High-frequency performance of subthreshold SOI MESFETs

Results from silicon-on-insulator (SOI) MESFETs designed for subthreshold operation are presented. The transistors have subthreshold slopes as low as 78 mV/dec and off-state drain currents approaching 1 pA//spl mu/m. Drain current saturation can be achieved with drain voltages of less than 0.5 V and with output impedance>100 M/spl Omega//spl middot//spl mu/m. The cutoff frequency of a 500-nm gate length device exceeds 1 GHz at currents significantly less than 1 /spl mu/A//spl mu/m. These results suggest that subthreshold SOI MESFETs might have useful applications in mixed-signal, micropower circuit design.     

Micro-power low-offset instrumentation amplifier IC design for biomedical system applications

This work presents a micro-power low-offset CMOS instrumentation amplifier integrated circuit with a large operating range for biomedical system applications. The equivalent input offset voltage is improved using a new circuit technique of offset cancellation that involves a two-phase clocking scheme with a frequency of 20 kHz. Channel charge injection is cancelled by the symmetrical circuit topology. With the wide-swing cascode bias circuit design, this amplifier realizes a very high power-supply rejection ratio (PSRR), and can be operated at single supply voltage in the range between 2.5-7.5 V. It was fabricated using 0.5-/spl mu/m double-poly double-metal n-well CMOS technology, and occupies a die area of 0.2 mm/sup 2/. This amplifier achieves a 160-/spl mu/V typical input offset voltage, 0.05% gain linearity, greater than 102-dB PSRR, an input-referred rms noise voltage of 45 /spl mu/V, and a current consumption of 61 /spl mu/A at a low supply voltage of 2.5 V. Experimental results indicate that the proposed amplifier can process the input electrocardiogram signal of a patient monitoring system and other portable biomedical devices.     

A 230-nW 10-s time constant CMOS integrator for an adaptive nerve signal amplifier

This paper describes a micropower CMOS integrator with an extremely large time constant for use in a variety of low-frequency signal processing applications. The specific use of the integrator in an implantable biomedical integrated circuit is described. The integrator is based on the OTA-C approach and a very small transconductance of 100 pA/V was achieved by cascading a short chain of transconductance-transimpedance stages. The time constant of the integrator is tunable between about 0.2 and 10 s, and any offset voltages at the output terminal can be trimmed out. The circuit was fabricated in a 0.8-/spl mu/m CMOS process, dissipates 230 nW from /spl plusmn/1.5 V power supplies (excluding the bias circuitry and output buffers) and has a core area of 0.1 mm/sup 2/. The integrator offers superior performance in terms of power consumption, die area and time constant when compared to previously published work.     

Self-heating and trapping effects on the RF performance of GaN MESFETs

RF power performances of GaN MESFETs incorporating self-heating and trapping effects are reported. A physics-based large-signal model is used, which includes temperature dependences of transport and trapping parameters. Current collapse and dc-to-RF dispersion of output resistance and transconductance due to traps have been accounted for in the formulation. Calculated dc and pulsed I-V characteristics are in excellent agreement with the measured data. At 2 GHz, calculated maximum output power of a 0.3 /spl mu/m/spl times/100 /spl mu/m GaN MESFET is 22.8 dBm at the power gain of 6.1 dB and power-added efficiency of 28.5% are in excellent agreement with the corresponding measured values of 23 dBm, 5.8 dB, and 27.5%, respectively. Better thermal stability is observed for longer gate-length devices due to lower dissipation power density. At 2 GHz, gain compressions due to self-heating are 2.2, 1.9, and 0.75 dB for 0.30 /spl mu/m/spl times/100 /spl mu/m, 0.50 /spl mu/m/spl times/100 /spl mu/m, and 0.75 /spl mu/m/spl times/100 /spl mu/m GaN MESFETs, respectively. Significant increase in gain compression due to thermal effects is reported at elevated frequencies. At 2-GHz and 10-dBm output power, calculated third-order intermodulations (IM3s) of 0.30 /spl mu/m/spl times/100 /spl mu/m, 0.50 /spl mu/m/spl times/100 /spl mu/m, and 0.75 /spl mu/m/spl times/100 /spl mu/m GaN MESFETs are -61, -54, and - 45 dBc, respectively. For the same devices, the IM3 increases by 9, 6, and 3 dBc due to self-heating effects, respectively. Due to self-heating effects, the output referred third-order intercept point decreases by 4 dBm in a 0.30 /spl mu/m/spl times/100 /spl mu/m device.     

An MSI GaAs integrated circuit: 4 bit arithmetic and logic unit

A GaAs 4 bit arithmetic and logic unit (ALU) has been fabricated using a planar ion implantation technique with 2 /spl mu/m gate length FETs. The basic circuit is a buffered FET logic (BFL) circuit composed of normally on GaAs MESFETs and Schottky diodes. The active layers of the FETs and diodes are made by implanting Si into Cr-doped semi-insulating GaAs substrate. This ALU contains 629 FETs and 225 diodes within an area of 1.6/spl times/2.1 mm/SUP 2/. The ALU, capable of driving 50 /spl Omega/ transmission lines, is mounted on a 24 lead flat package. A delay time of 2.1 ns through the data path and a total power dissipation of 1.2 W with supply voltages of +5 V and -3 V have been obtained.     

A 0.85-ns 1-kbit ECL RAM

A 1-kb ECL RAM with an address access time of 0.85 ns is described. This excellent performance is achieved by combining super self-aligned technology (SST) with 1-/spl mu/m design rules and high-speed circuit design. SST provides a narrow emitter stripe width of 0.5 /spl mu/m and a high cutoff frequency of 12.4 GHz at V/SUB CE/=3 V. A two-level metallization process is used. The minimum metallization pitches are 3 /spl mu/m in the first layer and 6 /spl mu/m in the second one. The chip size is 2.5/spl times/2.5 mm/SUP 2/ and the power dissipation is 950 mW. This RAM is promising for use in super computers and/or high-speed digital systems.     

A physics-based frequency dispersion model of GaN MESFETs

A physics-based model for GaN MESFETs is developed to determine the frequency dispersion of output resistance and transconductance due to traps. The equivalent circuit parameters are obtained by considering the physical mechanisms for current collapse and the associated trap dynamics. Detrapping time extracted from drain-lag measurements are 1.55 and 58.42 s indicating trap levels at 0.69 and 0.79 eV, respectively. The dispersion frequency is in the range of megahertz at elevated temperature, where a typical GaN power device may operate, although at room temperature it may be few hertz. For a 1.5 /spl times/ 150 /spl mu/m GaN MESFET with drain and gate biases of 10 V and -1 V, respectively, 5% decrease in transconductance and 62% decrease in output resistance at radio frequencies (RFs) from their DC values are observed. The dispersion characteristics are found to be bias dependent. A significant decrease in transconductance is observed when the device operates in the region where detrapping is significant. As gate bias approaches toward cutoff, the difference between output resistance at dc and that at RF increases. For drain and gate biases of 10 and -5 V, output resistance decreases from 60.2 k/spl Omega/ at dc to 7.5 k/spl Omega/ at RF for a 1.5 /spl mu/m /spl times/ 150 GaN MESFET.     

Substrate-triggered ESD protection circuit without extra process modification

A substrate-triggered technique is proposed to improve electrostatic discharge (ESD) protection efficiency of ESD protection circuits without extra salicide blocking and ESD-implantation process modifications in a salicided shallow-trench-isolation CMOS process. By using the layout technique, the whole ESD protection circuit can be merged into a compact device structure to enhance the substrate-triggered efficiency. This substrate-triggered design can increase ESD robustness and reduce the trigger voltage of the ESD protection device. This substrate-triggered ESD protection circuit with a field oxide device of channel width of 150 /spl mu/m can sustain a human-body-model ESD level of 3250 V without any extra process modification. Comparing to the traditional ESD protection design of gate-grounded nMOS (ggnMOS) with silicide-blocking process modification in a 0.25-/spl mu/m salicided CMOS process, the proposed substrate-triggered design without extra process modification can improve ESD robustness per unit silicon area from the original 1.2 V//spl mu/m/sup 2/ of ggnMOS to 1.73 V//spl mu/m/sup 2/.     

SODEL FET: novel channel and source/drain profile engineering schemes by selective Si epitaxial growth technology

In this paper, novel channel and source/drain profile engineering schemes are proposed for sub-50-nm bulk CMOS applications. This device, referred to as the silicon-on-depletion layer FET (SODEL FET), has the depletion layer beneath the channel region, which works as an insulator like a buried oxide in a silicon-on-insulator MOSFET. Thanks to this channel structure, junction capacitance (C/sub j/) has been reduced in SODEL FET, i.e., C/sub j/ (area) was /spl sim/0.73 fF//spl mu/m/sup 2/ both in SODEL nFET and pFET at Vbias =0.0 V. The body effect coefficient /spl gamma/ is also reduced to less than 0.02 V/sup 1/2/. Nevertheless, current drives of 886 /spl mu/A//spl mu/m (I/sub off/=15 nA//spl mu/m) in nFET and -320 /spl mu/A//spl mu/m (I/sub off/=10 nA//spl mu/m) in pFET have been achieved in 70-nm gate length SODEL CMOS with |V/sub dd/|=1.2 V. New circuit design schemes are also proposed for high-performance and low-power CMOS applications using the combination of SODEL FETs and bulk FETs on the same chip for 90-nm-node generation and beyond.     

Micropower high-performance SC building block for integrated low-level signal processing

A switched-capacitor instrumentation amplifier which uses correlated-double sampling to reduce the amplifier offset is discussed. Additional offset caused by clock-related charge injection is cancelled by a symmetrical differential circuit topology and a three-phase clocking scheme. An experimental low-power test cell has been integrated, showing 100 /spl mu/V equivalent offset voltage and input noise equal to 270 /spl mu/V. For a fixed gain equal to 10- and 9-kHz sampling frequency, the power dissipation is 36 /spl mu/W (power supply: 5 V); the circuit measures only 0.2 mm/SUP 2/.     

Evanescently coupled photodiodes integrating a double-stage taper for 40-Gb/s applications-compared performance with side-illuminated photodiodes

The design, fabrication, and performance of double-stage taper photodiodes (DSTPs) are reported. The objective of this work is to develop devices compatible with 40-Gb/s applications. Such devices require high efficiency, ultrawide band, high optical power handling capability, and compatibility with low-cost module fabrication. The integration of mode size converters improves both the coupling efficiency and the responsivity with a large fiber mode diameter. Responsivity of 0.6 A/W and 0.45 A/W are achieved with a 6-/spl mu/m fiber mode diameter and cleaved fiber, respectively, providing relaxed alignment tolerances (/spl plusmn/1.6 /spl mu/m and /spl plusmn/2 /spl mu/m, respectively), compatible with cost-effective packaging techniques. DSTPs also offer a wide bandwidth greater than 40 GHz and transverse-electric/transverse-magnetic polarization dependence lower than 0.2 dB. Furthermore, a DSTP saturation current as high as 11 mA results in optical power handling greater than +10 dBm and a high output voltage of 0.8 V. These capabilities allow the photodiode to drive the decision circuit without the need of a broad-band electrical amplifier. The DSTP devices presented here demonstrate higher responsivities with large fiber mode diameter and better optical power handling capabilities and are compared with classical side-illuminated photodiodes.     

A 64-kbit dynamic MOS RAM

A 64-kbit dynamic MOS RAM is developed by using 2 /spl mu/m rule VLSI fabrication technology and low power circuit technology. The 2 /spl mu/m rule VLSI fabrication technology is achieved by improving various aspects of the ultraviolet photolithographic, thin-gate oxidation, arsenic ion implantation, and multilevel interconnection processes. Microminiaturization of the device structure has made the voltage requirements for its MOST threshold voltage and DC supply voltages low. A highly sensitive and low power dissipating sense circuit has been developed for the VLSI RAM. A new level-detecting circuit with a logic threshold which is independent of MOST threshold voltage is proposed. A dynamic address-buffer circuit is also shown. The fabricated 64K RAM has 200 ns of access time, 370 ns of minimum cycle time, and 150 mW of power dissipation under typical supply voltage conditions of V/SUB DD/=7 V and V/SUB BB/=-2 V.     

A high speed molybdenum gate MOS RAM

Describes a high speed 16K molybdenum gate (Mo-gate) dynamic MOS RAM using a single transistor cell. New circuit technologies, including a capacitive-coupled sense-refresh amplifier and a dummy sense circuit, enable the achievement of high speed performance in combination with reduced propagation delay in the molybdenum word line due to the low resistivity. The n-channel Mo-gate process was established by developing an evaporation apparatus and by an improved heat treatment to reduce surface charge density. Ultraviolet photolithography for 2 /spl mu/m patterns and HCl oxidation for 400 /spl Aring/ thick gate oxide are used. The 16K word/spl times/1 bit device is fabricated on a 3.2 mm/spl times/4.0 mm chip. Cell size is 16 /spl mu/m/spl times/16 /spl mu/m Access time is less than 65 ns at V/SUB DD/=7 V and V/SUB BB/=-2 V. Power dissipation is 210 mW at 170 ns read-modify-write (RMW) cycle.     

5-Gbit/s Si integrated regenerative demultiplexer and decision circuit

A silicon bipolar circuit is presented which may be used as either a 1:2 demultiplexer or a decision circuit up to the bit rate of 5 Gb/s. The circuit was fabricated with a standard bipolar technology with oxide-wall isolation, 2-/spl mu/m emitter stripe widths, and a transit frequency of about 9 GHz at V/SUB CE/=1 V. The high-speed performance of the circuit was achieved by applying a double sampling scheme. Clock phase margin (CPM) and decision ambiguity are 120/spl deg/ and 150 mV at 4 Gb/s, respectively. CPM at 5 Gbit/s is about 90%. Decision feedback equalization may be included in the circuit scheme for optional use.     

A 4 /spl mu/m NMOS NAND structure PLA

A PLA of NAND structure, using a NMOS Si gate process, has been developed to minimize chip area and maintain medium fast speed. The smallest memory cell size of 7/spl times/9 /spl mu/m is achieved by using ion implantation for PLA bit programming with 4 /spl mu/m design rules. Dynamic clocking scheme and self-timing circuits which are used in this PLA are described. With PLA size at 20/spl times/20/spl times/20, transistor size of 8 /spl mu/m/4 /spl mu/m, and cell size of 7/spl times/12 /spl mu/m, an internal access time of 150 ns is achieved with an external 4 MHz clock. Measured circuit power dissipation is 20 mW under normal conditions.     

Modeling of the bulk versus SOI CMOS performances for the optimal design of APS circuits in low-power low-voltage applications

In this paper, we present the top-down design of an active pixel sensor (APS) circuit using an analytical model of its architecture. The model is applied to compare the performances of bulk versus silicon-on-insulator (SOI) CMOS processes and devices on the designs and performance of several 50-frames/s imagers in 2-/spl mu/m and 0.25-/spl mu/m CMOS with different pixels array sizes. For 2-/spl mu/m SOI, results show a reduction by two of the power consumption and a dynamic range increase of 0.85 V under a 3-V supply. This results in an SNR of 79 dB instead of 76. Fixed pattern noise (FPN) is also reduced from 2.7 to 1.8 mV which represents 0.26% and 0.08% of the dynamic range, respectively. For 0.25-/spl mu/m CMOS SOI, results show a reduction by 6.5 of the power consumption, FPN more than five time better, and a dynamic range increase of 0.29 V under a 1.5-V supply. However, because of the increase of the thermal noise due to the particular design choice, an SNR of 60.3 dB is achieved compared to 63 in bulk. A better SNR in SOI than in bulk can be achieved but at the expense of power consumption and FPN. However, this could be combined with an increase in pixels number in SOI compared to bulk. Potential results achievable in SOI have to our knowledge never been reached by bulk APS imagers up to now.     

A 16-bit low-voltage CMOS A/D converter

A/D converters used in telemetry, instrumentation, and measurements require high accuracy, excellent linearity, and negligible DC offset, but need not be fast. A simple and robust instrumentation A/D converter, fabricated in a low-voltage 4-/spl mu/m CMOS technology, is described. The measured overall accuracy was 16 bits. Using a digital compensation for parasitic effects, both offset and nonlinearity were below 12 /spl mu/V. With analog compensation, the offset was 60 /spl mu/V and the nonlinearity below 15 /spl mu/V. These results indicate that even higher accuracy can be achieved using higher voltage technology.     

Nonlinear analysis of GaN MESFETs with Volterra series using large-signal models including trapping effects

Nonlinearities in GaN MESFETs are reported using a large-signal physics-based model. The model accounts for the observed current collapse to determine the frequency dispersion of output resistance and transconductance. Calculated f/sub T/ and f/sub max/ of a 0.8 /spl mu/m/spl times/150 /spl mu/m GaN MESFET are 6.5 and 13 GHz, respectively, which are in close agreement with their measured values of 6 and 14 GHz, respectively. A Volterra-series technique is used to calculate size and frequency-dependent nonlinearities. For a 1.5 /spl mu/m/spl times/150 /spl mu/m FET operating at 1 GHz, the 1 dB compression point and output-referred third-order intercept point are 16.3 and 22.2 dBm, respectively. At the same frequency, the corresponding quantities are 19.6 and 30.5 dBm for a. 0.6 /spl mu/m/spl times/150 /spl mu/m FET. Similar improvements in third-order intermodulation for shorter gatelength devices are observed.     

Investigation of breakdown and DC behavior in HBTs with (Al,Ga)As collector layer

We report on the realization of an InGaP-GaAs-based double heterojunction bipolar transistor with high breakdown voltages of up to 85 V using an Al/sub 0.2/Ga/sub 0.8/As collector. These results were achieved with devices with a 2.8 /spl mu/m collector doped to 6/spl times/10/sup 15/ cm/sup -3/ (with an emitter area of 60/spl times/60 /spl mu/m/sup 2/). They agree well with calculated data from a semi-analytical breakdown model. A /spl beta//R/sub SBI/ (intrinsic base sheet resistance) ratio of more than 0.5 by introducing a 150-nm-thick graded Al-content region at the base-collector heterojunction was achieved. This layer is needed to efficiently suppress current blocking, which is otherwise caused by the conduction band offset from GaAs to Al/sub 0.2/Ga/sub 0.8/As. The thickness of this region was determined by two-dimensional numerical device simulations that are in good agreement with the measured device properties.     

A Micropower CMOS-Instrumentation Amplifier

A CMOS switched capacitor instrumentation amplifier is presented. Offset is reduced by an auto-zero technique and effects due to charge injection are attenuated by a special amplifier configuration. The circuit which is realized in a 4-/spl mu/m double poly process has an offset (/spl tau/) of 370 /spl mu/V, an rms input referred integrated noise (0.5 -f/sub c//2) of 79 /spl mu/V, and consumes only 21 /spl mu/W (f/sub c/ = 8 kHz, V/sub DD/ = 3 V).