small-signal parameters     

Four-Level Pulse Width Modulation for Fiber Optic Communications

The performance of a digital fiber optical system employing four-level pulse width modulation (PWM) is considered. It is shown that PWM may be an attractive alternative to pulse amplitude modulation (PAM), especially if a wide band channel is used. The probability density function (pdf) of the timing error is obtained in terms of the pdf of the noise amplitude, taking into account the thermal noise of the amplifier and the signal-dependent shot noise introduced by the photodiode detector. Near-Gaussian optical pulse edges at the receiver are assumed. Using typical system parameters, the variation of bit error rate (BER) with power level, and the effects of different bit rates, mean photodetector gain, and system bandwidths are calculated. It is observed that minimum BER's are achieved at moderate values of mean photodetector gain and that increased system bandwidth results in increased receiver sensitivity.     

Depletion Mode HEMT with Refractory Metal Silicide WSi Gate

ＤｅｐｌｅｔｉｏｎＭｏｄｅＨＥＭＴｗｉｔｈＲｅｆｒａｃｔｏｒｙＭｅｔａｌＳｉｌｉｃｉｄｅＷＳｉＧａｔｅ￥ＣＨＥＮＤｉｎｇｑｉｎ；ＺＨＯＵＦａｎ（ＩｎｓｔｉｔｕｔｅｏｆＳｅｍｉｃｏｎｄｕｃｔｏｒｓ，ＡｃａｄｅｍｉａＳｉｎｉｃａ，Ｂｅｉｊｉｎｇ１０００８...     

3-V MSM-TIA for Gigabit Ethernet

The paper describes a 3-V monolithically integrated metal-semiconductor-metal photodetector (MSM-PD) and transimpedance amplifier (TIA) chip that is fully compliant with the Gigabit Ethernet receiver specification for the short-reach application (IEEE 802.3z 1000BASE-SX). Key typical performance specifications are -22 dBm sensitivity, 1200 MHz 3-dB bandwidth, 1300-V/W differential responsivity, and 120-mW power dissipation at 3 V. The chip is fabricated in a production 0.5-μm gate length GaAs MESFET technology and is packaged in a TO-46 header with a flat window and a ball-lens cap option     

Polarization independent coherent optical receiver

This paper describes an optical heterodyne receiver for DPSK signals which can receive an optical signal having an arbitrary polarization state. This is achieved by splitting the received signal between two orthogonal polarization axes and processing the resulting two signals as in a conventional DPSK heterodyne receiver. The sum of the two demodulated signals provides a baseband signal independent of the polarization state of the received optical signal. When the receiver noise is dominated by the shot noise of the photodetectors, the receiver provides a BER of 10^-9for an average number of 22 photon/bit. In comparison, a conventional optical heterodyne receiver requires under the same noise condition 20 photon/bit to achieve the same BER for a received optical signal polarized along the polarization axis of the local optical signal.     

A 15 element focal plane array for 100 GHz

A focal plane imaging array receiver is described which covers the 86-115 GHz frequency range for radio astronomical observations. The 3×5 element array uses cryogenic Schottky diode mixers with integrated HEMT IF amplifiers. A cold quasi-optical filter selects the desired sideband, and terminates the image at 20 K. Polarization interleaving is used to minimize the array size on the sky. LO power is provided by a frequency tripled YIG tuned oscillator. The average receiver noise temperature of the array pixels varies from 250-350 K SSB depending on the frequency. Only three mechanical tuners are used in the system and all functions are under computer control     

Low-noise current optoelectronic integrated receiver with internalequalizer for gigabit-per-second long-wavelength optical communications

An equalizer, which is essential in order to improve the sensitivity of receiver optoelectronic integrated circuits (OEICs) at a gigabit-per-second data rate, has been monolithically integrated on an InP substrate with a p-i-n photodiode and a high-impedance high-electron-mobility-transistor (HEMT) amplifier. The receiver operated up to 1.6 Gb/s and showed low noise current characteristics. The minimum noise current is less than 4 pA/√Hz. The sensitivity calculated from the noise current characteristics is -28.4 dBm for 1.6-Gb/s signals. The receiver chip, which was assembled on a ceramic mount, exhibited a sensitivity of -30.4 dBm at 1.2 Gb/s and 1.3-μm wavelength. The performance is as good as those of receiver OEICs with an external equalizer and sufficient for practical use in gigabit-per-second optical communication system     

有源相控阵雷达系统技术参数测试

有源相控阵雷达采用分布式的发射机和接收机,雷达发射系统的辐射功率和接收系统的灵敏度等指标无法直接测量.通过测量有源相控阵雷达中雷达辐射功率与发射天线增益积和接收天线增益与系统噪声温度比来反映雷达系统指标是合理的,并给出了相应的测量方法.